Australian English: Its Evolution and Current State

T his paper provides a critical overview of research on Australian English (‘AusE’), and of the vexing questions that the research has grappled with. These include: What is the historical explanation for the homogeneity of the Australian accent? Was it formed by the ﬁ rst generation of native-born Australians in the ‘Sydney mixing bowl’, its spread subsequently facilitated by high population mobility? Or is the answer to be found in sociolinguistic reconstructions of the early colony suggesting that a uniform London English was transplanted to Australia in 1788 and that speakers of other dialects quickly adapted to it? How is Australia’s national identity embodied in its lexicon, and to what extent is it currently under the inﬂ uence of external pressure from American English? What are the most distinctive structural features of AusE phonology, morphosyntax and discourse? To what extent do allegedly unique Australian features such as sentence-ﬁ nal but and yeah-no in discourse serve the social role of indexing ‘Australianness’? What is the nature and extent of variation – regional, social and ethnic – in contemporary AusE? Are such regional phonological diﬀ erences as /æ/~/a/ variation increasing or diminishing? Does there exist a pan-ethnic variety of AusE that is particularly associated with younger Australians of second generation Middle Eastern and Mediterranean background? Has contemporary AusE consolidated its own norms as an independent national standard?

T his paper will concentrate on recent research on 'AusE'. Traditionally, AusE is thought of as the dialect spoken by native-born non-Aboriginal Australians (q.v. Ramson 1970;Collins & Blair 1989;Blair & Collins 2001), and contrasted with 'English in Australia', a term understood to encompass AusE along with varieties associated with the community groups of various non-English migrant backgrounds and the English of Aboriginal communities. The paper contains an overview of theories of the evolution of AusE (in Section 2), a discussion of the lexicon as it embodies the self-perception of the Australian people (Section 3), a description of the structure of AusE at the levels of phonology, morphology, grammar and discourse (Section 4), and fi nally an account of regional, social and ethnic variation in AusE.

Historical Evolution
I t would be diffi cult to think of any other country in which one could travel the vast distances that separate Sydney from Perth, and Darwin from Hobart, and encounter so relatively little regional variation. A number of historical explanations have been advanced to account for this homogeneity. The most plausible is generally considered to be the 'Sydney Mixing Bowl' theory (e.g. Bernard 1969, Trudgill 1986) supported by evidence of a high degree of population mobility. Bernard argues that AusE began with the fi rst generation of native-born Australians, a byproduct of the social situation in the early colony. The new dialect, whose accent Bernard labels 'proto-Broad', provided the basis for the milder accent varieties which developed subsequently in response to adverse social evaluations of proto-Broad and a perception by Australians that with growing prosperity and education their speech needed 'upgrading'. Bernard accounts for the homogeneity of AusE via the independent generation of proto-Broad in the regional centres that developed from coastal ports.
Alternative positions are argued by scholars like Hammarström (1980) and Gunn (1992), who believe that a uniform London English was transplanted to Australia in 1788 and that speakers of other dialects quickly adapted to it, and Horvath (1985), who provides a sociolinguistic reconstruction of AusE based on a historical reconstruction of the social conditions in the colony. Horvath observes the presence of sharp socio-economic diff erences in the colony, predicting that there would have been extensive linguistic variation from the outset. Horvath also rejects Bernard's explanation of the uniformity of AusE, arguing that the social circumstances in the major coastal centres would have varied greatly (New South Wales and Tasmania having been convict colonies while Victoria, South Australia, and Western Australia were not), and would have been unlikely to lead consistently to a unique set of linguistic features. Horvath's preferred explanation for the uniformity of AusE is based on the extensive population mobility that is attested to in the historical records.
While the study of the historical development of AusE fi rst began to gather momentum in the 1960s, it has more recently enjoyed an upsurge of interest with the posthumous editing of A.G. Mitchell's unfi nished ms by Yallop (see Yallop 2001) and such publications as Leitner (2004) and Fritz (2007). While the focus has traditionally been primarily lexical (e.g. Ramson 1966) and phonological (e.g. Mitchell & Delbridge 1965), some recent research has examined diachronic dimensions of AusE morphosyntax (see for example Peters, Collins & Smith 2009).

The Lexicon
T he Australian lexicon embodies the attitudes, values and self-perception of Australians, as evidenced in the preservation of value-laden words such as mateship (a code of conduct built upon solidarity and fellowship between males) and larrikin (a mischievous young person). A widely recognized feature of the lexicon is its informality, often manifested in an understated and irreverent humour (for instance in similes such as drunk as a skunk 'very drunk', and tight as a fi sh's arsehole 'parsimonious'). Australians are renowned for their colloquial creativity, sometimes generating sets of vulgar expressions built on a single stem: scared shitless 'very scared'; shit a brick! an expression of surprise; up shit creek 'in a diffi cult predicament'; built like a brick shithouse 'strongly built'; bullshit artist 'one who tells lies'; and shit-faced 'drunk'. It is observed by Leitner (2004: 338) that the Australian penchant for such colloquial usage transcends the merely covert prestige enjoyed by its counterparts in British and American English.
The affi nity shown by Australians for their vernacular can be traced back to the often coarse and irreverent language, originating in British dialectal slang, used by both the convicts and settlers from 1788. In Australia, as elsewhere, it has traditionally been males who swear more and use more obscene language than females. At the time of Taylor's (1976) study of swearing and abusive language in AusE, he was able to identify persistently marked gender diff erences, though it is clear that the gap between male and female use and attitudes has narrowed somewhat since the 1970s. Australians again evidence a good deal of creativity, as for example in the colourful compounds that are used to derogate others (e.g. shithead, deadshit, shitkicker, bullshitter; arsehole, arse-licker, smartarse, slackarse, tightarse) and the comparative expressions that are used to target people's physical appearance, mental ability, or various other character traits (e.g. ugly as a shithouse rat; as popular as a turd in a fruit salad; lower than a snake's belly; silly as a chook with its head cut off (see further Seal 1999: 122-123).
Swearing also serves as a means of reinforcing in-group solidarity in AusE. For example the use of bastard in aff ectionate phrases such as old bastard and silly bastard in AusE represents one application of the principle enunciated by Allan and Burridge (2009: 371) that "the more aff ectionate they [=speakers] feel towards someone, the more abusive the language can be towards that person". Perhaps the most extreme manifestation of the solidarity function is 'ritual insulting', as exemplifi ed in exchanges of the following type: A: If I had a pussy like yours I'd take it to the cat's home and have it put down. B: If I had brains like yours I'd ask for a refund (from Allen 1987: 62). Another common function of swearing in AusE is discourse-stylistic. The swearword most closely associated with this function in AusE is the 'great Australian adjective', bloody. In the following corpus examples from Allan & Burridge, bloody is used merely as an intensifi er, bleached of its taboo quality and without its standard force: It's a bloody crocodile!; You're driving too bloody fast!; It's turned bloody red! (from Allan & Burridge 2009: 376).
A signifi cant lexical development in recent decades has been the borrowing of words and expressions such as cookie, guy, and dude from American English, prompted by rapid developments in communication and American cultural imperialism. Opinions diff er as to the nature and extent of American infl uence on contemporary AusE. Members of the public and journalists regularly suggest that AusE is merely a passive receptacle for Americanisms (Taylor 2001), while linguists tend to regard Australian borrowings from American English as selective (Peters 1994). Taylor's (1989) research shows that American English infl uence 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. fi nANCE to FINance, reSEARCH to REsearch). Graphologically, simplifi cation of digraphs such as <ae> and <oe> as in medieval and fetal follows American practice. Syntactically, Taylor notes, amongst other things, the American-infl uenced elision of the in structures of the type I play (the) piano.
Another development that has aff ected the lexicon in recent decades is the infl ux of words associated with Aboriginal culture that refl ects the development of Aboriginal activism and a growing interest in Aboriginal languages and culture amongst white Australians (Moore 2008). Examples include native title and Mabo, which entered AusE following the High Court's decision in 1992 to recognize 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).

Phonology
P ioneering work on the phonology of AusE was conducted by Alexander G.
Mitchell in the 1940s, and subsequently in the early 1960s by Mitchell and his colleague Arthur Delbridge (Mitchell & Delbridge 1965). Mitchell and Delbridge identifi ed a spectrum of pronunciations and subsequently three points on the spectrum which they labelled 'Cultivated', 'General' and 'Broad' AusE and which have remained the standard descriptors of the range of phonological variation to this day. Cultivated, spoken by 11% of their subjects, was noted to be the most prestigious; Broad, spoken by about one third (34%), has the least prestige and has the most distinctively Australian characteristics; General, which falls between these, was used by the majority (55%) and appears to be expanding at the expense of the other two. According to Mitchell & Delbridge the three varieties are distinguished primarily in the realization of the FLEECE, GOOSE, FACE, GOAT, PRICE and MOUTH vowels.
Subsequent acoustic accounts of AusE, which have focused mainly on vowels, include those by Bernard (1989), which was the source of the pronunciations supplied in the Macquarie Dictionary, and by Harrington, Cox & Evans (1997). More recent studies have confi rmed that there are some consonantal pronunciations that diff erentiate speakers of AusE, including the vocalization of postvocalic /l/ (which is promoted by the backness of adjacent sounds in combination with syllable position: see Borowsky (2001).
Several instances of diachronic variation in the phonology of AusE have been identifi ed in recent work by Cox & Palethorpe (2001). One such change is /ae/-lowering, which Cox & Palethorpe suggest may be driven by sociolinguistic hypercorrection occurring in response to the availability of a perceived prestigious alternative, /a/, in words such as dance and chance. Other types of phonological variation that have attracted attention -regional, social and ethnic -are discussed in Section 5 below.
An intonational variable that has attracted a good deal of attention from linguists (e.g. Allan 1984;Guy & Vonwiller 1989) is the so-called 'Australian questioning intonation' ('AQI'), the use of a high rising tone used in declarative rather than interrogative clauses. The following text (from Guy and Vonwiller 1989: 23) was produced -as is often the case with the AQI -by a teenage girl: Oh, occasionally Mrs L used to blow up kids when they hadn't done anything. And once, a girl and I were walking down the stairs, and she touched a doorknob or something, 'cause she didn't realize what was wrong with it / And it fell off / and she got the cane for breaking it / And I knew very well she hadn't broken it / And I tried to tell the teacher. The teacher was really mean, you know.
The AQI has been suggested by some to indicate uncertainty and deference, implications deriving from stereotyped social evaluations of the speakers who use it and from the systematic meanings of intonational rises. There is general agreement, however, that its primary linguistic function is to seek verifi cation of the listener's comprehension (in much the same way that the discourse marker you know functions in speech). This function provides a clue as to why the AQI is relatively more popular in narratives and descriptions -where it appears to be used to monitor the listener's active engagement as the text unfolds -than it is in other genres.

Grammar and Discourse
In grammar and discourse there are rarely usages that are restricted to AusE. One that comes close is the use of the present perfect as a 'quasi-preterite'. The standard uses of the perfect aspect have been studied by Elsness (2009), who fi nds that it is losing ground to the preterite in both AmE and BrE (the former having the lower frequency), with AusE lagging behind in this development. However there is hypocoristic suffixation with -ie as in tummy 'stomach', and -o as in garbo another development in which AusE appears to be leading the way, the use of the present perfect in past time contexts where it indicates narrative progression, and either occurs with a defi nite past-referring adverbial or has the potential to do so, as in He walked up to her and then he's pushed her off the chair. In their study of this phenomenon in Australian radio chat shows Engel & Ritz (2000) comment on its comparatively high frequency and alternation with both the simple past tense (versus pushed, in the example above) and the 'historic present' (versus pushes).
Another category of the verb system that has attracted attention is modality. The modal auxiliaries and quasi-modals have been studied by Collins (2009), who cites independent evidence that while the quasi-modals have been on the rise in recent decades, the modals have been in decline. According to Collins, AusE is more advanced in these developments than British and New Zealand English (but less advanced than American English). Collins fi nds distinctiveness in the extreme distaste for the modals shall and ought in AusE compared to AmE and BrE, in the acceptance of negative epistemic mustn't (as in the attested example, He mustn't have wanted the coupons because he came up and give them to me), and in the tolerance of may as an alternative to might in the expression of past possibility (as in I suspect that she may have fallen asleep) and hypothetical possibility (as in She may be upset if you wake her).
A discourse usage that comes close to being distinctively Australian is sentencefi nal but, as exemplifi ed in: [I used to be the Under-17 champion.] I'm a bit out of practice but (from Mulder, Thompson & Penry-Williams 2009). Mulder et al., who distinguish 'fi nal particle but' from 'fi nal hanging but', claim that the former is a fully grammaticalized discourse particle marking contrast/concession, and that it serves the social role of indexing 'Australianness', a claim apparently supported by the extensive folklinguistic comment that fi nal but has attracted and by its common use in fi ctional dialogue evoking localness.
A discourse marker which is of relatively recent origins in AusE and which is not attested in other Englishes, is yeah-no. As the following attested example demonstrates, yeah-no is used where there is agreement yet the speaker wishes to make a negative response to remove any possibility of contradiction: he's a really good kid (…) he just comes over and we spoil him rotten. Yeah-no he's a good kid (from Burridge & Florey 2002: 150). Burridge & Florey argue that yeah-no operates to reinforce conversational solidarity on a number of levels simultaneously. Propositionally, it expresses a combination of assent and dissent, enabling speakers to comply with the conversational desiderata of agreement and compromise. Pragmatically, it softens the force of an utterance, enabling speakers to hedge and maintain face. Finally, discoursally, it links together speakers' turns, enabling them to maintain cohesion and rapport.

Regional Variation
R egional phonological diff erences in AusE are small but growing (see Cox & Palethorpe 2001). The variation between /ae/ and /a/ in words such as dance and chance respectively appears to refl ect the chronology of this change in southeastern England. Places such as Sydney, Hobart and Brisbane, settled by the early nineteenth century and mainly by people of lower socioeconomic status, use more /a/; Melbourne, settled in the mid-nineteenth century with a more mixed population, has a higher proportion of /ae/; and Adelaide, which was settled in the mid-nineteenth century by people of middle or higher socio-economic status, uses the highest proportion of /a/ (see Bradley 2008). Another regionally marked variant is the front-of-central rounded onset of the GOAT and GOOSE vowels found in Adelaide and elsewhere in South Australia (Bradley 2008;Oasa 1989). Another is the vowel merger of the FLEECE and KIT vowels resulting in homophony between, for example, deal and dill, which is most advanced in Adelaide and Hobart, less so in Sydney and Brisbane, and least in Melbourne. Finally, consider the frequency of /l/ vocalization across Australia's capital cities: according to Horvath & Horvath (2001), Sydney and Hobart are the next most frequent /l/ vocalizers behind Adelaide, with Brisbane and Melbourne the least frequent. A further interesting fi nding which has not been subjected to more recent (re)investigation is that by Mitchell & Delbridge (1965: 39) of a strong tendency for broad vowels to be used with greater frequency by their rural informants (43%) than by their urban counterparts (23%). Conversely, only 4% of their informants outside the capital cities were found to use cultivated vowels, as against 19% of their urban counterparts.
The best-known work on regional lexical usage is that conducted by Bryant (1989Bryant ( , 1991Bryant ( , 1997, who observes that because regional lexical variation in AusE is relatively unobtrusive, it has not attracted substantial research interest. Many of the items in question are mundane in nature and form sets of synonyms or nearsynonyms, one of the best-known examples being the various terms for the large, smooth, bland sausage with red skin that is usually thinly sliced and eaten cold: devon, German sausage, pork German, Strasburg/Stras, polony, Belgian sausage, and fritz. Another is the set of terms for a type of children's play equipment: slippery dip, slippery slide, and slide. Although speakers usually associate regionally distinctive words with interstate diff erences, Bryant's research identifi es major areas of lexical usage whose boundaries do not correspond exactly with state borders (for instance, her 'South East' area includes Victoria and Tasmania along with parts of South Australia and New South Wales).

Social Variation
A number of studies have documented the sensitivity of linguistic variation in AusE to socioeconomic factors. The fi rst linguist to systematically investigate the covariation between AusE phonology and social factors was Horvath (1985), in her study of the social class, gender and ethnic parameters of the Cultivated-Broad continuum. Using principal components analysis Horvath identifi ed a sociolectal continuum comprising four varieties, with social class and gender being the primary determinants. Working class men dominated the Broad end of the continuum and middle class women the Cultivated end.
Phonological variables have also been investigated, more recently, by Bradley (2008) and Horvath (2008), amongst others. Bradley's fi ndings include the discovery that the monophthongal variants of /r/-fi nal words such as near and cure -as opposed to off glided variants -are used more in casual speech, by lower socio-economic class speakers and by males, and furthermore that they are more frequent in Sydney than Melbourne. One of Horvath's fi ndings is that the palatalization of the phonemes /t d s z/ when they occur before /u/ in words such as assume (/əsjum/ -> /ə∫um/) is more commonly found with those of lower socio-economic status, especially younger males.
Diff erential practices among younger and older speakers have been noted in AusE. For instance, Collins & Peters (2008) report a survey of verb morphology which found that younger Australians are more likely than older Australians to simplify verb paradigms via use of a preterite form which is homonymous with the past participle. For more than two-thirds of the under-25s in this survey, shrunk, sunk, sprung were in use as preterite forms, as opposed to about half of those under 45. As another example consider the age-based variation observed by Taylor (1989) in his study of changes prompted in AusE by American English infl uence, including the tendency for younger Australians to pronounce schedule with /sk-/ rather than /∫-/, and address with the stress on the fi rst syllable.
Another infl uential variable in AusE is gender. For instance, studies by Shnukal (1982) and Shopen (1978) confi rm that it is women who favour the standard /ŋ/ variant in the variable pronunciation of the present participial suffi x -ing with /ŋ/ as in running or /n/ as in runnin', and of -thing in the compounds anything, everything, nothing and something with a fi nal /ŋ/ or /k/. Chevalier (2006) detects gender-related variation in some features of lexical morphology: for instance, the suffi x x -i is used far more commonly with female than male given names, and conversely for the suffix -o.
One type of AusE speaker in whom the most essential embodiment of traditional Australian ideals is to be found is the lower socio-economic class male. Stereotyped portrayals are regularly propagated by actors and comedians, such as the naive larrikins Crocodile Dundee (Paul Hogan) and Bazza McKenzie (Barry Humphries). Such speakers are the most paradigmatic users of non-standard AusE (see, e.g., Eisikovits 1989, Pawley 2008. Characteristic non-standard features include the normally plural but occasionally singular second person pronoun yous; feminine gender assignment to inanimate entities (e.g. She's a rough sea today), use of the negative auxiliary don't for standard doesn't (e.g. It don't fi t our plan); dropping of the perfect auxiliary have (e.g. I only been there a coupla times); and double negatives (e.g. I never said nothing for a while). Working class males are also most frequently associated with a number of phonological variables, such as the palatalization of the phonemes /t d s z/ when they occur before /u/ in words such as assume (Bradley 2008).

Ethnic Variation
The ethnicity of many second and later generation Australians is refl ected in their adoption of 'ethnolects', varieties which enable them to express their linguistic identity and demonstrate solidarity with their ethnic group, and which display varying levels of interference from the native language (such as loss of plural infl ections on nouns and tense infl ections on verbs): see Clyne, Eisikovits & Tollfree (2001).
the Despite the fact that migrants of non-English-speaking background do not constitute a single group, a number of commentators have identifi ed a pan-ethnic variety popularly known as 'Wogspeak' (Warren 1999;Kiesling 2005). It is associated particularly with young Australians of second generation Middle Eastern and Mediterranean background who use it to diff erentiate themselves from both their parents' values and those of the Anglo host culture. Characteristic phonological features include the avoidance of reduced vowels (as in the use of [a] in the fi nal syllable of a word such as pleasure), the replacement of /θ/ and /ð/ by /t/ and /d/ (see Warren 1999;Clyne, Eisikovits & Tollfree 2001).

Conclusion
A usE is, in its origins, a set of dialects transported by British convicts and immigrants which interacted with indigenous and migrant languages, American English and other varieties of English, ultimately emerging as an independent variety embodying the country's national identity. The distinctive features of AusE suggest that it has consolidated its own norms as an independent national standard. AusE is today recognized as a major variety of English, one for which there may be an increasing role as an epicentre in the Asia-Pacifi c region (Leitner 2004).