The Macquarie Dictionary , its History and its Editorial Practices

The Macquarie Dictionary , first published in Sydney in 1981, was intended to be the first comprehensive dictionary of Australian English. Now in its third edition it has been widely adopted by institutions and the general community as the national dictionary. This paper traces its development from a difficult birth to its present maturity, from a large set of cards to an electronic database, from a single book to a lexicographic library. The rationale and the methodology are laid out along with an account of the reception given to the dictionary in Australia and internationally. Keywords: dictionary; lexicography; macquarie dictionary; australianness; australianise; australian english; national dictionaries; phonology; international phonetic alphabet; lexical labels; lexicographic style; language style; colloquial; aboriginal words; electronic database; corpus; citation

being one of them.The concept Australian English is relatively new, first introduced as recently as the 1940s.But by the 1990s it was securely entrenched in popular usage and a basic reference point even in the language policies of the Australian Federal Government.An account of this development may be found in two articles by Delbridge (1992 and1999).idiom in passages of dialogue, but then to explain to readers in England what these strange expressions might mean (Webby 1988).One dictionary was published, with the aim of listing "all the new words and new uses of words that have been added to the English language by reason of the fact that those who speak English have taken up their abode in Australia, Tasmania, and New Zealand" (Morris 1898: xi).
That dictionary was not enough.Well into the 20th century, Australians have been without a comprehensive dictionary that focused on their own variety of English, not just its slangy idioms, but on all of its English, a book in which Australianisms did not have to be labelled as if they were peculiarities outside the range of normal English, and in which Australian English might become the basis of comparison with other national varieties of English.The dictionaries available in Australian bookshops for general use in the adult population were predominantly British -those from Oxford, Chambers, Longmans, etc.American dictionaries were not in wide use, since American English itself was largely but ignorantly deprecated by the Australian public, as was their own local variety.The history of the Macquarie Dictionary is the story of an attempt to supply what had been lacking, and to promote acknowledgment of the local variety, with a comprehensive dictionary.Not the first attempt, certainly, but the only one of the early attempts which actually attained publication.

How it began
As with so many projects which come to fill if not dominate the lives of those who participate, this one began with a fairly casual offer from a publisher.It came, late in 1969, from Brian Clouston, the head of the Jacaranda Press, Brisbane, a publishing house dealing principally with educational books.It was casual in the sense that he did not consult his board before making an offer, and the offer did not include a contract.Made to Arthur Delbridge, who was the foundation professor of linguistics in Macquarie University, it proposed the writing of "an aggressively Australian dictionary", to be finished and ready for publication in two years' time, with funds enough to encourage the participation of a small group of academic editors, a support staff, and office accommodation off campus.The university itself was not involved in the initial arrangement, though clearly it needed to give permission to its own academic staff who became involved.More importantly, the offer included the use of a dictionary of suitable size, the Encyclopedic World Dictionary (EWD), as a base on which the new work could be developed.This dictionary had itself been based on the American College Dictionary (ACD) , with Patrick Hanks as editor and the Hamlyn Publishing Group Ltd as publisher.It was published in London in 1971.The ACD was never on sale within Australia, and the subsequent work on the proposed Jacaranda Dictionary (to give the Macquarie Dictionary its first intended name) ignored it completely in favour of EWD.EWD itself had been available to Australian users only through a successful mail-order campaign of short duration in 1972.It was a revision of an American dictionary by a British lexicographer whose first concern was "to describe the vocabulary common to all or most brands of English in the fullest and clearest possible form" (Hanks 1971: Introduction).It was not in any sense a national dictionary, but it proved to be Australia's indispensable base dictionary.

Which way to go?
Work began with a group of four academic editors, Arthur Delbridge as chairman of an editorial committee, and J.R.L. Bernard, D. Blair and W.S. Ramson the other members, the first three all teachers of linguistics at Macquarie University, Sydney, and Ramson at the Australian National University, Canberra.All had had lexicographic experience, especially as consultants to English or American publishers, contributing Australian lexical material to more or less international dictionaries whose editors wished to have an Australian component within their wordlists.Now for the first time the members were confronted with the job of initiating a policy, considering a style, working out a method of data collection, assuring authenticity, and securing a readership.The possession of a base dictionary, as of right, both simplified and complicated these tasks.Nobody believed the would-be publisher's time scale of two years to be achievable for a book modelled in some degree on another which was almost 2 000 pages long.And nobody wanted it to be patently aggressive in delineating Australian English.Indeed, when the work was finished, and a publisher found, he refused to have the words Australian English in the title, fearing that buyers might be put off by an expectation that this was nothing more than an account of local Australian slang (which, though widely used, was commonly deprecated).
There was a view current among scholars working in the field of Australian English that the first priority was for a historical dictionary to be written.The Sydney University Australian Language Research Centre, founded in 1962, believed the only excuse for writing a general descriptive dictionary of current Australian English was that money from its sales would finance the research needed for a scholarly historical account of Australianisms.But the Macquarie team took a different view: the most urgent need was for a people's dictionary that would hold up the mirror directly to contemporary Australian speech and writing as its first object of study.When eventually a historical dictionary was produced, it was edited solely by Ramson (1988), whose contribution to the first edition of the Macquarie Dictionary had, as he put it in a prefatory essay, presented "the fullest available record of Australian additions to and modifications of the vocabulary of English" (Delbridge et al. 1981: 37).This meant that the users of that edition, as of subsequent editions, had more than an account of contemporary Australian English: they had also a key to the colonial, 19thcentury vocabulary to be met in public records and works of literature.
Towards the first edition  Closer inspection convinced the editors that EWD was a very good dictionary, likely to serve well as a base.Since it was obvious that much of it would be retained in the new dictionary, the editors' first decision was that the EWD style would be largely followed in matters of layout.In those early days English lexicography had not been influenced by the COBUILD defining style, and in any case this was not to be a leamer's dictionary, but one to be used by those members of the Australian community who felt the need for a comprehensive coverage of Australian English in dictionary form.So the EWD defining style was also adopted, with modifications to be made later in the light of experience.

(a)
The word list Those parts of its wordlist that belonged to the common core of English around the world could well be taken over into the new dictionary, updated or otherwise modified.But many of the common words of world English have uses in Australia that were not adequately covered in EWD.Also individual judgments had to be made about deleting some entries relating to aspects of life in England, Scotland, South Africa, etc.The EWD had apparently had an unbridled African consultant whose enthusiasm for geelslangs (Cape cobras), geelbeks (yellow bill ducks) and maasbankers (horse mackerel) now had to be restrained, in order to make room for geebungs, wombats and echnidasnot out of invincible parochialism, but for space.The publisher had specified a single volume, with a page size like EWD's, and an eye-friendly typography, so to fit in a comprehensive account of Australian words, idioms and meanings, room had to be created, by excision.
Then there were the specialist areas: the sciences, the technologies, the arts, law, sport, language and languages (especially Aboriginal languages), economics, politics, education, religion, and many others.All of these had undergone local modification throughout Australian history, and some had uniquely Australian features.For editorial purposes all the related words and phrases had to be recovered from the base dictionary, whether adequately labelled there as belonging to one of the special fields or not, and submitted to properly instructed local specialists, for judgments for or against their retention, with or without updating and other modifications.And specifically Australian words and phrases in these specialist fields had to be added.The committee's major task, however, was to assemble the ordinary words, phrases and idioms of Australian English, attending to the whole range of social styles and registers, using both written and speech texts as data.For any headword entered, policy decisions had to be made as to pronunciation, spelling, grammatical status, etymology and labelling.

Pronunciation
All the editors had been involved in major studies of the phonological patterns of Australian English across the community.So, contrary to general practice except in leamer's dictionaries, they decided to supply an Australian pronunciation, with variants, for every headword entry, using a suitable version of the International Phonetic Alphabet.The transcription symbols were to be phonemic, each accommodating within a single symbol the whole spectrum of variation found among Australian speakers.Thus not one of the three major categories of Australian pronunciation, Broad, General and Cultivated, would be favoured over the others, and every user could interpret the transcription in terms of their own position in the spectrum (Mitchell and Delbridge 1965).In the first edition of the dictionary (Delbridge et al. 1981) a detailed account of the pronunciation of Australian English and its treatment in the dictionary is given in a prefatory essay by J.R.L. Bernard (Delbridge et al. 1981: 18-27). (c)

Spelling
Nineteenth-century British spelling prevailed in the Australian colonies, and has continued to prevail in classrooms throughout Australia ever since.However, American influence has also become strong, with the result that Australian spelling shows the influence of two schools.But it is not a patterned influence, and individual words yield to it, or not, in a way that betokens a sort of independence in Australian writers.One result of this is that the Macquarie editors need to acknowledge variants in spelling whenever they are established in competent use.To do this securely they need to collect a lot of data to support their acknowledgment of variants which might, even so, displease the conservatives of the community.It was only after the publication of the first edition that they saw that the picture was not as clear as they had assumed.As subsequent corpus studies then showed, the effects of American English on British spelling norms in Australia were extensive but by no means regular.Every word had to be treated individually as to its variants.The elaboration of their treatment of this variance is evident from the second edition onwards.

Labelling
The EWD used a set of three types of labels: subject labels, regional labels and language labels.Attempting to follow this pattern the editors found that the subject labels presented a problem in the open-endedness of the list.They economised by distinguishing between single-definition and multi-definition entries.For the latter, labels were used to act as signposts to alert the reader of a change from one context or field to another.They rewrote the single-definition entries so that they could obviously stand without a subject label.
The editors were not persuaded that the EWD language labels slang, taboo, and taboo slang were capable of consistent use.They did not always reflect contemporary mores, at any rate in Australia where the wilder shores of vocabulary tend to be quite familiar territory.They felt that the users of this dictionary, who would not be learners, should not have too specific a categorisation applied to some of their familiar words.The word slang is not easily defined, and (to quote the editor of Webster's Third New International Dictionary) "no word is invariably slang" (Gove 1971: 19a).The editors decided instead to use the label colloquial for a wide range of informal expressions, leaving it to the user to decide appropriateness in particular social settings.(This view was modified but not disowned in the third edition.)Some headwords have restrictive labelling in the form of the symbol *.
This indicates that the word itself may give offence essentially because of its taboo nature.The use of two symbols (* *), applied to a small group of words, indicates an extreme degree of this restriction.A single symbol m is also used if there is a particularly crass and offensive meaning given to a usually neutral word.Taboo words are to be differentiated from words which are intended to offend, in particular racist terms, which are labelled derogatory (Delbridge et al. 1997: xxi). (e)

Grammar
The editors agreed that the grammatical categorisations in EWD, even though they did not reflect the mid-century grammatical stirrings among linguists, were nevertheless safe, useful and familiar.And with so much else to do (in two years?)they gratefully deferred modifying the EWD system, at least for the time being.The EWD practice of giving part-of-speech labels only to singleword entries was extended to other entries such as compound nouns. (f)

Etymology
The EWD etymologies were as supplied by Kemp Malone, and with some modifications of style they were retained for the words taken into the new dictionary from the base dictionary.Among the added words, many of them Australianisms, etymologies were written in the same style.But for the words taken into English through contact with the many Aboriginal tribes encountered throughout Australia in the processes of settlement there was a difficulty.Only some of these languages had been closely studied, either by linguists or missionaries; some no longer had any living speakers; none of them were written languages.So writing the etymology of words like myall, coolibah, brigalow and boong presented difficulties -so often there was no information.The editors had to wait until the late 1980s for research that could specify the Australian language from which each of such borrowings had come (Dixon et al. 1990).So for the first edition their etymology consisted only of the word [Aboriginal], but in the second and subsequent editions it has been possible, in the light of new research in Australian languages, to name the source language of each of the Aboriginal borrowings.

History of the first edition up to publication
The primary research task was to find the new material to be added to the base dictionary so as to produce a comprehensive record of the language.The primary editorial task was to slot the research material into a dictionary structure and a typographical plan.In later times the editors separated the dictionary structure from the ultimate appearance of the text on a printed page, but in this early stage the two were indistinguishable: their concept of the structure of an entry came from an analysis of the printed form of the base dictionary.Typeface, positioning, punctuation, etc. -all had lexicographic meaning.For example, there was a pattern in the etymologies comprising a language name followed by a colon followed by a gloss, indicating that the word in the language of origin had the same form as the headword and therefore did not need to be repeated, a subtlety that has escaped the notice of many readers of the dictionary over the years.Belief in the acceptability of the total structure was challenged once the transfer was made from cards to computer typesetting.What had seemed perfectly comprehensible to the editors seemed quite illogical to the programmer, who nevertheless felt that his task was to database the dictionary as it had been structured.Later programmers were not so respectful.It was not until the third edition that pressure came for the editors to comply with a logical database structure that was uniform throughout the whole range of Macquarie dictionaries.The original database was servant to the book, but by the mid-90s the book has become the incidental by-product of the database.
When work was finished for the first edition, it was still all on cards, but photocopied for security.In 1974 Jacaranda had a new British owner who was pressing for immediate publication in spite of the editorial committee's assertion that the work was not yet ready.But the message was clear: to publish immediately.The cards and the photocopy were sent to the publishing house next to the Brisbane River which, in a record flood, came within inches of destroying the material.A portion of the manuscript was sent to three international referees for judgment.All three commented favourably, commending the decision on the use of IPA for pronunciations, recognising the value of the strong Australian component, accepting the justification for the colloquial label, and agreeing that this was essentially a new text being prepared to meet a welldefined need.Useful suggestions were made and in time followed.In the meantime Jacaranda decided to import an English editor experienced in lexicography, who would take over the manuscript and prepare it for immediate publication.But in 1975 the whole project was closed down, its support staff dismissed, and the manuscript stored away in Jacaranda's Brisbane premises.
There were two resolves: to get the work finished, and to find a publisher for it, Jacaranda or some other.For the first of these, Macquarie University stepped in.It had earlier approved of the dictionary involvement of Delbridge, Bernard and Blair over the years, on the understanding that the work was part of the research commitment expected of academic staff.There had never been an expectation on either side that any of them would be relieved of any portion of their teaching and administrative duties.But the University did not want to see the years of work go for nothing, and it had begun to see ultimate publication as good for the university too.So the university made a formal agreement with Jacaranda to become joint owners of the manuscript with them, and to provide funds at least for the employment of Susan Butler as senior research assistant on the project for two years, for no one knew the state of the manuscript better than she already did.The agreement gave Jacaranda the option to publish at the end of the two years that now seemed necessary to finish the work.If Jacaranda did not publish then, another publisher would be sought.
It became apparent that Jacaranda would not proceed.It now had an American owner, Wiley of New York, who was surprised but pleased to discover that what he had just bought included a dictionary manuscript almost ready for publication.But the estimated cost of producing the book was one and a half million Australian dollars, and the Jacaranda board of management found the terms offered for an American commitment of this size unacceptable.Jacaranda indicated that it had to decline to publish a work it had supported since 1970.
In the meantime Jacaranda had been making unsuccessful approaches for financial support elsewhere, including one to the Federal Government.Its response was negative.So it was left to the editorial committee to find its own publisher, with useful input from both Macquarie University and the Australian National University, where Ramson was a teacher.Among those approached there was no lack of enthusiasm for the notion of a locally produced dictionary focused on the use of English in Australia.The fundamental problem was a doubt among publishers that a new and local dictionary could find favour from a public already accustomed to buying overseas dictionaries, especially the Oxford dictionaries of various sizes.Market research showed that it would take five years to recover from sales the initial cost of paper, printing, binding and distribution.The book was unlikely to have any significant international sales, and the financial returns would be too small and too slow.This was the verdict of all the publishers approached, whether they were Australian branches of international companies, or purely local.Finally, a consortium was attempted, including the Australian Broadcasting Commission (ABC), the Literature Board of the Australia Council for the Arts, the Australian National University Press (whose current manager was the same Brian Clouston who had first approached Delbridge with the original idea), and the Macquarie University.Under this plan the Literature Board would give some funds, the Broadcasting Commission would arrange distribution, the AND Press would produce the book, and Macquarie University would provide computing equipment with which the editors could tum its thousands of cards into electronic form and produce the dictionary camera-ready (all 2000 pages of it) for printing and binding in Canberra, with all the expected royalties from sales to go towards the work of producing a historical dictionary of Australian English also in Canberra.This plan failed, initially because the Broadcasting Commission found that its charter allowed book publication only of text arising directly from material it had already broadcast.Academic editors and staff could also not produce by a new technology a large typographically complex book which established publishers had already rejected.
But in 1978 a new player arrived on the scene.He was Kevin Weldon, who had been manager of Hamlyn Australia in Brisbane in time to make that brief mail-order promotion of the EWD.He knew of the work and was about to launch himself as a publishing proprietor.He was, and indeed still is, a strong supporter of Australian enterprise, and it suited his plans as an entrepreneurial publisher to have in his portfolio a major reference work associated with a respected Australian university.He asked for some months to look at the possibilities, then dealing directly with the university he agreed to publish the work, with Macquarie Dictionary as its name, and to set up a company which, would distribute it by mail order through one major capital city newspaper in each state of Australia.A period of readying the work followed, concentrating on local development of electronic programs for a dictionary database that would not only facilitate printing but also the work of revision and .correctionfor later editions, as well as the production of spin-off dictionaries.

Method of compilation
For the first edition of the dictionary the starting point was a huge card file created from the base dictionary.A team of schoolchildren for whom this was a summer holiday job had laboriously cut and pasted each entry of the dictionary onto.itsown card.It worked well enough although there were occasional obvious lapses of application to the task.
The major additional file was the editors' own citations collection file which grew almost as rapidly as they sifted through the obvious secondary sources and then began a program of reading of primary sources.These were largely contemporary materials -books, magazines, newspapers, etc.
Other files grew 'as editing progressed.One initial task was to provide specialists with material that related to their special field so that these lexicons could be considered from an Australian perspective as well as brought up to date.Since the meaning that related to the specialist field was often one of many unmarked meanings, the only way to do this was to duplicate the entry and set up a separate card file for each specialist area.Not every specialist set of definitions in the dictionary was revised in this way but the major ones were, and the ones which the editors felt had great significance to Australians, and therefore to Australian English.There was not much they wanted to say about the sport of curling for example but quite a lot to add under the headings of agriculture, botany, history and politics, and zoology.In terms of general updating economics, computing and medicine were as much in need of updating then as they would be for an edition produced today.
It was a slow procedure made inevitably cumbersome by the card files.The day-to-day writing of the dictionary was a matter of transferring data from one card to another.A process that affected the whole dictionary was the amount of filing involved.The collection of new material relied largely on an individual's knowledge, or on the happenstance of the reading programs.It helped to have a good memory in these circumstances to be able to say with some certainty what was in the file and what was not, and which particular file it was in.
Everyone who worked on the dictionary became attuned to capturing words on the run -as you read them in the paper in the morning, as you listened to the radio in the car, as you conversed with others.
The fear that drove all on at that stage was that some obvious lexical item of Australian English would be left out.The secondary sources available took the editors up to World War II.There had been some work on particular areas of Australian English at Sydney University but these were, on the whole, historical studies.The editors had to document Australian English from World War II onwards from scratch and had to sieve through the existing material selecting what was appropriate for the dictionary.In all this it was quite possible that something might be missed.
The basic commitment of this dictionary was that it would have ,all the words in it that a speaker of Australian English would expect to find.What if the first edition appeared and it was without the entry for fair dinkum or dingo or some such?It was this anticipated possibility that produced the scraps of paper, the backs of chequebooks, the anxious before-I-forget phone calls, just to check that a word was where it should be and had not been overlooked.
This was the research aspect of the task that confronted the first editorial team.This had to be complemented by an editing process that took the new material and slotted it into a dictionary structure.The principles for selection were as they have ever been: first, establish that the word has currency, or has had currency, in the language community; then try to establish its meaning in accordance with the evidence of its use; then provide information on morphology, etymology, and so on, for the final dictionary entry.Although the method is timeless, the means by which currency is established has changed over the years with access to huge corpora as instant suppliers of citational evidence.

publication, Reviews
The first edition of the Macquarie Dictionary was launched in Sydney on 21 September 1981 by Professor Manning Clark, Australia's foremost and most venerated historian, who in his printed Introduction to the work had described it as "a landmark in the history of this great change which has come over intellectual and cultural life in Australia" (Delbridge et al. 1981: 11).The first print run of 50 000 copies sold out within three months and the up-front bills were all paid.It was reprinted seven times with additions and corrections before the second edition appeared in 1991.
The reviews were welcoming and (especially the non-professional ones) very enthusiastic.The academic and specifically lexicographic reviews, both overseas and local, were not discouraging.Writing in A Bibliography of Writings on Varieties of English, 1965-1983, Broder Carstensen expressed a favorable view: Dem Leser dieser Zeitschrift sei verraten, daB dieses W6rterbuch des australischen Englisch eine der gr6Bten (die bedeutendste?)Leistungen der australischen Anglistik und der modernen englischsprachigen Lexikographie iiberhaupt ist.The reasons one might adduce for believing that it is a strictly local triumph and not an international one lie in two main directions, its primary derivativeness and its occasional charming unawareness of the standard professional requirements of reputable lexicography outside Australia.
He did not hold the editors' use of the base dictionary to be reprehensible, but he did expatiate on the extent of their retention of English and American material from it, which he estimated at 93%.But English as a world language is principally formed.in the populations of England and America, and the editors did not wish to exclude the English and American material.Dr Burchfield appears not to have taken into account, in his equation, the extent of Australianisation done even on his estimated 93%.For example, of the first 100 botanical entries under the letter B in the first Macquarie edition there were only 43 which also appeared in the ACD, and of these only 13 in Macquarie Dictionary are as they stand in ACD, disregarding the pronunciations and etymologies which are in each case altered: In the area of linguistics the editors deleted all but the most important American Indian languages, and replaced them with the names of Aboriginal languages.In law, parJiamentary procedure and industrial relations they put Australian instead of American definitions of the words that happened to occur in both varieties, and added those that are unique to Australia.Pronunciation and spelling variants current in Australian English are entered with a priority based on observation and record.Every headword has a pronunciation based on wide personal observation and census-taking, on the accumulated files of the ABC's Standing Committee on Spoken English (SCOSE), and on the published studies of the phonology of Australian English.The publisher of the dictionary eventually paid many thousands of dollars in a single royalty sum to Random House, publisher of the ACD, which had not been used as a base, and a lesser sum to Hamlyn.The use of the Hamlyn book is gratefully acknowledged early in the first preface, though not on the title page, as Dr Burchfield would have preferred.Such reference would have meant very little to the Australian readers because only a very small number of copies of EWD were still to be found in Australia, and, except perhaps in university libraries, none of the ACD.No purchaser of the Macquarie Dictionary has complained of having been misled by the notice given in the preface.In the publishing context in which Macquarie Dictionary appeared, the reference to the history of the book seemed to the editors to be both accurate and appropriate.The publisher had no plans to offer the book for sale in Europe or America, though people there who wanted it as an Australian dictionary could, and did, find ways of securing their copy.

Community acceptance
The publication of the first edition of the dictionary in 1981 was rewarded with instant acceptance by the majority of Australians.The penetration of the dictiomiry was mapped by the responses the editors received to a newsletter which came with the book inviting dictionary users to become contributors to ongoing work.This was the first appearance of the Macquarie Dictionary Society Newsletter which was designed to harness the enthusiasm of Macquarie Dictionary owners around Australia.The first editorial (Macquarie Dictionary Society Newsletter 1(1): I, October 1981) read as follows: It is impossible to have been everywhere, seen everything, and spoken to everyone, particularly in a country as big as Australia.So we know that while the Macquarie Dictionary is a comprehensive account of Australian English, there will be words we have missed, pronunciations which we have not recorded.People can add to our store of knowledge by pointing out to us these areas for future research.Perhaps it is just a single word, perhaps a whole feature of Australian life, which should be investigated and recorded.The Macquarie Dictionary Society will, we hope, be supported by people who will map out the aspects of Australian English yet to be explored.
The newsletter under a new title, Australian Style, has a national distribution, and is published by the Style Council Centre of Macquarie University with funding assistance from Language Australia, an organisation set up by the Federal Government.
It took a while to establish the claim that the Macquarie Dictionary was the national dictionary.The position of the dictionary was made more secure by a 1986 conference, which became the first of many, on aspects of Australian English.The conference was called Style Council since it addressed matters of style and usage, and since it was thought of as bringing together different people who had an interest in those issues as a kind of representative gathering.Linguists and lexicographers were present but outnumbered by teachers, writers, and editors.The first Style Council addressed topics which were of immediate concern to the dictionary editors -spelling, hyphenation, capitalisation and so on -but later conferences took themes of a more general nature but still focused on language and on Australian English in particular.The Macquarie Dictionary has been accepted as the arbiter of Australian English by many newspapers where it is the bedrock of a house style; in the law courts where Macquarie definitions are given priority over those of all other dictionaries; in Hansard, whose editors have declared it to be their "bible"; in the ABC for pronunciations; in the Commonwealth Government Style Manual (Grayston 1994); and in the education system where the Australian Education Council (now caUed the Council for Education, Employment, Training and Youth Affairs), comprising the state and territory Ministers of Education, agreed to take Macquarie as the standard reference on Australian spellings.

The range of dictionaries
If the editors had been tempted not to continue when the first dictionary was published, the publisher prevented this.Kevin Weldon had always envisaged that the Macquarie Dictionary was the starting point for many other dictionaries for different kinds of users.The first in the series had been decided in the original contract -a Pocket Dictionary for Australian schools.But there were to be others in the family for the general market, the education market and the children's market.Each one had its own users.The Pocket Dictionary was prepared in consultation with a range of teachers who advised on the vocabulary of different subject areas.It was a first stab at a controlled defining vocabulary of sorts, a procedure that the editors have refined for the ESL dictionaries, now in preparation.The Concise Dictionary on the other hand needed a greater range of subject areas, from components of the formal curriculum, such as physics, geography, etc., to activities which by their nature are considered informal, such as horseracing, gambling, surfing, etc.It also needed a broader sampling of colloquialisms.Each dictionary was derived from the comprehensive dictionary in that a headword selection from that database produced the first draft of each new dictionary.But from that point on much had to change, so much so that the editors opted eventually for a set of dictionary databases related by a common structure rather than just one dictionary database with multiple options.
On top of this dictionary program the work of creating a thesaurus was begun by using the comprehensive dictionary as the starting point.And this again led to a sequence of thesauruses to match the dictionaries.
It took ten years to establish the family of dictionaries and thesauruses.It was a process of coming to terms with the obvious from both a publishing and a lexicographical point of view: the publishing company would not survive on just one dictionary when its competition had the depth and strength of a range of dictionaries, internationally marketed.And not all the people who wanted an Australian English dictionary necessarily wanted it in its first format: different users needed different kinds of dictionaries.

The computer system
Card files had been the basis of the work from the time that the project began in 1970 to the typesetting of the first edition for publication in 1981.With the typesetting of the first edition this was not necessary.They transferred the data from card files to a computer database, but one to which they did not have direct access.The computer that stored the database was held by the typesetters at a location remote from the editors in Adelaide, the capital of South Australia.It was a cumbersome way of doing things, but preferable to the card system.
The great advantage at this stage was that searches of the whole database could be conducted, giving the editors a secure knowledge of precisely what was in and out of the dictionary.
The editors wished to have direct access to the database and developments in computer technology soon made this possible.The huge computers which housed the original database became museum pieces and Macquarie Dictionary was able to obtain a relatively tiny computer which was the central server to a local network.The software system used was one called Titan, developed by RMIT University computing department in Melbourne.It had been designed as a library system, so its strength was that it could search huge amounts of data very quiCkly and keep track of what happened to that data as it underwent various processes.
It was not particularly user-friendly, and the staff themselves had to write all the special software that they needed for dictionary use.For example, they set up a corpus of Australian texts on this system which they called Ozcorp.Ozcorp is a database of Australian English collected for lexicographical purposes.It contains at present approximately 23 million words, with significant representation from Australian fiction, non-fiction and newspapers.Although there is a bias towards contemporary materiat the database includes many texts from the 19th century through to W orId War II.There is also a small amount of material derived from letters, advertisements, and some spoken transcriptions.
The searching facilities of Titan on this corpus were swift so that it became an easy tool to use in the day-to-day work of dictionary editing.The facility to line up the corpus and the dictionary side by side meant that any entry that attracted the editor's eye for whatever reason could also, as part of the editing routine, be checked against Ozcorp.Suddenly the editors found themselves in the position to reshape the dictionary in a way that has greatly reduced their dependency on the base dictionary.
They still kept a kind of computer notebook alongside Ozcorp for the things that they noticed in their daily reading and listening, for human beings were still best at picking up useful items for the dictionary.They also experimented with using the computer to identify new words.As each text was sca1U1ed for Ozcorp they ran the dictionary spellchecker over it, in effect asking what words were in the text that were not in the dictionary.The results were, overwhelming and, they realised, mostly full of material that was not needed in the dictionary anyway.It is a procedure that needs to be refined.They found, as Rundell (1998: 323) says, that "good corpus data is merely a prerequisite for better dictionaries; it does not in itself guarantee that good dictionaries will actually be produced".But as a starting point it is far ahead of the resources available when lexicographical efforts were begun in the 1970s.
As well as ease of access to dictionary data, the new computer system gave the editors a system for tracking the history of each record -who had edited it on what date and for what reason.This allowed them to manage the process by which an entry written by one editor might be reviewed by another and finally passed to the whole editorial committee as part of a final read before publication.The process of review is an essential part of the compilation of new material for the dictionary.
As part of this new way of managing the nitty-gritty of editorial work, the editors established a classification of the kinds of errors that occurred -typos, definition content, pronunciation, etc. -which they used to code each correction.The advantage of this was that it was possible to sort the corrections in their different categories, a process which facilitated proofing, and to identify quickly which part of the entry should be scrutinised.They initially identified a large set of correction types which were later reduced to a smaller and more manageable set with which editors could become thoroughly familiar.
This system allowed the incorporation of computer-generated safety features into the system -certain fields in each record had to be filled in before the computer would let go, so to speak, of corrected cards which were automatically identified as candidates for proof cycles for the editorial committee.The computer would not let data progress to the next edition before it was assured that the new material had been checked and approved.

The second edition (1991)
The main feature of the second edition was the introduction of encyclopedic material.Again the editors drew on the base dictionary for world material, but this had to be updated, and the Australian material had to be developed with the help of consultants.There was a bias towards Australian material -about 5000 of the 20 000 entries were Australian content.As John Bernard said in the Introduction to the second edition: Our selection of encyclopedic entries caters specifically for the interests of Australians, and we see those interests as ranging intelligently across the whole world but having a special concern for all that is peculiarly Australian.For this reason the threshold for entry is rather lower for Australian names than for names from other countries, though of course many Australian names could get in over very high thresholds.There were broad principles for selection, such as that overseas cities were automatically included if their population exceeded 100000; for Australian towns the threshold was just one thousand.But exceptions were allowed either way for places with some special claim for inclusion.

Towards the third edition
With the publication of the second edition in 1991 the Macquarie Dictionary moved into yet another computer system.This change was driven by the need to create MacquarieNet -an online source of Australian reference materials.The data needed to be in SGML (Standard Generalised Markup Language) so that it could be output both in typeset format and in HTML (Hypertext Markup Language) for the web site.The software chosen was SIM, also developed by RMIT University in Melbourne.
The new computer format meant that editors could now see the dictionary entries on screen as they would appear on the page, a luxury which before this had been reserved for final proofs.Until this format came into use, editors had worked with a coded output right up to final page proofs.They had learnt to read these codes, but understandably this was not as safe for editing purposes as the ultimate page format.
These final proofs had formerly come at the very end when publishing deadlines put editors under great pressure.Now they could see current work in its proper format at any time.
The new system produced its own pressure for a rewrite of the structural elements of all our dictionaries so that they would conform to one standard Document Type Definition (DTD) for programming purposes.This was a major undertaking which is now completed so that an interesting new set of possibilities opens up.
Overall the advantages which editors have gained in the progress from a card system to a computer system are: the capacity to see the entry at any time as it would appear in the book, safety checks on data built into the system, the capacity to find any entry or any sets of entries with great speed and to impose various filters and categorisations on the data, an editing record for each entry built into the system, the capacity to transfer material from one database to another with comparative ease, and the capacity to link all the related entries in each dictionary so that a change in one entry can be carried through all the related entries in the other dictionaries with comparative ease.Ultimately there will be just one database with all the information interlinked within it.The kind of power over the data that this will give the editors is something to look forward to.

Features of the third edition (1997)
The main feature of the third edition was the link that it established between Australian English and Australian literature, described in the preface: One of the editorial policies, even for the first edition of this dictionary, was to assist readers of earlier Australian literature who would inevitably in their reading come upon words and phrases that 'are no longer current.Such words are listed and described in the dictionary, but in this third edition their meanings can also be illustrated by a selection of citations, each usually of a complete sentence, taken from a named work by a named author, with the date of first publication given as part of the entry.
Another feature of the third edition was its coverage of other language varieties of interest to Australians -Aboriginal English, and some of the Englishes of South-East Asia.The coverage of regionalisms in Australian English was revised by Dr Pauline Bryant on the basis of her research in this field.
The lack of usage notes in the dictionary had long been felt and it was decided that the third edition should remedy this.This was also in response to pressure from readers of the dictionary who sometimes did not read as much into coded labelling as the editors wanted them to do, or who raised questions such as whether the item should be included at all, which could not be answered from the material with which they were presented.
For example, why had the word youse been included?There were those who felt that this lowered the prestige of the dictionary.An editorial presence was needed to justify and explain, and that came with the usage note.The explanation made use of a labelling phrase (in non-standard use), a definition (plural form of you), an illustrative phrase (youse are all mad) and a dated citation from an Australian novel (Frost, the publican, desperately afraid of losing good customers, said to the four of us Englishmen: 'Can one of youse blokes play the piano?'The bush was full of pubs with pianos with no one able to play them.-HAROLD LEWIS, 1973.).All of these devices are used throughout the third edition, in order that the reader may see how, when, and why such a word might appear in 'a text or be used in an utterance.

The future of the Macquarie Dictionary
The basic set of dictionaries and thesauruses for Australian English requires constant updating and revision.Setting that aside, the editors have to ask themselves where else they might go and what else they might do.Within Australia there is a need for locally-produced ESL material and the first of a series of leamer's dictionaries will be launched later this year.Because of the increased links with South-East Asia, the editors see the need for English language reference in that region that acknowledges the varieties of English that exist there.It is not possible to do the complete account, as has been done for Australian English, but at least the editors can do their best to produce research on these Englishes and accord them a status in the dictionary that they have not previously had.
The other new field that presents itself is electronic publishing.Macquarie Dictionary has established a site, MacquarieNet, that provides Australian reference to Australian schools.It is in its infancy at present but is gaining support as more and more schools come online.
Since work began in 1970, this dictionary enterprise has maintained objectives but has extensively changed in methods and range.Three of the original academic editors are still working on the project (Arthur Delbridge, John Bernard, David Blair), and two others, Pamela Peters (editor of The Cambridge Australian English Style Guide) and Colin Yallop (now the Editor-in-Chief), have joined more recently.When the first edition was published, Macquarie University agreed to the formation of a publishing company, The Macquarie Library Pty Ltd, to carry forward the publishing program.It is housed on the University campus, for close communication with University staff and services.The publishing editor of the Editorial Committee, Susan Butler, has been with the project since 1970 and has the role of Publisher of The Macquarie Library Pty Ltd.The senior editing staff of that company join with the Editorial Committee to form the text of each edition and ensure its accuracy and authenticity.The institutional support and public acceptance that it enjoys well and truly justify the aspirations of those who had the ambition to produce a truly national dictionary for its primary users, a privileged possession which some other major varieties of English had already long enjoyed.
The most sobering of the reviews was by Dr RW.Burchfield, published  in the Monthly Review supplement of the Melbourne paper The Age (1 March 1982: lOll):