Miriam Corris, Christopher Manning, Susan Poetsch, Jane Simpson
5th April 2000 VERY DRAFT-Y
"The perfect dictionary is one in which you can find the thing you are looking for preferably in the very first place you look" Mary Haas.
We will report on the results of preliminary investigations of dictionary usage by speakers, semi-speakers and learners of Australian Aboriginal languages. In 1999 Corris, Poetsch and Simpson investigated actual and potential uses of paper and electronic dictionaries by speakers, semi-speakers and learners of Aboriginal languages. The dictionaries included: bilingual Warlpiri-English dictionaries (electronic and printed), a bilingual Warumungu-English dictionary, and a trilingual Alawa- Kriol-English dictionary (printed). The purpose of the investigation was to improve an electronic interface for a large bilingual Warlpiri dictionary designed by Manning and Kevin Jansz. The problem lies in reconciling the competing demands for:
• completeness of coverage
• ease of access
• accommodating low levels of literacy in English and the vernacular (including unfamiliarity with alphabetic order, and with the concept of citation form)
• shortage of lexicographic labour and of publication resources
• range of users' knowledge of the vernacular
• accommodating variation (whether due to dialect or to language change) in entries. This paper will focus on the various ways in which computer dictionaries can assist in reconciling some of these demands.
TO BE REWRITTEN
--STATE OF DICTIONARIES IN AUSTRALIA (O'GRADY, AUSTIN ED GODDARD AND THIEBERGER)
--WHAT PEOPLE HAVE HAD TO SAY ABOUT USEABILITY (LINDSTROM, SIL,HANSFORD, USP)
-THE GAP BETWEEN EXISTING USEABILITY STUDIES AND WHAT WE NEEDED TO KNOW
-THE NEED TO GET QUALITATIVE DATA BEFORE A QUANTITATIVE PROJECT COULD BE DESIGNED
EL dictionaries are almost always bilingual, because the makers are usually not speakers. The main body is usually arranged as EL-LWC. Occasionally, such dictionaries have LWC-EL finderlists, sometimes separated into semantic domains. The EL-LWC direction is important for two reasons.
(i) This arrangement is typically most useful for speakers of LWC (including the lexicographer) trying to learn, understand or explicate the EL, in other words for decoding EL.
(ii) It can also be put down to the symbolic function of the dictionary. Speakers sometimes feel that EL-LWC is the only direction that could truly be described as a dictionary of the EL. Having the EL first gives it a kind of primacy.
As Goddard and Thieberger (1997) note, while some compilations of Australian languages contain thesauruses (Heath 1982, Evans 1992), alphabetically ordered dictionaries are by far the most common. So far, all of these are bilingual or trilingual, with English, the language of wider communication, being one of the languages. A very few dictionaries have definitions in the vernacular as well; these include the big Warlpiri dictionary and the Arrernte dictionary (Henderson and Dobson 1994). Most of the dictionaries are ordered by Language-English, often with an English finderlist.
The Language-English order has a symbolic value, putting the Australian indigenous language first is a claim that it is important. But it also has its origins in the fact that almost all dictionaries of Australian indigenous languages are produced or coordinated by English speakers. A Language-English dictionary is useful for an English speaker as a decoding dictionary. The order also reflects the desire of many lexicographers to produce documentation dictionaries, recording every word they can of endangered languages.
The microstructure of EL dictionaries differ according to how big the dictionaries are. Most of the bigger ones include vernacular definitions and example sentences for some words; these are useful because they can contain cultural and grammatical information. This information is also useful for further study and documentation as well as for speakers maintaining the language. Actual definitional practice varies from one or two LWC glosses, to structured entries. Part of speech information is usually included.
Many of these properties of the macro- and microstructure have been taken for granted by lexicographers. Now, the emerging literacy among EL speakers means that these properties now have to be reconsidered.
Linguists and lexicographers hope that EL dictionaries can free learners (both of language and of literacy) from dependence on teachers, allowing them to learn independently. To some extent this view is shared by literate speakers of indigenous languages. It seems that EL speakers often agree that documentation and maintenance are important functions of a dictionary (Carroll 1999). There is also anecdotal evidence to suggest that the dictionary is important in the minds of speakers as symbolic of the status of the language. But there is little record2 of negotiations between speakers of endangered languages and dictionary makers, in particular of speakers’ views on dictionary structure (perhaps because in some cases the speakers were not previously aware of dictionaries) (but see Hansford,xx, McConvell et al 1983, Carroll to appear, and Stebbins 1999).
2. History of project
This project began when Chris Manning and Jane Simpson started a working group on dictionaries at the University of Sydney to examine possible electronic interfaces to dictionaries of indigenous Australian languages, both for browsing and for creating dictionary entries. Among the projects we started was to design a browser interface for the Warlpiri Dictionary, the biggest machine-readable dictionary of an Australian language ((how big is it?). Chris Manning and a computer science student, Kevin Jansz, designed "Kirrkirr", an interface written in Java. (Manning & Jansz,1998)
Once we had a trial version of "Kirrkirr" established, we needed to establish:
• who would use the electronic interface
• how they would use it
• for what purpose
• whether they would be able to use it
Corris's survey of useability studies (Corris, in preparation) found that such studies concentrated almost entirely on the use of paper dictionaries by children and learners of world languages, or on surveys of committed dictionary users (such as members of dictionary societies).
In February 1999
(i) Miriam Corris took the Warlpiri electronic dictionary and demonstrated it to a range of potential users in Alice Springs, and in two Warlpiri speaking communities, Yuendumu and Willowra. She observed their reactions to it, and also to a partial printout of the Warlpiri dictionary.
(ii) Susan Poetsch accompanied Margaret Sharpe to Minyerri, a community in the vicinity of Katherine for a workshop introducing the Alawa community to a new 3-way Alawa-English-Kriol3 paper dictionary that Sharpe had compiled. Susan designed 13 task-based activities, to be used in workshops with potential users for seeing how efficiently people could find information in dictionaries.
(iii) Armed with Poetsch's and Corris's reports, inIn April 1999 Jane Simpson tested use of the electronic dictionary (Warlpiri) and paper dictionaries (Warlpiri and Warumungu) with Warumungu and Warlpiri students in 3 adult education courses, and then in a Warlpiri-speaking community, Lajamanu. She used a mixture of observation and task-based activities.