12. SETTING UP A BAYALA DATABASE
1. FIELDS
On deciding to create a database, the first thing to do is to create the fields. In FilemakerPro version 19 this is done by
file>Manage>Database or ⇧⌘D.
Fig. 12.1 A few of the fields in the SOUTH database
After the fields have been created, changes by addition, amendment and deletion can occur at any time.
2. LAYOUTS
Layouts can then be made, to display the fields in any desired arrangement. In FilemakerPro version 19 this is done by
file>Manage>Layouts or ⇧⌘L.
Two of the principal layouts, OVERVIEW and LINKS, were described in 2. The Big Picture. See also 10. Database Layouts.
3. RELATED DATABASES
Many databases are simple with perhaps around a dozen fields. They are ‘flat file' databases. The Bayala databases, however, may not only have a hundred or so fields but were able to become increasingly useful by drawing upon related subsidiary databases. They are ‘relational' databases.
If it were to be possible to put databases of such complexity as the Bayala series onto the internet it would still be necessary for any User to have a copy of Filemaker Pro in order to make them work.
Fig. 12.2 A portion of the relationships structure of the ALLSYD database: ALLSYD fields in the centre, related databases at the sides
Fig. 12.2 reveals the relationships that enable the ALLSYD database to work.
Each of the rectangles on the left is connected to one of three fields in the ALLSYD database, all the fields for which are listed in the tall grey column in the centre of the illustration.
—The yellow rectangles on the upper left control the JSM links in the LINKS layout described in 2. The Big Picture and 10. Database Layouts.
—The brown rectangles at the bottom left control the NoH links in the same LINKS layout.
​
On the right are various other relationships where a rectangle is linked to a single ALLSYD field. It does not matter whether the linked rectangles are placed on the left- or right-hand sides; each could be placed anywhere. They are arranged as they are in order to present the linking system in a way that is as simple as possible to comprehend.
—The purple rectangles on the upper right control the search features described in 6. Search Portals.
—The mauve rectangles towards the upper right are part of a discontinued ‘search engine’ superseded by better systems already described.
—The green rectangles halfway down on the right control the eight elaboration bars.
—Other rectangles on the right are for geographical purposes, transcription keys, source details and the like, drawing in information from subsidiary related databases.
​
Some of these subsidiary related databases are described below in turn.
CAROL VOCAB DETAILS
In the earliest days of the Bayala series there was just a simple list of Sydney language words with their translations, together with page and line numbers and a record of the source. Then categories (fauna, body parts etc.) were added, and parts of speech.
Then came the idea to include a ‘story’ (i.e. provide further details) about the source, say about William Dawes or other of the early Sydney language word list compilers. To write the same story for every Dawes record meant repetition and the possibility of variation between the story entries. It was then learnt that a story could be written once somewhere else and made to appear by means of a link without the need to write it afresh every time. A colleague named Carol explained how this worked. So it was that the Carol VOCAB DETAILS listing of ‘stories’ became the first of the related databases to be developed: see Fig 3.2. Other related databases were to follow.
MAC CHT SETS
In case there is a need to see, or use, special characters connected with the database concerned there is a collection of these provided in the navy blue 'Mac cht' field situated below the Occurrences JSM portal in the top right of the OVERVIEW layout. It can be seen at the very bottom of Fig. 12.3 below, in the centre.
Fig. 12.3 The navy blue Mac cht related field is below the Occurrences JSM portal
Clicking on the 'Mac cht' field opens up the full display from the Mac cht sets related database of special letters or ‘characters’ in various point sizes. Any of these can be copied when needed in compiling any entry requiring such a character.
Fig. 12.4 Mac cht sets: displayed collection of special characters
ELABORATION
The elaboration bars at the top of the OVERVIEW layout were described in 4. Elaboration Bars. The information they contain is drawn from the Elaboration related database.
TRANSCRIPTION KEY
Two major features of the Bayala databases are:
1. the respelling of original Australian language records:
—using only three vowels ('a', 'i' and 'u')
—avoiding doubling of letters, as well as hyphens and diacritics; and
2. the provision of a simplified modern standard English translation.
For example, in relation to the first point 1, original recorders spelt the Biyal Biyal (Sydney language) word for ‘good’ in at least 26 different ways. From these records collectively it can be deduced that the word must have sounded something like ‘boodjery’. Even so it cannot be known today exactly how it really was pronounced. Consequently in the Bayala respellings it was rendered as bidyari (1), budyari (19) and budyiri (6) as best reflecting the originally recorded pronunciations. A search for b@dy@ri, where ‘@‘ represents ‘any letter’, finds all of the original 26 variations.
Fig. 12.5 The explanation of the Tr1 and Tr2 transcription indicators for the ‘boodgeri’ (11th) record can be seen in the grey bars at the top
The rules shown in the transcription bars (described in 7.4 Transcription bars) at the top of Fig. 12.5 are determined by the black and dark grey 'Transcription indicator' fields (described in 3.2 Word Fields), and are drawn from the Transcription Key related database. These rules are to inform any User of the Bayala databases as to why a particular respelling was adopted in a word. When particular codes are entered into these fields explanations of the rules appertaining to the codes appear in two black and grey bars nearby in most layouts.
One of the original words for ‘good’ was ‘buggery’ (second last in the list in Fig. 12.5). When this spelling was viewed alongside all the others the intended pronunciation was apparent; but the word when seen on its own suggested something else. Three transcription keys could have been used for this word to reveal the reasons why the eventual respelling was used. However, there are usually only two code fields available, and as a result the first of the rules below was not able to be presented:
—u=oo /u/ pronounced as 'oo' in 'moon'. TRANSCRIBE AS /u/
—g=J -g- PRONOUNCED '-j-': TRANSCRIBE AS /dy/
—-er- -er- ASSUME ENGLISH as in 'robber', 'robbery': TRANSCRIBE /a[r]/
Consistent with these explanations, ‘buggery’ became budyari.
In the Transcription Key subsidiary related database that stores these explanations there are about 280 explanatory rules.
JMS COINED WORDS
The second major feature of the Bayala databases, alluded to above under Transcription Key as point 2, is providing a standardised English translation (called EngJSM: for English JS Main word). As an example, ‘creek’, ‘brook’, ‘torrent’, ‘river’, ‘stream’ all become ‘stream’. A search for ‘stream’ finds all these, as well as still other variations of the idea of water flowing in a channel.
When a common basic translation of an idea, such as ‘stream’, is used, many more matches for original Aboriginal words recorded are the result.
Fig. 12.6 Various original English entries all become ‘stream’
When common standard words are used in the EngJSM field in this way they link to the JMS Coined words related database, and it is this that generates the series of equivalents shown in the purple bar at the top of Fig.12.6.
CATEGORIES
Virtually all words can be assigned to a category and most words are easy to categorise (e.g. ‘body parts’. ‘fauna’). When an EngJSM entry matches an entry in the Categories related database, entries appear automatically in the ‘category’ and ‘subcategory’ fields when these have been set up in a database.
From the same database come suggested ‘parts of speech’ for words.
Fig. 12.7 In relation to the yellow (EngJSM) column, categories appear on the left and parts of speech on the right, both from the Categories database
DIXON LANGUAGE LIST
The linguist and writer about Australian Aboriginal languages, R.M.W. Dixon, in his 2002 book Australian Languages (Cambridge Language Surveys. Cambridge, U.K.: Cambridge University Press) included a list of Australia’s Aboriginal languages with a code number for each. This language list and coding system, with minor variations, has been used with his permission in the Bayala database series. A language such as Wiradhuri, occupying a substantial part of New South Wales, was given the coding Nc2 by Dixon. But as the language covered many Geoscience rectangles (described in 7. Supplementary Information), alphabetical additions were made to the basic Dixon code number to separately identify each Geoscience rectangle (for example, Nc2.h, Nc2.t for the Nyngan and Narrandera rectangles respectively, and so on: see Figs. 7.7 and 7.8).
Fig. 12.8 The Dixon codes modified as described above are in the green 'Dixon code 02 JS' field on the left. Other fields are self-explanatory apart from ‘code AIATSIS’: AIATSIS (Australian Institute of Aboriginal and Torres Strait Islander Studies) has its own set of codes for Australian languages.
The modified Dixon codes are used as the link to other related databases such as Language location and Vocabs Pix, and to operate fields such as ‘LangShort’, ‘state’, ‘Geoscience map area’, ‘name language location’, ‘dialect’ and others.
VOCABS PIX and LANGUAGE LOCATION
Section 9. Picture Fields introduced the fact that more information, including pictures, can be found to the right of the OVERVIEW layout, as shown below, taken from the INTERSTATE database. Moving the screen image—as viewed on most computers—to the right usually reveals six large picture fields, along with other incidental information.
Fig. 12.9 OVERVIEW layout, left; pictures and sometimes additional descriptive information, right
The Fig. 12.9 display relates to the Macumba River region of South Australia north-west of Lake Eyre in Aranda (Arrente) country. The six picture fields visible in Fig. 12.9, right, draw information from the Vocabs Pix related database to which they are linked by ‘VocPix link’ fields below any pictures that appear. Nearly all the pictures used in the Bayala databases, some of which are used repeatedly, come from the Vocabs Pix related database.
Fig. 12.10 A picture (e.g. Cooper Creek), inserted into all six possible locations in the Vocabs Pix database
There are six positions for entering a picture, useful in case one of the positions available in a standard Bayala database is already occupied by another picture. The Cooper Creek illustration is used many times.
​
There are many languages across the country. Some of the Bayala databases, including INTERSTATE as featured in Fig. 12.9, provide location information about where the language concerned is located, drawn from the Language location related database. There are over 400 descriptions in that database of which those below are a sample:
Fig. 12.11 Language area information from the CURR LOCATIONS layout in the Language location related database. The description in the light blue 'language area detail' field in the middle relates to the blue 'language area' field on the left. If there is an entry in the ‘VocPix link 1’ field on the right it means there will be further information in the form of a picture.
The Language location related database, via the ‘Dixon code 02 JS’ link (for the Macumba River region, WAa3.24), is the source for the information above and to the right of the mauve picture field 6 (blue and cream coloured fields) in Fig. 12.9, right.
Below is a detail of the description featured in Fig. 12.9, where only the top few lines are shown (above mauve picture 6)..
Fig. 12.12 Language location related database description for the Macumba River region as it appears to the right of the OVERVIEW layout
A detail from the Language location database in PIX layout is featured in Fig. 12.13 below, for the same ‘Dixon code 02 JS’ (WAa3.24). This shows the additional 'language area detail' information in Fig. 12.12 above in a paler blue 'language area detail' field as well as an illustration of the area.
Fig. 12.13 Detail of the Language location PIX layout for the Macumba River region
Fig. 12.14 below presents the full Language location related database record for the Macumba River region in Arabana country, including location maps. This particular layout is only available directly through the Language location database, which is to say that it does not appear in other databases through a linking process.
Fig. 12.14 Full view of Language location PIX layout for the Macumba River region (Fig. 12.9)
In addition to the illustration of Lake Eyre with the Macumba River flowing into it in the first picture, there are three other pictures seen above in Fig. 12.14 and in detail below:
Fig. 12.15 Detail from Fig. 12.14 showing the location of the Aranda language, from the Vocabs Pix database
The picture on the left shows the languages in the region (South Australia), in which the underlying map is from Dixon 2002, with abbreviated language names added in colour. The middle picture shows the Aranda (Arrente) language group in red with adjacent language groups, all to the northward of Lake Eyre. On the right is a language location map available on the AIATSIS website <https://collection.aiatsis.gov.au/austlang/language/c8>.
NSW COUNTIES
A small number of entries mainly in the ALLSYD and ANTHROPSOC databases, perhaps 30 or so, mention a county in relation to the word concerned. One of these was Borromarree as recorded in Fig. 12.16:
Fig. 12.16 Borromarree, in the County of Cumberland, recorded by geologist W.B. Clarke in 1840 and included in a published article by James Jervis in 1944 (JERVIS, J. (1944) 'Rev. W.B. Clarke, M.A., F.R.S., F.G.S., F.R.G.S.: The Father of Australian Geology'. Journal of the Royal Australian Historical Society, XXX, 345-458.)
Drawing on the NSW Counties related database and linked through the ‘VocPix link’ seen just below each illustration, the ALLSYD database displays the following, drawn from the Vocabs Pix related database:
Fig. 12.17 The two middle pictures relate to NSW counties
There are 141 counties in New South Wales, of which the first to be established was Cumberland, in the region of present-day greater Sydney. Once of considerable significance, counties, while still a current concept, are limited now to such special purposes as land titles.
SPELLING SYSTEM
Word list compilers, especially those who compiled large lists, sometimes explained how and why they spelt the words they recorded in the way they did.
When attempting to respell the words in one of these lists in a standard way it is useful to be aware of the compiler’s principles when he or she was determining how to render the original spelling of the Australian word heard. This being the case it is also useful for the orthographical rules determined and followed by the compiler to be close at hand. And they are, thanks to the Spelling System subsidiary related database, and the ‘spell key’ field, which is generally to be found somewhere to the right of the OVERVIEW display.
Fig. 12.18 Spelling system followed by Mathews and Everitt in their 1900 paper ‘The Organisation, Language and Initiation Ceremonies of the Aborigines of the South-East Coast of N.S. Wales’
In the SOUTH database, for the following record:
"Bul´-lân" balan "woman" woman M&E: GGA 1900 [:266:3] [Gga] [nsw]
the spelling system featured in Fig. 12.18 was followed by authors M&E (Mathews and Everitt) for each of the 18 occurrences of this word in their paper, and their spelling system is available for viewing for every one of the 332 records in the list.