This
is an assemblage of documents written texts, diagrams,
and voice recordings, that were produced during the
days 24-27th September, 2003. It exhibits a specific
time and place.
At the point we began, the institutional base for our project was named as the Faculty of Indigenous Research and Education FIRE, located in Northern Territory University at Darwin. In January 2004 FIRE became the School of Australian Indigenous Knowledge Systems, SAIKS, located within a much larger Faculty of Law, Business and Arts in the now renamed Charles Darwin University.
This is not the full
set of documents that were produced at our meeting in
September ’03, the exhibit has been curated. There
are other documents, and there was much messy talk and
writing on white boards that you don’t see. It
might seem odd to put these working documents on our
website. They are probably of very little interest to
anyone who was not there. They are here because we are
committed to showing the workings of the complex ‘microworld’
of our project. By ‘microworld’ we mean
those ‘holding-together collectives’ involved
in getting the job done whatever the job happens
to be. Not only people Aboriginal and non-Aboriginal,
but also the materials, and social resources make up
the collectives of a working microworld. This is part
of doing responsible Nyiknyik Djarrma or native rat
methodology. The notion of microworld is more fully
discussed in a paper written by Helen Verran. ‘Ninteenth
Century British Explorers and Twenty first Century Australian
Databasers’ paper by Helen Verran in
the publications section
Like all modern projects
we had a written version of a more-or-less coherent
set of ideas about what we thought we might achieve.
Michael Christie came up with a short sharp outline:
| |
COMPUTER
ASSISTED KNOWLEDGE MAKING
Initial ideas with respect to user-friendly interfaces
for databases owned by Aboriginal communities -
for ARCLinkage project, M Christie FIRE, NTU, 18
July 2003 |
|
| |
Most databases
used by indigenous people today have four general
problems |
|
| 1. |
upload.
If indigenous people are going to use databases
for their own purposes, we need to improve the mechanisms
whereby the text, graphic, audio or video file is
brought into the database and metadata is attached. |
|
| 2. |
metadata. Firstly conventional databases have metadata protocols which embed taxonomic and structural assumptions about the nature of the world itself which reflect western understandings of knowledge but which may preclude indigenous ways of doing knowledge. (Theorising these differences collaboratively with the knowledge owners and users is a significant part of the research). Secondly the requirements of metadata input may preclude indigenous people (particularly old people and children) from using databases effectively for their own uses because they greatly complicate upload. |
|
| 3. |
Text dependence.
While the text-free interface remains a dream, we
can find ways in which upload, search, and display
can employ icons, lemmatised texts, ‘fuzzy’
searches, and voice recognition |
|
| 4. |
epistemology.
There is a disjunction which is often ignored, between
digital ‘information’ on computer, and
lived indigenous knowledge on land and in community.
We aim to develop databases with a keen eye to how
the digital representations will eventually be used
by indigenous groups as they celebrate their identifying
knowledges together, and they go about the work
of growing up the new generations. |
|
| |
Possible technical
solutions: |
|
| 1. |
make optional the distribution of
metadata into pre-existing categories (ie title,
source, file type). Metadata can be entered either
through dialogue screens or simply ‘abstract’
text entry (ie keyed or selected text in a single
field). (Hidden categories can be maintained) |
|
| 2. |
make all text in the data and the
metadata searchable |
|
| 3. |
develop a lemmatizer to produce a
glossary list of all words in the data and metadata.
The production and action of these lists is crucial
to the user-friendliness of the database. |
|
| 4. |
develop search mechanisms which focus
upon the ‘key word’ as the primary constituent
of meaning (ie a text string is the primary available
search input, while limited category searches, and
pull-down menus might be available) |
|
| 5. |
develop ‘fuzzy spelling’
mechanisms to render the glossary more searchable |
|
| 6. |
voice recognition which through the
fuzzy find, identifies a range of options from the
lemmatised filter list, from a spoken one-word input. |
|
| 7. |
develop lists of place names with
coordinates and a searchable map interface. |
|
| 8. |
develop intuitive interfaces for
data upload and metadata input. |
|
| |
|
|
44K
and a written agenda was developed to mediate our meeting
