Chapter 15 – Thick interfaces: Mobilizing language documentation
371
case study illustrating the emergence of an interface as a result of a collabo-
rative process involving the linguist, multimedia developer, and Paakantyi
community members.
At every stage of the project we prepared and delivered product drafts
representing the ongoing state of the CD in order to:
– make it easier for people to give feedback about concrete products
rather than react to abstractions;
– demonstrate our ongoing commitment to the project;
– help create a community “story” or “biography” for the CD.
We also workshopped some project participants in techniques for recording,
digitizing, editing, and linking the sounds they contributed.
Key graphic and navigational systems of the CD resulted from this collab-
orative work on its design and its linguistic and graphic content. Initially, we
were given permissions to use several artworks to decorate the CD (the Paa-
kantyi community boasts several accomplished artists). Under the guidance
of Badger Bates – a Paakantyi speaker, Park Ranger, and nationally-recog-
nized artist and sculptor – the design evolved so that recordings of impor-
tant speakers from previous generations of speakers became systematically
accessed via artwork created by their respective living descendants. This,
we believe, contributed to the aesthetic balance of the CD, which ultimately
gained keen acceptance within the community. The navigational structures
are summarized in Figure 2. The top row lists the “old time” speakers from
previous generations whose texts and songs feature on the CD. These people
have all passed away some time ago. The oval disks in the middle row cor-
respond to main resources in the CD – two stories (Mutawintyi, Anteater),
songs (including Emu), and the talking dictionary. The bottom row lists
project participants who contributed art (the two left boxes) and linguistic
content (the two right boxes, with Badger Bates contributing both). The
arrows represent linguistic input; the other bullet-head lines represent the
provision of artwork. Vertical alignment indicates ancestry (both Dumbo
Dutton and the Bates family descend from ‘Gunsmoke’ Johnson).
The example shows how an interface design “emerged” as a result of
collaboration, to result in a design that was not only more aesthetically at-
tractive, but also actually communicated more information – information
about the genealogical relationships between past and present participants –
even though this information may not be apparent to non-community mem-
bers.
372
David Nathan
Figure 2. Participants, art and language in the Paakantyi CD (Hercus and Nathan
2002)
Figure 3. Environmental objects as controls in Bunuba Yarrangi Thanani
(KLRC nd)
Granny Moisey
‘Gunsmoke’ Johnson
George Dutton
Emu
Anteater
Mutawintyi
Badger & Phillip
Bates
‘Dumbo’ Dutton
Talking Dictionary
Renie & Lottie
Mitchell
Chapter 15 – Thick interfaces: Mobilizing language documentation
373
4.2. Appropriate interfaces
Interfaces should help users perform their tasks while respecting local cul-
tural and aesthetic styles. But they should not be based on stylized impres-
sions of a culture.
10
Some earlier productions confused the two aims by, for
example, using rocks, animals and other environmental objects as buttons
and menus, as shown in Figure 3. These objects do not function well as
navigational metaphors because the animals displayed are unlikely to be
associated with buttons or navigation. Instead, they require memorization
and draw the user’s attention to the interface itself rather than support the
navigational task.
By contrast, the Paakantyi CD (see previous section, and Figure 4 be-
low) uses a mainly text-driven interface created by a graphic designer who
was given a brief to create a crisp, contemporary feel. It has been very well
received and found easy to use; Paakantyi students transcend their everyday
literacy levels in using it, because it allows them to focus on their chosen
tasks, most especially to navigate to spoken entries in the CD’s talking dic-
tionary (cf. Goodall and Flick [1996], who urge avoidance of text elements
in interfaces for Aboriginal children).
11
Although the Paakantyi interface largely utilizes text, it is not a “thin”
interface. Displayed text is not a simple projection of the underlying data,
which is considerably more complex and incorporates thousands of links
between various text and sound objects. Furthermore, much of the text
functions not as content but as navigation controls that users can click on.
Notice also that in Figure 4 there are variant spellings; these are the result
of differences between Hercus’ original research in the 1960s and the new
data we recorded in 2000. The CD simply presents both variants and allows
the users to draw their own conclusions.
With regard to this example, it may be worth noting that text-based in-
terfaces can work well even in societies that do not have traditions of liter-
acy in their own languages. Many such people do have literacies, of course,
in other dominant languages (such as English in this case; it may be Arabic,
Chinese, Russian, or other languages elsewhere). In addition, there is not
just a single type of literacy. Other literacies, such as computer literacies,
have quickly developed over the last decade across much of the world, es-
pecially as a result of familiarity with software used for internet access, and
these literacies involve changes in the functional balance between text, lay-
out, and graphics in comparison to standard “book” literacy (Nathan
2000a). At the limit, we can say that consistent text-based screen navigation