Indigenous Languages Tech

Learning about languages in Australia and worldwide at Puliima

Nick Doiron
6 min readAug 24, 2019

Recently I was fortunate to attend Puliima, an annual conference about recording, revitalizing, and educating in indigenous languages. It’s the first conference that I’ve been to which put indigenous groups’ experiences and speakers front and center. There was a lot of information, new to me, about exchanging information with linguists and researchers, ownership of a language, and the history of Australia.

Most of the languages represented at the conference originated from the First Peoples of Australia and Torres Strait Islanders, but there were presentations about Maori (New Zealand), Hawaiian, and First Nations of Canada.

There were many, many languages discussed here. In sessions which taught us the basics in one language or in linguistics, one after another person would come up and share their own language. Of 150+ indigenous languages in Australia, reportedly only 13 are taught to children ‘as a matter of course’ (unclear what this means), so every language is seeking funding, educational curricula and resources, and a technical home for their content.

Acknowledgements, taboos, and cultural sensitivity

Sessions always opened with a statement acknowledging that our conference was on Larrakia country. I took note of how this was phrased; in the US I recall only one speaker giving a similar specific statement (Sumana Harihareswara at LibrePlanet). The statement acknowledged the Larrakia people as traditional owners of the land, and brought it to the present day by thanking elders - past, present, and future. You can read guidelines for a similar statement at commonground.org.au/learn/acknowledgement-of-country and look up approximate territories in Australia and the Americas at native-land.ca

Though indigenous people of Australia are often referred to as Aboriginal, there is an ongoing conversation about changing this to First Peoples. In public comments on recent legislation for indigenous language, there was a strong push to use First Peoples term, but the government kept Aboriginal to stay consistent with their previous laws. This process is also careful to avoid using the word protection, as many measures in Australian history claimed to be protecting the land, families, children, when actually separating.

Presentations with older photos and videos carried a variation of a warning used on Australian TV: Aboriginal and Torres Strait Islander people are advised that this video may contain names and images of people who have passed away. One project explained that they had a national TV broadcast of their project, but the video was now offline to respect a participant who had passed away.

This tradition varies between different regions and groups, and is poorly understood even within Australia. From what I read, the TV warning is applied to any historical image, when virtually every tradition limits the taboo to months to 1–3 years after someone’s death. When celebrities pass away, the media is sometimes careless or uses full names ‘to avoid confusion’.

Open knowledge and owning a language

In the open knowledge / wiki / library sciences community, it’s common to think of online resources as a place to make your full resources available to the public. Technical speakers at the conference would make it clear that the people own the content and language, and not the library, company, or government which collects the content.

Many indigenous languages are in the middle of being researched and restored by language centres. Some communities decided to limit who should learn their language, and that’s reflected in who gets online access. Other communities (particularly the traditional owners of land now in major cities, such as Sydney) have experimented with large classes open to the public.

Translation and Intentions

A language centre might charge a consulting fee for non-Aboriginal people, or invite a singer/songwriter to attend classes as part of their translation process. The most common translations at one centre were for ‘meeting’ or ‘together’, and these words are now used in many public meeting spaces or conference rooms.
For many indigenous languages in Australia, there were only 3,000–4,000 words recorded and no living fluent speakers when language revitalization programs began. Others were suppressed before audio recordings existed. The organizers consult historic linguistics texts, and reach out to elders as much as they can. With so few words, focused on animals and grammatical rules, the translators often simplify phrasing (for a UN document, “people have the right to _” might be phrased “people want _, let it be done”).

One speaker told us that she often replies to translation requests with one question: What is the intention?
And so many companies are doing lip service, or think that it would be cool to have a Hawaiian theme, so the question totally dissuades them. The speaker urged the use of translation to teach through a language, and not to teach about the language as an abstract thing.
I’ve struggled with this in the past when seeking translations for Blockly or OpenStreetMap, so I am still reflecting on this.

Strengthening a language through creative expression

Maybe the most interesting presentations were those which covered creative expression. We heard about a translated Shakespearean sonnet which was presented at the Globe Theatre in London, the first Ngandi film, and one session featured a giant, handmade, plush crocodile with a Raspberry Pi inside for language lessons.
From a practical perspective, the Kaurna children’s book came with a video to help with pronunciation:

There was a session recapping these creative works and using it as proof that these languages are stronger than the statistics say. The census numbers were confusing (for example, numbers of speakers jumping anywhere between 400–800 each year). But either way, exposure and awareness are only one part of revitalizing a language. I don’t think we have enough information about endangered languages in the social media / sharing economy to say how this all fits together.

Sign languages

There were two Auslan interpreters at the conference, and a presenter at two sessions who delivered his presentation in Auslan. Much of this topic is above my head, but he did mention Plains Indian Sign Language, which was once a lingua franca across dozens of Native American languages.

Academics and Technologists

The conference was more education and academia-focused than I’d expected. A session on remote learning had us participate in a lesson on Adobe Connect, showing us the features, and explaining the logistic need for it (in a country where locations are given as ‘three hours drive from __’, you can’t visit a dozen schools in person each week).

A Maori audio-storybook app was shown which used some useful technology: tapping or highlighting a word repeated it, and double-tapping a word recited it syllable-by-syllable. Each scene had a little animation to go with it.

I also learned about the Austlang codeathon, which is not writing code, but adding language codes to library records, which previously labeled documents as simply ‘Australian languages’ or ‘Other’.

There were three open source software projects featured: SIL’s KeyMan keyboards, VACL’s upcoming Limba project (based on Drupal), and the National Library’s Trove (they are about to release a new version, but this code is private on GitHub).
SIL is looking for frequency word lists, to help build predictive text. The other two were language resource libraries.

Takeaways

As a technologist, it was informative to see the tools and resources which linguists and language teachers use in real life. There wasn’t an abundance of fancy new technology, but it would have been extra. Libraries and language centres are working on properly uploading and archiving the basics, the same standards that art institutions (GLAM acronym on the Wiki) have been working on for several years.

I remain open to helping anyone with adding an indigenous language to OpenStreetMap, Blockly (coding app for kids), or MetaMask (cryptocurrency wallet). Just get in touch here or message @mapmeld on Twitter!

--

--

Nick Doiron
Nick Doiron

Written by Nick Doiron

Web->ML developer and mapmaker.

No responses yet