Islands of language enter virtual reality

Thousands of the world’s languages are endangered, but linguists are breathing new life into them with virtual reality technology

There are almost 7,000 human languages spoken around the world, but by the end of the century almost half could be extinct, existing only as preserved specimens in obscure databases.

The survival, and even revival, of these endangered languages could well depend on these same databases, but only if they become a lot less obscure and a lot more accessible.

Enter virtual reality.

Imagine a virtual reality catalogue where looking up a language is the first step to immersing yourself in the sensory world of that language. You could experience street scenes, see and hear people speak, call up songs, listen to stories, learn the words for different things.

We aren’t quite there yet, but some Australian linguists are showing the way. They have created a virtual reality fly-through of the South Pacific Islands including Vanuatu and Papua New Guinea, home to an astonishing 130 and 900 languages respectively.

As you fly over Vanuatu’s 80 islands, courtesy of a virtual reality headset, shards of light beam up from each language locality, guiding you where to go. As you approach each one, the hubbub of voices subside and is replaced by people speaking the specific local language, with information provided on how many people are still speaking it. For some it is thousands, for others there are less than a 100 speakers. Looking out across Vanuatu, the islands appear as a forest of light beams reminding you of what we are in danger of losing.

“This could become the way people search and engage with language catalogues in the future,” says linguist Dr Nick Thieberger.

“At some point in the future rather than just watching a video file or listening to an audio file, you may be able to enter into the virtual world of that language.”

The VR display is the brainchild of Dr Thieberger, an Australian Research Council (ARC) Future Fellow at the University of Melbourne’s School of Languages and Linguistics, and Dr Rachel Hendery, Senior lecturer in Digital Humanities at Western Sydney University. Developed through the ARC’s Centre of Excellence for the Dynamics of Language, the VR was made by new media artist and creative developer Andrew Burrell, and uses audio and other information stored in the Pacific And Regional Archive for Digital Sources in Endangered Cultures (PARADISEC).

The PARADISEC virtual reality system on display at University of Melbourne. Picture: Paul Burston/University of Melbourne

PARADISEC is a collaboration between the Australian National University, the University of Melbourne and the University of Sydney and was set up to collect, digitise and catalogue precious audio-visual recordings from the Asia/Pacific where hundreds of languages are endangered. More than 2,000 languages are spoken across Australia, the South Pacific Islands and South East Asia, but by 2100 the number of languages spoken in the region is forecast to shrink to just a few hundred.

“That is why it is so important that we explore new ways of getting the material that is stuck away inside PARADISEC out to the broader public in an interesting way,” says Dr Thieberger. “It is even more important that we get the material back into the communities where the recordings were made and where the languages are endangered.”

Of course, delivering a full-blown virtual reality experience in the often-remote places these languages are spoken is a challenge given the lack of broadband. But the researchers are looking at low-cost options like using cheap Google Cardboard headsets that can work from mobile phones. Other alternatives include installing dedicated computers and having them loaded with particular language VR offerings.

In Australia researchers such as the University of Sydney’s Professor Linda Barwick have installed computers in local communities that host archival audio and songs on iTunes for locals to access and download. Dr Thieberger has done the same in Vanuatu.

Such initiatives provide communities with the opportunity to access material recorded years before, often by their own family members. It is all part of what Dr Thieberger says is the obligation of researchers to ensure that what is recorded is actually given back.

“As researchers we go out into the field and make recordings and earn our degrees, but we then have a responsibility to make sure that the material we collect isn’t just stuck away in an office somewhere and forgotten. The communities we go to have a right to expect access to the recordings and we have an obligation to future researchers to store them in a way that is readily accessible.”

It was this double obligation that motivated the establishment of PARADISEC. In Australia the Institute of Aboriginal and Torres Strait Islanders Studies in Canberra is responsible for collecting and storing Australian language material, but until PARADISEC started 14 years ago there was no such body collecting and digitising the hundreds of languages in the Pacific. There are now 1,085 languages catalogued at PARADISEC comprising 150,000 individual files and amounting to 25 terabytes of data.

“Before PARADISEC all this material - hundred of tapes - was just sitting in filing cabinets at different universities. There was no agency that was responsible for doing anything with the material so we did it ourselves,” says Dr Thieberger, who along with Professor Barwick is the joint director of PARADISEC.

It is archives like PARADISEC that combine the audiovisual files with metadata on where the recordings were made and by whom, that makes initiatives like the VR project possible.

“What we have shown is that once you build a good research collection you can have multiple outputs. The fact that we have collected together all this audio and geographic information meant that to produce the VR experience all we needed to do was export the information in the right format.”

Dr Nick Thieberger (R) plays audio of the Koita language near Port Moresby to Koitabuan E’ava Geita (L). “The feeling of knowing that your language has been documented or recorded in a structured way, kept safely somewhere in the world, and then hearing it spoken from 50 - 60 years ago and by people you haven’t seen but whose names you only hear in history, is quite incredible,” says Mr Geita. There are only some 2,000 speakers of Koita. Picture: Rachel Nordlinger

And what’s at stake is no less than the preservation of human understanding.

“Each language contains a whole universe of understandings,” says Dr Thieberger. “Knowledge can be separated out from one language into another, but there is some knowledge that is embedded in the language itself. That knowledge can for example be encoded in the similar names that may be given to different things that then expresses a relationship between them, such as perhaps the position of constellations indicating the timing of seasons when a particular animal is good to hunt.”

For example, in the Nafsan language of Vanuatu’s main island of Efate, Dr Thieberger says the importance of exchange in small island societies is underlined by the complex vocabulary Nafsan developed around gift giving. In Nafsan, they have a word for a gift that is given with no expectation of a return gift - tingpiel. It is distinct from a gift given in return for a gift, which is siriu. There is also a word for a gift given in exchange for a service - sautong, and there is a special verb for gift giving - ptu, which is distinct from tu - the verb for giving anything else.

It’s this complexity and knowledge that we’re in danger of losing.

“Languages with a long connection to the local environment like those in Australia and the Asia Pacific display a special knowledge of how interconnected the world is, and having that knowledge was crucial for these communities being able to survive. If we lose these languages we all lose other ways of understanding our world.”

Banner Image: Shutterstock