News

Global mission to save dying languages

[ UniNews Vol. 12, No. 2  24 February - 10 March 2003 ]

A University of Melbourne computer scientist has embarked on a global mission to save thousands of endangered languages in the world from disappearing forever.

Linguists estimate there are more than 6000 languages spoken worldwide. They also say that 90 per cent of these are likely to die before the end of this century, and 440 are within only a generation or two of extinction.

Most of these languages are not written down and when they disappear, any record of them will disappear along with unique oral histories and cultural heritage.

Associate Professor Steven Bird, of the University’s Department of Computer Science and Software Engineering, and a team of international colleagues, have launched a worldwide project to preserve threatened languages by designing and building an open language archives community (OLAC).

OLAC is an international network of 25 digital archives in six countries that already hold more than 30,000 items in the form of dictionaries, grammars, field notes, text collections and recordings of the world’s languages.

“We are sitting between the onset of the digital era and the mass extinction of the world’s languages. The window of opportunity to record these endangered languages is small and shutting fast,” says Associate Professor Bird.

“Unfortunately, digital preservation faces huge obstacles. Most digital media and storage formats are obsolete within five years and data must be continually transferred over to new formats.

“A lot of linguistic data is stored in Microsoft formats and secret binary formats that cannot be opened (even by Microsoft software) just five years later.

“Linguists also use a range of tools including word processing documents, spreadsheets, different forms of digital sound and video capture technology and end up in a terrible mess when they try to collate their data with other linguists’ research because of the incompatibility or unsuitability of the software used.”

One of Associate Professor Bird’s main tasks is to develop a standardised OLAC template that researchers will use to collect linguistic information, and a management system to ensure the data is accessible to future technologies.

“Without these tools we are simply turning endangered languages into endangered data,” he says.

It is not just cultural heritage that dies with a language. Thousands of years of history and a unique point of view on the world are lost along with knowledge of plants, animals and their uses as food and medicines, and their role in a healthy environment.

Two thousand languages (a third of the world’s linguistic heritage) are in the Australasian region, which includes South-East Asia and the South Pacific.

“It is a living museum of language and culture and Australia is the best placed to preserve this heritage.”

Associate Professor Bird says Australia has more critically endangered languages than all other countries combined. An estimated 80 Indigenous Australian languages have only five or fewer speakers.

The project is a race against the clock. While still desperate for funding, Associate Professor Bird has managed to secure $17 million in grants from the USA and UK in the past decade. Late last year, the universities of Sydney and Melbourne received a $275,000 Australian Research Council grant for hardware to digitise hundreds of hours of archived recordings of Indigenous languages.

“The ARC grant will help us rescue deteriorating tape archives of many endangered and extinct languages from our region making them accessible in perpetuity,” says Associate Professor Bird.

He says the Victorian Partnership for Advanced Computing has also given $30,000 for high-performance computing involving large linguistic databases. The grant will be applied to language technologies of commercial interest including spoken dialogue systems, networked archive systems such as OLAC and the analysis of data from endangered languages.

---
top of page