Institutional Partners

The Australian Institute for Aboriginal and Torres Strait Islander Studies (AIATSIS) is the national institution focused on the diverse history, cultures and heritage of Australia’s First Nations people. CoEDL established a partnership with AIATISIS from the outset to facilitate the Centre’s work on analysis, documentation and archiving of language data and materials. Research Associate Doug Marmion, CoEDL AIATSIS liaison, supported this partnership.

AIATSIS worked with CoEDL members to improve access to the vast collections of Australian language materials and to support community groups to deposit materials with the AIATSIS archive. This assisted CoEDL’s research and corpus building projects, made the archived language resources more accessible for Indigenous communities and provided safekeeping for precious language material.

From 2018, CoEDL also began collaborating with AIATSIS on the National Indigenous Languages Report (NILR) for the Australian Government Department of Infrastructure, Transport, Regional Development and Communications (Office for the Arts). It was published in August 2020 and followed by an online forum, co-hosted by CoEDL and First Languages Australia on the implications of NILR for the Closing the Gap targets. This work contributed to the Centre’s efforts to translate research into policy.

CoEDL’s partnership with AIATSIS resulted in several other collaborations, including the Paper and Talk Workshop, held with AIATSIS and Living Languages in September 2019. Centre members provided feedback into the 2019 revision of the AIATSIS Guidelines for Ethical Research in Australian Indigenous Studies, which informed CoEDL’s own Indigenous Linguistics and Cultural Heritage Ethics Policy. Doug facilitated AIATSIS support for research placements and internships working on Indigenous language. For example, Sarah Lattimore and Yuka Morinaga helped develop a Ngunnawal language learners grammar requested by Ngunnawal community members.

Doug also coordinates the AIATSIS Dictionary Project, which facilitates existing efforts across Australia publish dictionaries on Australian Indigenous languages and provide copies of the dictionaries to community members. As well as supporting the project, several CoEDL members compiled and submitted material to be made into dictionaries. By 2022, the Dictionary Project had published resources for over 15 languages.

Founded in Australia in 1996, Appen is a global company providing high-quality speech and search technology services. CoEDL established a partnership with Appen to facilitate the use of technology for speech and natural language processing and to access Appen’s skill in language data transcription and annotation. When CoEDL began in 2014, Appen provided language technology data and services in 140 languages and dialects; as of 2022, they have expertise in over 235 languages.

Appen shared its expertise with the CoEDL community in a number of ways. The Appen Internship Program paired CoEDL-affiliated research assistants and students with experienced Appen linguists, who introduced mentees to best-practice processes for language resource development and provided access to tools and technical support.

After a successful first run at Appen’s Sydney location in 2015, the internship program was made available to participants from across Australia and other countries. Examples of projects undertaken by Appen interns include Sasha Wilmoth’s work on mor-coding to refine the searchability of the Gurindji Kriol corpus led by CI Felicity Meakins and Ben Purser’s work to prepare transcripts for phone alignment and find ways to automate elements of the Sydney Speaks Project. Appen also sponsored summer research scholars to work on Sydney Speaks at ANU in 2017.

In 2017, this partnership led to an effort to localise tablet interfaces for members of a remote community in Papua New Guinea to use the devices in their local language. Appen engineer Jason Johnston supported CoEDL Research Fellow Hannah Sarvasy and Nungon speakers to localise a word processing software and Start menu for computers to be housed at the Nungon Community Technology Centre in Towet, a village in PNG’s Morobe Province. The goal of the project is to promote local language sustainability by introducing technology to the region in the Nungon language first, rather than wait for digital devices to usher in neologisms.

Corpus Manager Wolfgang Barth, CoEDL-Appen liaison, also connected many CoEDL researchers with Appen, who provided technical support for automating data workflow processes. In some cases, the cooperation focused on a single task, like standardising data or removing a bottleneck to accelerate data processing. In other cases, support continued over years through continuous dialogue. CoEDL researchers benefitted from Appen’s technical expertise and from exposure to the practical nature of work in the language industry.

CoEDL established a partnership with the Max Planck Institute for the Science of Human History (MPISHH) in Jena, Germany in 2015. From then, the Centre collaborated closely with PI Russell Gray and AI Simon Greenhill, both based at MPISHH, on the ambitious Glottobank project.

Glottobank is an international research consortium established to measure and understand the world’s linguistic diversity against a common set of yardsticks. Of the five global databases Glottobank has created, CoEDL members had a foundational role in three that document variation in language structure (Grambank),  lexicon (Lexibank) and paradigm systems (Parabank). The Glottobank team — which also included Director Nick Evans, PhD students Hedvig Skirgård and Kyla Quinn and Corpus Manager Wolfgang Barth — participated in and coordinated several workshops in Australia, Germany and New Zealand to assist the establishment of these databases.

The project seeks to develop new methods in language documentation, to compile comparable data on the world’s languages and to make this data accessible and useful, particularly to facilitate the study of human prehistory, relationships between languages and processes of language change.

This work was a major part of CoEDL’s investigation of deep relationships between languages and the development of a global phylogeny — family tree — for language evolution, which would shed light on the origins and dispersals of diverse peoples. Rigorous methods for estimating divergence and change were initially developed in evolutionary biology; CoEDL researchers have helped to re-engineer these models for linguistics.

For example, in 2018 Russell and Simon were part of an international team that uncovered the origin of the Dravidian language family, consisting of about 80 varieties spoken by 220 million people across southern and central India and surrounding countries; they estimated that it is at least 4500 years old [1]. Simon was also part of a study led by Laurent Sagart that estimated the Sino-Tibetan language family is at least 7200 years old [2].

Despite the increase in linguistic databases worldwide, the independent nature of many of these databases creates real difficulties for conducting comparative studies. Another important outcome of this collaboration, therefore, was Russell and Simon’s role in the MPISHH Cross-Linguistic Data Formats Initiative, which in 2018 helped to develop new guidelines on cross-linguistic data formats that facilitate sharing and data comparisons.