We are pleased to announce that Dr. Gregory D. S. Anderson and Dr. Luke Horo presented their work at the conference, Linguistic Issues in Speech Processing Research of Under-Resourced Languages (LISPRUL 2022) on March 2-3, 2022. The title of their presentation is “Under-resourced Languages and Documentation in India: The Living Tongues Approach.” Their abstract can be read below.

View all presentations and slides

ABOUT LISPRUL 2022

From the conference website: “There are 22 scheduled languages in India and several hundreds of under-resourced languages. Development of speech technologies like speech recognition, machine translation, speech synthesis and speech to speech translation systems for these languages is a resource and time intensive task. Apart from technology development, speech and language data collection in the under-resourced languages may also be aimed towards creating linguistic archives for community use and for linguistic analysis in digital, easily usable and open access manner. In this workshop, we intend to learn how linguistic data archiving, analysis and technology development can be accomplished in a synergetic manner.

Eminent researchers from all over the world, specialising in the fields of speech and language processing, linguistic analysis, archiving and technology development will share their experience and expertise in this workshop. Apart from in-depth discussions, the sessions are designed to have ample time for the participants to have open discussions with the speakers.”

---

TITLE OF PRESENTATION
“Under-resourced Languages and Documentation in India: The Living Tongues Approach”

by Gregory D. S. Anderson and Luke Horo

ABSTRACT
In this presentation we offer some details about our work documenting the under-resourced languages of India. We begin with a discussion of the two sets of languages we have worked on over the past decades in India which belong to the Munda and various subgroups of the Trans-Himalayan (Tibeto-Burman) language families. We then draw attention to how we have gone about classifying these languages and situate this against computational phylogenetics based on Swadesh lists that dominate the field today. We then give an overview of the Munda Languages Initiative and detail the types of data we collect and how we analyze it and why we feel this is the correct way to do so. We then detail some of the past and ongoing scientific and applied outcomes of this work including an introduction to a powerful tool the Living Dictionary app that we have developed that aids linguists and citizen-scientists alike in creating high-quality and free documentation records. We conclude with a look to ongoing and future projects and ways that interested and qualified participants can find roles in the furthering of these projects and the development of skills in best practice in language documentation.