A Pedagogical Corpus to Support a Language Teaching Curriculum to Revitalize an Endangered Language: The Case of Labrador Inuttitut

Abstract

An obstacle to revitalizing an endangered language is the shortage of authentic speech samples for learners to use as models. Digital recordings of community elders performing traditional chores and special rituals or narrating legends and myths are often made to overcome this obstacle. These recordings, however, contain speech that lacks the crucial features of conversational speech that make them appropriate instructional models. Effective model utterances should be short, have a stand-alone format, and have similar structures to utterances used in everyday transactions, which must be labeled and tagged and organized into a searchable corpus. To date, however, no such corpus exists for indigenous languages, and compiling one is an enormous task. To provide native speech models for adult Labrador Inuit learning their endangered language, Inuttitut, the authors explored the feasibility of building a specialized corpus potentially useful for aiding classroom instruction, using an internationally recognized open-source search and retrieval system called Topic Maps to create its database.