Nêhiyawi-pîkiskwêwina maskwacîsihk : Spoken Dictionary of Maskwacîs Cree

Abstract

ABSTRACT: This paper details the development of nêhiyawi-pîkiskwêwina maskwacîsihk: Spoken Dictionary of Maskwacîs Cree (in progress). Since 2014, this joint project between the Maskwacîs Education and Schools Commission (MESC) and the Alberta Language Technology Lab (ALTLab) has sought to record carefully pronounced, isolated spoken audio for the approximately 9,000 entries in the Maskwacîs Dictionary of Cree Words (Maskwachees Cultural College 2009), as well as to fill lexical gaps through elicitation, to record example sentences for as many of these entries as possible, and to make these recordings publicly available online. Between 2014 and 2018, approximately 700 hours of audio and close to 120,000 recordings for 20,300 carefully spoken word and phrase types were gathered in elicitation sessions. After extracting and annotating the relevant Cree vocabulary, these audio clips were compiled in a novel, publicly accessible online Speech Database as well as through itwêwina , the intelligent bilingual online Cree–English dictionary. The entries in this database are currently in the process of orthographic standardization, gloss standardization, and linguistic analysis. Simultaneously, native speakers of Cree are re-reviewing the database's entries to ensure pronunciation quality and verify definitions where needed. In this paper, we discuss the origins of this project; the original elicitation sessions; the postprocessing, standardization, and validation of the recordings; and means by which these recordings can be publicly accessed online.