Modeling Northern Haida Verb Morphology

Modeling Northern Haida Verb Morphology Jordan Lachler1, Lene Antonsen2, Trond Trosterud2, Sjur N. Moshagen2, Antti Arppe1 1University of Alberta 2UIT Arctic Univerity of Norway Email: {lachler, arppe}@ualberta.ca {lene.antonsen, trond.trosterud, sjur.n.moshagen}@uit.no Abstract This paper describes the development of a computational model of the morphology of Northern Haida based on finite state machines (FSMs), with a focus on verbs. Northern Haida is highly endangered, and a member of the isolate Haida macrolanguage, spoken in British Columbia and Alaska. Northern Haida is a highly-inflecting language whose verbal morphology relies largely on suffixes, with a limited number of prefixes. The suffixes trigger morphophonological changes in the stem, participate in blocking, and exhibit variable ordering in certain constructions. The computational model of Northern Haida verb morphology is capable of handling these complex affixation patterns and the morphophonological alternations that they engender. In this paper, we describe the challenges we encountered and the solutions we propose, while contextualizing the endeavour in the description, documentation and revitalization of First Nations Languages in Canada. Keywords: less-resourced/endangered languages, language modeling, morphology 1. Introduction While the study of the Indigenous languages of North America has been a mainstay of descriptive linguistics over the past century or more, these languages have received much less attention in the computational linguistic literature. This has begun to change in recent years, e.g. through work initiated on Algonquian languages such as Plains Cree (Snoek et. al., 2014; Harrigan et. al., 2017), East Cree (Arppe et al., 2017a), Odawa (Bowers et al. 2017), as well as Dene languages such as Tsuut’ina (Arppe et al. 2017b). This work has come about through computational linguists, field linguists and Indigenous communities working together to create tools and applications, which can support the continued use and learning of these often threatened languages (Arppe et al., 2016). In this paper, we provide a description of a computational model of the verbal morphology of one highly endangered North American Indigenous language, Northern Haida, which was created in furtherance of these goals. 2. Background 2.1. Northern Haida Northern Haida, known as Xaad Kíl, is one of two members of the macrolanguage Haida (Simons and Fennig, 2017). Speakers of Northern Haida live in a small number of communities on the Haida Gwaii archipelago off the coast of British Columbia, and on the islands of far southeastern Alaska. Haida is an isolate, with no known linguistic relatives. Today, there are fewer than a dozen fluent speakers of Northern Haida remaining, all of them over the age of 70, but also an active and growing cohort of younger learners of the language. 2.2. Morphological Structure of Verbs Most verbs in Northern Haida are entirely suffixing in a relatively neatly agglutinative manner. They inflect for a range of grammatical categories including Tense, Aspect, Mood, Evidentiality, and Polarity among others, but not for the person and number values of their subject or object. The only exception to this is the Third Person Plural suffix -'wa-, which is used to indicate that there is a definite third person plural …

Content truncated to 500 words