A Morphological Analyzer for St. Lawrence Island / Central Siberian Yupik

A Morphological Analyzer for St. Lawrence Island / Central Siberian Yupik Emily Chen, Lane Schwartz Department of Linguistics University of Illinois Urbana-Champaign, Champaign, Illinois {echen41, lanes}@illinois.edu Abstract St. Lawrence Island / Central Siberian Yupik is an endangered language, indigenous to St. Lawrence Island in Alaska and the Chukotka Peninsula of Russia, that exhibits pervasive agglutinative and polysynthetic properties. This paper discusses an implementation of a finite-state morphological analyzer for Yupik that was developed in accordance with the grammatical standards and phenomena documented in Jacobson’s 2001 reference grammar for Yupik. The analyzer was written in foma, an open source framework for constructing finite-state grammars of morphology. The approach presented here cyclically interweaves morphology and phonology to account for the language’s intricate morphophonological system, an approach that may be applicable to typologically similar languages. The morphological analyzer has been designed to serve as foundational resource that will eventually underpin a suite of computational tools for Yupik to assist in the process of linguistic documentation and revitalization. Keywords: computational morphology, finite-state grammar, morphological analyzer 1. Introduction We introduce in this paper an implementation of a morpho- logical analyzer for St. Lawrence Island / Central Siberian Yupik (ISO 639-3: ess), an agglutinative, polysynthetic lan- guage of the Inuit-Yupik language family.1 This analyzer is implemented in foma (Hulden, 2009b), and represents a faithful adaptation of the grammatical and morphophono- logical rules documented in A Practical Grammar of the St. Lawrence Island / Siberian Yupik Eskimo Language (Jacob- son, 2001). We intend for the morphological analyzer to be a foundational resource that underpins a suite of com- putational tools to be shared with the Yupik community for purposes of language preservation and revitalization. 2. Language Description Yupik is the westernmost variety of the Inuit-Yupik lan- guage family (Krauss et al., 2010). Most of the 2400–2500 Yupik people reside in two villages on St. Lawrence Island, Alaska and two villages on the Chukotka Peninsula of far eastern Russia (Krupnik and Chlenov, 2013). Out of that population, fewer than 1000 are estimated to be L1 Yupik speakers (Koonooka, 2005; Schwalbe, 2017). The Yupik phonemic inventory comprises 31 consonants and 4 vowels, /ə/, /i/, /a/, /u/ which can be lengthened (with the exception of /ə/) for a total of seven vocalic phonemes. Of the 31 consonants, there are 8 pairs of continuants and 4 pairs of nasals where each pair differs only in voicing. In the standard Latin-based orthography, this difference is marked in 5 of the 8 continuant pairs and in all of the nasal pairs through graphemic doubling, that is, l and ll, r and rr, g and gg, gh and ghh, ghw and ghhw, m and mm, n and nn, ng and ngng, ngw and ngngw. In each pair, the doubled grapheme represents the voiceless phoneme. 2.1. Morphology Like all languages in the Inuit-Yupik family, the morphol- ogy of Yupik is remarkably generative. Yupik nouns and 1The term Yupik will be used henceforth to refer to the St. Lawrence Island / Central Siberian Yupik variety. verbs …

Content truncated to 500 words