Re-evaluating the reconstruction of Proto-Eskimo-Aleut

Abstract

Abstract The relationship between Unangam Tunuu (Aleut) and Eskimo was established in the early 19th century, and the 20th century especially saw a number of efforts on the reconstruction of Proto-Eskimo-Aleut (PEA). Reconstruction has supported assumptions of a largely genealogical relationship between the EA languages, assumptions which include a long history of independent development in isolation from other languages and language families. The reconstruction of PEA, however, is incomplete; many apparent cognates have irregular or imperfectly understood sound correspondences. Furthermore, advances in archaeology and genetics have called into question many assumptions about EA prehistory and about the isolation or lack thereof of Unangam Tunuu. In this study, I re-examine the proposed cognates and evaluate them based on the strength of their correspondences and their distribution within the lexicon, with reference to new findings regarding technological innovations and periods of cultural contact. Several patterns emerge, including a large group of proposed cognates with overly-specific semantic correlations relating to technologies or cultural practices post-dating the split of EA languages, a gender-based difference in the number of cognates relating to cultural activities, and a correlation between known borrowings and high levels of cognates in certain semantic domains. Results suggest extensive language contact, especially in the past millennium.