Alaska Native Language Center

ABS = absolutive, AN = anaphoric, AUG = augmentative, AUX = auxiliary, CAUS = causative, CESS = cessative, COND = conditional, CONJ = conjunctive, DAT = dative, DEM.DIST = demonstrative distal, DEM.PROX = demonstrative proximal, DIM = diminuative, DU = dual, DUB = dubitative, ENCL = enclitic, FUT = future, HAB = habitual, IND = indicative, IMP = imperative, INTEN = intentional, INTER = interrogative (particle), NEG = negative, PART = participial, PASS = passive, POS = possessor, POSM = possessum, PL = plural, REL = relative, SG = singular, SUP = superlative, TRANS = transitive, ZERO = participial with zero morphology

Aleut does not appear to have antipassive constructions, as the related Eskimo languages so ubiquitously do. This lack may be the result of fundamental changes in the Aleut inflectional system, as suggested by Bergsland ( , 1997. However, antipassivization is not the only possible means of decreasing valency by means of object-reduction, and a close examination of Aleut suggests at least three strategies for doing so. These include the use of a valency-decreasing suffix, simple omission, and reflexivization. In this paper, I explore the uses of each of these strategies.

In the mid-19th century, a Central West Greenlandic catechist named Hans Egede wrote an account of his family’s oral history in Greenlandic. The unpublished manuscript is extensive and gives fascinating information about daily life, including hunting practices, religious traditions, and local traditions and superstitions, spanning the period from the 1780’s to the 1850’s. Working with this manuscript, I have transliterated Egede’s writing into the modern Greenlandic orthography and added a morphemic breakdown with English translation. This process has brought to light interesting linguistic data, including differences between 19th and 20th century Greenlandic, oral narrative style in Greenlandic, and the importance of contextual information in decoding discourse. This paper presents a brief analysis of some of these points.

The Eskimo-Aleut language family is well known for its extreme polysynthesis; in general, however, descriptions of this extreme polysynthesis more accurately describe the situation in Eskimo languages. In fact, both the degree and the nature of polysynthesis in Aleut are very different. In this paper, I present a detailed discussion of the ways in which Aleut polysynthesis differs from that of Eskimo. In particular, I show that it is not only much reduced, but also more templatic and more fluid than polysynthesis in Eskimo.

Handout from Conference Presentation
Using data from both published texts and original fieldwork, I examine neutral and non-neutral word order, the indexing and expression or lack thereof of arguments, and topic and topic/comment structure in Aleut.  I explore some challenges that Aleut presents to the theory of information structure as a result of features such as its polysynthesis, verb indexing, and clause chaining.

The Eskimo-Aleut language family is well known for its extreme polysynthesis; in general, however, descriptions of this extreme polysynthesis more accurately describe the situation in Eskimo languages. In fact, both the degree and the nature of polysynthesis in Aleut are very different. In this paper, I present a detailed discussion of the ways in which Aleut polysynthesis differs from that of Eskimo. In particular, I show that it is not only much reduced, but also more templatic and more fluid than polysynthesis in Eskimo.

Aleut is the only language in its branch of the Eskimo-Aleut language family.  It is quite divergent from Eskimo languages and is traditionally considered to have developed in isolation, both from Eskimo and neighboring languages, until the Russians colonized the Aleutians in the 18th century (Bergsland 1986).  Many assumptions concerning, e.g. the time and place of the proposed split between the two branches of the language family, features present in the proto-language, and subsequent independent developments, have in the past been used to support interpretations of archaeological findings (e.g. Dumond 1965, 1979, 1984).  These assumptions, based primarily on rather few and far from finished comparative studies, have not yet been adequately reexamined in decades, despite the subsequent broadening of the field of historical linguistics. Meanwhile, recent advances in the archaeology and genetic investigations of the Aleut and a growing body of linguistic research all increasingly suggest extensive contact with neighboring groups (Crawford, M.H. et al. 2010, inter alia, Leer 1991, Fortescue 1998, 2002, Berge 2012, 2014). In this paper, I argue for the need for a comprehensive reexamination of the linguistic prehistory of Aleut.

A re-examination of the genesis and development of the Eskimo-Aleut language family, starting with an in investigation into the distribution of Eskimo cognates to non-cognates in Aleut (Unangam Tunuu); Eventual goal:  Understanding migration and language contact in the Aleutians and Pacific Coast of Alaska.

Unangam Tunuu is the language of the people indigenous to the Aleutian Islands; it is highly endangered, but speaker communities are found in the Aleutian and Pribilof Islands of Alaska and the Commander Islands of Eastern Russia. Pribilof Anĝaĝigan Tunungin / The Way We Talk in the Pribilofs is a serious and in-depth introduction to one of the less-well documented varieties of Unangam Tunuu, that spoken on the Pribilof Islands. The book is intended for adult learners and provides some of the first college-level materials for conversational language learning of Unangam Tunuu, although this also may be used in high schools or for self-study. The book is organized in three units of five lessons each; the first four lessons of each unit introduce vocabulary, grammar, cultural notes, and (in Unit 1 only) pronunciation notes, while the fifth lesson is geared toward review and practice. Most lessons are accompanied by sound files for self-study and practice. The materials lead the learner toward conversational ability, with many examples of individual speaker preferences and different ways of expressing the same thing. All examples were collected from Native speakers from 2002 to 2012. The book is unique in presenting grammatical, lexical, and cultural information specific to the Pribilofs; differences with other documented dialects are noted. In addition, and for ease of reference, the book includes a number of appendices such as a glossary of linguistic terms used in the lessons, a grammatical sketch with many charts of grammatical forms, and an Unangam Tunuu–English / English–Unangam Tunuu lexicon. This is an entry-level introduction to conversational Unangam Tunuu; it is not a comprehensive grammar of the language. Upon completion of these units, a learner will have enough basic language skills to continue learning from Elders and from published materials and recordings.

Unangam Tunuu (Aleut) is member of the Eskimo-Aleut (EA) language family; its extreme divergence has long been thought to be a result of its relative isolation over an extended period of time (Bergsland 1986, Krauss 1990). Recent work in archaeology and genetics, however, is suggestive of long-term cultural contact with both neighboring Eskimo and non-Eskimo groups. The linguistic evidence for contact is not yet clearly established, although there is strongly circumstantial evidence of contact from both lexical studies of the distribution of cognates and borrowings in different semantic domains (Berge 2016a, submitted; Berge and Holton 2015) and grammatical studies of innovations in Unangam Tunuu shared with neighboring non-EA peoples (Leer 1991, Fortescue 1998, Berge 2016b).
Language contact effects are apparent in both the lexicon and the grammar, reflecting contact primarily with the Alutiiq Eskimo and the Dena’ina Athabaskans.  Lexically, there are distributional differences in the number of cognates and borrowings in certain semantic domains between Unangam Tunuu and Eskimo; and Unangam Tunuu, but not Eskimo, has features more characteristic of Athabaskan languages, such as multiple synonyms for basic terms, multiple phonological variants for many words, and signs of a once-active replacement strategy for nouns.  Grammatically, Unangam Tunuu has EA structure:  basic word formation, nominal and verbal inflection, verbal mood, deictic terms, particles, etc. appear to reconstruct to EA; however, there have been substantial changes in this grammatical system.  Some features of Unangax̂ grammar look like importations from Yupik Eskimo grammar; the deictic system, for example, likes like a Yupik deictic system that has been reorganized and paradigmatically leveled. On the other hand, many features of Unangax̂ grammar are found in neighboring non-EA languages, such as auxiliary verb formations, widespread use of positional nouns, stem-stem compounding (Berge 2016b).
These two contact experiences occurred during the same time period, from about 1000-300 BP; however, they had vastly different effects.  The arrival of the Alutiiq Eskimos at the easternmost edge of Unangax̂-speaking territory may have resulted in a shift from Unangam Tunuu to Alutiiq, pushing the Unangan further east; but in Unangam Tunuu, the effects of contact were relatively superficial, perhaps limited to certain types of borrowings suggestive of an influx of Alutiiq men and most probably as a result of warfare caused by economic or ecological stress (Berge, submitted).  Around the same time, the whole Pacific Coast region shows a radical shift in social complexity, involving social stratification, warfare motivated by the need for social status and slaves (Misarti and Maschner 2015). Slaves, mostly women or children, were often from non-Unangan speaking communities such as the newly arrived Alutiit and Dena’ina.  This type of contact shows substratum influences from non-Unangan languages, as a result of imperfect learning during language shift.  This explains the few numbers of attested borrowings from other languages, but the extensive borrowing of grammatical structures. 
In this paper, I discuss the non-linguistic evidence for late prehistoric linguistic contact in Unangam Tunuu, the linguistic evidence for each type of contact discussed above, the methods used to determine linguistic contact as opposed to direct inheritance, and the implications for the study of EA and the identification of substratum effects.

This paper is available to read online at https://benjamins.com/#catalog/books/z.215/toc.
The Eskimo-Aleut are arctic and subarctic hunter-gatherers known for their geographic
spread and successful adaptation to a harsh climate; they are one of the
canonical examples of a people that spread without agriculture. One of the most
prehistoric recent spreads in this language family occurred about 1000 years
ago, with effects felt throughout coastal Alaska. One area of language contact
and possible spread was in Southeast Alaska, between the Pacific Coast Yupik
language Alutiiq and the Aleutian language Unangam Tunuu. In this paper,
I look at the distribution of cognates and borrowings of subsistence terminology
in Unangam Tunuu, and I show that Alutiiq must have spread into a previously
Unangax! area as a result of warfare rather than subsistence activities.

Unangam Tunuu (UT, a.k.a. Aleut; ISO 639-3 ale) is member of the Eskimo-Aleut (ISO 639-5 esx) language family; its extreme divergence was long thought to be a result of its assumed isolation from other languages, the length of time since the branches of the family split, and the effects of prehistoric contact with an unknown substrate language  (Bergsland 1986, Krauss 1990, Fortescue et al. 2012).  Recent work in archaeology and genetics, however, does not support these assumptions:  there appears to have been long-term cultural contact with both neighboring Eskimo and non-Eskimo groups, with several periods of more intense contact, particularly around the historic boundary of the eastern Unangax̂ and Alutiiq peoples.  The most recent period, from about 1000 BP, involved the entire Pacific Coast area and may have involved some degree of population replacement (Smith et al., 2009). The nature of this contact is still unclear:  Maschner (2016) argues for a former affinity between the cultures of the Aleutian Islands and Kodiak, and the recent arrival of Alutiiq culture on Kodiak, which displaced an indigenous population closely related to the Unangan.  Steffian et al. (2016) argue for long-term continuity on Kodiak and thus several thousand years of indigenous Alutiiq culture, recent external influences from the north, and extensive trade and warfare with neighboring cultures, including the Unangan.
A number of linguistic studies have noted UT features shared with neighboring languages.  Bergsland (1986, 1994) and Berge (2017) have noted lexical borrowings between UT and neighboring Eskimo languages, predominantly from UT into Alutiiq (Alutiiq, ISO 693-3 ems).  Leer 1991, Fortescue 1998, 2002, etc., and Berge 2016 have noted grammatical features shared with Athabaskan-Eyak-Tlingit (AET; ISO 639-5 ath, eya, tli) languages, some of which are typologically unusual and most of which are non-trivial.
In most cases, however, no specific period or mechanism of contact have been proposed; these are, however, crucial to any discussion of Eskimo-Aleut, since UT shows substantial contact effects, even extending to signs of language mixing or shift (Berge 2017).  UT must have had contact with an AET language or languages and yet these are thought to have been relatively late arrivals on the northern Pacific Coast (ca. 1300 BP, cf. Kari 1989, Workman and Workman 2010), and, furthermore, there are several hundred miles and the Alutiiq language separating UT from the closest AET language.  This contact must have been substantial enough for the effects thereof to permeate all UT dialects by the time they are first recorded, in the 18th century.  There are suggestions of a late prehistoric westward wave of language features along the Aleutian Islands (Woodbury 1984, Bergsland 1994, Berge 2010) and some indication of an influx of new people (Black 2003, Smith 2009), probably from Kodiak Island (Black 2003, Maschner 2016), but there is no compelling evidence of wholesale population replacement, either on Kodiak Island or along the Aleutians.
In this paper, I discuss a) evidence for the former presence of UT in currently Alutiiq areas (especially Kodiak Island and the Kenai Peninsula), which provides a geographical justification for UT-AET language contact effects, b) linguistic evidence suggestive of a language shift from AET to UT and a later language replacement of UT with Alutiiq on Kodiak Island, and c) the problem of apparent language shift(s) without obvious population replacement (cf. Heggarty 2015).

Previous studies have noted lexical and grammatical features shared between Unangam Tunuu and neighboring Eskimo (esp. Alutiiq) and Athabaskan-Eyak-Tlingit (AET) languages, suggesting prehistoric language contact, without, however, proposing a specific period or mechanism of contact.  In this paper, I provide evidence from likely Unangan place names to support archaeological findings suggesting the former presence of UT in contemporary Alutiiq areas; and further, that Unangax̂ presence along the eastern Pacific Coast provides a locus of contact to explain AET-like grammatical features in UT

Assimilation is common. Two of the more common types of assimilation involve voicing assimilation. Thus, voiced approximants and nasals are devoiced near voiceless consonants ((14); n.b. voiceless nasals are not phonemic), including when postbases are added, and likewise, voiceless ones are voiced in voiced environments (13). There are some exceptions, such as uɣtsəx 'bearded seal,' where the voiced velar is not devoiced; a full analysis of the exceptions remains to be done. The sibilant /s/ is underlyingly voiceless and may be voiced through assimilation (Menovshchikov 1964:26). [ʃ] appears to be an allophone of /s/, although a schwa may intervene between [ʃ] and a voiceless consonant (15). (14) kitn̥ əχ 'heel' puvəx 'wind' à pufkəʃiʁapantaχ 'hurricane' (15) təkəʃəx 'corner' naʁəɫiqʃista 'inhabitant' An interesting exception to these assimilation rules arises in combinations of <tʃ> (variously pronounced as /ts/ or /tʃ/) and consonants that normally assimilate. Menovshchikov (1964:18) notes that it seems to correspond to voiced approximants in other Yupik and Inuit languages. Fortescue et al. (2010:xvii, xix) propose multiple PE sources for Sirenikski /tʃ/, but these appear limited to PE *ty and *l (but PE *ɫ > Sir. /s/; some historical details remain to be worked out). Another common type of assimilation is assibilation with or without affricativization. Some of the more frequently found examples include t à ts after a voiced continuant or schwa (16) or j à s before a dental stop (17; the applicative suffix-ə(t)-loses the dental stop) (Vakhtin 2000). (16) t à ts / ə_ iqə-tsəqəχ-tə-χ fall-PRS-IND-3SG 'he is falling, sinking' t à ts /_voiced continuant yiv-tsəqəχ-tə-χ shimmer-PRS-IND-3SG 'he is shimmering' t elsewhere asəvaχ-təqəχ-tə-χ ascend-PRS-IND-3SG 'he is descending' (17) j à s /dental stop _ aʁaʁ-ə(t)-suk-a! go-APPL-OPT-3SG 'let me lead him!' j elsewhere aʁaʁ-juɣ-tsə-χ go-want-IND-3SG 'he must go, he wants to go' Stem-final-tə always entails some special rules.-tə may be dropped before a suffix (18); it may entail the assibilation of a following suffix, as in the transitivizing suffix-t-(19); or it may be assibilated (20): (18) mitə-'to grow'-tə# dropped before suffix mi-təqəχ-tə-χ grow-PRS-IND-3SG 'he is growing' naʁətə-'to live' naʁə-təqəχ-t-uj live-PRS-IND-3PL 'they live/are living' (19) naʁətə-'to live' assibilation of suffix following-tə# naʁətə-s-qə-l-ki live-CAUS-PRS-CONJ-3PL 'they, being able to live' (20) mitə-'to grow'-tə# à-s-mis-əmə-tsə-χ grow-PST-IND-3SG 'he grew' 2.2.2.2 Deletion Deletion occurs especially stem-finally (Menovshchikov 1964:25). Thus, stem-final consonants are lost following another consonant when the resulting combination would be disallowed (21). Stem final vowels are lost before the initial vowel of a postbase (22). As mentioned above, stem-final-te tends to be dropped (18). (21) jaquχ-n arm-2SG.POSS/PL.POSM 'your arms' à jaqun (22) asa-i land-LOC 'on land' à asi 2.2.2.3 Epenthesis Epenthesis may not be significant in Sirenikski. Menovshchikov (1964) assumes a schwa may be inserted to break up consonant clusters resulting from the addition of consonant-initial suffixes. However, many of his examples of epenthetic phonemes appear to be misparsed. Thus, in (23), he analyzes the passive form as consisting of an epenthetic schwa (Menovshchikov 1964:80), whereas it actually derives from the past morpheme-əma-, ultimately from PE *-(u)ma-'perfective,' see section 4.2.4.1). Many schwas drop as a result of some morphophonological affixation, and reappear as a result of others ((24); Vakhtin 2000:508). (23) *qatsiɣ-ə-matsaχ-tə-χ put-EPEN-PASS-IND-3SG 'it was put' à qatsiɣ-əmatsaχ-tə-χ put-PASS-IND-3SG 'it was put' (< qatsiɣ-əma-tsaχ-tə-χ) put-PST-PST.PRT-IND-3SG (24) əvə-'to pour' əv-lataʁ-əmə-rax pour-again-PST-3SG.PAST.IND 'he poured it аgain' vs. əvə-pəstaχ-təʁ-a pour-not.for.nothing-IND-3SG/3SG 'he didn't pour it for nothing' 7 Stress assignment in Sirenikski is quite unlike Yupik stress assignment, cf. Krauss 1985:177 and the chapters on Yupik languages in this volume. Different characteristics include the assignment of stress on the second syllable regardless of syllable shape and stress on reduced vowels. 8 Schwas are also stressed under specific circumstances in Alutiiq, Norton Sound CAY, and CSY (e.g. Jacobson 1990:280 atə́pik 'real name'). 16 Menovshchikov (1964) calls it a gerund; but this is term used differently in the other languages as well as in general linguistics today. It is the mood known as 'subordinative' in Central Alaskan Yup'ik; appositional in most Inuit languages; contemporative in Greenlandic; and conjunctive in Unangam Tunuu. 17 Menovshchikov calls the anterior a participial; but it fulfills the same function as the anterior in other languages. I am using the term 'anterior' because of its function; it is also known as the 'consequential' in descriptions of Alaskan Yupik and Inuit languages, and as the causative in Greenlandic.

The Aleut language, currently spoken along the Aleutian chain and the Pribilof and Commander islands, is the only language in its branch of the Eskimo-Aleut language family, and traditional methods of linguistic reconstruction have neither satisfactorily explained its relationship with languages on the Asian continent nor its development from Proto-Eskimo-Aleut. Linguistic reconstruction has always been important in understanding the prehistory and history of the Aleuts, and new approaches in comparative linguistics, more comprehensive information on typological features of neighboring languages, and continuing language documentation allow us to propose a rich and continuous history of contact with various groups of people. I evaluate evidence that the Aleut language may have been shaped by contact with neighbors in Asia and Alaska, eventually giving rise to its differentiation from the Eskimo languages. I look at dialect differentiation along the Aleutian chain and what this differentiation reveals about the migration trends of the Aleut along the chain. I look at the colonial expansion of the Aleut-speaking area and resulting additional varieties of Aleut in the historical period. Finally, I review the effects of the Russian and American colonial periods on the Aleut language and the severe endangerment that the language faces today as a result. I conclude that there is evidence of possible Aleut contact with both neighboring peoples; however, much of this evidence has not yet been subjected to systematic comparative reconstructions. Linguistic evidence supports theories of at least two westward expansions of Aleuts along the island chain, but it is not yet clear what motivated the dialect differentiations. Finally, I offer some thoughts on directions for future dialect studies and the continuing documentation of Aleut.