Aleut language

The consonant phonemes of the various Aleut dialects are represented below. Each cell indicates the International Phonetic Alphabet (IPA) representation of the phoneme; consonants existing only in loanwords are in parentheses. Note that some phonemes are unique to specific dialects of Aleut.

The palato-alveolar affricate /tʃ/ and uvular stop /q/ are pronounced with strong aspiration.

Attuan labial fricative /v/ is pronounced voiced or devoiced.[3]

Voiceless approximants and devoiced nasals are preaspirated. The preaspiration of approximants causes very little friction and may pronounced more as a breathy voice. The preaspiration of devoiced nasals starts with a voiceless airstream through the nose and may end voiced before a vowel. The preaspiration feature is represented orthographically with an ⟨h⟩ preceding the given sound. For example, a devoiced, preaspirated labial nasal would be written ⟨hm⟩.[4]

Voiced approximants and nasals may be partly devoiced in contact with a voiceless consonant and at the end of a word.

The voiceless glottal approximant /h/ functions as an initial aspiration before a vowel. In Atkan and Attuan, the prevocalic /h/ aspiration contrasts with an audible but not written glottal stop initiation of the vowel. Compare halal ('to turn the head') and alal ('to need'). This contrast has been lost in Eastern Aleut.[3]

Modern Eastern Aleut has a much simpler consonant inventory because the voice contrasts among nasals, sibilants and approximants have been lost.[25]

Consonants listed in the dental column have varied places of articulation. The stop, nasal, and lateral dentals commonly have a laminal articulation. The voiced dental fricative is pronounced interdentally.[25]

The pronunciation of the sibilant /s/ varies from an alveolar articulation to a retracted articulation like a palato-alveolar consonant. There is no contrast between /s/ and ʃ in Aleut. Many Aleut speakers experience difficulties with this distinction while learning English.[25]

Aleut has a basic three-vowel system including the high front /i/, low /ɑ/, and high back /u/. Aleut vowels contrast with their long counterparts /iː/, /ɑː/, and /uː/.

Notably, Aleut /u/ is pronounced slightly lower than /i/ in the vowel space.

The long vowel /ɑː/ is pronounced retracted in the vowel space creating a significant distinction relative to the vowel length of /ɑ/. The two high vowels are pronounced with the same vowel quality regardless of vowel length.[25]

In contact with a uvular, /i/ is lowered to [e] and /u/ is lowered to [o]. In contact with a coronal, /ɑ/ is raised to [e] or [ɛ], and /u/ is fronted to [ʉ].[4][25]

The mid-vowels [ɔ] and [ɛ] occur only in family names like Nevzorof and very recently introduced Russian loanwords.

(C)(C)V(V) ± {C(C)(C)V(V)} ± C

An Aleut word may contain one to about a dozen syllables, all syllables with a vocalic nucleus. In Atkan and Attuan, there is a word-final CC due to apocopation. There also exist word initial CCC in loanwords.[3]

A word may begin or end in a vowel, both short and long, with few exceptions. Due to apocopation, short /u/ is not found in the final position. The same is true for short /i/, except in some obsolete suffixes, such as -chi 'your' (pl.) which is realized as -chin and -chix in modern Eastern and Atkan Aleut.

Vowels within a word are separated by at least one consonant. All single consonants can appear in an intervocalic position, with the following exceptions:

• /ʍ/ and /h/ do not occur in intervocalic positions
• /w/ does not occur in contact with /u/
• /ç/ does not occur in contact with /i/

Words begin with any consonant except /θ/ and preaspirated consonants (with the exception of the preaspirated /ŋ̥/ in Atkan Aleut). Only in loanwords do /v/, /z/, and the borrowed consonants (p, b, f, d, g, ɹ/ɾ) appear word-initially.

The word-initial CC can take many forms, with various restrictions on the distribution of consonants:

• a stop or /s/, followed by a continuant other than /s/ or /z/
• a coronal stop or /s/, followed by a postlingual continuant (velar, uvular, or glottal).
• postlingual stop or /tʃ/, followed by /j/
• /k/ or /s/, followed by /n/

Intervocalic CC can occur in normal structure or as the result of syncopation.

In CC clusters of two voiced continuants, there is often a short transitional vowel. For example, qilĝix̂ 'umbilical cord' is pronounced [-lⁱĝ-] similar to qiliĝi-n 'brain'.

Almost all possible combinations of coronal and postlingual consonants are attested.

The combination of two postlingual or two coronal consonants is rare, but attested, such as hux̂xix 'rain pants', aliĝngix̂ 'wolf', asliming 'fit for me', iistalix 'to say; to tell; to call'.

In CCC clusters, the middle consonant is either /t/, /tʃ/, /s/. For example, taxtxix̂ 'pulse', huxsx̂ilix 'to wrap up', chamchxix̂ 'short fishline'.

The most common single consonants to appear word-final are /x/, /χ/, /m/, /n/, /ŋ/, and /j/.

Through apocopation, word-final single consonants /l/ and /s/ occur, and word-final consonant clusters ending with /t/ or /s/.

The details of the extensive syncopation characteristic of the Eastern Aleut dialect are described below. In the examples, the syncopated vowel will be shown in parentheses.

A word medial short vowel may be syncopated between single consonants, except after an initial open short syllable and/or before a final open short syllable. For example, ìx̂am(a)nákux̂ 'it is good' and alqut(a)maan 'what for?'.

The syncopation often creates consonant clusters beyond those prescribed by the general rules of Aleut phonotactics. The resulting clusters include:

• clusters of two stops: asx̂at(i)kuu 'he killed it'
• geminate consonants: yuug(i)gaadakux̂ 'is young'
• a regular three consonant cluster: ingam(a)sxakum 'after a short while (he)'

In some frequently encountered syncopated forms, which otherwise result in irregular three consonant clusters, the middle uvular fricative is deleted along with the preceding vowel. For example, (h)iis(ax̂)talix 'saying' and (h)iil(ax̂)talix 'being said'.

At slower speeds of speech, the syncopation may not be realized. Compare txin saakutikux̂txin 'you are getting skinny' beside ting saak(u)tikuqing 'I am getting skinny'.

[4]

Aleut stress is indeterminate and often difficult to define. Stress varies based on the relation to the beginning or end of the word form, the length of the vowels, the sonority of the consonants, open- or closeness of the syllables, or the number of syllables in sentential rhythm and intonation. Stress affects the length of both vowels and consonants. Stress underlies the distinctive syncopation characteristics of Eastern Aleut. In the following discussion, the acute accent (á) indicates the stronger stress and the grave accent (à) indicates the weaker.

In Eastern Aleut, stronger stress tends to fall on the penultimate syllable if it is short (has a short vowel), or on the last syllable if it is long (has a long vowel). The weaker stress commonly falls on the first syllable. For example, úlax̂ 'house', tùnúnax̂ 'talked', tùnulákan 'without talking', ìnaqáam 'he himself'.

Eastern Aleut words with more than two syllables exhibit a wider variety of stress patterns. Stress may be attracted to another syllable by a long vowel or relatively sonorant consonant, or by a closed syllable. It is possible the stress can be determined by rhythmic factors so that one word will have different stress in different contexts, such as áĝadax̂ compared to àĝádax̂, both meaning 'arrow'.

In Atkan and Attuan Aleut, stronger stress more commonly falls on the first syllable. However, long vowels, sonorants, etc. have similar effects on stress as in Eastern Aleut. For example, qánáang 'how many' vs qánang 'where'; ùĝálux̂ 'spear' vs álaĝux̂ 'sea'.

Stress may also be expressive, as with exclamations or polite requests. Stronger stress falls on the last syllable and is accompanied by a lengthening of a short vowel. For example, kúufyax̂ àqakúx̂! 'coffee is coming'. There is a similar structure for polite requests: qadá 'please eat!' vs qáda 'eat'.

Under ordinary strong stress, a short syllable tends to be lengthened, either by lengthening the vowel or geminating the following single consonant. Lengthening of the vowel is most common in Eastern, but is found in Atkan before a voiced consonant. In all dialects gemination is common between an initial stressed syllable with a short vowel and a following stressed syllable. For example, ìláan 'from him' pronounced [-ll-] and làkáayax̂ 'a boy' pronounced [-kk-].

[3][4]

The following descriptions involve phonological processes that occur in connected speech.

Word-final velar and uvular fricatives are voiceless when followed by a word-initial voiceless consonant and are voiced when followed by a word-initial voiced consonant or a vowel.

Word-final nasals /m/ and /n/ are frequently deleted before an initial consonant other than /h/. For example, tana(m) kugan 'on the ground' and ula(m) naga 'the interior of the house'.

In Eastern Aleut, the final vowel may be elided before or contracted with the initial vowel of the following word, as in aamg(ii) iĝanalix 'he is bleeding awfully'.

In Atkan, the final syllable of a word form may be clipped off in fast speech. This is even frequent at slow speed in certain constructions with auxiliary verbs. The result will be a sequence of vowels or full contraction:

• waaĝaaĝan aĝikux̂, waaĝaa-aĝikux̂, waaĝaaĝikux̂ 'he's about to come'
• ixchiinhan aĝikuq, ixchii-aĝikuq, ixchiiĝikuq 'I'll go home'

Ordinary nouns have suffixes for [26]

• number: singular, dual, and plural
• relational case: absolutive and relative
• person: first, second, anaphoric third, reflexive third

The anaphoric third person refers to a proceeding term, specified by being marked in the relative case or from context. For example, tayaĝu-m ula-a 'the man's house' and ula-a 'his house'. [26]

As free forms, the pronouns are used primarily as an object, like fully specified nouns. As enclitics they function as subject markers.

Positional nouns indicate positional, directional, or some more abstract relation to a definite referent (a person or proceeding noun in relative case). Positional nouns have possessive suffixes but no inherent number.

Unlike ordinary nouns, positional nouns have two adverbial cases: locative and/or ablative. The most important stem i-, called the dative, has only a locative form (largely irregular) meaning 'to, at, for-'.

Most are used in absolutive case, as in ula-m agal-a agikux̂ 'he passed behind the house'. They can also be used in the relative case, as in laavki-m agal-a-n ula-a 'the house behind the store'.[4]

The numeral system is decimal with hatix̂ 'ten' and sisax̂ 'hundred' as the basic higher terms. The higher tens numerals are derived through multiplication (e.g. 2 x10 for 'twenty'). The multiplicative numerals are derived with the suffix -di-m on the base numeral followed by hatix̂ 'ten'. For example, qankudim hatix̂ 'forty'.[3]

Verbs differ from nouns morphologically by having mood/tense suffixes. Like nominal stems, verb stems may end in a short vowel or consonant. Many stems ending in a consonant had auxiliary vowels which have largely become part of the stem itself.

Negation is sometimes suffixal, preceding or combining with the mood/tense suffix. In some cases the negation will be followed by the enclitic subject pronoun.

There are 570 derivational suffixes (postbases) including many composite ones, but about two thirds are found only in a small number of words. There are approximately 175 more common suffixes, considerably less than the Eskimo branch of the family.

A postbase may be nominal or verbal, yielding nouns derived from nouns or verbs, or verbs derived from verbs or nouns, or from nominal phrases. Many stems are ambivalent, being either nominal or verbal and even some derivatives can be ambivalent.

Difficulties distinguishing between nominal and verbal parts of speech arise because the parts of speech in Aleut are not easy to distinguish. A verbal stem may be used as a verbal predicate, and quite often as a noun. The verbal use of nouns is also very common.

The derivational suffixes may combine in strings of up to about six components, some belonging together to form composites. In sequences, each successive suffix often modifies the preceding string.

The majority of derivatives have a single stem that occurs also without the suffixes in question. While some stems are bound, only occurring with some derivational suffix. For example, compare iĝa-t- 'to scare, frighten' iĝa-x̂ta- 'to fear, be afraid of' iĝa-na- 'to be terrible, frightening'.

Most Aleut words can be classified as nouns or verbs. Notions which in English are expressed by means of adjectives and adverbs are generally expressed in Aleut using verbs or postbases (derivational suffixes).

Aleut's canonical word order is subject–object–verb (SOV).

Nouns are obligatorily marked for grammatical number (singular, dual, or plural) and for absolutive case or relative case (some researchers, notably Anna Berge, dispute both the characterization of this feature as "case" and the names absolutive and relative. This approach to Aleut nouns comes from Eskimo linguistics, but these terms can be misleading when applied to Aleut). The absolutive form is the default form, while the relative form communicates a relationship (such as possessive or contrastive) between the noun and another member of the sentence, possibly one that has been omitted. Absolutive and relative are identical in most combinations of person and number.

In possessive constructions, Aleut marks both possessor and possessum, with the possessor preceding the possessum:

The verbal predicate of a simple sentence of the final clause of a complex sentence carries the temporal and modal marking in relation to the speech act. Verbs of non-final clauses are marked in relation to the following clause. A complex sentence may contain an unlimited number of clauses.

Simple sentences may include a subject or no subject. The predicate may be a verb with no complement, a predicate noun with a copula, or a verb with a preceding direct object in the absolutive case and/or an oblique term or local complement.

The number of arguments may be increased or decreased by verbal derivative suffixes. Arguments of a clause may be explicitly specified or anaphoric.

The verb of a simple sentence or final clause may have a nominal subject in the absolutive case or a first-/second-person subject marker. If the nominal subject is left out, as known from context, the verb implies an anaphoric reference to the subject:

Positional nouns are a special, closed set of nouns which may take the locative or ablative noun cases; in these cases they behave essentially as postpositions. Morphosyntactically, positional noun phrases are almost identical to possessive phrases:

Verbs are inflected for mood and, if finite, for person and number. Person/number endings agree with the subject of the verb if all nominal participants of a sentence are overt:

If a 3rd person complement or subordinate part of it is omitted, as known from context, there is an anaphoric suffixal reference to it in the final verb and the nominal subject is in the relative case:

When more than one piece of information is omitted, the verb agrees with the element whose grammatical number is greatest. This can lead to ambiguity:

Although Aleut derives from the same parent language as the Eskimo languages, the two language groups (Aleut and Eskimo) have evolved in distinct ways, resulting in significant typological differences. Aleut inflectional morphology is greatly reduced from the system that must have been present in Proto-Eskimo–Aleut, and where the Eskimo languages mark a verb's arguments morphologically, Aleut relies more heavily on a fixed word order.

Unlike the Eskimo languages, Aleut is not an ergative-absolutive language. Subjects and objects in Aleut are not marked differently depending on the transitivity of the verb (i.e. whether the verb is transitive or intransitive); by default, both are marked with a so-called absolutive noun ending. However, if an understood complement (which may either be a complement of the verb or of some other element in the sentence) is absent, the verb takes an "anaphoric" marking and the subject noun takes a "relative" noun ending.

A typological feature shared by Aleut and Eskimo is polysynthetic derivational morphology, which can lead to some rather long words: