Aleut Grammar

16 — Aleut Grammar — Phonology

1. Phonology
1.1. Phonemes

1.1.1. Notation

The writing system used in this grammar, as also in Aleut Dictionary 1994,
is the practical orthography designed in 1972 for the bilingual program in the Aleut
region of what was then the Alaska State Operated School System, used in many
school books and other publications since 1973. It is a modified transliteration of
the Aleut Cyrillics developed by Ioann Veniaminov, for which see Aleut Dictionary,
General Introduction 2.7. It includes single roman letters, several digraphs and one
trigraph, to be explained in 1 . 1.2. below. A hyphen is used in certain phrases ( 1 .4. 1 . ;
2.1.9.4.2.3.). In the syntax a hyphen is sometimes inserted to show the morpheme
border.

In the charts, items in parentheses are found only in Russian or English
loanwords, the one in italics only in Eastern Aleut and the underlined one only in
Atkan and in loanwords, while the Attuan consonants are tabulated separately.

Consonants

Labial

dental

Prelingual
alveolar

palatal

Postlingual
velar uvular

Stops
voiceless

(p)

t

ch

k

q

(voiced)
Fricatives

(b)

(d)

(g)

voiceless

(0

hd

s

X

X

voiced

(v)

d

z

g

g

Nasals

aspirated
voiced

hm
m

hn
n

hng
ng

Approximants
aspirated
voiced

hw
w

hi
1

(r)

hy

y

Attuan

Stops
Fricatives

(P)

V

t

V

ch

s

k

g/x

q

g/x

Nasals

m

n

ng

Approximants
Vowels, short and long
Rounded

1

(r)

y

Unrounded

front

back

High

(Mid)

Low

uuu

(ooo)

iii

(eee)
[aci]

aaa

Glottal

22 — Aleut Grammar — Phonology

or whispered (cf. 1.5.). For example, the u and a in the last syllable of the Atkan
variants Amilaayus and Amilaayas 'Americans' may both come close to an i (gen-
erally not found after y, see 1 .2. 1 .2. 1 .)* In Eastern, through syncopation ( 1 .2.2.), a u
may be left as a labialization of the preceding consonant, written as u in parenthe-
ses, e.g. asa£uta& 'good luck', asag(u)talix [x w t] 'to be lucky'.

The long vowels are mostly somewhat lower than the respective short ones
but in contact with uvular consonants less retracted, e.g. in uuquchiing 'blue fox',
qiiqi& 'storm-petrel', qaaqaan 'eat it!'. In contact with front consonants, as in
chaaya& 'musical instrument', yaagilix 'to move', the aa is generally closer to [a:]
than to [as:].

In Eastern, an aa may be fronted by a following i, also across a postlingual
consonant, e.g. qach£inaagiku£ [qachxina:gikux] 'tries to feed (her child)' . Through
syncopation the fronted vowel may become distinctive, e.g. qach$inaa§(i)kuu
[qachxin&xkuu] 'tries to feed it'. Likewise in Atkan influenced by Eastern, e.g.
alqu£ maamis [ma:mis] 'what are you going to do?' ; by apocopationaqadaam(ing)
[aqada:m] 'after I had been ...' vs. aqadaam [aqada:m] 'after he/she had been ...'. The
front low vowel is found also in recent English loanwords such as A daading 'my dad'.

The mid vowels o, oo (about as in English off) and e (about as in English
bed), ee, are used only in family names such as Nevzo(o)rof and in a very few recent
loanwords, e.g. A 1973 goornicha& beside guurnicha$ 'living room' (Russian
gornitsa).

1.2. Phonotactics

1.2.1. Syllabic structure

The basic syllabic structure of Aleut may be summarized as follows (C =
consonant, V = short vowel, VV = long vowel):

C

cc

+

IvvJ ±

cc
ccc

[U

1-n

In loanwords there is also an initial CCC. In Attuan the medial CCC is reduced to
CC. In Atkan and Attuan there is also, through apocopation, a final CC.

A word form may contain from one to about a dozen syllables, each with a
vocalic nucleus.

1.2.1.1. Word initial

1.2.1.1.1. Zero and single consonants

All the consonants except the Eastern hd and the aspirated hng are found in
word-initial position. Initial labial obstruents, voiced stops, r and z are found only in
loanwords.

There are restrictions in relation to the following vowel only for the
semivowels. The labial semivowels occur only before the unrounded low vowel,

30 — Aleut Grammar — Phonology

a£six 'putting, giving'; A kims, Au kivs, E kimsix 'descending'. The place name
En Qiigangams is obscure.

1.2.2. Syllabic sequences
1.2.2.1. Stress and quantity

A word form may contain from one to some dozen syllables, each with a
vocalic nucleus, short or long, in Atkan and Attuan without systematic restrictions.
In Eastern the sequences are restricted by the syncopation of a short vowel under
certain conditions (see 1.2.2.2.).

Stress is not phonemic. In so far as the initial syllable of polysyllabic words
has frequently more than weakest stress, stress has a certain delimiting function. But
it is largely determined by the relative sonority or weight of the syllables and/or by
rhythmic factors, such as the number and character of the syllables within the words
as well as in phrases.

In phrases and sentences the rhythmic factors may be decisive but they are
difficult to determine. For example, in an Atkan text recorded on tape in 1952 the
word ataqan 'one' in the phrase ataqan tayagui 'one man' is stressed in two ways
( A = strongest stress; ' = medium stress; v = weak stress; no pauses between the
words): ataqan tayagu£ ilingiin agulaqas 'one man of them was left'; ataqan
tayagui imax amgignaasal ... 'taking one man as sentinel for themselves' (Bergsland
1959:60 f., 4 (81), (114) = N.M.3:26, 37). There could have been a difference of
emphasis (cf. 1.6.1.) but this is not easy to weigh.

The most important factor of sonority is the vowel length. In disyllabic word
forms with a short final syllable the initial syllable, short or long, has the stronger
stress in all the dialects, e.g. ulafc 'house', aalu& 'wave'. In forms with a long vowel
in the second syllable the stronger stress shifts to the final syllable, while a preced-
ing syllable with a short vowel has a stronger stress in Atkan (1952) than in Eastern,
e.g. ulaa, A also ulaa 'his house'; A qanaang 'how many' vs. qanang 'where'.
With long vowels in both syllables the stress is rather even, e.g. A qaaqaan or
qaaqaan 'eat it' vs. qaqaan 'you ate it'.

A stressed open syllable with a phonemically short vowel tends to be length-
ened. Either the vowel is lengthened, more or less, in Atkan especially before voiced
consonants, in Eastern perhaps more generally, or the intervocalic consonant is length-
ened or geminated, e.g. A imis [f.mis] 'for you' vs. iimis [f:mis] 'choosing', but
qilai [qfl.ax,qflla£] 'morning', etc.; Eu 1909 hinaS [hf.nax] or [hi:nax] 'went out',
akum [a.kum] (A [ak.um, akkum]) 'but then he' (J 34:84, 113); Ea 1983uda£ [u:dax]
(fricative d) 'bay', suka^ [su.kax] 'spray-skirt', iqa£ [f.qax, f:qax] 'kayak'. Between
the short vowel of an initial syllable and a long vowel a phonemically single conso-
nant is lengthened in all the dialects, e.g. A qanaang [qannaang] 'how many'; Eu
1909 ilaan [llla:n] 'from him', lakaayaS [lakka:ya£] 'boy' (J 34:19, 61); Ea 1909
tanaan hulix [tanna:n hulix] 'reaching his settlement' (J 6:31); E 1870 catechism
ms. chi$&i[i]lugim rel. 'baptismal font'. In Eastern a stressed final long vowel is
shortened, e.g. En 1982 ichaa [fcca.] 'go out!'.

Phonology — Aleut Grammar — 31

The effect of the sonority of the consonants, that is, of sonorants vs.
obstruents, is seen best in forms with three or four short syllables, not affected by
the Eastern syncopation.

In Eastern the stronger stress tends to fall on the next to the last syllable if
the last one is short, but it may also be attracted by a sonorous consonant to the
preceding syllable, e.g. tununa^ 'talked'; taya§u£ but also (e.g. Eu 1909) taya§u£
'man' (J 34:10, 18, etc.); Ea 1983 alamax 'humpback whale'. In Atkan the stronger
stress tends to fall on the initial syllable of a trisyllabic form but it may also be
attracted by a sonorous consonant to the following syllable, e.g. taya§u£ 'man' but
tagayu£ beside ta§ayu& 'salt' ; alagu£ (E alagu£) 'sea' but u§alu$ 'spear' . In East-
ern too, however, the initial stress as such may be predominant, resulting in varia-
tions such as Eu 1909 Waangilafc beside Waangila& (a person's name, J 34:142,
153) andagada* besidea£ada£ 'arrow' (J 34:43, 44); Ea 1983 laxtaka* 'seal hide'.

Forms with four short syllables tend to have a trochaic rhythm but relatively
sonorous consonants may break it, e.g. Ea 1983 asanaqax 'daughter-in-law' beside
asanaqang 'my daughter-in-law'; Eu 1984 qagilaka& 'there are no fish in it',
inulakuS 'is eating fast'; Eb 1984 kachfgikuS beside kachigikux 'it is blowing real
hard'; A qalagada 'don't eat it' but sanalakan 'being insufficient'. The effect of
sonority is seen also in longer forms, e.g. A 1952 haqatalakaq 'I don't know',
haqatalakaging *I don't know it', haqamasulaka$ 'perhaps he does not know',
haqatamasulaka& 'perhaps he does not know', hingamatalagada 'don't be that
way'.

Long vowels attract stress also in longer forms, e.g. A 1952 Tanaangis
(name of bay) vs. tanangis 'their island'; taangaatul 'wants to drink' vs. taangatul
'has much water in it'; qalagaaqaan 'don't eat it'; qanaagikiiu 'he tried to eat it';
qanaagiiqaan 'try to eat it'; aaluusaaqaan (very evenly stressed) 'laugh at him';
Eu 1984 asag(u)taasaaqaan (the middle aa half long) 'be lucky with it = you're
welcome'; A 1952 ting adaluusanaa§iigutamasu$taku& 'perhaps he tried to fool
me again'. In some of these forms a short unstressed vowel would be syncopated in
Eastern.

1.2.2.2. Eastern syncopation

The syncopation of short vowels in Eastern is tied up with stress but also
has a grammatical aspect.

Given the phonological conditions to be specified below, the syncopation
may be optional, possibly depending on speed. In the texts recorded on cylinders by
Jochelson in 1909 the syncopations are regular while the texts written by his Aleut
helpers, as their own stories or at the storyteller's dictation, have mostly the full
forms. In 1860 Shayashnikov wrote the syncopated form hila[a]&salix 'confessing'
but ten years later corrected it to hila[a]§asalix (Mark 1.5.). Later examples are En
1950 ukudigaasaaqaan 'be happy with it = you're welcome', where the i could
have been syncopated; Eu 1984 spoken slowly txin saakutiku$txin 'you are get-
ting skinny' vs. faster ting saak(u)tikuqing 'I am getting skinny' (on this one see

32 — Aleut Grammar — Phonology

below). Or the syncopation may be obligatory, possibly an irretrievable historical
process, e.g. E 1791 (and A 1840-1950) ayaqudaax > E 1909- ayuqdaax, En 1982
ayuqdaagft 'sea otter spear' . Only the comparison with the old Eastern or the Atkan
form could show that the middle u, probably unstressed (A ayaqudaax), was syn-
copated, quite regularly, between the two short initial syllables and the final long
one and rounded the preceding short vowel; in Eastern the suffix -qudaax is known
only from this word.

The phonological conditions come about by suffixation. A suffix may have
an initial long vowel or a short vowel assimilated to the stem vowel (see 1 .3.) and so
introduce a long vowel that attracts the stress, e.g. (chagi&) as&atiku-& 'he killed
(the halibut)' - as&at(i)kuu (-ku-a) = as&atkuu 'he killed it'. Or the suffix may add
one or more syllables with short vowels and so change the stress conditions, e.g.
i£amana-£ '(something) good' - i&am(a)na-ku& = Eu 1987 i£amnaku£ 'is good'.
But the stem to which the suffix is added may or may not be syncopated. If to the stem
i&amana- one adds the disyllabic suffix -laka& rather than the monosyllabic -ku£,
the rules will turn an underlying formi$amanalaka& (Atkan i§amanalaka£) 'is not
good' into i$amanlaka&, but also i$amnalaka& is possible (Eu 1984). The synco-
pation in the latter variant might possibly be due to a stress on the vowel before the
lateral (cf. 1.2.2.1.). But the form is rather the negation of the already syncopated
i$amnaku$ 'is good', by substitution of -Iaka& for -ku&, the stem being actually
i$amna-, with a consonant cluster that blocks the syncopation of the following
vowel.

The syncopation is possible only after a syllable with a long vowel or after
an initial sequence of two syllables with short vowels. With this proviso a short
vowel in an open syllable is (or may be) syncopated before a syllable with a long
vowel or before a final sequence of two syllables with short vowels, e.g. E 1791
Aklig(a)yuux 'Poor harpoon' (Census 3.2.28, man of Unimak); En 1982
taang(a)cha(a) 'fetch water' (stem taangat-), taang(a)kuqing 'I am drinking';
a§at(i)kuu 'killed it';i£am(a)naku£ 'is good'; Ea 1909haqan(i)salix 'waiting (for
them) to come' (J 1:36); 1805 [h]ingam(a)taasada 'do so with it = wait a bit (Rus-
sian postoy!)', la§uud(a)kuqing *I beg = please'. Having conditioned the syncopa-
tion of the following short vowel a long vowel may be shortened, e.g. Ep 1983
(Amrikaanchi-tunuu ilkiingin agach) agu&ta(a)s(a)ku&txin 'you speak (Ameri-
can better than us)'.

The syncopation of a short vowel in an open syllable may yield a consonant
cluster different from the pattern described in 1 .2. 1 .2.2., as in taangkuqing, as&atkuu,
etc. (see further below). In a closed syllable a short vowel is syncopated only if the
resulting three-consonant cluster agrees with the general pattern, e.g. Ea 1952
ingam(a)sxakum 'after a short while he'.

As observed by Marsh (1956) the short vowel is syncopated before one
final syllable with a short vowel in En 1952 qa§aas(a)kung 'I thank for it', cf.
qa§aas(a)kuqing 'I thank (you)'; anaaq(i)sa£ 'counterpart (in a special kind of
mutual social relationship)'; likewise En 1982 sentence final maas(a)kan 'because ...';

Phonology — Aleut Grammar — 33

Ep 1983 agitaasaq(a)sing 'my neighbor', rather than agitaas(a)qasing (probably
also possible). Between the voiceless consonants in a final syllable a short vowel
would be voiceless, whispered, and so could disappear.

If there is/was a sequence of two or more syllables with short vowels be-
tween the syllables in question, the last one is regularly syncopated and the resulting
consonant cluster blocks the syncopation of the next preceding syllable, e.g. 1805
adalul(a)gada 'don't lie'; Ea 1952 uku£taqal(i)kuu 'got to see it' vs.
uku&taq(a)liku$ 'got to see (something)' ; Ea 1 974 kayugiq(a)dalix 'getting weak' ,
kayug(i)qad(a)kuqing 'I am getting weak'; Ea 1983 qaatuq(a)lil(a)ka§a 'she didn't
want to eat it', mataad(a)lil(a)gaaq(i)laa&txin 'don't you by any means do so'; En
1978 ugunuq(a)tud(a)kuqing 'I am very forgetful'; En 1983 qam angunaada-
k(u)chang(i)nulux 'some small (lit. not big) fish', ilan maas(a)qad(a)qaa 'where
it got lost'.

As mentioned above, the variant Eu 1987 i£amnalaka& 'is not good' beside
i$amanlaka& appears to be based on i£amnaku£ 'is good', by substitution of the
disyllabic suffix -laka& for the monosyllabic -ku£; likewise iiam(a)naku£txin 'you
are good', Eu 1984 ting saak(u)tikuqing 'I am getting skinny', probably based on
txin saak(u)tiku& 'he/she is getting skinny'. The form En 1983 (i&tanangin)
aqat(a)namad(a)nulux '(his sayings) are never understandable either' appears to
be based on aqat(a)nad(a)nulux 'are never understandable', the suffix -ma- 'ei-
ther' being inserted in the syncopated form; likewise En 1982 uyaasaq(a)Iidana&
'usually started bringing (something)' with insertion of the suffix -da- 'usually' into
the very common complex ending -q(a)lina$ 'started -', present -q(a)liku£ A gram-
matical relation of this sort may perhaps explain also the following forms, where the
syncopation appears to start one syllable farther from the end of the word than usual:
Eu 1909 husuii maas(a)qadasxalix, tayagug(i)qadasxalix 'all of it (his party) be-
ing annihilated, having no more men' (J 34: 18 1 f.), rather than (*) maasaq(a)dasxalix,
tayag(u)|iq(a)dasxalix

A syncopated rounded vowel may leave no trace, especially if the following
vowel is rounded, as in Eu 1984 qaat(u)kuqing 'I want to eat'. Or it may labialize
the preceding consonant, e.g. qaat(u)naqing [-t w n-] 'I wanted to eat' , saak(u)naku&
[-k w n-] 'is skinny', whence also saak(u)naku£txin 'you are skinny'; Ea 1983, Eu
1984 asaguta& 'good luck' - asa§(u)talix [-& w t-] 'being lucky' vs. asa&talix 'having
as name' ; Ea 1983 agal(u)§i&taku& [-l w g-] 'seems to have teeth' ; chaay(u)sixcha(a)
[-y w s-] 'drink lot of tea! '. A preceding or following velar fricative may be labialized
or pass to w, e.g. Eu 1984 angiig(u)saming [-x w s-], angiiwsaming 'quite close to
me'; Ea 1983 igadwikuX < igadugiku& 'snow is drifting'; unguchiywaadada
[-y u w-] < unguchiyugaadada 'sit down for a while'. Or the syncopated rounded
vowel may replace a preceding or following low vowel, e.g. Enk 1982 agiitusnaku£
'is easy to get along with' vs. agiit(a)sun(a)Iaka$ 'is not easy to get along with'; Ea
1983 ayagu&tikuqing 'I look for a wife' , but En 1982 ayaga$tiku& < ayagagutiku£
'he looks for a wife'; Ea 1974 igiim qagadsulkakan < qagadusalakakan 'being
displeased with'; Ea 1909 aygasqudalix < aygasuqadalix 'going more slowly' (J

34 — Aleut Grammar — Phonology

6:30). The syncopation of a conditions the vowel assimilation in Ea 1909 ayugluku$

< ayugalaku& 'was gone out with' (J 6:46).

A syncopated palatal vowel may leave no trace, as inas&at(i)kuu 'he killed
it'. Or it may palatalize a preceding velar fricative or a preceding long aa, e.g. En
1984 as£amaag(i)kuqing [-x'k-] 'I am lonely'; Eu 1984 amaykuqing [~ey x k-] be-
side amaxkuqing < amagikuqing 'I did (so and so) last night'; [a:xta:dalix] beside
aaxtaadalix < aagitaadalix 'keeping (his eye) barely open' ; qach$inaag(i)kuu
[-na:xk-] 'tried to feed it'. Or the palatal vowel may replace a preceding short a, e.g.
Ea 1909 masxilgaan < masxaligaan 'in what way, a way to' (J 6:41); En 1982
angignaqing < angaginaqing 'I lived'.

The syncopation creates many two-consonant clusters beyond the basic pat-
terns: two stops, two prelingual consonants, two postlingual consonants, etc., as in
the examples above. Among the secondary clusters are also double consonants, genu-
ine geminates, e.g. Ea 1983yuug(i)gaadaku£ 'is young'; Eu 19841akaay(a)yuuging
'my poor boy ' ; nuqal(i)Iaagna£txin '(he said that) he had reached (the kayak)' ; Ea
1983 aakul(a)lakan 'being smart'; with dissimilation of the two laterals Ea 1909
haqayaadlakan < haqayaalalakan 'being treated fearlessly' (J 6:34). A double
stop is shortened to a single one in the frequent type En 1982 taang(a)duu(ka)kuqing
T am going to drink' . Two syllables are eliminated in Ep 1983 (agalkiingin) aqadniin

< aqadananiin 'for those who come (after us)' . The syncopated sequence -iy(a)- is
adjusted as a long vowel in Eu 1984 agiilaaganang < agiyalaaganang 'I tried to
wake him up'.

A uvular fricative has disappeared together with the preceding short vowel
in the extremely frequent invariable stem 1 909- iistalix (borrowed in A, Au ashiistal)

< 1826-70 (A 1838-) hi-isa-£ta-lix 'to say, etc.'; passive 1909- iiltalix beside
iila&talix, 1804-34 hiilafctalix. Likewise En 1982 amustuku£, A a-masu-£ta-kux
'perhaps is ...'; Ea 1970 ulaal(a&)takugaan 'when (fine weather) dawned' (cf. J
40:11,73:26).

1.3. Internal sandhi

1.3.0. General remarks

A word form may consist of a stem (base) only or contain one or several
suffixes, possibly also an enclitic. By definition, a suffix cannot be separated by a pause
from its stem, except in cases of self-correction, marked by a rapid onset of the
corrected suffixal part, e.g. A 1952 suun'gidix -ngis suun'giisal 'taking for themselves
their (dix reflexive) - their (-ngis non-reflexive) things', correction tosuun'gi-ngis 'their
(the other ones') things'; ay&aasiin -Iigaan ila$ta£igii[m] atxa&s a£taku& '(said that)
he [b] had made ready for him [a] also his [a's] - for him [a] a boat', correction to
ayiaasi-liga-an 4 a boat for himself (-liga- 'a future'); hfikus ila&tas as&aasal -laasal
'dying of - in great number of that too', correction to as&a-la-asa-l 'dying in great
number (-la-) of (1959, text 4:281, 429, 4:57 = N.M.3:70, 90, 111). The correc-
tions, ending as demanded by the context, have an initial consonant like a stem (cf.
ngiin 'for them', lil 'letting out', lal 'gathering') and are cut at a morpheme border.

44 — Aleut Grammar — Phonology

fqag(an) nagan in his kayak' (J 6:44); En 1982 qiix (< qiigam) sinigingin in the
grass'; Eu 1909 hawa(kun) Kanaagin those Koniags', Kanaa& (< Kanaagim)
tiikuu the Koniag chief (J 34:179, 174); E 1764- Tachiqala < *Ti*chigim qala
'Wislow Point (former village)'; Eu 1984 ala£ w kugan, A alagu(m) kugan 'on the
ocean'; Ea 1983 alax w saa < alagu(m) saa 'ocean bird'; Ea 1983 angi$-qalaadaa <
angagim qalaadaa 'many people'; Ea 1909chaayik tugulg ...< chaayakin tugalix
'beating his drum' (J 2:15).

In Eastern the deletion of a syllable may take place also in other cases, e.g.
Ea 1909 ha$s(ix) tunugnaan ... '(said that) it got up and said', am(an) qalngaa&
'that raven' (J 1:23,71).

1.4.3. Vowel lengthening in phrases

In ordinals formed with the transitive noun hiisi-$ (2. 1 .5.6.) the short vowel
of the initial syllable of the numerals A alax (E aalax) 'two' , A qankus (E qaankun)
'three' , and A siching, E sichin 'four' is lengthened as in a single word ( 1 .3 .0- 1 .): E
1834-, A 1840- aalax hiisft 'the second'; E 1870- qaankun hiisiS, A 1840-qaankus
hiisi* 'the third' ; E 1 834- siichin hiisi* , A 1 952- siiching hiisi* 'the fourth' . These
phrases appear to have originated in Eastern, where also the long vowel of aalax and
qaankun is generalized, while old Atkan had a different formation of ordinals (see
2.1.5.1.).

1.5. Contours

By a contour is understood a stretch without an intonational break. In the
1990 edition of Jochelson's cylinder texts transcribed by Moses Dirks (J 1-2, 5-6,
34, 76-79) and in the 1959 edition of the Atkan andAttuan texts recorded on tape in
1 952, the contours range from a single word of one syllable to a sequence of a dozen
words of more than thirty syllables (in both collections the contours are numbered
separately, in the Jochelson collection by lines). There may be internal pauses, some-
times due to hesitation or change of mind, but even at slow speed, with pauses be-
tween most words, the intonational unity of a contour is observable (e.g. 1959:74
(1-42)).

In these narrative texts the contours are in general rising-falling and appear
to end in three different ways: (a) by a slightly falling intonation, an intonational
break before the next contour, marked by the number of the latter; (b) by a moder-
ately falling intonation, in Atkan with unvoicing (whispering) of one or a couple of
the final syllables, marked by a comma; (c) by a pronounced falling intonation, in
Atkan a steep descent and/or, in the case of longer contours, unvoicing of several of
the last syllables; in Eastern the unvoicing is less extensive, comprising only one or
two syllables (Marsh 1956). The descent from the top, marked by relatively strong
stress ( A ), is of course short in a short contour like A 1952 (said by a child) TayaguS
haqaku&. 'The man is coming.' In the longer contours, especially in Atkan, the
descent may comprise a fairly large number of syllables, ending in a series of whis-
pered (sometimes barely audible) syllables.

Phonology — Aleut Grammar — 45

The pronounced falling intonation (c) clearly marks the end of a sentence,
simple or complex. The moderately falling intonation (b) may mark the end of a
clause but is used also to amplify or make an expression more accurate, e.g. A 1952
... Qawalangis udaagaxtal, udan agiidal, Niigugis agiidal, ...'... the Fox Islanders
coming here, visiting here, visiting the Andreanof Islanders, ...' (1959:58 (7-9) =
N.M. 3:5). The syntactic function of the slightly falling intonation (a) is less distinc-
tive. In general, the contours are important for the understanding of complex sen-
tences but do not always cover syntactic units.

A question may be marked by a final rising pitch (?), e.g. late A
Chaayuuna$? (fairly even stress) 'Is he going to drink tea?' vs. Chaayuuna£. (un-
stressed final syllable) 'He's going to drink tea.' The interrogative marker ii?, E hip
(2.1.8.5.) appears to have an inherent rising pitch and may come at the end of a
sentence as well as within a sentence, after the term that the question concerns,
while the following part ends like a declarative sentence, e.g. A Txin kixs ii? 'Did it
(really) bite you?'; Txin ii? kixs. 'Did it bite you?'. An interrogative pronoun or
adverb does not entail any specific intonation, e.g. A Kiin a&. 'Who is he?' like Ting
a£. 'It's me.' (talking about a picture). With the interrogative intonation a sentence
with an interrogative pronoun or adverb is a question about the question : Kiin a$
ii? 'Who he is? (is that what you ask about?).'

1.6. Expressive features

1.6.1. Emphatic stress

For emphasis, the initial syllable of a polysyllabic word form may have the
main stress, possibly with gemination of the following intervocalic consonant, also
in Eastern, e.g. En 1982 (spontaneous) kayyagnaku& 'it is heavy'.

In Atkan 1952, strong emphatic stress (marked " before the stressed syl-
lable), with lengthening of a syllable-initial consonant (if word-initial, with an
overshort anticipatory vowel), was used as an expression of the effort or feeling
experienced by the person(s) in the story, e.g. (from a story about a fight within a
cave) "kudukiim [ u k:udduki:m] ikin aqadusaqadaam, adum akangan hakang
"qidusal, 'having stretched both of his legs with him (on his back), crushing (jam-
ming) him against the ceiling of the cave,' ; "agalan hamang sanakuun sanal, 'do-
ing his utmost there after him (to catch up with him in his kayak)'; qidal [pause]
tu"taatuna§ulax [pause] agu&taqalikus tutalgaqaliqas a$takus. 'they were heard
crying and starting to make an a w f u 1 noise.' (1959, text 4:357-358, 344, 148 =
N.M. 3:76,92,41).

1.6.2. Final exclamatory lengthening

In Eastern, an exclamation is marked by a raised level tonality (marked !)
and lengthening (marked :) of the short vowel of the last syllable or overlengthening
of a long one, e.g. En 1952 chingliku:$! 'my, it's hot', awal(a)kaqi:ng! 'I am not
working !', uku&takuu:! 'he does see him!' (Marsh 1956); Ea 1983 kuufya$
aqaku:$! 'coffee is coming !' (said when coffee was brought in);aqataq(a)daku:ng!

46 — Aleut Grammar — Phonology

'I've forgotten it!' (lit. 'I no more know it'); slaachxi& aqa$tali:x! 'a storm seems to
be coming!'; chi£taku:$ ii! (high rising intonation) 'is it raining ?!'; "sa£ ulu:g
alga& ulu:g" iistalix 'saying: "it is not a duck! it is not an animal!'"; Ea 1909 qus(aa)
agnaa ginad(a)laka:n! 'passing above he (Raven) is not attractive!' (J 1:28); Eu
1909 wan agad(a)x ika:!, (rising-falling intonation) '(he started thinking,) the sun
up there, ...' (J 34:46).

The same gesture may be used in a polite request, e.g. A 1973 (by some
Atkans considered Eastern) qada: 'please eat!' vs. qada 'eat!' (possibly a rude com-
mand); tutada:! 'listen (I'll tell you something)'.

In Atkan (1973), the optative haqaaxt 'come now (to eat)' as an impatient
mother's shout is truncated to a vowel-final haqaa:!

Aleut Grammar
Unangam Tunuganaan Achixaasix

by

Knut Bergsland

© 1997 Alaska Native Language Center, University of Alaska Fairbanks;
and Knut Bergsland

Printed in the United States of America
All rights reserved

Library of Congress Cataloging-in-Publication data

Bergsland, Knut, 1914-

Aleut grammar = Unangam Tunuganaan Achixaasix / by Knut Bergsland.
p. cm. - (Alaska Native Language Center Research Papers ; no. 10)
Includes bibliographical references.
ISBN 1-55500-064-9 (alk. paper)

1. Aleut language—Grammar. I. Title. II. Series: Alaska Native Language
Center Research Papers ; no. 10.
PM32.B46 1997

497'. 195-dc21 97-38814

CIP

Alaska Native Language Center

Research Paper Number 10

University of Alaska Fairbanks

1997

First Printing 1997 400 copies

Address correspondence to

Alaska Native Language Center
University of Alaska Fairbanks
P.O. Box 757680
Fairbanks, AK 99775-7680