Adelaar Willem, Muysken Pieter. The languages of the Andes
Подождите немного. Документ загружается.
294 3 The Inca Sphere
meanings. The availability of several verbalisation devices, which interact with nom-
inalisation processes and which sometimes can be used recursively, further stimulates
the formation of idiomatic expressions.
Even to a greater extent than Quechua, Aymara is exceptionally rich in lexical terms
referring to ways of carrying (cf. Tate 1981). The overall term for ‘to carry’ is apa-, as in
Quechua, but a number of other terms, exclusively Aymara in their majority, occur as
well.
(190) aˇcu- ‘to carry with the teeth’
asa- ‘to carry open recipients’ (plates, vessels)
aya- ‘to carry something long’ (sticks, pencils)
haˇc’i- ‘to carry a handful’ (grain, rice)
harp
h
i-, t’imp
h
i- ‘to carry in apron or skirt’
iˇcu- ‘to carry in the arms’ (children, small animals)
iqa- ‘to carry something flexible’ (cloth, rope)
ira- ‘to carry objects that fit in the hand’ (bread, fruit, money)
itu- ‘to carry something big with both hands’ (stone, cooking-pot)
kal
y
a- ‘to carry a large object, alone or with others’ (trees, tables)
p
h
ux
.
tu- ‘to carry with two hands’, ‘to carry a double handful’
(also in Quechua)
q’ipi- ‘to carry on the back’ (also in Quechua)
wayu- ‘to carry with a handle’ (baskets, buckets, suitcases)
Kinship terminology in present-day Aymara is rather similar to European kinship
terminology (cf. Pyle 1981). It distinguishes gender of referent, rather than gender of ego.
Unlike the case of Quechua, the terms for ‘son’, yuqa, and ‘daughter’, p
h
uˇca, are kept
apart. In the terms for ‘brother’ and ‘sister’, respectively, hila(ta) and kul
y
aka, gender
of ego is no longer indicated. However, Bertonio’s dictionary (1612a) gives evidence
of a more complex system, in which not only gender of ego, but also relative age play
a role. For instance, alu (Bertonio alo)is‘younger brother of woman’, ˇcinki (Bertonio
chinqui)is‘younger sister’, whereas kul
y
aka (Bertonio collaca)istranslated as ‘elder
sister’. ‘Elder brother’ and ‘younger brother’ are listed as hila and sul
y
ka (Bertonio
sullca), respectively. Affinal kinship terms are well elaborated in present-day Aymara:
tul
y
qa ‘son-in-law’, yux
.
ˇc’a ‘daughter-in-law’, lari ‘relative of wife’. Parents-in-law are
indicated by means of a vowel-suppressing suffix -ˇc’i as in awk-ˇc’i ‘father-in-law’ (from
awki ‘father’) and tayk-ˇc’i ‘mother-in-law’ (from tayka ‘mother’).
The Aymara numeral system is decimal and parallel to that of Quechua in the way
complex numerals are formed; see above (150) and (151). From an etymological point
of view, however, the Aymara facts seem to betray a less elaborate system. The words
maya ‘one’, paya ‘two’ and pusi ‘four’ are exclusive to the Aymaran languages. A term
3.3 The Aymaran language family 295
for ‘two human beings’ or ‘both’ is pa(:)ni. The words for ‘seven’ paqal
y
qu, and ‘eight’
kimsaqal
y
qu, contain the elements for ‘two’ pa(ya) and ‘three’ kimsa, respectively, which
leaves the element qal
y
qu as the possible remnant of an old word for ‘five’. The words
for ‘three’ kimsa, ‘five’ p
h
isqa, ‘six’ sux
.
ta, ‘ten’ tunka, ‘hundred’ pataka and ‘thousand’
waranqa, are shared with Quechua. Although the direction of borrowing is not traceable
in all cases, the added vowel in the word for ‘hundred’ (Quechua paˇc
.
ak) suggests a
Quechua origin.
87
The word for ‘nine’ l
y
a(:)tunka can be interpreted as a reflex of
*l
y
al
y
a tunka ‘almost ten’; llalla tunca is the form listed in Bertonio (1612a).
The spatial deictic elements exhibit a fourfold distinction based on distance from the
speaker. The proximate term is aka ‘this’, the non-proximate term is uka ‘that’; for further
distance k
h
aya ∼ k
h
a: and k
h
uyu ∼ k
h
uri can be used, the latter indicating a greater
distance than the former. Substantive-like use of the demonstratives can be brought
about by consecutive verbalisation and nominalisation as in ak-i:ri ‘the one here’, ‘this
one here’. Local deictics are case-marked; e.g. aka-na ‘here’. Further combinations are
possible with the vowel-suppressing affixes -ha ‘size’ and -hama ‘like’ (e.g. ak
h
a ‘this
much’, ak
h
ama ‘like this’).
Temporal deixis consists of special elements that are obligatory in temporal expres-
sions, such as hiˇc
h
a ‘now’, which appears in hiˇc
h
u:ru ‘today’, hiˇc
h
ayp’u ‘tonight’ (uru
‘day’, hayp’u ‘evening’), and q
h
ara ‘tomorrow’, which is more often found as q
h
aru:ru.
Some of these elements do not occur alone; cf. masu:ru ‘yesterday’ and walu:ru ‘day
before yesterday’.
Of the interrogative roots kama and kuna (meaning ‘what?’), the former is restricted
to derivations such as kam-sa- ‘to say what?’, kama-ˇca- ‘to do/happen what?’ and kamisa
‘how?’, whereas the latter occurs alone or in temporal expressions, such as kuna paˇca
‘when?’. Further interrogatives are k
h
iti ‘who?’, kawki ‘what place?’ and q
(h)
awq
h
a ‘how
much?’, ‘how many?’. Several forms for ‘which?’ are derived from kawki by consecutive
verbalisation and nominalisation: kawki(:)ri, kawkni(:)ri. The latter form appears to be
derived from kawki-na ‘where?’ rather than from kawki itself (Briggs 1993: 93).
Aymara has root names for the colours: ‘black’ ˇc’iyara,‘white’ hanq’u,‘grey’
ˇc’ik
(h)
u, ˇc’ix
.
i, ‘red to brown’ ˇcupika, wila,
88
‘coffee brown’ ˇc’ump
(h)
i,‘yellow’ q’il
y
u
(cf. Quechua), ‘light blue’ sahuna, ‘dark blue’ larama, and ‘green’ ˇc’ux
.
n
y
a.
As in Quechua, root reduplication plays a role in Aymara as well. Most frequent is the
reduplication of substantives in order to indicate a dispersed quantity, e.g. in qala=qala
87
The interpretation of the direction of borrowing is dictated by the fact that Aymara has no
nominal stems ending in a consonant, whereas conversely Quechua has many nominal stems
ending in a vowel. There is no reason why Quechua should lose such a vowel in the process of
borrowing.
88
The status of wila as an exclusive colour term cannot be upheld, because it also refers to ‘blood’.
In the related Jaqaru language ‘blood’ is the sole meaning of wila (Belleza Castro 1995: 196).
296 3 The Inca Sphere
‘stony place’ (from qala ‘stone’). The existence of reduplicated elements which together
constitute a single root was already observed by Bertonio (1612a), as appears from his
commentary on the gloss of the word huarahuara [warawara] ‘star’ (author’s translation):
‘And it is not a repeated noun, such as calacala [qalaqala], which means “stony place”
or “heap of stones”, but it is a single noun. Maya huarahuara, one star.’
3.3.8 Literary production in Aymara
In the colonial period Aymara played a more modest role than Quechua, and only very
few texts have been preserved. Some remnants of the pre-Columbian oral tradition in
the form of song texts are found in the chronicle of Guaman Poma de Ayala (1615); see
section 3.3.1. As a lengua general, the Aymara language formed part of the Doctrina
Christiana programme of 1584 (see section 3.2.2). Bertonio (1612b), assisted by his
Aymara consultant Don Mart´ın de Santacruz, published a Libro de la vida y milagros
de Nuestro Se˜nor Iesu Cristo en dos lenguas, aymara y romance (Book of the Life
and Miracles of Our Lord Jesus Christ in Two Languages, Aymara and Spanish). A
specimen of this text in a modernised version can be found in the anthology of Aymara
literature published by Alb´o and Layme (1992), which furthermore contains several
good examples of traditional narratives, myths, etc., along with political discourse and
historical accounts.
As in the case of Quechua, most of contemporary Aymara literature is traditional, and
consists of myths, rituals, narratives, autobiographical accounts, song texts and riddles.
Several texts were recorded in the 1940s in Chucuito by the anthropologist Tschopik
(1948). One of the texts he collected was the autobiographical account of an Aymara
woman, named Manuela Ari, which was eventually published by Briggs and Dedenbach
(1995). It illustrates social injustice and the hardships of life on the altiplano.Acollection
of stories entitled Jichha na: parlt’a (Now I Am Going to Tell), told by Elvira Espejo
Ayka (1994), an eleven-year-old girl from Qaqachaka (province of Avaroa, Oruro), was
published by Arnold and Yapita.
Finally, we should mention the production of the Proyecto Experimental de Educaci´on
Biling¨ue (Experimental Project for Bilingual Education), which operated in Puno during
the 1980s. This programme has generated several publications, among them, collections
of myths and folktale Wi ˜nay Pacha (Eternal Time) I and II (L´opez and Sayritupac 1985,
1990), and a compilation of song texts Qala Chuyma (Heart of Stone) (Paniagua Loza
1986).
3.3.9 Aymara sample text
The following sample text is an excerpt from a narrative relating the sufferings of
the altiplano Indians during the Chaco War (1932–5), between Bolivia and Paraguay. The
account, narrated by Esteban Yapu Mamani from La Paz, was first published in the
3.3 The Aymaran language family 297
trilingual magazine Jayma (Community Field) in La Paz (1984–6) and, subsequently,
reprinted in Alb´o and Layme (1992: 120–4). The original orthography has been adapted
to the transcription used in the present work.
1. wali ˇc’umi-raki-:n-wa.
{wali ˇc’umi-raki-:n(a)-wa}
very forested-AD-3S.NR-AF
‘It was all covered with bushes there.’
The original transcription of sentence 1 has a short vowel in the verb form. Since the
other instances of the near-remote past ending -:na do contain an indication of vowel
length, we treat this as an oversight and restore the long vowel mark.
2. uma-t-x
.
sint-pin wan
y
-kata-p-x
.
-ita-:n-x
.
a, uk
h
ama-rak manq’a-t-x
.
awt-ha-p-x
.
-ita-:n-x
.
a.
{uma-t(a)-x
.
(a) sint(i)-pin(i) wan
y
(a)-kata-p-x
.
(a)-ita-:n(a)-x
.
a, uk(a)-
hama-rak(i) manq’a-t(a)-x
.
(a) awt(i)-ha-p-x
.
(a)-ita-:n(a)-x
.
a}
water-AB-TO strongly-EM dry-PT-MA-CM-3S.1O-NR-TO,
that-CP-AD food-AB-TO hunger-LS-PL-CM-3S.1O-NR-TO
‘The lack of water made us terribly thirsty, and, likewise, the lack of
food made us suffer from hunger.’
The expressions uma and manq’a followed by the ablative case ending -t(a) must
be translated with the expression ‘for lack of’ (cf. Quechua mikuy-manta ‘for lack of
food’). The verbs wan
y
-kata- ‘to become dry’ and awt-ha- ‘to be hungry’ are used with
an impersonal third-person subject. Both show idiomatic use of derivational affixes.
3. na-naka-x
.
ˇc’ax
.
wa-n
y
hak’a-n-ka-p-x
.
a-ya-t-wa, uka-t-x
.
q
h
ux
.
uq-ir-x
.
is.t’-x
.
a-p-x
.
a-rak-t-wa.
{na-naka-x
.
(a) ˇc’ax
.
wa-n
y
(a) hak’a-n(a)-ka-p-x
.
a-ya-t(a)-wa, uka-t(a)-x
.
(a)
q
h
ux
.
uq(i)-ir(i)-x
.
(a) is(a).t’(a)-x
.
a-p-x
.
a-rak(i)-t(a)-wa}
I-PL-TO fight-IF nearby-L-LV-PL-CM-NR-1S-AF that-AB-TO
thunder-AG-Z-TO hear-CM-PL-CM-AD-1S-AF
‘We were close to where the fighting was, so we could even hear the
explosions.’
The near-remote suffix -ya:- does not contain a vowel-length marker in hak’a-n-ka-
p-x
.
a-ya-t-wa. This may be a consequence of the foreshortening effect produced by the
first-person subject marker -t(a).However, most sources do write vowel length in such
cases (see, in particular, Yapita 1991). The verb is.t’a- ‘to hear’ is derived from a root
*isa-, which is not found without a derivational affix.
298 3 The Inca Sphere
4. uka-t nayra-qat ˇc’ax
.
w-ir sultaru-w un
y
-s-ta-ni-p-x
.
-itu, q’ala
t
h
ant
h
a-w puri-ni-:na.
{uka-t(a) nayra-qat(a) ˇc’ax
.
w(a)-ir(i) sultaru-w(a) un
y
(a)-s(i)-ta-ni-p
-x
.
(a)-itu, q’ala t
h
ant
h
a- w(a) puri-ni-:na}
that-AB first-LS fight-AG soldier-AF see-RF-UW-H-P-CM-3S.1O,
totally ragged-AF arrive-H-3S.NR
‘Then, soldiers who had been fighting earlier appeared before us, they
arrived totally in rags.’
The suffix -qata is restricted to the word nayra-qata ‘first’, ‘before’ exemplified here
(Briggs 1988: 200). The meaning of nayra is ‘eye’, ‘first’ or ‘ancient’. One could consider
the possibility that -ta in -qata was originally the ablative case marker (‘from’, ‘after’).
The verb un
y
-s-ta- ‘to appear before someone’ is here used transitively. It contains the
reflexive suffix -si-, which preserves a preceding vowel except in this particular case.
There is an alternative form un
y
a-si- without vowel suppression, which has the same
meaning.
5. na-naka-r un
y
-ha-sa-x
.
haˇca-raki-:n-wa.
{na-naka-r(u) un
y
(a)-ha-sa-x
.
(a) haˇca-raki-:n(a)-wa}
I-PL-AL see-LS-SU-TO cry-AD-3S.NR-AF
‘They even cried when they saw us.’
Note that the allative case, rather than the zero marker, is used with human
complements.
6. uka-t-x
.
sa-p-x
.
a-rak-itu-wa, kuna-ti pas-ka-:n uk-x
.
a.
{uka-t(a)-x
.
(a) sa-p-x
.
a-rak(i)-itu-wa, kuna-ti pas(a)-ka-:n(a) uk(a)-x
.
a}
that-AB-TO say-PL-CM-AD-3S.1O-AF what-IR happen-AN-3S.1O
that-Z-TO
‘Then, they told us more about what was happening over there.’
The word kuna-ti ‘what + interrogative suffix’ is used here as a connective element
introducing a complement clause. The clause is closed by uka,which summarises the
complement clause and identifies it as the direct object of the main clause (by zero mark-
ing). The verb pasa- ‘to happen’ is from Spanish pasar. Note the use of the incompletive
marker -ka-, which indicates an anticipated event.
7. t’aq
h
i-si-ta-naka-p-s awis-t’a-p-x
.
a-rak-ita-:n-wa.
{t’aq
h
i-si-ta-naka-p(a)-s(a) awis(a)-t’a-p-x
.
a-raki-ta-:n(a)-wa}
suffering-RF-SN-PL-3P-Z-AD inform-M-PL-CM-AD-3S.1O-3S.NR-AF
‘And they also informed us about how they had suffered.’
The verb awisa- ‘to inform’ is from Spanish avisar. The element -t’a- adds politeness.
When used as a verb, t’aq
h
i-si- ‘to suffer’ obligatorily contains the reflexive suffix -si-.
3.3 The Aymaran language family 299
In sentences 4 to 7 we have followed the source edition, which speaks of the arrival of
several soldiers from the front. However, the use of the plural markers in this passage does
not preclude the alternative interpretation that only one soldier arrived. Plural markers
can apply to the object (if not third-person), as well as to the subject, and the text excerpt
presented here does not contain cases of verbal plural marking in which the object (‘us’)
is not encoded (cf. England 1988: 112).
8. ˇcaku-n-x
.
wali uma-t hiwa-n
y
a-:n-wa.
{ˇcaku-n(a)-x
.
(a) wali uma-t(a) hiwa-n
y
a-(:-)-:n(a)-wa}
Chaco-L-TO a.lot water-AB die-IF-CV-3S.NR-AF
‘In the Chaco one had to die for lack of water.’
Sentence 8 contains a reverbalised infinitive, which indicates obligation. The required
vowel length is not discernible, because the following near-remote marker -:na also
begins with a long vowel. See also sentence 10.
9. ˇc
h
ux
.
-sa hani-w ina-k ˇc
h
ux
.
u-rpaya-p-ka-:n-ti, ma:.ki-w
uma-nt-x
.
a-p-x
.
a-:n-x
.
a.
{ˇc
h
ux
.
(u)-sa hani-w(a) ina-k(i) ˇc
h
ux
.
u-rpaya-p-ka-:n(a)-ti, ma(y)a.ki-w(a)
uma-nt(a)-x
.
a-p-x
.
a-:n(a)-x
.
a}
urine-Z-AD not-AF in.vain-DL urinate-LB-PL-PR-AN-3S.NR-NE
at.once-AF drink-IW-CM-PL-CM-3S.NR-TO
‘Even urine they would not throw away just like that, they would drink
it on the spot.’
The suffix -rpaya- has been described as a marker of ‘multiple action’, ‘reversed
action’, as well as an intensifier (England 1988: 107; Briggs 1993: 168). In sentence 9
it seems to coincide with the meaning of Quechua -rpari-inverbs of abandonment (‘to
dispose of and leave behind’). The common expression ma:ki ‘at once’ is derivable from
ma(y)a ‘one’ and the delimitative suffix -ki ‘just’. Note the double occurrence of the
completive marker -x
.
a in the verb uma-nt-x
.
a-p-x
.
a-:n(a). This is a common phenomenon
due to the fact that the second occurrence of the suffix is part of the plural marking,
whereas the first instance indicates completive aspect (‘eventually’, ‘in the end’). See
also the sentences 11, 13 and 15, and England (1988: 112).
10. uka-t-sti, pastu-s manq’a-n
y
a-ki-:-n-wa.
{uka-t(a)-sti, pastu-s(a) manq’a-n
y
a-ki-(:-):n(a)-wa}
that-AB-CT grass-AD eat-IF-DL-CV-3S.NR-AF
‘But then, it was also necessary to eat grass.’
11. kuriya-naka, t
h
ant
h
a sapatu-nak-s manq’-x
.
a-p-x
.
a-ki-ya-t-wa.
{kuriya-naka, t
h
ant
h
a sapatu-nak(a)-s(a) manq’(a)-x
.
a-p-x
.
a-ki-ya-t(a)-wa}
belt-PL ragged shoe-PL-Z-AD eat-CM-PL-CM-DL-NR-1S-AF
‘We ended up eating belts and old shoes.’
300 3 The Inca Sphere
The word kuriya-naka lacks both zero complement marking and the additive suffix
-sa,which would mark its coordination with t
h
ant
h
a sapatu-nak-s.Apossible association
with pastu-s in sentence 10 would also require the presence of the additive suffix. Here
we interpret the forms kuriya-naka and t
h
ant
h
a sapatu-nak-s as coordinated objects on
the basis of the Spanish translation accompanying the text.
In the original text, a long vowel is written in the element -ki-oftheverb manq’-x
.
a-
p-x
.
a-ki-ya-t-wa. Since there seems to be no motivation for a long vowel, we have left it
out. For the shortness of the vowel in -ya:- see the commentary on sentence 3.
12. um t
h
aqa-sa-x
.
luq
h
i-w tuk-x
.
a-p-x
.
a-:n-x
.
a.
{um(a) t
h
aq
(h)
a-sa-x
.
(a) luq
h
i-w(a) tuk(u)-x
.
a-p-x
.
a-:n(a)-x
.
a}
water-Z search-SU-TO mad-AF become-CM-PL-CM-3S.TO
‘Searching for water, people eventually went mad.’
13. um t
h
aq
h
a-sa-x
.
uraq p’iya-p-x
.
a-:n-x
.
a, hani-pini-w
uma-x
.
ut-h-iri-:-k-i-ti.
{um(a) t
h
aq
(h)
a-sa-x
.
(a) uraq(i) p’iya-p-x
.
a-:n(a)-x
.
a, hani-pini-w(a)
uma-x
.
(a) ut(a)-h(a)-iri-:-k(a)-i-ti}
water-Z search-SU-TO earth dig-PL-CM-3S.NR-TO, not-EM-AF
water-TO exist-LS-AG-CV-AN-3S-NE
‘Searching for water, they dug into the earth, but there was never any
water to be found.’
The reverbalised agentive form in the verb form ut-h-iri-:-k-i-ti indicates a habit-
ual past. A third-person-subject habitual past without reverbalisation is illustrated in
sentence 14.
14. uma-layk uraq p’iya-s ma.nta-sa-x
.
haqi-mp-paˇca-ki-w ˇc
h
aq
h
-x
.
-iri.
{uma-layk(u) uraq(i) p’iya-s(a) ma.nta-sa-x
.
(a) haqi-mp(i)-paˇca-ki-w(a)
ˇc
h
aq
h
(a)-x
.
(a)-iri}
water-cause earth dig-SU enter-SU-TO man-CO-IN-DL-AF
get.lost-CM-AG
‘Digging into the earth for the sake of water and entering the pits, it
even happened that men disappeared.’
The verb manta- ‘to enter’ contains a non-productive root ma- ‘to go’ (Jaqaru
maha-), which is obligatorily accompanied by a derivational suffix (here -nta- ‘in-
ward motion’). The common combination -mp(i)-paˇca-isdiscussed by Briggs (1988:
208). It indicates inclusion and consists of the coordinative marker -mpi (Quechua
-wan) and the inclusive marker -paˇca (Quechua -nti-), which indicates total inclusion.
3.3 The Aymaran language family 301
The suffix -paˇca also conveys the meanings ‘(it)self’ and ‘inferential knowledge’ (see
above).
15. ma: hila.ta-x
.
pitrulyu un
y
-ha-raki-tayna, uk uma-sa-s
hiw-x
.
a-p-x
.
a-raki-:n-wa.
{ma(y)a hila.ta-x
.
(a) pitrulyu un
y
(a)-ha-raki-tayna, uk(a) uma-sa-s(a)
hiw(a)-x
.
a-p-x
.
a-raki-:n(a)-wa}
one brother petrol see-LS-AD-3S.FR, that-Z drink-SU-AD
die-CM-PL-CM-AD-3S.NR-AF
‘They say one of our comrades even found petrol, and then he drank
it and died.’
The word hilata ‘brother’ may have been derived from hila- ‘to grow’; an alternative
term for brother is hila (originally ‘elder brother’). There is no ready explanation for the
absence of zero complement marking in petrulyu; possibly the fact that the (underlying)
vowels on both sides of the word boundary were identical may have caused confusion.
3.3.10 The Jaqaru language
89
As we noted in passing, the sound system of the Jaqaru language is highly complex.
The vowel system is similar to that of Aymara and Central Peruvian Quechua. It com-
prises three vowels (a, i, u) with additional distinctive length. As in the neighbouring
languages, the high vowels are subject to obligatory lowering in a uvular environment.
Non-obligatory lowering is found in other environments. The status of length has been
a matter of debate. Although it has been general practice to classify the vowels as plain
(short) versus long, Hardman (1983a, b) proposes a distinction between plain (short) and
extra-short vowels. In this approach, the plain vowels correspond to the longer option in
pairs differentiated by length, whereas extra-short vowels (˘a, ˘ı, ˘u) represent the shorter
option in such pairs. Cerr´on-Palomino (1994b) criticises this analysis on empirical and
general linguistic grounds. He prefers to treat the shorter vowels as plain and the longer
ones as long in pairs differentiated by length. Even though Hardman’s approach may
seem unusual from a general linguistic point of view, it should be observed that if in a
minimal pair differentiated by length one of the elements is a lexical item shared with
Quechua, it often, though not always, has the longer vowel option. Compare in (191)
Quechua ˇcaki ‘(to) dry’, and in (192) Quechua kaka ‘maternal uncle’.
90
89
The main sources of the grammatical information given in this paragraph are Hardman (1966),
Hardman (1983a) and Belleza Castro (1995). It should be noted, however, that our analysis of
the Jaqaru data may differ from the one given in these sources.
90
Our own observation of Jaqaru speech indicates that the vowels can be pronounced long, even
when they are not part of a minimal pair differentiated by length (e.g. nar-ma [na:rma] ‘laugh!’,
ut-ma [o:tma] ‘your house’). The matter requires further investigation.
302 3 The Inca Sphere
Table 3.13 Jaqaru consonant inventory
Alveolar
Labial affricates Alveolar Alveopalatal Palatal Retroflex Velar Uvular
Plain p c t t
y
ˇcˇc
.
kq
Obstruents Glottalised p’ c’ t’ t
y
’ˇc’ ˇc
.
’k’q’
Aspirated p
h
c
h
t
h
t
yh
ˇc
h
ˇc
.
h
k
h
q
h
Fricatives s ˇsh
Nasals m n n
y
ŋ
Laterals l l
y
Vibrant r
Glides w y
(191) ˇc˘ak-k-i ˇcak-k-i (Hardman 1983a, b)
{ˇc˘ak(i)-k-i}{ˇcak(i)-k-i}
ˇcak-k-i ˇca:k-k-i
(Cerr´on-Palomino 1994b)
{ˇcak(i)-k-i}{ˇca:k(i)-k-i}
look.for-SM-3S dry-SM-3S
‘He is looking for it.’ ‘It is drying.’
(192) k˘aka kaka
(Hardman 1983a, b)
kaka ka:ka (Cerr´on-Palomino 1994b)
‘uncle’ ‘wing’
The Jaqaru consonant inventory (cf. also section 3.3.3) is shown in table 3.13.
91
Although all these consonants are attested, some exhibit a low frequency of occur-
rence. Of the alveolar affricates only the glottalised affricate is found in word-initial
position (e.g. c’irara ‘black’; Aymara: ˇc’iyara). The aspirated affricate ˇc
h
is limited to
word-internal position (in iˇc
h
u ‘straw’; Aymara: hiˇc
h
u). For the limited distribution of
the velar nasal see section 3.3.3. Initial liquids and r in Aymara cognates and Quechua
loans are usually matched by a nasal in Jaqaru; for instance, in n
y
aki ‘sadness’ (Quechua
l
y
aki), nuri ‘inside’ (Quechua I ruri) and nura- ‘to do’ (Aymara lura-, Quechua I rura-).
92
Vowel suppression is common in Jaqaru. Both morphophonemic and syntactically
based rules are represented (cf. sections 3.3.3 and 3.3.5). Interestingly, vowel suppression
91
Hardman (1983a) and Belleza Castro (1995) use different spellings for the alveolar affricate
series (respectively, tz and ts, etc.), for the alveopalatal series (tx and ty, etc.) and for the retroflex
series (cx and tr, etc.). Both authors use Peruvian Spanish spellings for the palatal series (ch,
etc.), as well as for the (alveo)palatal nasal, lateral and fricative (˜n, ll, sh) and the velar fricative
( j). The velar nasal is written nh. Aspiration is written “ by Hardman and h by Belleza Castro.
92
Surprisingly, Jaqaru also has a number of words with initial l and l
y
, such as l
y
uqal
y
a ‘boy’,
possibly a loan from Aymara or from an extinct Aymaran language.
3.3 The Aymaran language family 303
can also affect vowels inside a root or ending, in contrast to Aymara, where this is
exceptional. Furthermore, non-contiguous, as well as contiguous affixes can exert an
influence upon the shape of a morpheme (Hardman 1966: 32–9; 1983a: 55–72). Vowel
suppression in Jaqaru is dominated by a bewildering variety of ad hoc rules, which are
much more difficult to formulate in general terms than in the case of Aymara. In the
following example the word for ‘fox’ atuqu is subject to root-internal vowel suppression,
whereas kuntiri ‘condor’ has root-final vowel suppression (but compare kuntri-wˇsqa in
example (198) below).
(193) amur aruma atqu-qa sa-k-i kuntir-ha
{amur(u) aruma at(u)qu-qa sa-k-i
93
kuntir(i)-ha}
well night fox-TO say-SM-3S condor-AC
‘At midnight the fox says to the condor:...’
(Farf´an 1952: 80)
The way in which vowel suppression takes place, as well as its occurrence or non-
occurrence, can lead to differences in interpretation of the resulting forms. In (194) the
first-person possessive suffix and the first-person-subject future marker (both -
ŋ
a)donot
exert the same influence on the preceding root (Hardman 1966: 33; 1983a: 56). In (195)
the absence of vowel suppression before the possessive marker -
ŋ
a indicates object
function (zero complement) under specific conditions. The suffix -
ŋ
a itself must be
followed by a word-boundary, and the preceding substantive base can only be extended
by a limited number of affixes (Hardman 1966: 89; 1983a:148). For the possessive
markers see section 3.3.4.
(194) i
ŋŋac-ŋŋaiŋŋca-ŋŋa
{i
ŋac(a)-ŋa}{iŋ(a)ca-ŋa}
serf-1P hire.a.serf-1S.F
‘my serf’ ‘I will hire a serf.’
(195) tat-
ŋŋa tata-ŋŋa
{tat(a)-
ŋa}{tata-ŋa}
father-1P father.Z-1P
‘my father (subject)’ ‘my father (object)’
Some of the Jaqaru case markers are identical to those of Aymara, namely, -hama
‘comparative’, -kama ‘limitative’, -na ‘locative-genitive’, -ru ‘allative’, -taki ‘benefac-
tive’ and -t
h
a ‘ablative’. The accusative case can remain unmarked, or it can be indicated
by a functional absence of vowel suppression (195) or by a special (vowel suppressing)
marker -ha (193)(196). The latter can be reduced to an aspiration and indicates object
93
The full form of the verb ‘to say’ is saha but the root in inflected forms is sa-.