Adelaar Willem, Muysken Pieter. The languages of the Andes
Подождите немного. Документ загружается.
374 3 The Inca Sphere
(370) t
h
un
y
i qat-tan waxta t
h
on-ˇcam-ˇc
.
a
sun fall-PV.DS town come-PA.2.SG-DV
‘When the sun had fallen, you came to the village.’
(Cerr´on-Palomino, MS)
Apart from the genitive (see above), the Chipaya case system comprises an
instrumental–comitative -(ˇs
.
)tan, illative -(ˇs
.
)ki¸s,ablative (also in comparisons) -ki¸stan,
two locatives -ki¸s and -kin, perlative -nuˇs
.
and benefactive -xapa. Cerr´on-Palomino re-
ports that where there is a choice the shorter forms are used with feminine nouns.
The benefactive must be preceded by a genitive marker. (The same may hold for the
instrumental–comitative and illative endings.) There are some additional case markers
of Quechua origin.
The two locative case markers are distinguished as follows: -ki¸s is used when the
location referred to coincides with that of the speaker (371). If not, -kin is used (372).
(371) wer-ki urur-ki¸s qam-u-ˇc
.
a
I-TO Oruro-L.PX live-PN.1.SG-DV
‘I live here in Oruro.’
(Cerr´on-Palomino, personal communication)
(372) wer-ki urur-kin qam-u-ˇc
.
a
I-TO Oruro-L.DT live-PN.1SG-DV
‘I live in Oruro (not here).’
(Cerr´on-Palomino, personal communication)
The Uchumataqu case markers only partly resemble their Chipaya counterparts, e.g.
comitative -st´a(ni), instrumental -st´a ∼ -n´a,ablative -kistani, allative -k´ı(na) and perla-
tive -nis. Illative is -w´ınta ∼ -wint´ani and locative is -(kos)t´a (Vellard 1967: 27; Muysken
2000). Other locative endings are (with motion) -ki (kw´as-ki ‘into the water’) and -nik
(uhˇc-nik ‘into the fire’) and (without motion) -na/-nu (kw´as-na, kw´as-nu ‘in the water’,
Vellard 1967: 4–5). Some Uchumataqu examples are given below:
(373) w´ır-il ´o¸kwa-ˇca ki-st´ani
I-1 go-DV he-C
‘I go with him.’
(Vellard 1967: 28)
(374) w´ır-il w´at
ta-k´ına ´o¸kwa-ˇcay
I-1 village-AL go-DV
‘I go to the village.’
(Vellard 1967: 29)
(375) w´ır-il w´at
ta-kistani p´ıˇca-ˇcay
I-1 village-AB come-DV
‘I come from the village.’
(Vellard 1967: 29)
(376) w´at
-nis ´o¸kw-´a-ˇcay
village-PT go-F-DV
‘I shall go through the village.’
(Vellard 1967: 29)
3.7 The Atacame˜no language 375
The personal pronouns of Chipaya are wer ‘I’, wer-naka ‘we (exclusive)’, uˇc
.
um-
naka ‘we (inclusive)’, am ‘you (singular)’, am-ˇc
.
uk ‘you (plural)’, ni: ‘he’, ni:-naka
‘they (masculine)’, na: ‘she’, na:-naka ‘they (feminine)’; the Uchumataqu pronouns are
wir/wer ‘I’, uˇcum(i) ‘we (exclusive)’, am ‘you (singular)’, am-ˇcuka ‘you (plural)’, ni
‘he’, ‘she’, ni-naka ‘they’. For the inclusive first person Vellard (1951: 16) gives a form
w´aq
pa ˇc´umi, probably from waqpaˇca ‘all’ and uˇcumi ‘we’. The pluraliser suffix -naka
coincides with Aymara and may be a borrowed element.
The interrogative pronouns h´e
k
‘who’, ˇcul(˜u) ‘what’, ks´ım ‘where’ and kas´ut
‘when’
were recorded by Vellard (1967) for Uchumataqu; cf. Chipaya hek ‘who’ (Porterie-
Guti´errez 1990) and ˇc
.
h
ulu ‘what’ (Olson 1965: 31).
Vellard (1951, 1967) recorded a full decimal set of numerals for Uchumataqu: t¸i
‘one’, p´ıˇsk
i ‘two’, ˇc´e
.
p
i ‘three’, p´axk
up´ıki ‘four’, t´ax
sn´uko
.
‘five’, t´ax-t-n´uko
.
‘six’,
t
´
˜o
.
ko
.
‘seven’, k
´
˜o
.
ko
.
‘eight’, s´ar-n´uko
.
‘nine’ and kalu/kalo
.
‘ten’. Chipaya has preserved
only four of its original numerals: ti: ‘one’ (Olson 1965: c
h
i:), piˇska ‘two’, ˇc
h
ep ‘three’,
paqpik ‘four’; the remaining numbers are borrowed from Aymara. There is a special term
for ‘two people’: Uchumataqu p´ıkilt´a(ni) (Vellard 1967), Chipaya pukultan (Porterie-
Guti´errez 1990). Vellard, furthermore, mentions the existence of an ancient numeral
system recorded near Tiahuanaco, which in the 1940s was known to some of the Uru.
Some of its components bear a resemblance to Aymara, others to Uchumataqu.
Olson (1964: 313) reports that in his unpublished vocabulary of Chipaya 33 per
cent of the items are loans from Aymara, Quechua and Spanish. We may assume that a
similar figure holds for Uchumataqu. A high percentage of borrowings does not surprise,
considering the sociolinguistic situation. There seem to be cases, however, where Uru–
Chipaya must have been the giving language. Aymara speakers in Solajo near Carumas
(Moquegua, Peru) use a verb lul(u)- ‘to eat’, which is similar in form to the Uru–Chipaya
term (Chipaya lul-).
163
It has not been found in other Aymaran dialects. Furthermore,
it would be a mistake to limit the attention to the three languages mentioned above.
Uru–Chipaya also has some vocabulary in common with Puquina and with Moset´en,
e.g. Chipaya ¸si¸s-, Uchumataqu ˇsiˇs-, Puquina sisca ‘to know’; Uchumataqu t´ara, Moset´en
(Sakel 2003) t
y
˜ar˜a
ʔ
‘maize’. For more cases of Uru–Puquina contact see Torero (1992).
3.7 The Atacame˜no language
The people of the Atacama desert in northern Chile have attracted the attention of
travellers and scientists by their relatively isolated position and the impressive natural
setting that surrounds their picturesque old villages. The habitat of the Atacame˜nos is
situated in the Chilean region of Antofagasta, between the river Loa (east of Calama)
and the border with Argentina and Bolivia, which is marked by a series of 6,000 metre
163
Recorded by the author during fieldwork in Moquegua in 1984.
376 3 The Inca Sphere
high volcanoes. The main oasis San Pedro de Atacama is dominated by the Licancabur
(‘the mountain of the community’, by which San Pedro is understood). Most Atacame˜no
villages, such as C´amar, Peine, Socaire, S´oncor, Tilomonte, Tilopozo and Toconao, are
situated to the south and southeast of San Pedro, between the Atacama salt-lake (Salar
de Atacama) and the border. A second area is situated at a distance north of San Pedro
near the banks of the Loa and Salado rivers; it comprises the villages of Caspana and
Chiu Chiu.
The language of the Atacame˜no people is also known as Kunza,aword meaning ‘ours’
in Atacame˜no, or as Lican antai ‘language of the community’. Several sources report
that its area of influence, and possibly the language itself, once extended further east into
adjacent parts of Argentina (Philippi 1860) and the southwestern tip of Bolivia (Ibarra
Grasso 1958). A genetic relation of Atacame˜no with what was originally its eastern and
southern neighbour, the Diaguita or Kak´an language group, was long considered likely
(Schuller 1908; Mason 1950), but the absence of any reliable Diaguita data (cf. Nardi
1979) precludes the verification of this hypothesis.
164
Kunza received a substantial amount of attention in publications of the nineteenth
century. They give an interesting view of the protracted process of extinction that has af-
fected the language. Foreign scientists, Chilean state officials and local parish priests
165
collaborated in collecting word-lists and grammatical notes, which are useful but in-
sufficient to obtain a full picture of the language (Philippi 1860; von Tschudi 1866–9;
Moore 1878). The best sources date from the 1890s. They consist of a grammatical
sketch by the Chilean engineer Francisco San Rom´an (1890), based on data collected
after 1883, and a word list (glosario) with pronunciation notes, the product of a joint
effort of several interested scholars (Va¨ısse, Hoyos and Echeverr´ıa y Reyes 1896). A
drawback to the study of Atacame˜no is the absence of text. Several ritual songs that are
performed without being understood any longer and two versions of the Lord’s Prayer,
written down by the Swiss Americanist Johann Jakob von Tschudi, make up the entire
corpus (cf. Lehnert Santander 1976).
At the end of the eighteenth century the Atacame˜no language was still in use, although
the community itself was probably never greater than a few thousand. Since colonial
times many Atacame˜no people were experts in leading pack-animals across the icy and
desolate Puna de Atacama (cf. Bowman 1924). This activity took them to places far
164
Swadesh (1959: 18) saw a connection between Atacame˜no and two native languages of
Rondˆonia, Brazil: Kapishana (now known as Kanoˆe) and Mashub´ı. To date, Mashub´ıisclas-
sified as a Jabut´ıan language and considered unrelated to Kanoˆe. A rapid comparison of an
unpublished vocabulary of the Kanoˆe language (Bacelar 1996) with the Atacame˜no lexical ma-
terial has brought no evidence that would support a relationship between these two languages.
165
At the church of San Pedro a tablet proudly commemorates the efforts of those who studied the
Kunza language, in particular, the parish priests Benito Maglio and Emilio Va¨ısse.
3.7 The Atacame˜no language 377
away from their native oases, a circumstance which may have speeded up the demise of
the language. In 1858 Tschudi estimated the number of Atacame˜no speakers at less than
200. At the end of the nineteenth century, San Rom´an reported that the language was
known only by a few octogenarians and that it would soon be gone. Va¨ısse et al. (1896)
estimated the number of Kunza speakers at less than two dozen. It comes, therefore,
as a surprise that sixty years later Mostny was still able to locate (semi-)speakers in
the village of Peine (Mostny 1954). Mostny’s data, however limited, are relevant for
the understanding and interpretation of the nineteenth-century material. Among other
things, she published a local version (in Kunza) of the tal´atur,aritual song related to the
annual ceremony of cleaning the irrigation canals, and made a valuable effort to interpret
its contents. Later on, another version of the tal´atur, recorded in 1976 and originally
from the village of Socaire, also appeared in press (Rodr´ıguez 1991).
166
Atacame˜no is fairly well documented as far as its lexicon goes. All the available
lexical data have been brought together by a research group from the University of
Valpara´ıso in an (unpublished) computerised dictionary (S´aez Godoy et al. 1974). On
the other hand, the grammatical information is extremely limited and sketchy, because
San Rom´an only gives information on a few selected aspects of the grammar. The case
system, for instance, is not presented at all. As for the sounds of Atacame˜no, the sources
coincide in saying that it was a harsh-sounding language with many unpronounceable
sounds, which could not be represented easily with the symbols of the Latin alphabet. The
orthography which is used by the different sources is far from consistent. Nevertheless,
both San Rom´an and Va¨ısse et al.offer interesting observations on the pronunciation,
which make it possible to reconstruct several aspects of the Atacame˜no sound inventory.
In contrast to Aymara and Quechua, Atacame˜no had five distinct (short) vowels: a,
e, i, o, u.
167
Contrastive pairs for the front and back vowels, respectively, are sem(m)a
‘one’ vs. sim(m)a ‘man’, and potor ‘landslide’ vs. putchur ‘flower’. Cases of distinctive
vowel length are found in some environments at least, e.g. ckacka ‘forehead’, ckaacka
‘stutterer’. The hyphen in forms such as cki-itur ‘to fight’ and ma-istur ‘to find’ suggests
the presence of an intervocalic glottal stop. The spelling of the verb balt-hitur ‘to run’
in Va¨ısse et al. possibly indicates that the glottal stop could occur in other positions
as well.
Interestingly, few lexical items in the glossary of Va¨ısse et al.begin with a vowel;
a and i are the only vowels found in that position. Items that do have an initial vowel
166
The tal´atur and another Kunza ritual song, the cauz´ulor,have been brought out on music
albums: Le Chant du Monde: Chez les indiens du d´esert d’Atacama LDZ-S-4287 (no date);
and Unesco Collection Musical Sources: Amerindian Ceremonial Music from Chile 6586-026
(1975).
167
Mostny (1954) recorded a word containing schwa-type vowels (ˇc
ə
rəsnər ‘narrow opening in
the rocks that lets through water’).
378 3 The Inca Sphere
are mainly grammatical words (e.g. ackcka ‘I’; inti ‘that much’) and loan words (e.g.
atitur ‘to win’, ‘to be superior’, from Quechua ati-; astatur ‘to whip’, from Spanish
azotar). Most of the non-grammatical native words with initial a that are listed in the
Va¨ısse glossary have alternative forms beginning with an aspiration (h), e.g. ara ∼
hara ‘lodging place’, ‘temporary shelter’, atta ∼ hatta ∼ haata ‘yesterday’.
168
The
interrelation between aspiration and vowel length, as illustrated in the latter example, is
of particular interest. San Rom´an states explicitly that aspirations in Kunza are followed
by ‘a notable lengthening of the vowel’. In his conjugation of the verb ‘to eat’, San
Rom´an registers ohlmtur (with the aspiration written after the vowel), whereas Va¨ısse
et al.haveholmtur. This suggests that the aspiration may have been simultaneous with
the vowel, rather than previous or successive to it. Cases of postvocalic aspiration as
recorded by San Rom´an (e.g. in ohlmtur ‘to eat’ and pahni ‘child’) have been interpreted
as a phenomenon accessory to phonemic vowel length (Lehnert Santander 1987). On the
other hand, as we have just seen, there are also cases where aspiration triggers length,
rather than the opposite.
The consonant inventory of Atacame˜no, as represented in the glossary of Va¨ısse et
al., is surprisingly small. There is no doubt that glottalisation could be distinctive in the
labial and in the alveolar positions (e.g. poi.ya
169
[poyya] ‘calf of leg’, ppoya [p’oya]
‘two’; tilir ‘spicy’, ttelir [t’elir] ‘vicu˜na’). Nevertheless, many forms are listed with both
a glottalised and a plain initial, indicating that the distinctive value of glottalisation may
have been limited. For the palatal series the glossary employs only one symbol: tch.
The added explanation (Va¨ısse et al. 1896: 33) suggests that it referred to a glottalised,
rather than to a plain alveopalatal affricate comparable to Spanish ch.However, San
Rom´an makes a distinction between the symbols ch (as in choraca [ˇcoraka] ‘ostrich’)
and chch (as in chchoya [ˇc’oya] ‘seven’), thereby indicating a glottalisation distinction
in the palatal series as well.
Strangely, the Va¨ısse glossary contains no evidence of glottalisation in the velar or
postvelar series, nor does it distinguish between these two series at all. It employs the
symbol ck to cover all of the velar and postvelar areas of articulation (except for h).
The sound corresponding to ck is described as comparable to ch in German followed
byasound similar to r, thereby creating the suggestion of a postvelar fricative (Va¨ısse
et al. 1896: 15).
170
However, San Rom´an affirms that there were several k-like sounds
in Atacame˜no, ‘to be written c or k,inaccordance with their degree of strength’. He
168
There are a few non-grammatical items with initial i in the Va¨ısse glossary: ipnatur ‘to stick
to’ and ittin(tur) ‘(to put) straight’; they have no alternative forms with initial h.
169
From here on, the hyphen in the glossary is replaced by a dot, in order to avoid confusion with
morpheme boundary markers.
170
The symbol ck is used in the descriptive tradition of the Argentinian Quechua dialect of Santiago
del Estero for the representation of a uvular stop.
3.7 The Atacame˜no language 379
assigns a special status to a sound which he writes qc or q-c,inl´ıqcau ‘woman’ (Va¨ısse
et al.: lickau), and which is supposed to be pronounced with ‘a certain tenderness’. His
characterisation of the initial consonant in khˆuro ‘wind’ (ckuri ∼ ckuru´ı.ya in Va¨ısse
et al.) seems to refer to a strongly aspirated postvelar stop.
171
Other spellings used by
San Rom´an are c-k in v´ac-ka ‘river’ (back.cka in Va¨ısse et al.), cj in cjara
tur ‘to break’,
cc in ccara
tur ‘to cut’ (both ckaratur in Va¨ısse et al.), and qqu in qquepe ‘eye’ (ckepe or
ckepi in Va¨ısse et al.) The initial sound in this last word also drew the attention of von
Tschudi, who wrote igkjepe and described a deeply guttural sound, both preceded and
followed by affrication (S´aez Godoy 1971). In sum, there can be no other conclusion than
that Va¨ısse et al. fell short in establishing the relevant distinctions that existed between
Atacame˜no velar and postvelar sounds.
172
The other symbols that are found in the Va¨ısse glossary are b (sometimes written v),
l, m, n, r, s and y. All are described as having similar functions as their equivalents in
Spanish or other languages. The sound b is historically related to w in other Andean
languages, as can be seen in integrated loan words, such as baina ‘boy-friend’, ‘lover’
(Quechua wayna) and backtcha ‘poor’, ‘orphan’ (Quechua wakˇca).
173
The sound [w] it-
self, written hu,islimited to loans (e.g. hualcka ‘necklace’, from Quechua wal
y
qa; huata
‘belly’ from Chilean Spanish guata) and onomatopoeic expressions. In contradistinction
to Aymara and Quechua, there were no palatal laterals and nasals (l
y
, n
y
). Philippi (1860)
and von Tschudi (1866–9) recorded what may have been a voiceless, aspirated lateral
in the word h(j)lacse ‘head’ (Va¨ısse et al.: lacksi). In addition to the consonants just
mentioned, there was an affricate ts ∼ tz,which may have had the status of a separate
phoneme. It was apparently not very frequent; e.g. aytzir ‘vizcacha (an Andean rodent)’;
tserar ∼ serar ‘cold’, ‘winter’; tsimir ∼ tchimir ‘snow’.
Table 3.28 represents a tentative, if not speculative, overview of the speech sounds
that may have existed in Atacame˜no. Consonants for which we have no direct evidence
are given between square brackets. Round brackets indicate that the sound is limited to
loan words.
Clusters of up to three consonants, in medial position, and of two consonants, in initial
or final position, are not uncommon, e.g. ckolcktur ‘to fall’, icks ‘like that’. Remarkable
is the frequent occurrence of geminate consonants in medial position (e.g. ackcka ‘I’).
171
Note that Va¨ısse et al.haveckuru for ‘mountain-lion’, whereas San Rom´an has k´uhri (cf.
Lehnert Santander 1987). Confusion may have arisen in one or the other of the two sources.
172
In 1981 Bill Harrison (personal communication) recorded names of fields, plants and a few
lexical items in the village of Caspana. His taped recording contains ample evidence of glot-
talised as well as aspirated velars and uvulars. Only two conclusions can be reached: either the
Atacame˜no of the Salado river basin differed considerably from that of the San Pedro area, or
Va¨ısse et al. somehow failed to recognise the distinctions.
173
A similar development (*w >β) can be observed in the Aymara of northern Chile
(cf. Clair-Vassiliadis 1976) and in Chipaya (section 3.6).
380 3 The Inca Sphere
Table 3.28 Tentative inventory of the Atacame˜no sounds
Labial Alveolar Palatal Velar Postvelar Glottal
Plain stops p t k q ʔ
Glottalised stops p’ t’ [k’] [q’]
Plain affricates c ˇc
Glottalised affricates ˇc’
Fricatives b s x, γ x
.
h
Nasals m n
Plain lateral l
Voiceless lateral l
Vibrant r
Glides (w) y
Vow els: a, e, i, o, u, ə(?)
length
aspiration (= h)
‘These consonants are pronounced more strongly and more separate than in Italian’, says
San Rom´an (1967: 88). Gemination may have been the result of an active phonological
process, because it could affect loan words from Quechua and Spanish as well as native
words, e.g. ckas.sa ‘mountain-pass’ (from Quechua q’asa), am.mu ‘boss’, ‘master’ (from
Spanish amo). However, the existence of gemination has not been recorded for all lexical
items containing an intervocalic consonant. Theoretically, the orthographic gemination
may have represented glottalisation, as in ttelir ‘vicu˜na’, but the inventory of consonants
affected (stops, fricatives, resonants) and the frequent presence of a hyphen between two
identical consonants does not seem to favour such a conclusion.
As far as the rudimentary data available allow a general characterisation of the
language, the following grammatical facts emerge. Atacame˜no presents a mixture of
prefixation and suffixation. Personal reference affixes were prefixed to both verb and
noun. Tense, mood, nominalisation and negation with verbs, as well as nominal case
were indicated by means of suffixes. The morphology does not seem to be very elab-
orate. Perhaps this can be attributed to the state of decay in which the language found
itself when it was recorded. The general order was subject–object–verb, and modifiers
preceded heads with one notable exception: adjectives were located after substantives,
as in puri lari ‘red water’ (from puri ‘water’ and lari ‘red’).
174
174
This situation is characteristic of a whole range of languages native to northern Argentina
and northern Chile, such as Atacame˜no, Lule, Santiago del Estero Quechua and possibly also
Diaguita (Nardi 1979). It corresponds to type 24 in Greenberg’s classification of basic word-
order types (Greenberg 1966: 109).
3.7 The Atacame˜no language 381
Table 3.29 Possessive nominal paradigm in
Atacame˜no (based on San Rom´an 1967)
locjma ‘dog’
1 pers. singular c’-locjma-ia ‘my dog’
plural cun-locjma-ia ‘our dog’
2 pers. singular s’-locjma-ia ‘your (sing.) dog’
plural chin-locjma-ia ‘your (plur.) dog’
3 pers. singular ai-locjma-ia ‘his/her dog’
plural c’-locjma-ia ‘their dog’
The possessive nominal paradigm as given by San Rom´an is illustrated in table 3.29
with the noun l´ocjma (Va¨ısse et al.: lockma) ‘dog’. Note that a relational element -ia ∼
-ya is added to the root. The significance of the apostrophe that occurs with some of the
prefixes remains unexplained. It may indicate the absence of a syllabic vowel.
In Va¨ısse et al. many grammatical elements, including bound morphemes, are listed as
separate lexical items. The prefixes for first- and second-person possessor are recorded
as cki.i and iss, respectively. The latter form can be recognised in one of the ver-
sions of the Lord’s Prayer collected by von Tschudi (377). (In this and in the fol-
lowing examples the square brackets indicate the forms such as they occur in Va¨ısse
et al.)
(377) is chea
[tchei ‘name’]
{is-che-ya}
2P.SG-name-RL
‘Thy Name’
175
(von Tschudi 1866–9, in Echeverr´ıa y Reyes 1967)
San Rom´an assigns two other functions to the element ia ∼ ya. One of these functions
is that of an affixed article with nouns (378). It is possible that this so-called article
indicated possession of a noun by a non-identified possessor.
(378) sil´a-ya
llama-RL
[sil.la, sila ‘llama’]
‘the llama’ (maybe: ‘someone’s llama’) (San Rom´an 1967: 79)
175
The full sentence is santi.hijia is-che-a ya-clo ‘Hallowed be Thy Name!’ (hallowed 2P.SG-name-
RL be-IM). At the present state of our knowledge it is not possible to analyse the expression
santi hijia,but note its resemblance to Spanish santificar (‘to hallow’).
382 3 The Inca Sphere
Table 3.30 Atacame˜no personal and possessive pronouns (San Rom´an 1967;
Va ¨ısse et al. 1896)
San Rom´an Va¨ısse et al. San Rom´an Va¨ısse et al.
I acca ackcka mine acsa, ´ajsaya* acksaya
you (sing.) chema tchema yours (sing.) chienza, ch´ensaya tch´ensaya
we cuna ckunna ours cunza ckunsa
you (plur.) chime tch´ımi yours (plur.) chienza, chinzaya tch´ensaya
he ia, ya ´ı.ya his isa, is´aya issi.ya
*
The j sounds as ch in German (San Rom´an 1967: 83).
The other function of ia ∼ ya is that of a predication marker (copula) with pronominal
predicates (379). Whenever two nominals (i.e. a noun and an adjective) were connected
predicatively, a copula or predication marker was apparently not required (380).
(379) chema ya
[tchema ‘you’]
you be
‘You are.’ ‘It’s you.’
(San Rom´an 1967: 85)
(380) pauna v´alchar [pauna ‘child’; baltchar ‘bad’]
child bad
‘The child is bad.’
(San Rom´an 1967: 87)
There are two sets of pronouns to match the possessive prefixes: a set of personal
pronouns and a derived set of possessive prounouns, containing the genitive element
-sa. The pronouns that can be extracted from San Rom´an’s work and the corresponding
forms from the Va¨ısse glossary are given in table 3.30. A number of the possessive
pronouns have only been found followed by the element -ya,which possibly indicates a
predicative function (‘it is mine’, etc.).
In a possessive construction in which the pronoun is expressed explicitly, both the
possessive pronoun and a possessive prefix are required:
(381) chien-za chin-tic.han-ia
[tickan ‘father’]
you.PL-G 2P.PL-father-RL
‘your (plural) father’
(San Rom´an 1967: 80)
Nominal plural is indicated by means of a suffix -ckota,
176
which can be attached both
to nouns and to the third-person pronoun root i-:
176
Compare the Puquina nominal pluraliser -gata (cf. section 3.5).
3.7 The Atacame˜no language 383
Table 3.31 Verbal past-tense paradigm in Atacame˜no
(based on San Rom´an 1967)
yocon-tur ‘to speak’
1 pers. sing. acca q’-yocon-a ‘I spoke.’
plur. cuna cun-y´ocon-a ‘We spoke.’
2 pers. sing. chema se-y´ocon-a ‘You (sing.) spoke.’
plur. chime chin-y´ocon-a ‘You (plur.) spoke.’
3 pers. sing. ya s’-y´ocon-a ‘He spoke.’
plur. cota et’-y´ocon-a ‘They spoke.’
(382) a. ´ı-cota ∼ i-ckota ‘they’ [´ı.ya ‘he’]
b. l´ıqcau-cota ∼ lickau-ckota ‘women’ (San Rom´an 1967: 79, 85)
The prefixes that identify person of subject in a verb are partly coincident with the
possessive personal prefixes of the noun. The past-tense paradigm of the verb yockon-
tur ‘to speak’, as given by San Rom´an (1967: 86), is presented in table 3.31. The full
personal pronouns that precede each verb form were probably optional.
As in the case of the possessive prefixes the significance of the apostrophe remains
unexplained. The evidence for a third-person-singular marker s’ is weak, because it is not
attested in the other third-person subject forms occurring in the data. (One either finds
zero, or q’,which may not be a personal person reference marker.) The prefix et’- (note
the initial e)isnot found in Va¨ısse et al. The pronominal form cota is an abbreviated
variant of ´ı-cota ‘they’. Note in passing the possible diachronic link between the
pronouns and the person prefixes.
The endings of the past tense are said to be unpredictable (San Rom´an 1967: 86).
Other tense and mood categories are indicated by specific endings, where necessary
accompanied by morphophonemic adjustments, such as -ma for present tense, -(o)lo for
future tense and -(l)s for obligation.
(383) a. acca q’-minij-ma ‘I see’
[minck-tur ‘to see’]
b. acca q’-c´olc-olo ‘I will fall’ [ckolck-tur ‘to fall’]
c. acca q’-lan-s ‘I will go’ [lan-tur ‘to go’]
Examples of the Atacame˜no imperative appear in Moore (1878) and in the Lord’s
Prayer versions transmitted by von Tschudi, but we owe the identification of the ending
-k
.
al(o) as a marker of the imperative to Mostny (1954: 141).
177
Another imperative
ending found in the Lord’s Prayer versions is -alo. The examples suggest that the former
177
It is likely that a uvular stop is intended to be represented by the symbol k
.
.