Adelaar Willem, Muysken Pieter. The languages of the Andes

Подождите немного. Документ загружается.

144 2 The Chibcha Sphere

Table 2.19 Awa Pit (Cuaiquer) consonant inventory

(based on Curnow 1997)

Labial Alveolar Palatal Velar

Stops p t k

Voiceless fricatives s ˇs

Voiced fricatives z ˇz

Lateral fricative l



Lateral approximant l

Nasals m n ŋ

Glides w y

Table 2.20 Tsaﬁki (Colorado) consonant inventory (based on

Moore 1966)

Labial Alveolar Palatal Velar Glottal

Voiceless stops p t k ʔ

Voiced stops b d

Affricates c

Fricatives ϕ sh

Nasals m n

Laterals l

Vibrants r

Glides w y

As we have seen, the velar nasal in the Awa Pit consonant inventory (Curnow 1997)

compensates for the absence of a nasality contrast in the vowels. A palatal nasal [n

]

occurs but is analysed as a nasal+glide sequence ny. The phoneme l



varies between

[h], [x] and [l



], either within or between dialects. Several dialects only have [h] ∼ [x],

instead of the lateral fricative, hence the sounds in question have also been categorised

as a velar or glottal fricative (e.g. in Calvache Due˜nas 2000). Awa Pit has affricates ([t

[ˇc]), which can be analysed as allophones of s and ˇs (Curnow 1997: 29–32). When found

intervocalically, they are to be interpreted as ambisyllabic sequences (ts, tˇs); Calvache

Due˜nas prefers to treat them as separate phonemes. Geminated stops also occur in

Awa Pit. Stops that are not geminated are subject to a set of allophonic rules, such as

voicing between a voiced consonant and a voiced vowel, and voicing combined with

fricativisation between vowels. Word-internal intervocalic /t/isrealised as [r].

In Tsaﬁki the phonemes c and s have palatal allophones [ˇc] and [ˇs] before high vowels.

The vibrant r is sometimes pronounced [dr] in word-initial position. The consonants b

and d are preglottalised in word-internal position. Frequently, voiceless consonants are

2.17 Barbacoan languages 145

Table 2.21 Cha



palaachi (Cayapa) consonant inventory (based on

Lindskoog and Lindskoog 1964; Vittadello 1988)

Labial Alveolar Palatal Velar Glottal

Voiceless stops p t t

k ʔ

Voiced stops b d d

Affricates c ˇc

Voiceless fricatives f s ˇsx

Voiced fricative β

Nasals m n n

Laterals l l

Glides w y

preaspirated in word-internal position. If the phonological context should not automat-

ically account for all the cases of preaspiration, it may be necessary to add additional

series of preaspirated stops and fricatives to the inventory. The status of the glottal stop

is uncertain. A characteristic feature of Tsaﬁki is that stress is phonemic, e.g. in mer´ano

‘to hear’ and m´erano ‘to wait’ (Moore 1966: 96).

From a general typological point of view, the Barbacoan languages belong to the cen-

tral Andean sphere, rather than to the Chibchan domain (cf. Constenla Uma˜na 1991). The

inﬂuence of Quechua in the area must have been a relatively recent phenomenon, which

may go back approximately to AD 1400 (cf. chapter 3). Therefore, the Central Andean

characteristics of Barbacoan may be attributed either to an earlier period of interaction

or to genetic connections. Curnow and Liddicoat (1998: 387) observe the following

similarities in the grammatical typology of the Barbacoan languages. All the Barbacoan

languages are verb-ﬁnal (SOV). Modiﬁers (adjectives, adverbs) precede their heads.

The main morphological device is sufﬁxation. Preﬁxes occur in Cha



palaachi and in

Tsaﬁki, but to a limited extent. The Barbacoan languages are case-marking languages of

the nominative–accusative type. Nominative is zero, but accusative case is marked with

a sufﬁx (although all languages reserve this for human or deﬁnite objects). In most cases,

the accusative markers have locative functions as well. Curnow and Liddicoat emphasise

the lack of cognacy between the accusative markers in the Barbacoan languages, which

suggests a relatively recent development. All these characteristics are reminiscent of the

situation in Quechua and in Aymaran. The locative–accusative marker -ta of Awa Pit is

even formally similar to the corresponding case sufﬁx in Quechua, e.g. Awa Pit pastu-ta

[pastura], Quechua pastu-ta ‘to Pasto’. The genitive case marker of Awa Pit coincides

with conservative Quechua dialects in two of its allomorphs (-pa, -p), e.g. in santos-pa

Calvache Due˜nas (2000: 108) holds that Awa Pit has noun–adjective order. In Guambiano both

orders are possible, although the adjective–noun order is the preferred one (Long 1985: 19).

146 2 The Chibcha Sphere

kuˇzu ‘Santos’s pig’, a-p ‘my’, ‘mine’, ‘our(s)’ (Curnow 1997: 124–5); cf. section 3.2 on

Quechua. To this list we could add the prominent position of switch-reference and the

richly developed nominalisation strategies of the southern Barbacoan languages.

On the other hand, it should be observed that the Barbacoan languages also share

typological phenomena with the Chibchan languages. An example is the existence of

declarative and interrogative markers and their intimate coalescence with verbal ﬂection.

Constructions involving auxiliary verbs and nominalisation are also well represented in

Guambiano, Tsaﬁki and Awa Pit.

A typological element that is characteristic for the Barbacoan family in particular is

the way speech act participants are encoded. All the languages make a basic distinction

between speaker and non-speaker. The verbal inﬂection reﬂects this distinction, rather

than the usual division into ﬁrst, second and third person. Number distinctions are

often reserved for ﬁrst person. In Guambiano verbal inﬂection encodes three person and

number categories: speaker singular, speaker plural and non-speaker. In (222), (223)

and (224) three forms of the auxiliary verb ‘to be’ are illustrated. The forms ku-r and

k-er (with endings -r and -er) refer to a ﬁrst person singular and plural, respectively.

For second and third person only one option is available, namely k



-n (with ending -n).

Example (225) illustrates the use of the ending -n with a second-person subject, but with

a different auxiliary verb, wa- ‘to sit’.

(222) na-p´e empresa-yu k

-´ıp-ik ku-r

I-TO factory-L work-N-AJ.SG be-SR.SG

‘I am working in the factory.’

(V´asquez de Ruiz 1988: 120)

(223) na-m mis´ak k-er

I-PL people be-SR.PL

‘We are Guambianos.’

(Trivi˜no Garz´on1994: 609)

(224) un´y´au-wan m´a-p-ik k-n

child meat-AC eat-N-AJ.SG be-NS

‘The child is eating the meat.’

(V´asquez de Ruiz 1988: 69)

(225) n

i pur´ak

ac-´p wa-n

you maize husk-N sit-NS

‘You are (sit) husking maize.’

(V´asquez de Ruiz 2000: 159)

In Tsaﬁki, Cha



palaachi and Awa Pit the distinction between speaker and non-

speaker is equally important. Furthermore, there is the additional complication that a

According to V´asquez de Ru´ız (1988), the pronoun n

i can also be interpreted as third person,

when the subject is located at a short distance of the addressee. It refers to a close non-speaker,

rather than to the addressee in particular.

Note that p is a voiced fricative [β] intervocalically and a voiceless fricative [ϕ]inword-ﬁnal

position.

2.17 Barbacoan languages 147

second-person subject has to be classiﬁed with the speaker category in an interrog-

ative context. In Tsaﬁki the speaker sufﬁx -yo-/-yu-isused in ﬁrst-person subject

declarative forms (226) and in second-person subject interrogative forms (227); second-

person declarative and third person are indicated by a zero form (228). Additionally,

there is a special sufﬁx -i- for ﬁrst-person subject interrogative (Dickinson 1999, 2000;

Curnow MS).

(226) an´o ϕi-yo-ʔʔ´e

banana

eat-SR-DV

‘I ate.’

(Moore 1966: 100)

(227) an´o ϕi-y ´u-n

banana eat-SR-IR

‘Did you eat?’

(Moore 1966: 100)

(228) an´o ϕ´ı-wiya-∅∅-ʔʔe

banana eat-almost-NS-DV

‘He almost ate.’

(Moore 1966: 100)

In Cha



palaachi we ﬁnd the same opposition in ruku-yu ‘I am a man’, ‘are you a

man?’, ruku-de:-yu ‘we are men’, as opposed to ruku-

βe ‘you are a man’, ‘he is a man’,

ruku-de:-

βe ‘you/they are men’ (Lindskoog and Lindskoog 1964: 123). The sufﬁx -de:-

is a plural marker.

Personal reference in the strict sense is not indicated morphologically in any of the

Barbacoan languages. Free pronouns are used instead. In addition to the basic pronouns

that refer to subject, there may be special pronouns to be used in the object role, as

well as possessive pronouns. Possessive pronouns may have different shapes depending

on whether they are used attributively or independently (and as predicates). In all the

Barbacoan languages the pronominal system distinguishes ﬁrst, second and third person,

as well as number. Awa Pit has distinct pronoun inventories for subject, object/beneﬁciary

and possessor. The sufﬁxes that characterise the object pronouns appear to have only that

function. Although the subject and object pronouns are marked for number (singular

or plural), this is not he case of the possessive adjectives. A plural possessor is either

indicated by means of the corresponding singular form, or by the plural nominative.

In Tsaﬁki and Cha



palaachi the word for ‘banana’ (Tsaﬁki an´o, Cha



palaachi panda) coincides

with the word for ‘food’ in general. The concept ‘to eat’ without a speciﬁed object is translated

as an´o ϕi- and panda ﬁ-, respectively. Compare the use of Mandarin Chinese f`an ‘rice’ as an

unspeciﬁed object in ch˘



f`an ‘to eat’.

Note that syllable-ﬁnal n in Tsaﬁki normally serves to indicate nasality in the preceding vowel

[ϕiy

˜u]. This observation and the notation of a glottal stop in the declarative sufﬁx -

e are based

on Moore (1961, 1962).

148 2 The Chibcha Sphere

Table 2.22 Awa Pit pronouns (based on Curnow 1997: 86, 94)

Nominative Accusative Possessive

1 pers. sing. na na-wa a-p

1 pers. plur. au au-m



za –

2 pers. sing. nu nu-wa u-p

2 pers. plur. u u-m



za –

3 pers. sing. us us-a payn

3 pers. plur. uspa uspa-tuza –

Table 2.23 Cha



palaachi pronouns (based on Lindskoog and

Lindskoog 1964)

Personal Adjectival Independent

pronouns possessive pronouns possessive pronouns

1 sing. iin in-ˇci

1 plur. la-la la-la-

la:-ˇci

2 sing. n

u-ˇci

2 plur. n

u-l

u-i-ˇci

3 sing. ya ya-

ya-ˇci

3 plur. ya-la ya-la-

ya-i-ˇci

The following examples illustrate the use of some of these forms. The endings -w and

-y refer to speaker (subject) and non-speaker, respectively.

(229) ap yal au yal

1.G house we house

‘my house’, ‘our house’ ‘our house’

(Curnow 1997: 124)

(230) uspa uz-puta-y

they sit-S.PL.NS-NS

‘They are sitting.’

(Curnow 1997: 185)

(231) nu-ne na-wa pak-pyan-tu-y

you-TO I-AC cure-know-IA-NS

‘You are curing me.’

(Calvache Due˜nas 2000: 111)

(232) na-na uspa-tuza ttkyan-na-ta-w

I-TO they-PL.AC cut throw-PL.O-PA-SR.S

‘I stabbed them.’

(Curnow 1997: 183)

The difference between adjectival and independent possessive pronouns is found in

Cha



palaachi, as shown in table 2.23. The genitive markers -

and -ˇci,which can be

2.17 Barbacoan languages 149

recognised in the pronominal forms, are also used to indicate a genitive relation between

nouns.

(233) kuˇca-ʔʔ panda

dog-G food

‘the dog’s food’

(Lindskoog and Lindskoog 1964: 122)

(234) kuˇca-ˇci

dog-G

‘the dog’s’, ‘belonging to the dog’

(Lindskoog and Lindskoog 1964: 122)

Tsaﬁki is exceptional in that it encodes a gender distinction in the ﬁrst-person pronoun.

The pronoun la ‘I’ is used by men, whereas women use ˇci

k´e.For children’s use there is

yet another form: ˇce. Each of these elements forms its possessive counterpart by adding

the genitive marker -ˇci.For the plural (‘we’) there appears to be no distinction because

Moore (1966) recorded a single form, ˇci

ke-l´a, for both men and women.

Both Cha



palaachi and Tsaﬁki have an extensive verbal morphology. They both have

a rich variety of sufﬁxes that indicate different types of verbal subordination. Many of

these subordination types allow for a further distinction between subordinate verbs that

have the same subject as the verb to which they are subordinate and those that do not

have the same subject. Since personal reference is not normally visible in a subordinate

verb, the alternation of ‘same subjects’ and ‘different subjects’ must help the listener to

keep track of the role of the actors in a succession of events. The situation in Tsaﬁki, for

instance, is reminiscent of the use of switch-reference in Ecuadorian Highland Quechua

(see chapter 3). Since the latter is not typical for Quechua as a whole, the inﬂuence of a

South Barbacoan substratum may be suspected. Consider the following examples from

Tsaﬁki.

(235) h´a-namin-nan pin´ı kir´a-yo-ʔʔe

come-SM.SS-AD

snake see-SR-DV

‘While I was coming, I saw a snake.’

(Moore 1966: 102)

(236) h´a-nasa-nan pin´ı kir´a-yo-ʔʔe

come-SM.DS-AD snake see-SR-DV

‘I saw a snake, while it was coming.’

(Moore 1966: 102)

(237) w´alpa ϕi-

ˇc´un k´a-yo-ʔʔe

hen eat-F.SS catch-SR-DV

‘I caught the hen in order to eat it.’

(Moore 1966: 101)

(238) an´o ϕi-

s´awar´e-∅∅-na-ʔʔe

food eat-F.DS cry-NS-PR-DV

‘He is crying so that someone else might eat.’

(Moore 1966: 101)

The element -nan [n˜a] is translated as ‘also’ (tambi´en).

150 2 The Chibcha Sphere

Tsaﬁki is the only Barbacoan language in which an unambiguous system of grammat-

ical classiﬁers has been recorded. They refer to shape and are used both with numerals

and with adjectives. When used with adjectives they are followed by a sufﬁx -n (that is,

vowel nasalisation). The classiﬁer -de,which refers to long objects, illustrates the use

of classiﬁers in (239) and (240).

(239) palu-d´ean´o

two-CL banana

‘two bananas’

(Moore 1966: 99)

(240) n´a-de-n an´o

child-CL-AJ banana

‘small banana’

(Moore 1966: 99)

As we anticipated, Guambiano has an elaborate system of auxiliary verbs. These

auxiliary verbs are inﬂected for person and number, whereas the (nominalised) verbs

they accompany can carry different types of aspectual distinctions. Guambiano has a

series of auxiliary verbs that refer to bodily positions; one of them, wa- ‘to sit’, was

illustrated in example (225). Other positional auxiliaries are paˇs

a- ‘to stand’, cu- ‘to

lie’, meka- ‘to hang’ (241). These verbs can also be used as existential verbs ‘to be

somewhere’ (Spanish estar), as in (242).

(241) y´e-wan yan´a-m-ik paˇs

´a-n

potato-AC sell-F-N stand-NS

‘I will sell the potatoes.’ (lit. ‘It stands to sell the potatoes.’)

(Trivi˜no Garz´on 1994: 614)

(242) ul-p´ek´aw-yu cu-n

snake-TO grass-L lie-NS

‘The snake is in the grass.’

(Long 1985: 28)

Awa Pit, Cha



palaachi and Tsaﬁki often use the verb stem ‘to do’ (*ki-) with nouns,

adjectives and Spanish loan roots in order to create new verbs (cf. Curnow and Liddicoat

1998: 403). An example from Cha



palaachi is (243).

(243) iŋŋbi ke-nu

spittle do-IF

‘to spit’

(Lindskoog and Lindskoog 1964: 80)

Among the morphological characteristics of the southern Barbacoan languages we

may mention a morphological passive in Cha



palaachi, the existence of an extensive set

of nominalisations (comparable to Quechua) and the use of verbal preﬁxes. Tsaﬁki has

the following nominalisations: agentive in -min, inﬁnitive in -no, stative participle in -ka

and instrumental in -nun.

2.18 Kams´a 151

(244) m´usika ke-n´un

music make-IS.N

‘something to make music with’

(Moore 1966: 103)

Verbal preﬁxes have different functions. Cha



palaachi has two preﬁxes de-, one with

the meaning ‘to ﬁnish doing something’ and one with the meaning ‘plural subject’. The

preﬁx ma(n)- ‘repetition’ has an equivalent in Tsaﬁki, where it is man- (compare man

‘one’).

(245) an´o man-ϕi-∅∅-ʔʔ´e

food RP-eat-NS-DV

‘I ate again.’

(Moore 1966: 104)

Avery interesting preﬁx is Tsaﬁki pe-, which indicates help or joint action (246).

(246) pe-ayud´ai-de

[Spanish ayudar ‘to help’]

AS-help-IM.NS

‘Help!’

(Moore 1996: 104)

(247) pe-h´ı-ˇcina-yo-ʔʔe

AS-go-F-SR-DV

‘I will go with you/him.’

(Moore 1996: 104)

2.18 Kams´a

The language called Kams´a (Kam



nˇc

´a)orSibundoy has often been identiﬁed with the

language of an ethnic group known historically by the Quechua name of Quillacinga

(‘moon-nose’). For the time being, Kams´aisbest treated as a linguistic isolate, as former

associations with the Chibchan family have proved unsubstantial. The Quillacinga were

a highland people, settled in the area northeast of the town of Pasto. The present-day

Kams´a, who according to Arango and S´anchez (1998) number approximately 4,000

people, inhabit the valley of Sibundoy between Pasto and Mocoa at the upper reaches

of the Putumayo river. They share the area with their more numerous neighbours, the

Quechua speaking Inga (cf. chapter 3), who may be (Quechuanised) descendants of

the Quillacinga. The Indians of the area are known for their persistent struggle against

monopolistic missionary activity during most of the twentieth century. The history of

oppressive domination by the Capuchin mission among the Kams´a and the Inga is told

by Bonilla (1972).

The Kams´a language with its extraordinarily complex morphology has only been

described to a limited extent, most of the published work available so far being phono-

logical or on special topics, such as the structure of discourse (Howard 1977a, b). A

text collection with a phonological introduction has been published by Mongu´ıS´anchez

152 2 The Chibcha Sphere

Table 2.24 Kams´a consonant phonemes (based on Howard 1979)

Labial Alveolar Palatal Retroﬂex Velar

Voiceless stops p t k

Voiced stops b d [

d] g [

Voiceless affricates c ˇcˇc

Voiceless fricatives (ϕ)s ˇsˇs

Nasals m n n

Laterals l l

Vibrant r [ˇr]

Glides w y [d

ˇz

]

(1981). An encouraging aspect of Kams´a studies is the active participation of members

of the Kams´a community itself (Juajibioy Chindoy 1962; Juajibioy Chindoy and Wheeler

1973; Jamioy Muchavisoy 1992, 1995). Mel´endez Lozano (2000c) provides a short ty-

pological sketch of the language, partly based on unpublished sources. Regrettably, we

have not been able to locate a dictionary of the language.

According to Howard (1979), Kams´a distinguishes six vowels (a, e, i,



, o, u). The



is a high central vowel. An inventory of the consonant phonemes is represented in

table 2.24.

The inventory in table 2.24 calls for some observations. The special realisations of r

and y (given between square brackets) occur, respectively, in word-initial position (e.g. in

r´ala [ˇr´ala] ‘money’, from Spanish real ) and after n (e.g. in bom´ınyi [bom´ınd

ˇz

i] ‘eye’). The

Kams´a language has been particularly receptive to borrowing from Spanish, a fact which

has profoundly inﬂuenced the sound system as it is today. The sound [ϕ] (alternatively

[β]) may be treated as a preconsonantal allophone of either p or b if borrowings from

Spanish are not counted (e.g. in skon´uϕta [skon´uϕta ∼ skon´uβta] ‘nine’). In borrowings

it replaces Spanish f (e.g. in ϕlako-x´ema ‘slim’, from Spanish ﬂaco ‘slim’ and -xema

‘old’), a fact which justiﬁes its status as a separate phoneme. In native words the voiced

stops d and g are found after nasals (e.g. in

d´on

e ‘no’, ϕce

g´a ‘black’). They have

additional fricative allophones ([ð], [γ ]) that occur in other environments, but only

in loan words, e.g. the root-initial d in xa-dema

d´a-na

‘to demand’, from Spanish

demandar.From the viewpoint of the native sound system, the clusters

d and

g may

very well be treated as unit phonemes, due to their frequent occurrence in word-initial

position. In some well-established loan words, such as

de´olp ‘suddenly’, from Spanish

de golpe (Mongu´ıS´anchez 1981), a nasal element that was not even there originally has

been introduced.

The frame x(a)-...-nais characteristic for inﬁnitives in Kams´a.

2.18 Kams´a 153

Table 2.25 Kams´a personal reference markers

(after Mel´endez Lozano 2000c)

1 pers. sing. c(i)-, s(



)-, i-

1 and 3 pers. dual bo-, b



-, bn-

1 pers. plur. exclusive ϕc(i)-, ϕs(



1 pers. plur. inclusive and 3 pers. plur. mo-, m(



2 pers. sing. ko-, k(



2 pers. dual ˇs

o-, ˇs



-, ˇs

2 pers. plur. ˇs

mo-, ˇs



A characteristic feature of the Kams´a language are its consonant clusters. Up to three

consonants (rarely even four) can be found in medial and initial position, e.g. in ˇsknen´a

‘wooden plate’, x



ϕcbeˇc

´ana ‘to carry a child’,

goϕˇsn´a ‘green’ (Howard 1979: 89). The

place of the accent is contrastive. It can be either on the last or on the penultimate vowel

of a word.

Typologically, Kams´a differs from any other language in the area by its abundant use of

preﬁxes. Nevertheless, preﬁxation is not the only morphological device of the language.

Case and number with nouns, as well as several verbal categories, are indicated by means

of sufﬁxes. A set of eighteen classiﬁers based on shape (e.g. -b´e for round objects, -ˇc

for oval objects, -ϕxa for rigid cylindrical objects, -xa for ﬂexible cylindrical objects, -ˇs

´a

for hairy and/or composed elements, -ye for liquids) can be sufﬁxed to demonstrative

pronouns, numerals and adjectives (cf. Mel´endez Lozano 2000c: 136). They also occur

as sufﬁxes with nouns; e.g. in ﬂor-ˇs

´a ‘ﬂower’, from Spanish ﬂor, and in eskardon´a-ˇc

e ‘an

insect’ (Howard 1979: 88), presumably from a local Spanish word escardona (compare

escardar ‘to weed’).

Howard (1977a) distinguishes eight distinct slots of afﬁxes that can precede the verbal

root in narrative discourse. The functions of the preﬁxes occurring in these slots belong to

the domain of narrative type (historical versus legendary event), evidentiality (presence

and/or participation of the speaker in the event), personal reference, movement, tense

and aspect. Mood (imperative, contrary-to-fact, potential), negation and interrogation

are also expressed by means of preﬁxes. Mel´endez Lozano (2000c) lists nine modal

preﬁxes (including negative and evidential) and ten aspectual preﬁxes. He reports the

personal reference markers used for identiﬁcation of the subject (based on Juajibioy

Chindoy and Wheeler 1973), as indicated in table 2.25.

His enumeration does not contain any markers referring to a third person singular.

The non-singular third-person markers coincide with non-singular ﬁrst-person markers.

Howard (1977a: 7) gives o- for third-person-singular subject. However, many third-

person-singular subject forms occurring in her examples seem to lack this preﬁx,