Adelaar Willem, Muysken Pieter. The languages of the Andes
Подождите немного. Документ загружается.
114 2 The Chibcha Sphere
2000). Constenla Uma˜na (1991: 60) distils several other typological considerations from
the scanty data. Yukpa is an SVO language, in which genitives and demonstratives
precede the head noun, whereas numerals and adjectives follow the head. Leaving aside
the genitive word order, this is the same pattern as the one found in the Chibchan
languages.
The survival well into the twentieth century of indigenous groups in the Op´on and
Carare river areas in the Colombian department of Santander constitutes unequivocal
proof of the advance of Cariban-speaking peoples along the Magdalena valley. Separate
word lists for Op´on and Carare were published by von Lengerke (1878), and further
(undifferentiated)Op´on–Carare material wascollected in 1944 by Pineda and Fornaguera
(1958). The latter source is also presented in Landaburu (1998). It includes an account
of a deadly feud between the Op´on and the Carare, which brought both groups to the
vergeofextinction in 1914. Durbin and Seijas (1973a) report that there must have been
at least five speakers in 1944, one of whom was in his twenties.
Durbin and Seijas (1973a) noted considerable differences between the Op´on and
Carare lists of von Lengerke, on the one hand, and the Op´on–Carare lists of Pineda and
Fornaguera, on the other. The latter seems to represent a divergent dialect with respect
to the other two. An interesting feature of this dialect is the widespread occurrence of
a suffix -id
/-in
/-in
y
,which may be historically identical to the relational-possessive
suffix of other Cariban languages, e.g. in pot´a-id
‘mouth’ (Macoita Yukpa p´ota-
ʔ
)
and in pan´a-in
y
‘ear’ (Macoita Yukpa p´ana-
ʔ
). In several respects Op´on–Carare is
more conservative than Yukpa. It conserves the Cariban root *tuna in tun´a-in
y
‘water’,
where Yukpa varieties have k´una(-
ʔ
), and the r in yor-id
‘tooth’, where Yukpa has
yi(-
ʔ
), y
or d
ˇz
ʔ
; cf. also Pineda Giraldo’s comparative word list in Landaburu (1998:
531–5).
The existence of a Cariban speaking group in an Inter-Andean setting, such as the
Magdalena valley, opens the possibility of a wider distribution of Cariban peoples in
the area. For several nations of great historical importance, namely, the Colima, the
Muzo, the Panche, the Pant´agora and the Pijao, a Cariban linguistic affiliation has often
been proposed. All these peoples have long been extinct, except for the Pijao of the
department of Tolima. In 1943 Pijao word lists were collected in the municipality of
Ortega by Alicia and Gerardo Reichel-Dolmatoff, as well as by Roberto Pineda Giraldo
and Milc´ıades Chaves (Durbin and Seijas 1973b). The Pijao language is now considered
extinct as well.
Durbin and Seijas (1973b) suggest that all these languages should be left unclassi-
fied, because the extremely limited data do not provide enough evidence for a Cariban
affiliation. This is certainly true of Panche, for which there are almost no data. The main
reason to assume a Cariban affiliation for Panche is the existence of a large number of
place names in -aima, -oima and -ima,which are highly suggestive of Cariban toponymy
2.12 Arawakan languages of the Caribbean coast 115
in Venezuela and the Guyanas.
53
The Panche terms acaima ‘important personage’and
colima ‘cruel assassin’, mentioned by Durbin and Seijas (1973b: 51), show that this end-
ing was also used for human beings. Similar features can be found in the Muzo–Colima
and Pijao domains.
54
The Colima, Muzo and Pijao word lists contain a few items of
basic vocabulary that point at a Cariban connection (cf. Constenla Uma˜na 1991: 62–3).
It appears possible to detect a common sound innovation in the three languages, when
some of these lexical items are compared to their counterparts in other Cariban languages,
such as Galibi (Hoff 1968).
(146) Pijao: t´ana Galibi: tu:na ‘water’
Pijao: t´ape Colima, Muzo: tapa Galibi: to:pu ‘stone’
The Cariban elements found in Colima, Muzo and Pijao do not suggest a specific
relationship with Op´on–Carare and Yukpa. They may reflect an older Cariban invasion
of the Magdalena valley, or they may represent conservative traits that have not been
preserved in the northern languages. For instance, the Pijao word for ‘moon’ n´una
is found in many Cariban languages, but not in Op´on–Carare and Yukpa, where it is
kan´o-n
y
and k´unu, respectively (Durbin and Seijas 1973b: 49).
2.12 Arawakan languages of the Caribbean coast
Two closely related Arawakan languages are located in the area separating the northern
Andes from the Caribbean coast, Guajiro and Paraujano.With more than 300,000
speakers Guajiro or Wayuunaiki (‘language of the Wayuu people’) may very well be
the fastest growing indigenous language of the area covered by this book. Its original
homeland is the Guajira peninsula, shared by Colombia and Venezuela. The Guajiro
language can be subdivided into two main dialects, a northern peninsular dialect called
arribero or winpum
in ‘towards the waters’ and a southern inland dialect called abajero
or wopum
in ‘towards the roads’ (P´erez van Leenden 2000).
A large part of the Guajira peninsula belongs to Colombia and, consequently, a ma-
jority of the Guajiro people used to reside in that country. However, this situation has
been reversed in the past decades. A vigorous colonisation process is taking place in
the Venezuelan state of Zulia towards the shores of Lake Maracaibo and the town of
Maracaibo itself. Alvarez (1994: 10) records the following increase in the statistics of
the Guajiro speakers in Venezuela: 16,793 in 1950; 52,000 in 1982; 179,318 in 1992.
This last figure covers more than 50 per cent of the present-day indigenous population
53
The ending -im
is used as an augmentative in several Cariban languages of the Guyanas, e.g.
in Trio (Eithne Carlin, personal communication).
54
Personal names, which occur abundantly in the historical sources, are rarely exploited in linguis-
tic reconstruction. The Muzo area, for instance, is characterised by a great incidence of personal
names in -p´ı, e.g. Boquip´ı, Quinancep´ı (Rodr´ıguez Baquero 1995).
116 2 The Chibcha Sphere
in Venezuela. The Colombian Guajiro have been calculated at 144,000 (Arango and
S´anchez 1998). The spectacular growth in the number of Venezuelan Guajiro is dif-
ficult to explain by natural increase alone. Immigration from Colombia and statistical
underexposure in the past may provide an explanation.
By contrast, Paraujano or A˜n´un is on the verge of extinction. The Paraujano (‘beach
people’, from Guajiro palauhe ‘from by the beach’) inhabit the coast and islands between
Maracaibo and the Guajira peninsula. The last speakers of the language live in villages
of pile-dwellings located in the Lagoon of Sinamaica, north of Maracaibo.
55
In the 1980s
only a few aged people continued to speak the language (Patte 1986). Alvarez (1994)
estimates the number of Paraujano speakers at less than a dozen.
56
As a consequence of the Caribbean background of Guajiro and Paraujano, they do
not share many typological features with the languages of the Andean and Chibchan
spheres. Together with Lokono, the Arawakan language of the Guyanas, and the extinct
Arawakan languages of the Caribbean islands, Guajiro and Paraujano constitute a north-
ern, Caribbean branch of the Arawakan language family (Payne 1991a), referred to by
Payne as Ta-Arawak on the account of the shape of the first-person prefix, which is ta-
(or da-)inthese languages. This feature separates Guajiro and Paraujano from eastern
Colombian Arawakan languages, such as Achagua and Piapoco, which use the more
widespread Arawakan marker nu- for that purpose.
In relation to Lokono and the Arawakan languages of the Lesser Antilles (St Vincent,
Dominica), Guajiro is phonologically innovative. At least one innovation, the devel-
opment of *k to glottal stop in intervocalic position, has affected borrowings from
Spanish, e.g. pa:
ʔ
a ‘cow’ (from Spanish vaca). Some of the first Amerindian words
borrowed by the Spaniards after their occupation of the Caribbean islands, as well as
terms recorded in Hispaniola by the sixteenth-century chronicler Fern´andez de Oviedo,
have a shape that could be derived directly from Guajiro (cf. Taylor 1978: 123). It sug-
gests that Guajiro must have been closely related to Ta´ıno, the extinct native language
55
It is to settlements like those of the Paraujano that the country owes its name of Venezuela
(‘Little Venice’).
56
Durbin and Seijas (1973a), also Durbin (1985: 349), mention a third Arawakan language in
the northern Colombian Andes, Hacaritama, allegedly a close relative of Guajiro. Hacari-
tama was the name of the sixteenth-century inhabitants of the province of Oca˜na in the
department of Norte de Santander (online information 2001 ‘Rese˜na hist´orica de Oca˜na’
http://www.cocota.com/histcult/resea.htm). A word list collected near the town of Hacar´ı (La
Palma) was published by Justiniano P´aez in 1936. Rivet and Armellada (1950) elaborate on the
context in which this list was collected: in 1912 a settlement of agricultural workers was at-
tacked and all but wiped out by (Bar´ı?) Indians. Consequently, the police detained three Guajiro,
whowere travelling through the area. They probably provided the data for P´aez’s ‘Hacaritama’
list, which is clearly Guajiro. The real affiliation of the Hacaritama language, if it ever existed,
remains undetermined.
2.12 Arawakan languages of the Caribbean coast 117
Table 2.12 Guajiro consonant inventory (Alvarez 1994;
Mansen and Captain 2000)
Labial Alveolar Palatal Velar Glottal
Stops p t k ʔ
Affricate ˇc
Fricatives s ˇsh
Nasals m n
Flap l []
Trill r [rr]
Glides w y
of Hispaniola, even when compared to other documented Arawakan languages of the
Caribbean.
(147) Spanish aj´ı < **aˇs´ı Guajiro haˇs´ı ‘hot pepper’
cf. Lokono haˇci (< **h´at
h
i)
(148) Spanish aje < **´aˇse Guajiro h´aiˇsi ‘sweet potato’
cf. Lokono haliˇci (< **h´alit
h
i)
The existing linguistic literature on the Guajiro language is substantial. We shall only
mention a few studies, such as Celed´on (1878), Holmer (1949), Hildebrandt (1963),
Olza and Jusay´u (1978), Jusay´u and Olza (1981, 1988), Goulet (1981), Alvarez (1985,
1993, 1994, 1996), Mansen and Captain (2000), and Mosonyi and Mosonyi (2000).
ForParaujano see Patte (1978, 1981, 1989). The Guajiro language is taught at the
universities of Riohacha (Colombia) and Zulia (Maracaibo, Venezuela). There have
been several standardisation efforts, and the language benefits from an official status
in the state of Zulia, Venezuela. The following short overview of characteristics of the
Guajiro language is based mainly on Alvarez (1994).
The sound system of Guajiro is relatively limited and straightforward. The consonants
are shown in table 2.12.
In our presentation of Guajiro we follow the habit of writing l and r for two vibrants
that are characteristic of the language. The sound symbolised by l has been described as a
combination of a slightly retracted Spanish r with a lateral sound. It is, in the terminology
of Ladefoged (1975), a flap, rather than a tap. In order to produce this sound, the tip of
the tongue is at first folded backwards and then projected against the upper alveolars
before reaching a rest position (cf. Alvarez 1994: 61–2). The sound written r is a trill,
comparable to Spanish rr and the corresponding sound of Bar´ı, Damana and Chocoan.
The alternation between l, r and t plays an important role in Guajiro morphophonemics.
This morphophonemic alternation extends to loan words, where l is replaced by t in
118 2 The Chibcha Sphere
syllable-final position, e.g. p
:tp
ra, Spanish p´olvora ‘gunpowder’ (Taylor 1978: 125);
katsu:na, Spanish calz´on ‘trousers’.
Guajiro has six vowels, which can either be short or long: a, e, i, o, u,
. The
is a
high central vowel. Vowel length is contrastive. Guajiro has a predictable pitch accent.
Vowel length and accent are not interrelated. Word-final vowels, for instance, whether
long or short, can but need not be accented. The accent rules are described in Alvarez
(1994: 13–37). If the first syllable is heavy (that is, if it ends in a long vowel, a diph-
thong or a consonant), it is accented. If it is light (that is, if it ends in a short vowel),
the second syllable is accented. If the first syllable ends in a glottal stop, that sylla-
bleisnot counted for the assignment of the accent. In that case the second syllable is
accented when it is heavy. If it is light, the accent moves to a third syllable when avail-
able. Alvarez considers initial CV
ʔ
- sequences to be extrametrical syllables (cf. Hayes
1982).
(149) ´atpana: ‘rabbit’
ay ´aʔʔlaha: ‘to buy’
aʔʔyal ´aha: ‘to cry’
(Alvarez 1994: 15)
Apparent exceptions, such as ei
ʔ
r´aka: ‘to threaten’ and eikk´ahawa: ‘to transport’
may be re-analysed as regular e-
ʔ
ir´aka-: and e-
ʔ
ik´aha-wa:, respectively (Alvarez 1994:
32–3). Such an analysis is justified by the morphological structure of the forms in
question. They contain the ‘zero person’ prefix a- (e-/o-) followed by a root with an initial
glottal stop (-V: and -wa: are allomorphs of the infinitive ending). A comparison with
the related forms ni-
ʔ
ir´ak
-in ‘he threatens’and ni-
ʔ
ik´aha:-in ‘he transports’ (ni- ‘3rd
person masculine’; -in ‘present’) reveals the existence of a glottal stop in the underlying
root.
The word order of the Guajiro language is verb-initial with a preference for VSO
order. Adjectives and the possessor in a genitive construction follow the noun they
modify. In a comparison of word order in several northern Arawakan languages, Wise
(1991) observes that Guajiro is the only language that has verb-initial word order.
In an environment, such as Colombia, where almost all the indigenous languages
are verb-final, this is quite remarkable. In Guajiro, non-verbal elements that occur in
sentence-initial position, including interrogatives and adverbs, normally function as
predicates (there is no copula). Such predicative elements can be followed by a rel-
ative or a subordinate clause in a sort of cleft construction (150). The tense and as-
pect markers and gender-number markers (see below) of the subordinate verb are then
transferred to the predicate (151).
57
An example with a fronted interrogative pronoun
is (152).
57
The same holds for Paraujano (Patte, cited in Alvarez 1994: 105).
2.12 Arawakan languages of the Caribbean coast 119
(150) u:holu ta-s-e:-ka
chicha 1S.SG-drink-DE-DF
‘It is chicha I want to drink.’
(Alvarez 1994: 97)
(151) tay-e:-ˇci o-ʔʔuna-ka maikoʔʔu-min
I-F-S.MS ZP-go-DF Maicao-AL
‘I am the one who will go to Maicao.’
(Alvarez 1994: 96)
(152) hara-l-e:-r p-a:pira-ka ya:
which-FE-F-O.FE 2S.SG-warn-DF Q
‘Whom (feminine) are you going to warn?’
(Alvarez 1994: 103)
One of the most common ways to indicate negation is by means of the negative verb
nnoho-lu-.
58
It is followed by a verb with the subordinative ending -in; tense, aspect and
gender-number markers are added to the negative verb (153).
(153) nnoho-l-e:-ˇci ta-sa-k-in kami:r
not.be-MS-F-O.MS 1S.SG-greet-TS-SU Camilo
‘I shall not greet Camilo.’
(Alvarez 1994: 98)
From a morphological point of view, Guajiro is a complex language, even by Andean
standards. The morphological devices are prefixation, suffixation and root reduplication.
The suffixes may number more than a hundred all together. The prefixes are limited
both in number and in function. Most of them are personal reference markers, which
can either refer to the subject of a verb, or to the possessor of a noun. One prefix
indicates non-specified personal reference (‘zero person’). Two other prefixes, ka- and
ma-, indicate possession (‘having’) and non-possession (‘not having’), respectively. All
prefixes consist of a consonant and a short vowel (zero person consists of a vowel alone).
The vowels are variable and can be either non-high (basically a)orhigh (basically
).
Both are subject to different types of fusion (with an adjacent vowel) or harmony (with a
non-adjacent root vowel). An intervening consonant can also influence the choice of the
prefix vowel, depending on whether it is coronal (alveolar and palatal) or not (Alvarez
1990, 1994: 39–59). The rules of fusion and harmony are quite opaque.
Guajiro personal prefixes and pronouns encode person, number and gender, as
represented in table 2.13.
Two types of possession, alienable and inalienable, are distinguished. Nouns can
be marked for inalienable possession by the addition of a personal-reference prefix
(154). If the possession is alienable a special relational form of the noun must be used.
Relational forms normally contain one of a set of special suffixes (155). The existence
of such relational suffixes is frequent in the Amazonian region, in Central America and
58
The initial n in nnoho-lu- represents a syllabic nasal necessary to account for the fact that the
first (and not the second) o in this form is accented [
˚
nn´ohou...].
120 2 The Chibcha Sphere
Table 2.13 Personal prefixes and pronouns in Guajiro (based on
Jusay´u and Olza 1988; Alvarez 1994; Mansen and
Captain 2000)
Prefixes Pronouns
Singular Plural Singular Plural
1 pers. ta- wa- taya waya
2 pers. p
-h
- pia hia
3 pers. masc. n
- na- nia naya
3 pers. fem. h
- ∼ s
-* hia ∼ ˇsia*
zero pers. a-
* The s
- and ˇsia forms of the 3 pers. fem. are characteristic of the abajero
dialect.
in Mesoamerica. In the Andean region it is limited to a few languages (Atacame˜no,
Mochica). For a detailed treatment of relational forms in Guajiro see Alvarez (1996).
(154) ta-ye:
1P.SG-tongue
‘my tongue’
(Alvarez 1994: 87)
(155) ta-kulu:t-se [kulu:lu‘cloth’]
1P.SG-cloth-RL
‘my cloth’
(Alvarez 1994: 67)
In a genitive construction, in which the possessor follows the possessed, the latter is
referred to by a noun with a possessive prefix (156).
(156) n-ˇsi kami:ru
3P.SG.MS-father Camilo
‘Camilo’s father’
(Alvarez 1996: 31)
In Guajiro the subject and object roles remain unmarked. There is a set of relational
stems that function as oblique case markers. They can take possessive personal-reference
prefixes, forming a possessive phrase with their complement (157).
(157) a-tunku-ˇsi
59
kami:r hu-luʔʔut si-ka-l
ZP-sleep-S.MS Camilo 3P.FE-inside the.FE hammock.DF.FE
‘Camilo sleeps in the hammock.’
(Alvarez 1994: 125)
59
‘To sleep’ is -tunka- (infinitive a-tunka-:); the vowel change (a > u)istriggered by the high
vowelinthe suffix. Mosonyi (1993: 176) notes an unrounded vowel in at´unk¨ushi tay´a [a-tunk
-ˇsi
taya].
2.12 Arawakan languages of the Caribbean coast 121
Alternatively, a relational stem can take a zero person prefix and cluster syntactically
with the verb. The logical complement of the relational stem is then raised to argument
position (either subject or object), and the relational stem takes gender-number marking
(see below) in agreement with the raised complement. For instance, in (158) t
s
i-ka-l
‘the hammock’ can be seen as the direct object of -tunka a-lu
ʔ
u ‘to use something to
sleep in’.
(158) ta-tunka a-luʔʔu-lu t si-ka-l
1S-sleep ZP-inside-O.FE the.FE hammock.DF.FE
‘I sleep in the hammock.’
(Alvarez 1994: 163)
Finally, the relational stems can be affixed to nouns (159).
(159) pi-pia-luʔʔu
[pi:ˇci ‘house’; relational form: -pia]
2S.SG-house-inside
‘in your house’
(Alvarez 1994: 63)
One of the most remarkable features of Guajiro is its gender–number system, which
comprises three categories, masculine singular, feminine singular and plural. The fem-
inine singular is in reality the least marked category and does not exclusively refer to
female beings. A more appropriate qualification may be ‘non-masculine non-plural’.
These three categories are encoded morphologically by means of a set of markers that
emerge in various agreement positions in the sentence, as well as lexically. Demon-
stratives, personal pronouns (third person only), nouns marked for definiteness, verbal
subject prefixes and possessive prefixes (third person only), the (suffixed) object of a
transitive verb in a non-present tense, and the (suffixed) subject of a non-transitive verb
must be specified for gender–number. Some of these uses are illustrated in (160)–(163);
see also (157).
(160) t mayaht-ka-l
60
[mayahl ‘young lady’]
the.FE young.lady-DF-FE
‘the young lady’
(Alvarez 1994: 129)
(161) ta-sa-k-e:-ˇci pia
1S.SG-greet-F-O.MS you
‘I shall greet you (man).’
(Alvarez 1994: 100)
(162) haˇsiˇci-s tare:sa
be.angry-S.FE Teresa
‘Teresa is angry.’
(Alvarez 1994: 93)
60
In Mansen and Captain’s description (2000) of the Colombian abajero dialect this ending is
described as -ka-t.
122 2 The Chibcha Sphere
(163) koʔʔu=koʔʔu-t-ˇsi: tepiˇci-ka-na na-la-irua
silent-TS-S.PL child-DF-PL DC.PL-DT-PL
‘Those children are silent.’
(Alvarez 1994: 81)
Examples (157) and (160) furthermore exemplify the use of prenominal definite
articles: t
(feminine), ˇci (masculine) and na (plural). When following a noun, they are
interpreted as proximate demonstratives: ‘this’, ‘these’ (Jusay´u and Olza 1988). Other
demonstrative pronouns are derived by the affixation of one of the elements -la/-ra,-sa
and -(y)a,which express an ascending scale of remoteness, e.g. na-la ‘those’ in (163)
exemplifies the use of -la/-ra with the lowest degree of remoteness. Example (163),
furthermore, contains a plural marker -irua,which functions independently from the
gender–number system.
Another remarkable feature of the Guajiro language is the existence of two competing
verbal conjugation types. The first type involves the use of personal reference prefixes,
which identify the subject. If the verb is transitive, an object can be specified by means
of gender–number suffixes under certain restrictions. This type of conjugation has been
called prefixal (Hildebrandt 1963) or synthetic (Alvarez 1994). In the second conjugation
type, the subject of the verb is expressed by means of a free pronoun, located after the
verb in accordance with the word-order rules of the language. The verb may carry
gender–number suffixes, but these refer to the subject (not to the object) of the verb.
This type of conjugation has been characterised as analytic. Guajiro makes a formal
distinction between active and stative verbs. Stative verbs cannot take any personal
reference prefixes. As a result they can only be used in the analytic conjugation. Active
verbs that are used in the analytic conjugation must fill their prefix position with a
‘zero person’ marker. The difference between the two conjugation types is illustrated in
(164) and (165) with an active transitive verb (‘to buy’). The first example illustrates the
analytic type, whereas the second exemplifies the synthetic type.
(164) a-yaʔʔla-h-e:-ˇci pia ˇci ka:ʔʔula-ka-i
ZP-buy-TS-F-S.MS you the.MS goat-DF-MS
‘You will buy the he-goat.’
(Alvarez 1994: 115)
(165) p-yaʔʔla-h-e:-ˇci ˇci ka:ʔʔula-ka-i
2S.SG-buy-TS-F-O.MS the.MS goat-DF-MS
‘You will buy the he-goat.’
(Alvarez 1994: 115)
The division of labour between the synthetic and analytic conjugations is based on
the syntactic environment in which the verb occurs. In subordinate clauses active verbs
are conjugated synthetically, whereas stative verbs are in the analytic construction. In
main clauses, however, a more elusive situation obtains. The synthetic conjugation is
limited to transitive verbs with a definite object, whereas the analytic conjugation can be
used under any circumstances. It means that transitive verbs without a definite object,
2.12 Arawakan languages of the Caribbean coast 123
along with all intransitive and stative verbs, are conjugated analytically. When a verb is
transitive and its object is definite, the two conjugation types are in competition.
The numerous morphological devices of Guajiro include the formation of a passive
(suffix -na/-n
). Interestingly, Guajiro exhibits a proliferation of pseudo-passive con-
structions, in which not the patient but some oblique complement is raised to the subject
position. The nature of its function is made explicit by the presence of a relational stem
with a zero person prefix. Example (166) illustrates the use of such a pseudo-passive
construction, in which an original comitative complement plays the subject role. The
construction is part of a relative clause. The comitative function is indicated by a re-
lational stem with zero person prefix a-ma: ‘with’. The agent is introduced by another
relational stem: -tuma ‘by’.
(166) t-eʔʔr-in t mahayl a-yonna-h-na-ka-l a-ma:
nu-tuma kami:r
1S.SG-see-PR the.FE young.lady ZP-dance-TS-PS-DF-FE
ZP-with 3P.SG-by Camilo
‘I saw the girl that has been danced with by Camilo.’
(Alvarez 1994: 127)
Several types of root reduplication occur in the Guajiro language. Their functions
and shape are reminiscent of those of the reduplication types found in Quechua (see
chapter 3) and in Mapuche (see chapter 5). Alvarez (1994: 75–86) describes one type
of reduplication in detail. A large group of stative verbs consisting of a root followed by
a thematic suffix -ta- refer to bodily positions, attitudes, etc. Plurality of subject can be
expressed by the root in reduplicated form, followed by a thematic suffix -l
- or any of
its allomorphs (-lu-, -r
-). The allomorph -lu- is required after rounded back vowels.
(167) waya-ta- ‘to be stretched’, ‘to be spread out’
waya=waya-l- ‘to be stretched (several)’, ‘to be spread out (several)’
(168) koʔʔu-ta- ‘to be silent’
koʔʔu=koʔʔu-lu- ‘to be silent (several)’
The Guajiro language contains a number of interesting neologisms. The word for
‘horse’, a very important animal in Guajiro society, is ama. This word originally had
the meaning of ‘tapir’, which is now called lanta (from Spanish danta). The word for
‘aeroplane’ is presented and analysed in (169).
(169) ka-tna-s-ka-l
OS-arm-FE-DF-FE
‘she who has arms’
(Alvarez 1994: 129)
Guajiro has a fully developed decimal system. The unit numbers are: wane ‘one’,
piama ‘two’, ap
n
in ‘three’, pienˇci ‘four’, ha
ʔ
ra(l)i ‘five’, aipirua ‘six’, akaraiˇsi