Adelaar Willem, Muysken Pieter. The languages of the Andes
Подождите немного. Документ загружается.
394 3 The Inca Sphere
in modern Imbabura Quechua is the use of a labial fricative f (from earlier Quechua
*p
h
; cf. section 3.2.5). Place names with characteristic stems and endings, as well as a
fair number of family names mentioned in the colonial documents, constitute evidence
of the past existence of Cara and give an impression of its phonology. Some endings
(e.g. -mued, -pud) are shared with Pasto, suggesting some kind of relationship between
the two languages. River names in -p´ı/-b´ı (e.g. Calap´ı, Chulxab´ı) suggest a connection
with the Barbacoan languages, where this element has the meaning of ‘water’ or ‘river’.
PazyMi˜no (1941a) favours the view that Cara was related to Tsafiki (cf. section 2.17).
Among the most frequent endings in place names is -qu´ı [k´ı] (compare Pasto -quer), for
instance, in Cochasqu´ı, Pomasqu´ı and Tuntaqu´ı. Caillavet identifies further toponymical
elements, such as -puela/-buela ‘field’, -pigal ‘earth ridge’ and -yasel ‘artificial mound
(locally known as tola)’. A recurrent element in personal names is ango, identified as
‘lord’. It is found, for instance, in the name of the sixteenth-century cacique of Otavalo,
Otavalango,but it also occurs by itself (Caillavet 2000: 28). Unfortunately, there are
hardly any Cara forms of which the meaning is known with certainty. One of the very
few cases is the element -piro/-biro, identified as ‘lake’ or ‘pond’ in the Relaciones
geogr´aficas,which is found in the place names Pimampiro and Tumbabiro (Jim´enez de
la Espada 1965, II: 236, 248).
186
PazyMi˜no (1961a) refers to an (extinct) Cara population on the Ecuadorian coast, near
Car´aquez bay (Manab´ı). Presumably, this idea was inspired by Velasco’s controversial
account of the Cara or Scyri rulers, who would have invaded the highlands from the
coast in pre-Inca times (Velasco 1789). Regardless of the question whether Velasco’s
account has any historical basis, there is no reason to assume that the Cara language
as we know it from toponymy and historical sources was the product of such a coastal
invasion.
The Panzaleo language owes its name to the Inca settlement of the same name,
situated not far from Quito (cf. Salomon 1986). The Panzaleo area extended from Quito
in the north to the town of Mocha in the south, covering the provinces of Cotopaxi
and Tungurahua, as well as the southern part of Pichincha. The main centres in the
area, apart from Quito itself, are Ambato and Latacunga. The Panzaleo area may have
been the first in the Ecuadorian highlands to become thoroughly Quechuanised. Typical
toponymical endings of the Panzaleo area are -(h)al´o (Pilal´o, Mulahal´o), -leo (Tisaleo,
Pelileo) and -lagua/-ragua (Cutuglagua, Tunguragua). The possibility that the original
language of the Quijos Indians, an ethnic group inhabiting the Andean foothills in the
western part of the province of Napo, was actually a variety of Panzaleo (as suggested
by Cieza de Le´on) is discussed in Jij´on y Caama˜no (1940: 289–95). The evidence is not
186
Translated as ‘great lake’ and ‘bird lake’, respectively. Note, however, that pim´an means ‘bridge’
in Tsafiki (Moore 1962).
3.9 Extinct and poorly documented languages 395
conclusive. Loukotka (1968) in his classification presents Panzaleo as related to P´aez
(see section 2.15). One of the sources for this view is Jij´on y Caama˜no (1940: 395–6),
who recognises that the evidence is weak and adds that any observed similarities, when
valid at all, may be attributed to contact.
Puruh´a or Puruguay was once the dominant language in the present-day province of
Chimborazo, except in its southernmost section of Alaus´ı and Chunchi, where Puruh´a
overlapped with neighbouring Ca˜nar. Jij´on y Caama˜no (1940: 417) leaves open the
possibility of an extension into the Chimbos area further west (province of Bol´ıvar),
but this idea is not accepted by Paz y Mi˜no (1942). The main centre of the former
Puruh´a area is Riobamba. In spite of this strategic position, the amount of information
on the language that has been preserved is disappointingly small. A late seventeenth-
century grammar is reported lost. Modern Puruh´a descendants now generally speak
Quechua.
Family names frequently end in -cela or in -lema. Examples are Duchicela, the lineage
of Atahuallpa’s mother, reportedly of Puruh´a descent (Velasco 1789), and Daquilema,
the name of a nineteenth-century rebel. Frequent endings of Puruh´a place names are -shi
(e.g. Pilligshi), -tus (e.g. Guasunt´us), -bug (e.g. Tulubug). More complex endings are
-cahuan, -calpi and -tactu. Some of these elements occur in combination with Quechua
roots, as in Supaycahuan (Quechua supay ‘devil’).
Puruh´aisgenerally believed to be related to Ca˜nar. Nevertheless, the two languages
are kept distinctly apart in the colonial Spanish sources. Some phonological character-
istics not usually found in the languages further north are shared by Puruh´a and Ca˜nar,
for instance, the use of voiced stops in word-initial position. Another interesting com-
mon feature is the existence of a voiced palatal fricative [ˇz] (written zh), which is not
demonstrably a reflex of earlier l
y
(as it is in some Ecuadorian Quechua dialects). It is
found word-initially in Ca˜nar place names, e.g. Zhud (R. Howard, pers. comm.).
The Ca˜nar language was spoken in the Ecuadorian highlands, in the provincesof Ca˜nar
and Azuay, and in parts of southern Chimborazo (Alaus´ı). According to the Relaciones
geogr´aficas de Indias (Jim´enez de la Espada 1965, II: 279), it was also used in Saraguro
in the northern part of the province of Loja, in spite of the fact that the Indians of Saraguro
were hostile to the Ca˜nar people further north. The Ca˜nar people played a crucial role
in the history of the Inca empire. After they had been conquered, the Incas embellished
their principal town of Tomebamba (Cuenca), which was then destroyed during the
civil war between Atahuallpa and Huascar. Many Ca˜nar people were sent to other parts
of the empire as mitimaes.Ca˜nar troops were sought out for police tasks and special
assignments, which they continued to perform under Spanish rule. The influence of
Quechua became particularly strong in the Ca˜nar region. Nevertheless, Ca˜nar continued
to be used during the colonial period until it was eventually replaced by Quechua. Today,
the main centres of the area are Cuenca, Azogues and Ca˜nar.
396 3 The Inca Sphere
Ca˜nar toponymy is highly characteristic and easier to recognise than that of Puruh´a.
Nevertheless, some endings (-pala, -pud,-bug, -shi) are shared by the two languages.
Luckily, the Relaciones geogr´aficas de Indias (Jim´enez de la Espada 1965, II: 265, 274–
5, 278, 280) are somewhat more informative about the meaning of Ca˜nar place and family
names than about those of other Ecuadorian highland languages. For instance, Leoquina
is translated as ‘snake in a lake’; Paute as ‘stone’; Peleusi (or Pueleusi)as‘yellow field’;
Pue¸car as ‘broom’. The Jubones river was called Tamalaycha or Tanmalanecha ‘river that
eats Indians’. A particularly interesting explanation concerns that of the plain in which the
city of Cuenca is situated. It was called Guapdondelic ‘large plain resembling heaven’.
This explanation allows an interpretation of quite a few place names, such as Bayandeleg,
Chordeleg, etc., which contain the element deleg ‘large plain’. Other place names contain
an element del (e.g. in Bayandel), which may have had a related meaning.
187
The most
frequent ending in place names is -cay [kay]. It has been interpreted as ‘water’ or ‘river’,
presumably because of its occurrence in river names, such as Saucay and Yanuncay
(Paz y Mi˜no 1961b). Other examples of endings occurring in Ca˜nar place names are
-copte (Chorocopte), -hui˜na (Catahui˜na), -turo (Molleturo), -zhuma (Guagualzhuma)
and -zol (Capzol). In addition to the original Ca˜nar place names, Quechua names are
also frequent. The ending -pamba ‘plain’, for instance, is more common than the native
-deleg.
In the sixteenth century the province of Loja was inhabited by a number of small ethnic
groups, which the Spaniards found difficult to control. The language they spoke was
known as Palta. Another language, Malacato,was spoken in a limited area south of the
town of Loja. Caillavet (2000: 233) also mentions an enclave of Ca˜nar speakers (Amboca)
situated to the northwest of that town. The Palta language had a wide distribution with
eastern outliers in the Zamora region and in the area of Ja´en (department of Cajamarca,
Peru).
188
Furthermore, its use may have extended to the area of Zaruma in the present-
day province of El Oro (Caillavet 2000: 216). Today the area, which no longer has a
substantial indigenous population, is mainly Hispanophone, with the exception of the
Quechua-speaking community of Saraguro in the north.
Several observations in the Relaciones geogr´aficas de Indias seem to indicate that the
languages of the area were closely related. It is generally believed that they belonged to
the Jivaroan language family. The only Palta words that were recorded are not from Loja,
but from a mountainous area called Xoroca not far from Ja´en (Jim´enez de la Espada
1965, III: 143). Of these four words (yum´e ‘water’, xeme ‘maize’, capal ‘fire’, let ‘fire-
wood’) the three first ones have been recognised as Jivaroan (Gnerre 1975). Many place
187
Bay´an is the name of a shrub.
188
On Ecuadorian maps the area of Ja´en is normally represented as a part of Ecuador, as a result
of the boundary conflict between the two countries.
3.9 Extinct and poorly documented languages 397
names in the Loja area end in -anga (Cariamanga), in -numa (Purunuma) or in -nam´a
(Cangonam´a, Gonzanam´a). Gnerre (1975) interprets the two latter endings as instances
of a Jivaroan locative case marker -num ∼ -nam,whereas Torero (1993a) offers a dif-
ferent interpretation by associating the ending -nam´a with Aguaruna nam´ak(a) ‘river’.
189
Similar elements are found in place names further east in the area of Zamora, where
the Relaciones geogr´aficas de Indias (Jim´enez de la Espada 1965, III: 136–42) mention
the existence of three other languages, Bolona, Rabona and Xiroa (possibly a variant
of J´ıvaro). For the Rabona language a certain number of vocabulary items, mainly
plant names, are listed. Torero (1993a) has identified some of these items as Candoshi
(e.g. chicxi ‘sapodilla’, Candoshi ˇc´ıˇci(ri)), but others are reminiscent of Aguaruna (e.g.
guapuxi ‘large guava’, Aguaruna w´ampuˇsik,Wipio Deicat 1996: 140). In spite of the
fact that Rivet (1934), Loukotka (1968) and Torero (1993a) have classified the Rabona
language as a member of the Candoshi family, we prefer to leave the matter undecided.
The affiliations of the Bolona language cannot be known for lack of data; it has been
classified as Jivaroan by Loukotka, whereas Torero suggests a connection with Ca˜nar
based on the geographical location of Bolona. Few of the communities of the Zamora
region mentioned in the early sources have survived, as much of the area was devastated
by intensive colonisation during the sixteenth century and by the subsequent Jivaroan
uprising (cf. Velasco 1789; Taylor 1999).
As a final addition to this discussion of the pre-Quechuan Ecuadorian highland lan-
guages, we may observe that, with the exception of the languages of Loja and possibly
also of Panzaleo, these languages share a number of distinctive elements that seem to
point at a common origin or a long period of interaction at the least. Examples of such
elements are words ending in the voiced consonants -d and -g, the ending -pud, and sylla-
bles with a complex labial onset (mue, pue, bue).
190
The fact that Pasto, with its probable
Barbacoan affinity, is one of the languages exhibiting these shared characteristics seems
to suggest a Barbacoan connection for this whole group of languages. Although far
from being proven, such a connection is more likely than the alleged relationship with
Mochica advocated by Jij´on y Caama˜no.
3.9.2 Northern Peru
In seventeenth-century sources the northern Peruvian Andes and coast are singled out
as a multilingual area that resisted Quechuanisation. Blas Valera, cited in Garcilaso de
189
The usual meaning of nam´ak(a) in the Jivaroan languages is ‘fish’, but in Aguaruna it has the
additional meaning of ‘river’.
190
Complex labial onsets are also found in Awa Pit (Curnow 1977: 34), where labial consonants
have a labial off-glide before the high central vowel ([p
w
], [m
w
]). Allophonic labial off-glides
following labial consonants have furthermore been encountered in Muisca (section 2.9.2), in
Shuar and in Yanesha
(chapter 4).
398 3 The Inca Sphere
la Vega (1609, Book 7, chapter 3), reports that the Quechua language was ‘unknown in
the administrative domain of Trujillo and other provinces belonging to the jurisdiction
of Quito’. The present situation and the little we know of past linguistic developments
in northern Peru give support to Valera’s impression. A number of non-Quechuan native
languages survived the conquest, giving way to Spanish often as late as in the nineteenth
and twentieth centuries. On the other hand, Quechua is not absent from the area. It has
survived in three distinct areas of northern Peru (see sections 3.1, 3.2.3), and it was
certainly more widely used in the sixteenth century (Rivet 1949: 2–3).
The near absence of documentation concerning the non-Quechuan languages of north-
ern Peru is striking. Except for the cases of Mochica and Chol´on, there are neither gram-
mars, nor dictionaries of these languages, not even the sort of religious texts that Spanish
priests considered necessary for evangelisation (cf. Adelaar 1999). An essential source
is Bishop Mart´ınez Compa˜n´on (1985 [1782–90]), who provided word lists of 43 items
each for nine languages spoken in his diocese (figure 1): Castilian (Spanish); Quechua;
the Yunga language of the provinces of Trujillo and Sa˜na (Mochica, see section 3.4);
the languages of Sechura, Col´an and Catacaos in the province of Piura (Sechura and
Tall´an); the Culli language of the province of Huamachuco; and the languages of the
Hivitos and Cholones of the Huailillas missions (Hibito and Chol´on, see section 4.11).
The most comprehensive attempt to reconstruct the colonial language map of northern
Peru is found in Torero (1986, 1989, 1993a). In this context a distinction should be
made between languages effectively mentioned in the sources and those whose former
existence is merely assumed on the basis of clusters of place and family names with
shared characteristics.
Among the well-attested languages are those of the coastal plain of the department of
Piura. One of these languages, known as Sechura, has been associated with the port of
Sechura at the mouth of the Piura river. A second language or language group, generally
kown as Tall´an,was found along the Chira river, including the coastal towns of Col´an and
Paita, and along the middle course of the Piura river, where it was spoken, among other
places, in the important indigenous community of Catacaos (near the town of Piura).
Ramos Cabredo (1950) provides an extensive list of place names, family names and
native words still in use in the Tall´an area. Typical endings for place names are -l´a,
-r´a (e.g. Narigual´a, Tangarar´a) and -ura (as in Nonura, Piura, Sechura). Family names
often end in -lup´u (Belup´u, Sirlup´u), in -bal´u∼-gual´u (Cutibal´u, Mangual´u), in -naqu´e
(Lequernaqu´e, Yamunaqu´e) and in -cherre (Pacherre, Tupucherre).
The word lists included in Mart´ınez Compa˜n´on exemplify the so-called ‘languages of
Col´an and Catacaos’, as well as ‘the language of Sechura’. The Col´an and Catacaos lists
represent closely related varieties, possibly dialects of the same language. A notable
difference between the two is that Catacaos often features a final element -chim on
nouns, whereas the Col´an equivalents only have a nasal (e.g. Catacaos puruchim, Col´an
Figure 1 Mart´ınez Compa˜
n´on’s word lists of 43 items each for nine languages spoken
in his diocese. Copyright
C
Patrimonio Nacional, Madrid
399
400 3 The Inca Sphere
puru
˜
m ‘sister’). Another interesting correspondence is phonological. Col´an syllable-
initial dl (possibly a voiced lateral affricate) is either d or l in Catacaos (e.g. Col´an
dlacati, Catacaos lacatu ‘to die’; Col´an dlur˜um, Catacaos durum ‘earth’).
191
Mart´ınez
Compa˜n´on’s Sechura list represents a separate language, because only about 30 per cent
of the vocabulary items show a similarity with either Col´an or Catacaos (Torero 1986:
532). Among the possible cognates we find Sechura lactuc, Col´an dlacati, Catacaos
lacatu ‘to die’; Sechura colt, Catacaos ccol ‘meat’; Sechura ˜nangru, Col´an nag ([na
ŋ
]?),
Catacaos nam ‘moon’; Sechura fic, Catacaos vic ‘wind’; Sechura yaibab, Col´an yaiau,
Catacaos yeya ‘bird’.
In 1863 Richard Spruce obtained from a native speaker in Piura a list of 38 words
which was later published by von Buchwald (1918). It has been identified as a sam-
ple of the Sechura language on the basis of a comparison with Mart´ınez Compa˜n´on’s
word lists (Torero 1986: 542). The available data suggest that Sechura was a suffixing
language. Kinship terms often contain the elements -ma and -˜ni (e.g. Spruce ˜nosma,
Mart´ınez Compa˜n´on ˜nos˜ni ‘son or daughter’), which were possibly possessive endings.
Sechura verbs as listed in Mart´ınez Compa˜n´on usually end in -(u)c (e.g. unuc ‘to eat’, nic
‘to cry’).
Rivet (1949) treats Sechura and Tall´an as a single language, which he calls Sek,
adopting the denominaton Sec introduced by Calancha (1638). This procedure is justly
rejected by Torero (1986). However, the possibility of a language family, including
both Sechura and Tall´an (Col´an–Catacaos), must remain open, considering the limited
number of vocabulary items that are available.
A possible extension of the Sechura language was found in the desert oasis of Olmos
in the present-day department of Lambayeque. When referring to the Sec language,
Calancha observed that the people of Olmos ‘change letters and endings’ (mudan letras
y finales). Although the name ‘Sec’ could also have referred to the Tall´an language,
later ethnographic research by Br¨uning (1922) gives support to the idea of a special
connection between Olmos and Sechura suggested by oral tradition and the coincidence
of weaving terminology. For a detailed discussion see Torero (1986), who ventures the
idea that a Callahuaya-type situation may have existed in Olmos (cf. section 3.5). Finally,
the linguistic unity of Olmos and Sechura–Tall´an is confirmed in an archival document
of 1638 enumerating the administrative and religious divisions of the northern Peruvian
coastal region (Ramos Cabredo 1950: 53–5).
The original language situation in the interior highland part of the department of
Piura with its centres Ayavaca and Huancabamba is difficult to reconstruct. Today its
predominantly indigenous population is reported to be Spanish-speaking. During the
colonial period Quechua may have been the dominant language in that area, as was
191
It may be assumed that the difference between -u
˜
m and -
˜
umreflects an inconsistency in Mart´ınez
Compa˜n´on’s orthography.
3.9 Extinct and poorly documented languages 401
also the case further east in Tabaconas, in the area descending towards the Amazon
(Torero 1993a), and in the Huambos area further south in the department of Cajamarca.
A systematic investigation of place names and family names in eastern Piura remains a
task for the future.
Another well-attested language in northern Peru is Culli (or Culle). Culli was once the
dominant language in the Andean hinterland of the coastal town of Trujillo. Its former
area comprises four provinces of the department of La Libertad (Julc´an, Otuzco,
Santiago de Chuco and S´anchez Carri´on), the province of Cajabamba (department of
Cajamarca) and the province of Pallasca (department of Ancash). Pallasca was part
of the colonial province of Conchucos,which also included Quechua-speaking areas
in northern and eastern Ancash. The Culli translation of Conchucos is ‘water land’
(co˜n ‘water’, chuco ‘land’). The Mara˜n´on river valley constituted the eastern border
of the Culli-speaking area. To the southwest it bordered on the Quingnam language of
the Trujillo coastal plain (see section 3.4). Further extensions of the Culli language in
Ancash and Cajamarca, as well as the possibilities of overlap with other languages, are
controversial.
The main historical centre of the Culli-speaking people was the town of Huamachuco,
now the capital of S´anchez Carri´on province. It was the seat of an important religious cult
to the creator god Ataguju and the thunder god Catequil. In spite of violent persecution
by the Incas, the Huamachuco cult was still vigorous when the Spaniards arrived. The
Augustinian friars Juan de San Pedro and Juan del Canto, who soon after the conquest
attempted to convert the local population, wrote a detailed account of local beliefs and
customs, in which a Huamachuco language and several native names and terms are
mentioned (Castro de Trelles 1992). Towards the end of the sixteenth century a travel
account by Toribio de Mogrovejo, archbishop of Lima, of his visit to the area refers to
a language called Linga or Ilinga (cf. Rivet 1949). Notwithstanding the fact that these
terms seem to represent modified versions of lengua del Inga ‘language of the Inca’,
the geographic setting indicates that only Culli could be meant. The earliest mention of
the name Culli itself is found in the abovementioned document of 1638 published in
Ramos Cabredo (1950). For a historical overview of the Culli language and its speakers
see Silva Santisteban (1982).
The primary sources for Culli are two word lists. One list of 43 words, referring to the
Culli language of the province of Huamachuco, is found in Mart´ınez Compa˜n´on (1985
[1782–90]), whereas the other was collected around 1915 in a hamlet called Aija near
Cabana in the province of Pallasca by a local parish priest, father Gonz´alez. This list of
19 words was published in Rivet (1949: 4–5).
192
In recent years, a number of students
from the area (Flores Reyna 1996; Andrade Ciudad 1999; Cuba Manrique 2000; Pantoja
192
We have obtained the information concerning the place of origin of the Gonz´alez list from the
field notes of Walter Lehmann, kept at the Ibero-American Institute in Berlin.
402 3 The Inca Sphere
Alc´antara 2000) have succeeded in expanding the lexical data base of the Culli language
by collecting words and expressions still in use today. Flores Reyna reports that Culli was
spoken by at least one family in the town of Tauca (Pallasca province) until the middle
of the twentieth century. Although Culli has been replaced by Spanish, the possibility
that speakers survive in remote villages cannot be ruled out entirely (cf. Adelaar 1988).
The ethnolinguistic atlas published by Chirinos Rivera (2001) contains several mentions
of municipalities in the Culli area for which the census of 1993 reports the existence of
speakers of an unidentified native language.
In spite of the extremely scanty data it is possible to tentatively determine a number of
characteristics of the Culli language. In compounds a modifying element precedes the
head, like in Quechua. Verb forms as given by Mart´ınez Compa˜n´on all end in a suffix -`u
([u]∼ [w]), e.g. canqui`u ‘to laugh’, ˇcollap`u ‘to die’ (see below for a possible interpreta-
tion of the diacritic on ˇc ). The language may have had prefixes for possessive personal
reference considering the fact that several kinship terms feature an initial velar stop
(viz. quin`u ‘father’, quimit ‘brother’, ca˜ni ‘sister’). Possibly, the same element is found
in quiyaya,aform of ritual singing, which may contain the word yaya ‘god’.
193
Like
Quechua and Aymara, the language may have had a three-vowel system (a, i, u). The oc-
currence of mid vowels (e, o)islimited and seems to be favoured by the neighbourhood
of c, qu, g,nomatter if the latter occur in clusters or by themselves (e.g. mosˇc´ar ‘bone’,
ogˇoll ‘child’, co˜n ‘water’; Quesquenda ‘a place name’). It suggests that the language
may have had a velar–postvelar opposition, an assumption fed by the fact that Mart´ınez
Compa˜n´on often uses a haˇcek-style diacritic when c and g are adjacent to a mid vowel or
a. Other instances of mid vowels (especially o) are found in the environment of r (e.g. in
the place name Choroball) and in contexts in which vowel harmony appears to play a role
(e.g. in Chochoconday, the name of a mountain). Variation between mid and high vowels
is frequent in place names (e.g. Sanagor´an, Candigur´an; Corgorguida, Llaugueda). A
characteristic feature of Culli is that the second member in a compound often begins
with a voiced stop, producing internally dissimilar consonant clusters when the first
member ends in a voiceless consonant (for instance, in place names such as Chiracbal
and Ichocda). If the same element occurs independently or in initial position it has a
voiceless stop. Compare, for instance, Chusg´on (a river name) and Conchucos (both con-
taining the element con∼-gon ‘water’); also Parasive and Pushvara.Frequent endings of
Culli place names are -day∼-tay ‘mountain’, -chugo∼-chuco ‘earth’, -gon∼-go˜n ‘water’,
-gueda ∼-guida∼-queda ‘lake’, -gur´an∼-gor´an ‘river’, -pus ∼-pos ‘earth’; the elements
193
The analysis of qu(i)- as a prefix is, however, contradicted by the fact that some early Culli
family names contain the element quino (Pantoja Alc´antara 2000). In addition, Torero (1989:
225) recognises the element quin`u in the divinity name paiguinoc ‘lord of the guinea-pigs’
mentioned in the Augustinian chronicle. Note also that quiyaya appears as quillalla in Mart´ınez
Compa˜n´on.
3.9 Extinct and poorly documented languages 403
-bal∼-ball, -bara∼-huara∼-vara, -chal, -da∼-ta, -gal∼-galli∼-calli, -ganda, -maca,
-sicap∼-s´acap(e) ∼-ch´acap(e), all of undetermined meaning; and the borrowed ele-
ments -malca ‘village’ and -pamba ‘plain’ (from Quechua -marka and -pampa, respec-
tively). For an extensive list of Culli place names see Torero (1989). Place names com-
bining a Culli element with a Quechua or a Spanish element (Mumalca,
194
Naopamba,
Cruzmaca) are frequent and indicate that these languages must have coexisted for a
considerable time.
From a phonological point of view, it is possible to distinguish several dialect areas.
Place names in the province of Pallasca and neighbouring parts of Santiago de Chuco ex-
hibit endings with the palatal final consonants -˜n [n
y
] and -ll [l
y
] (e.g. Acogo˜n, Camball)
whereas non-palatal endings appear elsewhere (e.g. Chusg´on, Marcabal). The distribu-
tion of palatal and non-palatal endings coincides with the geographical distribution of
the lexical endings -day and -ganda as noted by Torero (1989: 226).
The Culli word lists show a relatively high incidence of Quechua loans (e.g. aycha, Qu.
ayˇca ‘meat’; challuˇa, Qu. ˇcal
y
wa ‘fish’; cuhi, Qu. kuˇsi ‘pleasure’). The item mi`u ‘to eat’
appears to reflect a hypothetical pre-Quechua *mi-,which can be reconstructed on the
basis of modern Quechua miku- ‘to eat’ and miˇci- ‘to herd’. Other items, however, show
no relationship with Quechua whatsoever and suggest that the language is genetically
independent. Similarities with the Piuran languages and Mochica are also limited (cum`u,
Col´an c˜um ‘to drink’; chuip, Catacaos chupuchup ‘star’; paihaˇc, Mochica pey ‘herb’).
They may be due to contact.
The department of Cajamarca has a dense, predominantly indigenous population. The
accounts of colonial inspection visits (visitas)tothis area (Remy and Rostworowski de
Diez Canseco 1992) feature a wide variety of family names reflecting the existence of
several unidentified languages. The toponymy is also highly varied and exotic. Never-
theless, only Quechua has survived in a small number of well-defined localities. Interest-
ingly, Cajamarca Quechua (Quesada Castillo 1976a, b) contains a substantial amount of
non-Quechuan vocabulary. In Pantoja Alc´antara’s study of the Culli linguistic heritage
in the Spanish of Santiago de Chuco, an area quite distant from Cajamarca, the follow-
ing correspondences between Santiago de Chuco Spanish and Cajamarca Quechua have
come to light:
Santiago de Chuco Spanish Cajamarca Quechua
ˇcukake ˇsukaki ‘headache with nausea’
ˇcurgap(e) ˇcurʁap ‘cricket’
dasdas dasdas ‘hurry up!’
inam inap ‘rainbow’
194
The meaning of mu is ‘fire’; the other Culli elements have not been identified.