Adelaar Willem, Muysken Pieter. The languages of the Andes
Подождите немного. Документ загружается.
214 3 The Inca Sphere
as Bolivian Quechua -qpa, occur after (short) vowels and reflect an original postvocalic
allomorph *-p.Inthe Quechua IIB branch -pa is sometimes replaced by reflexes of
the benefactive case marker -paq.InEcuador *-pa and *-paq have merged as -pax
(< *-paq). In addition, the benefactive marker -paq remains in use in virtually all dialects,
including those exhibiting the -pa/-paq alternation. Both the accusative (in Quechua II)
and the genitive case markers (in Quechua I) are used for indicating adverbial use of
adjectives.
The genitive case marker occurs on the dependent (possessor) member of a possessive
construction, whereas the (possessed) head is marked for third person. Only when the
possessor is a pronoun can the head be marked for first, second or fourth (inclusive)
person under a condition of agreement. It follows that in a full possessive construc-
tion the members are obligatorily double-marked; cf. the examples (51) and (52) from
Ayacucho.
(51) runa-pa wasi-n
man-G house-3P
‘a person’s house’
(52) qam-pa wasi-ki
you-G house-2P
‘your house’
Spatial case marking is limited to a general indication of the direction of a motion or
the absence of it. The locative case marker indicates location in rest regardless of the
nature of the position with respect to the case-marked object. The shape of the locative
marker is not dialectally uniform. It is -pi in most of Quechua II and -ˇc
.
aw (or its reflex)
in Quechua I (cf. section 3.2.3). Some dialects, in the area of Yauyos (department of
Lima) use -pa, rather than either -pi or -ˇc
.
aw for locative. This practice could be due to
language contact, judging from the fact that in Aymara the genitive and locative case
markers coincide as well (cf. section 3.3.6).
Two case markers can be used to indicate motion towards a goal. It is one of the
functions of the accusative marker -(k)ta and the principal function of the ‘allative’
marker -man. The functional demarcation between the two markers does not always
coincide in the dialects. For -(k)ta to be used the subject must be human or humanised,
the action must be deliberate and the goal must have a fixed position. Non-human subjects
and moving or non-local goals require the use of -man. The examples (53) and (54), also
from Ayacucho, illustrate this.
(53) yaku-man ri-ni
water-AL go-1S
‘I go for water (wherever I can find it).’
3.2 The Quechuan language family 215
(54) kay n
y
an-qa ayakuˇcu-man ri-n
this road-TO Ayacucho-AL go-3S
‘This road goes to Ayacucho.’
The directional use of the two case markers just discussed is extended metaphorically
to the function of indirect-object marking. In the southern Quechua II dialects there is
a tendency to use -man for indirect objects and -(k)ta for direct objects, whereas the
central Peruvian Quechua I dialects use -(k)ta for both purposes.
A special directional case marker -kama conveys the meaning ‘until’ both in a temporal
and in a spatial sense. This marker is related to the verb kama- ‘to fit’, ‘to animate’ (a
religious concept).
The concept of separation (‘ablative’) is indicated by means of a case marker which is
variable in shape. Most Quechua II dialects have -manta (or reflexes of it). The Quechua I
dialects use (reflexes of) -pita,-piq and -piqta (cf. section 3.2.3).
There can hardly be any doubt that the element ta in -manta,-pita and -piqta is histo-
rically related to the accusative marker -(k)ta,whereas man, pi (possibly also piq) can
be traced to the allative and locative case markers, respectively. From a synchronic point
of view, the inner structure of the ablative case marker is no longer relevant. Along
with its main function of referring to separation, it can convey other meanings such as
a topic of conversation or reflection (‘about’), or it can mark the second member in a
comparison.
There is no special case marker for the concept of motion through a space (‘perlative’).
For this purpose the genitive is used in some Quechua I dialects, and a combination -n-ta
in southern Peruvian Quechua II, as in Ayacucho:
(55) urqu-n-ta ri-n
mountain-3P-AC go-3S
‘It goes through the mountains.’
Another case marker which plays a central role in Quechua grammar is -wan.It
refers to the instrumental or comitative case (‘with’), but has the additional function of
coordinating two noun phrases (‘and’). The coordinating function of -wan is compatible
with other case markers. Interestingly, the instrumental and coordinative functions of
-wan cannot be separated easily. Of two noun phrases in a coordinative construction only
one must be marked with -wan,whereas the other need not be expressed lexically, as in
the following example from Tarma:
(56) punu-ˇci-ma-ra-ygi xuk wamra-ta-wan
sleep-CA-1O-PA-2S another girl-AC-IS
‘You made me sleep with another girl.’ (lit. ‘You made me and
another girl sleep.’)
216 3 The Inca Sphere
The instrumental case marker is frequently used to identify the causee in causative
constructions, in particular when the causative verb has a transitive base, as in (57) from
Ayacucho:
(57) parqu-ˇci-ni ˇcakra-ta pedru-wan
irrigate-CA-1S field-AC Peter-IS
‘I have the field irrigated by Peter.’
(Soto Ruiz 1979: 340)
There are a small number of other case markers, not found in every dialect of Quechua.
A frequent marker is -rayku ‘for the sake of’. Likeness is expressed by means of a case
marker -naw (-nuy,-nu:) in Quechua I. In Quechua II the notion of ‘like’ is expressed
by means of the adverb (or postposition) hina. Another widespread marker is -pura
‘among’.
As a consequence of their limited number, most Quechua case markers are multi-
functional. For each grammatical case a list of uses is required. Sequences of case
markers occur to a limited degree. Most often they involve either the genitive marker
(e.g. Ayacucho bisinti-pa-ta ‘to Vincent’s’, Parker 1969a: 44), or the instrumental marker
-wan in its coordinative and comitative functions. In addition, the case markers -kama
and -naw can also be found in combinations. In combination with specific categories
of nominalisation, the case markers show a tendency towards acquiring specialised
functions (see below).
Nominal suffixation is not limited to case, number and personal reference. Some
Quechua IIC dialects, such as Ayacucho, have diminutive and, occasionally, augmenta-
tive affixes (Parker 1969a: 60):
(58) wasi-ˇca
house-DI
‘little house’
(59) wasi-su
31
house-AU
‘big house’
Virtually all Quechua dialects have nominal derivational suffixes referring to own-
ership in a very broad sense. A very common suffix is -yuq (-ni-yuq after consonants,
diphthongs and long vowels), to be translated as ‘owner of’, ‘having’, or ‘belonging to
a place, a community’:
31
The suffix -su is said to be derived from the ending -azo in the local Spanish (e.g. perrazo ‘a
huge dog’). If this is true, it is probably the sole morphological element borrowed from Spanish
into Ayacucho Quechua. It has not been reported in other dialects.
3.2 The Quechuan language family 217
(60) wasi-yuq
house-OS
‘house owner’, ‘someone who has a house’
Other frequent suffixes are -sapa ‘owner of many’, ‘owner of something big’ and
-nti(n) ‘including’, ‘with...andall’. In some of the central Peruvian dialects (Hu´anuco,
Pacaraos) a suffix -nnaq or -:naq conveys the meaning ‘without’, ‘owner of none’. It
was also found in colonial Cuzco Quechua, and it may, therefore, be reconstructed for
the common proto-language. Example (61) is from Pacaraos.
(61) wawi-:naq
child-LA
‘childless’, ‘having no children’
Definiteness is not a general morphological category in the Quechua languages. How-
ever, the Huanca dialects have developed an affixed definite article, comparable to those
found in Rumanian or Swedish. The suffixal character of this marker of definiteness
makes it unlikely that its existence could be explained through the contact with Span-
ish. Formally, the Huanca definite marker has been derived from *ka-q, the agentive
nominalised form of the verb ka- ‘to be’. In the Chongos Bajo subdialect of Huanca its
reflexes are -ka: between consonants, -ka in word-final position and vowel length (-:)
after non-final short vowels, e.g.:
(62) wamla-kuna-ka
girl-PL-DF
‘the girls’
A characteristic feature of many Andean languages is the existence of a suffix referring
to limitation, which marks a noun as trivial, limited in number or size, or close in
distance to the speaker. The Quechua suffix which has this function, -l
y
a (-la, -la:-), may
originally have been an independent suffix. It became part of the nominal morphology
and, in Quechua I, also part of the verbal morphology. There are rather complex rules
determining the location of -l
y
a in relation to case, number and personal reference
markers in nouns. Whatever the status of this suffix, its semantic interpretation is stable.
In Andean Spanish it is characteristically reflected by the expressions no m´as or nada
m´as ‘no(thing) more’.
As in many other native American languages, the verb constitutes the richest part of
the morphology in Quechua. Verbal morphology is contained within a general frame-
work of formal restrictions, which are the following. All verb roots end in a vowel.
32
They
32
In dialects that distinguish vowel length root-final low vowels can be long underlyingly. In verbs
showing this characteristic length surfaces whenever the phonological context permits it, e.g.
(Tarma) ˇc
.
a:-ˇsun ‘we shall reach’, but ˇc
.
a-nki ‘you (shall) reach’.
218 3 The Inca Sphere
do not occur by themselves but must be followed by at least one suffix that qualifies
the verb as a syntactically usable unit. Suffixes that can fulfil this function are per-
sonal reference (subject) markers, whether or not in combination with tense and mood
markers.
Averbal root may be followed either by one, or by a string of affix extensions, which
all share the formal peculiarity of ending in a vowel. Like roots, these extensions cannot
occur word-finally but must be followed by one of the endings required to assure the
syntactic use of the verb form.
33
These root-extending affixes are commonly referred to
as derivational or modal suffixes in the literature on Quechua. Although they are subject
to shared formal restrictions (most of them are either CV or CCV), they represent a
wide array of functions and meanings ranging from changes in syntactic valence and
argument structure to subtle semantic and pragmatic shifts. As a rule, there is a one-to-
one relation between form and function, but certain combinations of suffixes can acquire
new and idiosyncratic meanings. Combinations of a verbal root and a derivational suffix
may become lexicalised as idiomatic units.
Tense, mood and personal reference markers occupy the final block in a Quechua verb
form. Full verb forms must contain one of the suffixes in question or a combination of
several. For the Quechua IIB and IIC dialects the number markers are to be added to the
above list, as they operate in close connection with the personal reference markers. The
place of tense and mood markers can also be occupied by nominalising and subordinating
(switch-reference)affixes.
In Quechua verbs personal reference is built on the same distinctions as for their
nominal counterparts. It consists of a four-term system based on the inclusion of speaker
and addressee. However, an additional dimension is brought in through the fact that not
only the subject, but also a (human) object can be identified for person. Third-person
objects remain unexpressed, as in (63) and (64) from Ayacucho:
(63) miku-nki-ˇcu
eat-(3O-)2S-IR
‘Do (did) you eat?’, or ‘Do (did) you eat it?’
(64) riku-n-ku
see-(3O-)3S-PL
‘They see (him/her/them/it).’
First-, second- and fourth-person objects that are expressed in the verb form can
have either an indirect or a direct-object function depending on the semantics of the
verb base. There are four endings involving subject reference only and an additional
33
The single exception may be the serial verb yal
y
i- ‘to exceed’ in Ecuadorian Quechua, which
occurs uninflected in comparative constructions.
3.2 The Quechuan language family 219
Table 3.4 Subject conjugation in Ayacucho Quechua
Present Future Imperative
1 pers. subject wata-ni wata-saq
2 pers. subject wata-nki wata-nki wata-y
3 pers. subject wata-n wata-nqa wata-ˇcun
4 pers. subject wata-nˇcik wata-sun wata-sun
five combinations that involve object marking as well. These combinations were called
transiciones (‘transitions’) by the Spanish colonial grammarians, a term still used in
many of today’s traditional-style grammars.
34
The subject endings are relatively stable
throughout the paradigms referring to tense and mood. The imperative mood and the
future tense both have special portmanteau endings. The Ayacucho Quechua subject
paradigms of the verb wata- ‘to tie’ in table 3.4 illustrate this.
Non-future, non-imperative subject endings resemble the nominal possessive endings.
However, in Quechua IIC dialects, such as Ayacucho, first- and second-person subject
endings differ from the possessive endings in all tenses of the indicative mood.
(65) wata-r(q)a-ni uma-y
tie-PA-1S head-1P
‘I tied.’ ‘my head’
(66) wata-r(q)a-nki uma-yki
tie-PA-2S head-2P
‘You tied.’ ‘your head’
In Quechua I there is no such difference between the first-person endings, and in many
of these dialects the second-person subject ending coincides with the possessive ending
in the past tenses. The examples (67) and (68) from northern Jun´ın illustrate this point:
(67) wata-r´a-: um´a-:
tie-PA-1S head-1P
‘I tied.’ ‘my head’
(68) wata-r´a-y(ki) um´a-y(ki)
tie-PA-2S head-2P
‘You tied.’ ‘your head’
34
According to our information, the term transiciones appeared for the first time in the anonymous
grammar published by Antonio Ricardo in 1586. Starting from Gonz´alez Holgu´ın (1607), the
transitions are subject to a numbering system, which is still in use here and there. This numbering
system fails to distinguish between the combinations 3S-10 and 3S-40.
220 3 The Inca Sphere
Of the five endings that conjointly refer to a subject and an object (1S-2O, 2S-1O,
3S-1O, 3S-2O, 3S-4O) four consist of potentially discontinuous sequences of suffixes,
the intervening element being either a tense marker, a nominaliser or a subordinator.
The second component of these combined endings is always identical to one of the
subject endings, although it need not convey the latter’s usual meaning. The shape
of this second component varies in accordance with the tense, mood or nominalisation
paradigm in which it occurs. The initial component in the combination remains the same
in all paradigms, leaving aside predictable morphophonemic variation if any. When the
initial component is -ma(:)- (in Quechua I, Pacaraos and Ferre˜nafe) or -wa- (in most
of Quechua II), its function as a first-person object marker is transparent whenever the
subject is second or third person. (The examples (69)–(74) are from Ayacucho Quechua.)
(69) muna-wa-nki
want-1O-2S
‘You want me.’
(70) muna-wa-n
want-1O-3S
‘He/she wants me.’
However, the first-person object marker -ma(:)-/-wa- can also co-occur with a fourth-
person subject marker, in which case the result is a semantically irregular sequence
referring to a third-person subject acting upon a fourth-person object.
(71) muna-wa-nˇcik
want-1O-4S
‘He/she wants us.
The second-person subject ending can co-occur with an internal suffix -su- (Quechua I
-ˇsu-). The resulting combination refers to a third-person subject acting upon a second-
person object.
(72) muna-su-nki
want-3S.2O-2S
‘He/she wants you.’
The combinations 2S-1O, 3S-1O, 3S-2O and 3S-4O mentioned above are found both
in Quechua I and in Quechua II dialects. The combination 1S-2O is a rather divergent
case. If it exists at all at the morphological level, it is either indicated by means of the
suffix -q or a reflex of it (in Quechua I), or by -yki (in most of Quechua II). For the
future tense there is a special portmanteau ending common to most dialects (*-ˇsqayki >
Ayacucho -s(q)ayki).
3.2 The Quechuan language family 221
(73) muna-yki
want-1S.2O
‘I want you.’
(74) riku-s(q)ayki
see-1S.2O.F
‘I shall see you.’
Several dialects situated at the periphery of the Quechua linguistic area have to a cer-
tain extent regularised the paradigms relating to personal reference marking (Cajamarca,
Ferre ˜nafe, Bolivian Quechua and Santiago del Estero), or lost part or all of it (Ecuador).
As a consequence of this regularising tendency, the suffix -su- became a straightforward
second-person object marker in Ferre˜nafe (Taylor 1994). The same occurred in Santiago
del Estero (Alderetes 1994), as illustrated in (75)–(76).
(75) tapu-su-ni
ask-2O-1S
‘I ask you.’
(76) tapu-su-n-ku
ask-2O-3S-PL
‘They ask you.’
In Quechua I number is indicated by means of derivational suffixes which are inserted
between the root and the personal reference endings (cf. section 3.2.3). In southern
Quechua I dialects there are several pluralisers, whose selection is determined by the
presence of other suffixes, especially aspect suffixes. (The examples (77)–(80) are from
Tarma Quechua.)
(77) wata-ba:ku-n
tie-PL-3S
‘They tie (it).’
(without aspect marker)
(78) wata-rga-ya-n
tie-PL-PR-3S
‘They are tying (it).’
(with progressive aspect)
(79) wata-ra-:ri-n
tie-PF-PL-3S
‘They have tied (it).’
(with perfective aspect)
Normally, the internal pluralisers refer to the number of the subject, but incidentally
they can also indicate the number of an object. Pluralisation of object may occur when
both subject and object are explicitly marked in the verb form and the object is not
222 3 The Inca Sphere
third person. Forms with an internal pluraliser and explicit object marking are therefore
ambiguous.
(80) ˇcay-la-ta ni-ba:ku-x lapa-y-ta
that-DL-AC say-PL-1S.2O all-2P-AC
‘I say that to all of you.’
The Quechua IIC dialects use the same strategy for pluralising verbal personal refer-
ence markers as is found in the nominal conjugation. It has the advantage of providing
an easy way of distinguishing between number of subject and number of object. The
pluraliser -ˇcik (-ˇcis)isreserved for second-person subjects and objects, whereas the plu-
raliser -ku is used for first- and third-person subjects, as well as for first-person objects.
(The examples (81)–(82) are from Ayacucho Quechua.)
(81) muna-su-nki-ku
want-3S.2O-PL.3S
‘They want you.’
(82) muna-su-nki-ˇcik
want-3S.2O-PL.2S
‘He/she/they want(s) you (plural).’
In forms such as (81) and (82) it is not possible to indicate morphologically that both
the subject and the object of the verb are plural. As a rule, the indication of plurality of a
first or second person is considered more essential than that of a third person. Ambiguity
can be removed by the addition of a noun containing a plural marker.
When plural marking is applied to first-person endings, the resulting forms refer to
an exclusive first-person plural. These forms stand in contrast with the fourth-person
category which can only denote an inclusive ‘we’. Ecuadorian Quechua, however, lost
the distinction between inclusive and exclusive first-person plural, as the original inclu-
sive (fourth-person) developed into a general first-person plural combining both func-
tions. As a result, the modern Ecuadorian personal reference system is based on a
three-way person distinction and a two-way number distinction, as in most European
languages.
The Quechua finite verb permits a three-way mood distinction involving indicative,
optative and imperative. The number of tense distinctions within the indicative mood
can vary according to the dialects, the richest system (of seven tenses) being found in
Ancash Quechua. The optative (also called conditional or potential) has two tenses in all
dialects. There are no tense distinctions in the imperative. Nevertheless, a future-tense
form which is not accompanied by an evidential suffix can be interpreted as a polite
command or recommendation.
3.2 The Quechuan language family 223
Verbs that do not belong to any of the three mood options listed above are either
subordinated or nominalised. Subordinate verb forms are marked for switch-reference
and could well be considered as a fourth mood if it were not for their non-finite character.
All Quechua dialects exhibit a sharp distinction between realised and non-realised
events. The future tense refers to non-realised events, as do the imperative and the present
optative. In the indicative mood, future tense is required for any event that takes place
after the moment of speaking. All dialects have an unmarked or present tense and at
least one past tense.
In the unmarked present tense the subject personal reference endings are added to
the verb stem directly. In the remaining tenses (except for the future), a tense-marking
suffix precedes the subject endings. Additionally, most Quechua dialects have one or
two compound tenses based on a combination of nominalisations and the auxiliary verb
ka- ‘to be’.
The most common marker for plain past tense can be reconstructed as *-rqa-.Itis
illustrated in the following examples from Ayacucho Quechua, where it contrasts with a
habitual past-tense form. The habitual past tense consists of the agentive nominalisation
form in -q and the verb ‘to be’.
(83) miku-r(q)a-ni
eat-PA-1S
‘I ate (it).’
(84) miku-q ka-ni
eat-HB be-1S
‘I used to eat (it).’
In Ayacucho Quechua, as in several other dialects, there are good reasons for consid-
ering the habitual past a separate tense. As the third-person subject form of the verb ‘to
be’ in its copula function is regularly deleted, the pluralising suffix -ku is added directly
to the nominalised form, thus emphasising the finite verb character of the construction.
When combinations of subject and object markers occur, these are normally distributed
over both components of the construction, as in (86) (Parker 1969a: 49).
(85) riku-q-ku
see-HB-(be.3S-)-PL.3S
‘They used to see (him/her/it).’
(86) riku-su-q ka-nki
see-3S.2O-HB be-2S
‘He used to see you.’
The existence of a special tense category denoting surprise or lack of previous aware-
ness on the side of the speaker (cf. section 3.2.3) has become an areal phenomenon,