Linguistic Discovery
Dartmouth College

Volume 13 Issue 2 (2015)        DOI:10.1349/PS1.1537-0852.A.461

Note: Linguistic Discovery uses Unicode characters to represent phonetic symbols. Please see Optimizing Display for requirements to accurately reproduce this page.

How Strong is the Case for Contact-Induced Grammatical Restructuring in Quechuan?

Fernando Zúñiga

University of Bern

 

1. Introduction[1]

 

Pieter Muysken’s work since the late 1970s on Northern Quechua has suggested the possibility that grammatical structure may be restructured due to contact in a gradual, rather than an abrupt, fashion (cf. Muysken 1977, 1980, 2000, but especially Muysken 2009, which develops ideas found in Arends 1993, 1996, and also Cardoso 2009).[2] Additionally, he has proposed that such a “gradual transformation of an expansion language, Incaic imperial Quechua, into a morphologically more simple variety as it spread northward into Ecuador” (Muysken 2009: 77) is best seen as showing not only contact-induced change without substrate influence (“koineization”) but also contact-induced change with substrate influence (“creolization”), and has offered some likely candidates for this development of Ecuadorian Quechua (henceforth EQ):

 

The Barbacoan family, spoken in coastal southern Colombia and northern Ecuador, and the Jivaroan family, spoken in the Andean foothills of southern Ecuador and northern Peru, would be the most likely candidates [where we could identify a major dominant substrate language that may have influenced EQ]. A detailed comparison of the structural features of these languages and the specific traits of [EQ] still needs to be made (Muysken 2009: 85).

 

Muysken also proposes specific traits of EQ where such an influence may be detected, viz. the simplification of some parts of the morphology when compared with other Quechua varieties and with Proto-Quechuan, “some Shuar and Barbacoan loans in flora/fauna [and] possible influence in local [EQ] grammar features” (Muysken 2012: 239). Other studies by this author consider the possibility of substrate influence on EQ as well (cf. especially Muysken 2010, 2011b).

 

The external history of EQ is known to some extent (cf. Muysken 2011a, 2011b); the Ecuadorian varieties emerged between the 15th and 18th centuries (i.e., the Incaic expansion and the colonial periods) as the product of transplanted (mostly southern) Peruvian varieties. The sierra was reportedly occupied by peoples speaking Barbacoan, Jivaroan and unclassified languages by the mid-16th century; Quechua seems to have been the language of an urban elite at that time. By the end of the colonial period, however, large parts of the sierra (both urban and rural) were reportedly EQ-speaking. EQ would then be caught up between the expansion of its superstrate (Spanish as the prestige language) and the marginalization or disappearance of most of its substrate (the non-Quechuan languages). Given this reconstruction of the region’s linguistic ecology, it is actually only natural to look for contact-induced phenomena in EQ.

 

The goal of the present paper is to preliminarily assess the available evidence in favor of some of Muysken’s thought-provoking claims mentioned at the beginning. Based on the extant descriptive literature on Quechuan, Barbacoan and Jivaroan, and concentrating on the loss and restructuring of EQ morphology, this study concludes that there is still a substantial amount of work to be done in order to render Muysken’s intuitively attractive and possibly even correct case stronger. Unfortunately, the current state of descriptive work on all the languages involved does not allow us to be more confident as to how much of present-day EQ morphology is really due to creolization possibly with a Jivaroan, or probably rather with a Barbacoan, substrate.

 

2. The morphology of EQ in genealogical and areal perspective

 

Four morphological developments in EQ and can easily be shown to represent deviations from Quechuan patterns and were identified in Muysken (2009) as innovations of this particular northern variety: some reduction in person and number marking, the form and function of an erstwhile benefactive applicative, and the make-up of (de-)verbal forms used in selected kinds of subordinate clauses.[3] (I will very briefly address selected tense-aspect-modality-evidentiality (TAME) markers at the end of this section.) The deviations are schematically summarized in Table 1 below and discussed in turn in what follows. Within Quechua, the relevant comparisons are made with the Ancash and Ayacucho varieties, which stand for Central and Southern Quechua, respectively; Imbabura Quechua (a variety from northern Ecuador) stands for EQ.[4] The non-Quechua languages chosen —because of both data availability and spatial proximity to EQ— are Awa Pit and Tsafiki (Barbacoan) on the one hand and Aguaruna and Shuar (Jivaroan) on the other.[5]

 

 

EQ

Other Quechuan

Person/number

- verbal.1

- verbal.2

- nominal

 

no clusivity

no portmanteaus

none (e/Pastaza)

 

clusivity

some portmanteaus

suffixal psr

Voice

- ben applicative

 

valency-neutral

loss of -pu

 

valency-relevant

Subordination

- Adverbial clauses

- Purpose clauses

 

SR (no person)

SR (no person)

 

SR (DS w/person)

nmlz+acc[6]

TAME

- PQ *-yku

- PQ *-yka

- PQ *-ku

- PQ *-ri

 

(lost)

(lost)

dur

refl

 

dyn/dirct

dur

mpass

inch

Table 1. Selected EQ morphology vis-à-vis Quechuan. Based on Muysken (2009: 80)

 

2.1 Person/Number Marking

 

First note that the inclusive vs. exclusive distinction found in pronouns, possessive suffixes and verbal markers in most Quechuan varieties (e.g. in Ancash in (1) and Ayacucho (2) in below) has been lost in EQ (3). Second, verb forms used with 1↔2 interactions in other Quechuan typically show portmanteaus (e.g. ‑q in Ancash and ‑yki in Ayacucho), whereas EQ has nothing of the sort. Moreover, the 1P marker -wa ~ -ma, still robust in most Quechua varieties, is reportedly becoming obsolescent in EQ;[7] the 2P marker ‑su ~ ‑shu is already lost.

 

(1) Ancash Quechua present verb forms (Parker 1976: 105f)[8]

 

P \ A

1sg

1pl.excl

2sg

2pl

3sg

1sg

-ma-nki

-ya-ma-nki

-ma-n

1pl.excl

-ya-ma-nki

-ya-ma-nki

 

1pl.incl

 

 

-ma-ntsik

2sg

-q

-ya-q

-shu-nki

2pl

-ya-q

-ya-q

 

S

-:

-ya-: excl

-ntsik incl

-nki

-ya-nki

-n

 

(2) Ayacucho Quechua present verb forms (Soto Ruiz 1976: 93f)

 

P \ A

1sg

1pl.excl

2sg

2pl

3sg

1sg

-wa-nki

-wa-nki-chik

-wa-n

1pl.excl

-wa-nki-ku

 

-wa-nku

1pl.incl

 

 

-wa-nchik

2sg

-yki

-yki-ku

-su-nki

2pl

-yki-chik

 

-su-nki-chik

S

-ni

-ni-ku excl

-nchik incl

-nki

-nki-chik

-n

 

(3) Imbabura EQ present verb forms (Cole 1982: 103-104,159-160)

 

P \ A

1sg

1pl

2sg

2pl

3(sg)

1sg

(-wa)-ngui

(-wa)-ngui-chi

(-wa)-n

1pl

-ngui

-ngui-chi

-n

2sg

-ni

-n-chi

-n

2pl

-ni

-n-chi

-n

S

-ni

-n-chi

-ngui

-ngui-chi

-n

 

Non-Quechuan languages of the region show both dissimilar and similar argument-marking patterns. Barbacoan is known to have an egophoric vs. allophoric distinction (also known under the names conjunct vs. disjunct, assertor vs. non-assertor, or congruent vs. non-congruent), either additionally distinguishing grammatical relations (Awa Pit) or limited to egophoricity (Tsafiki):

 

(4) Awa Pit argument marking: Egophoricity + GRs (Curnow 1997: 193f)

 

e.g. pst: -wegoph.sbj vs. -segoph.obj’ vs. -zialloph

 

(5) Tsafiki argument marking: Egophoricity (Dickinson 2002; also Moore 1979, Turner 1992)

 

e.g. decl -yoegoph’ (vs. Ø), inter -yuegoph’ (vs. Ø), etc.

 

Jivaroan languages, by contrast, has argument markers closer in function to the Quechuan ones — and note that both Aguaruna and Shuar actually have portmanteau morphemes for 1↔2 interactions:

 

(6) Aguaruna SAP↔SAP verbal suffixes (Overall 2007: 317)

 

-hamɨ1sg→2sg’ vs. -himɨ1pl→2sg’/‘2pl →1’

 

(7) Shuar SAP↔SAP verbal suffixes (Gnerre 1999: 68f)

 

(7a)

Tu-rám-tat-j-i.

 

say-3→1pl/2-fut-1pl-decl

 

‘S/he will tell us.’

 

(7b)

Su-kárta-r-me.

 

give-2→1pl-pl-2

 

‘You (sg) give [things] to us.’

 

(7c)

Tu-ru-t-.

 

say-2/3→1sg-imper-2

 

‘Tell me!’

 

Thus, even though contact with languages that have different argument-marking patterns may explain part or all of the loss of the relevant EQ verbal morphology, it is not easy to see how this would have worked in the details, particularly so regarding the 1↔2 portmanteaus and considering the quite different egophoricity-centered system of Barbacoan. The loss of clusivity in EQ, on the other hand, is more plausibly attributed to contact with clusivity-less Barbacoan and Jivaroan, even though a contact-independent explanation is certainly also possible.

 

2.2 Benefactive Applicative

 

There would be more to say about the evolution of Quechuan benefactive (quasi-) applicatives, especially regarding their form, but suffice it here to say that EQ only retains the marker ‑pa and, more importantly, that this marker is no longer an applicative or quasi-applicative.[9] Whereas in Ancash verbal valency is clearly affected by the suffixation of ‑pu/‑pa (8) (in other Quechua varieties, the applied object appears in the accusative instead of the benefactive illustrated in 8a), in EQ the suffix conveys an honorific meaning without altering the valency of the predicate and the syntax of the clause: 

 

(8) Ancash Quechua (Parker 1976: 84, 119)

 

(8a)

Wawqi-:-paq

apa-pu-shaq.

 

brother-1sg.psr-for

bring-ben.qappl-1sg.fut

 

‘I will carry (it) for my brother.’

 

(8b)

Tushu-na-yki-paq

toka-shaq.

 

dance-nmlz-2sg.psr-for

play-1sg.fut

 

‘I will play for you (SG) to dance.’

 

(8c)

Rura-pa-ma-rqa-n.

 

make-ben.qappl-1sg.obj-rec.pst-3sg

 

‘S/he made (it) for me.

 

(9) Imbabura EQ (Cole 1982: 113, 185)

 

(9a)

Wasi-ta

rura-rka-ni

ñuka

churi-paj.

 

house-acc

make-pst-1sg

1sg

son-for

 

‘I made a house for my son.’

 

(9b)

Miku-na-ta

muna-pa-ngui=chu?

 

eat-nmlz-acc

want-hon-2=irr

 

‘Do you (sg hon) want to eat?

 

Is contact helpful here as an explanatory factor? As far as I can judge from the extant descriptions, Barbacoan languages do not have benefactive applicative markers, but those found in Jivaroan (particularly in Aguaruna), are well-behaved (i.e., they are not valency-neutral like the present-day EQ morpheme), i.e. they conform to the expected behavior of applicatives. The EQ development might just as well have been contact-independent.

 

2.3 Subordinate Clauses

 

Two types of subordinate clauses are relevant in the present context: temporal adverbial clauses and purpose clauses. As to the former, both EQ and other Quechua varieties have switch-reference forms, but EQ does not mark person on them while other varieties do. The different-subject forms of both Ancash (10) and Ayacucho (11) take possessive suffixes to convey person and number of the different subject (illustrated here with 1st person singular forms), whereas those of EQ have an invariable suffix ‑jpi (12):

 

(10) Ancash Quechua (Cole 1983: 2-3): Switch-reference (SS -r ~ -shpa vs. DS -pti-{psr})

 

(10a)

Lima-ta

chaa-ri-r,

rikaari-shaq    

amigo-u-ta.

 

L.-acc    

arrive-after-SS1

see-1sg.fut

friend-1sg.psr-acc

 

‘After arriving in Lima, I will see my friend.’

 

(10b)

Chakra-chaw

urya-shpa

pallamu-rqu-u

wayta-kuna-ta.

 

field-loc

work-SS2

pick-rec.pst-1sg

flower-pl-acc

 

‘While I worked in the field, I picked flowers.’

 

(10c)

Chakra-chaw

urya-pti-y,

María

pallamu-rqu-n

wayta-kuna-ta.

 

field-loc

work-DS-1sg.psr

M.

pick-rec.pst-3sg

flower-pl-acc

 

‘While I worked in the field, Maria picked flowers.’

 

(11) Ayacucho Quechua (Hartmann 1994): Switch-reference (SS -stin ~ -spa vs. DS -pti-{psr})

 

(11a)

Tuma-stin

puriku-chka-nki.

 

drink-SS.sim

walk.around-dur-2sg

 

You (sg) are walking around drinking.’

 

(11b)

Miku-spa(-n)

lluqsi-rqa.

 

eat-SS.seq-3.psr

leave-pst

 

Having eaten, s/he left

 

 

(11c)

Ñuqa

ni-pti-y=mi    

ri-rqa.

 

1sg

say-DS-1sg.psr=evid

go-pst

 

S/he went because I said it.’

 

(12) Imbabura EQ (Cole 1983: 5): Switch-reference

 

(12a)

Utavalu-man

chaya-shpa,

ñuka

mama-ta

riku-rka-ni.[10]

 

O.-all

arrive-SS

1sg

mother-acc

see-pst-1sg

 

‘When I arrived in Otavalo, I saw my mother.’

 

(12b)

Juzi

Utavalu-man

chaya-jpi

pay-paj

wasi-man

ri-rka-ni.[11]

 

J.

O.-all

arrive-DS

3sg-gen

house-all

go-pst-1sg

 

‘When Jose arrived in Otavalo, I went to his house.’

 

Interestingly enough, Barbacoan languages have switch-reference forms in temporal adverbial clauses, but they do not mark person, therefore lending plausibility to the contact hypothesis. Jivaroan languages, by contrast, have switch-reference forms that do mark person in such clauses. More information on concrete contact scenarios, their intensity and duration, would be needed in order to come to an attractive solution here.

 

Regarding purpose clauses, note that other Quechua varieties have deverbal nominalized forms there; consider the Ancash (13) and Ayacucho (14) examples below with forms showing the nominalizerna, a possessive suffix, and the benefactive/purposive suffix ‑paq ‘for’. By contrast, EQ uses a switch-reference form in these clauses (15) (and, as expected, such forms do not mark person); note that the markers (‑ngapaj / ‑chun) differ from those employed in temporal clauses (‑shpa / ‑jpi):

 

(13) Ancash Quechua (Cole 1983: 3-4): Nominalized form

 

(13a)

Huaraz-ta

shamu-rqu-u

mamaa-nii-ta

rikaa-na-a-paq

 

H.-acc

come-rec.pst-1sg

mother-1sg.psr-acc

see-nmlz-1sg.psr-for

 

‘I came to Huaraz to see my mother.’

 

(13b)

Fuan-ta

Huaraz-ta

kacha-rqu-u

mamaa-nin-ta

rikaa-na-n-paq

 

J.-acc

H.-acc

send-rec.pst-1sg

mother-3sg.psr-acc

see-nmlz-3sg.psr-for

 

‘I sent Juan to Huaraz to see his mother.’

 

(14) Ayacucho Quechua (Soto Ruiz 1976: 156): Nominalized form

 

(14a)

Tarpuku-na-nku-paq=ña=m

allichaku-chka-nk

 

sow-nmlz-3pl.psr-for=already=evid

get.ready-dur-3pl

 

 ‘They are already getting ready to sow.’

 

(14b)

Yanuku-na-nku-paq=mi

mikuy-kuna-ta

apa-chka-ni

 

cook-nmlz-3pl.psr-for=evid

food-pl-acc

bring-dur-1sg

 

‘I am bringing food for them to cook.’

 

(15) Imbabura EQ (Cole 1983: 6-7): SR

 

(15a)

Utavalu-man

shamu-rka-ni

ñuka

mama-ta

visita-ngapaj.[12]

 

O.-all

come-pst-1sg

1sg

mother-acc

visit-SS

 

I came to Otavalo to see my mother.’

 

(15b)

Juzi-ta.

Utavalu-man

kacha-rka-ni

pay-paj

mama-ta

visita-chun

 

J.-acc

O.-all

send-pst-1sg

3sg-gen

mother-acc

visit-DS

 

I sent Jose to Otavalo to see his mother.’

 

These purposive switch-reference forms are also used in desiderative expressions in EQ. Here, again, other Quechua varieties employ nominalizations, like the accusative-marked infinitive of Ayacucho Quechua below:

 

(16) Desiderative (Soto Ruiz 1976: 157, Cole 1982: 38)

 

Ayacucho Quechua

 

(16a)

 Piñaku-pti-ki=m

mana    

yayku-mu-y-ta

muna-n=chu.

 

get.angry-DS-2sg.psr=evid

neg

enter-dirct-inf-acc

want-3sg=irr

 

‘S/he doesn’t want to get in because you (sg) get angry.’

 

Imbabura EQ

 

(16b)

Ñuka

muna-ni

miku-ngapaj.

 

1sg

want-1sg

eat-SS

 

‘I want to eat.’

 

(16c)

Ñuka

muna-ni

kan

miku-chun.

 

1sg

want-1sg

2sg

eat-DS

 

‘I want you (sg) to eat.’

 

Leaving aside the issue of what governs the distribution between same-subject and different-subject forms in EQ (the data in Cole 1983 suggest a direct/inverse-related 1/2 vs. 3 opposition playing an interesting role here), as well as the intriguing and unclear etymology of ‑chun (which is a 3rd person imperative in other Quechuan varieties), I will now turn to Bruil (2008) regarding the probable evolution of such forms.[13] This author hypothesizes that the loss of argument marking in subordination in general is due to Barbacoan influence (pp. 123f), that switch-reference forms in purpose clauses is also due to contact with Barbacoan (pp. 124, following a suggestion in Adelaar 2004: 149), and that the extension of such switch-reference forms to desiderative constructions is due to Spanish influence (p. 126). Limiting myself here to the second hypothesis, it is clear that both Barbacoan and Jivaroan languages have morphologically simple patterns with respect to subordinate clauses: adverbial and purposes clauses show switch-reference forms in all languages (like EQ), but these do not mark person in Barbacoan, even in different-subject forms (like EQ) and do mark person in Jivaroan, even in same-subject forms. Examples follow:

 

(17) Awa Pit purpose clauses (Curnow 1997: 268-269)

 

(17a)

Carmen

piya

kɨɨ-t

kway-zi

atal

pashpa

kwin-na.

 

C.

corn

mill-ser

drop-alloph

chicken

dim

give-inf

 

‘Carmen ground corn to give to the baby chickens.’

 

(17b)

Na=na

pyan--ma--mtu-s,

Carmen

ayna-t

kwa-npa

 

1sg=top

firewood

chop-asp-asp-asp-asp-egoph

C.

cook-ser

eat-DS

 

‘I’m going to chop firewood so that Carmen can cook and eat.’

 

(18) Tsafiki purpose clauses (Moore 1979: 48)

 

(18a)

Fi-chun

ka-yo-e.

 

eat-SS

take-egoph-decl

 

‘I bought (lit. took) (it) to eat (it).’

 

(18b)

Sona

mera-sa

ta-yo-e.

 

woman

hear-DS

have-egoph-decl

 

‘I have (it) so that my wife can hear (it).’

 

(19) Aguaruna purpose clauses (Overall 2007: 504)

 

(19a)

Hiina-aha-maia-hi-I

ii-na

batsama-taĩ-ka

mini-ika-ku.

 

exit.pfv-intm.pst-1pl-decl

1pl-acc

live-1pl/3-foc

arrive-ipfv.pl-SS.1pl

 

‘We set off to go back to our homes.’

 

(20) Shuar adverbial clauses (Gnerre 1999: 74,76)

 

(20a)

Wi

chichá-ku-n

paánta-n

chichá-ja-i.

 

1sg

speak-SS-1sg

clear-obj

speak-1sg-decl

 

‘When I speak, I do so (lit. speak) clearly.’

 

(20b)

Chichá-i    

nín-kia

-y-i.

 

speak-DS.1sg/3

3-top

arrive-3-decl

 

While I / s/hei was speaking, s/hej arrived.’

 

In the light of this evidence, Bruil’s claim as to the possibility of Barbacoan influence here is indeed a very attractive one.[14]

 

2.4 Selected TAME markers

 

Finally, it is in order to at least mention the TAME markers addressed by Muysken, viz. the reflexes of Proto-Quechuan *‑yku, *‑yka, *‑ku, and *‑ri. The first two have been lost in EQ; they mark dynamicity/direction and durativity in other Quechuan varieties, respectively. The marker ‑ku encodes durativity in EQ, but in other Quechuan varieties, and presumably also in Proto-Quechuan, it is a mediopassive morpheme. Awa Pit has an imperfective marker ‑(m)tu, which is somewhat similar to Quechuanku, but even though they might be considered linked (as two subtypes of diminished-transitivity markers), the unclear details regarding the semantic change and the rather modest formal similarity does not allow one to regard this mere finding as strong evidence in favor of a contact-induced change mediopassive > durative in EQ. Lastly, ‑ri is an inchoative marker in Proto-Quechuan and other Quechuan varieties but a reflexive marker in EQ — a hitherto unexplained development. The connection between inchoativity and reflexivity is, at least formally, well attested elsewhere (cf. Spanish anticausative se, evolved from a reflexive marker), but it is far from clear how Barbacoan influence would have worked here: neither Awa Pit nor Tsafiki have a verbal reflexive proper, let alone any related marker (e.g. a pronominal one) that would formally resemble EQ -ri (but Tsafiki has an inchoative suffix ‑di, which might have played a role). I was not able to see any other connection between these developments, especially the latter two listed above, and possible Jivaroan or Barbacoan models.  

 

2.5 Summary

 

The above discussion of EQ deviations from the patterns prevalent in Quechuan and possible contact-induced developments can be schematically summarized in Table 2 below.

 

 

EQ

Other Quechuan

Jivaroan

Barbacoan

Person/number

- verbal.1

- verbal.2

- nominal

 

no clusivity

no portmanteaus

none (e/Pastaza)

 

clusivity

some portmanteaus

suffixal psr

 

no clusivity

some portmanteaus

suffixal psr

 

no clusivity

no portmanteaus

none

Voice

- ben applicative

 

valency-neutral

loss of -pu

 

valency-relevant

 

(valency-relevant)

 

(none)

Subordination

- Adverbial clauses

- Purpose clauses

 

SR (no person)

SR (no person)

 

SR (DS w/person)

nmlz+acc

 

SR (w/person)

SR (w/person)

 

SR (no person)

SR (no person)

TAME

- PQ *-yku

- PQ *-yka

- PQ *-ku

- PQ *-ri

 

(lost)

(lost)

dur

refl

 

dyn/dirct

dur

mpass

inch

 

 

 

 

 

 

(AP ipfv -(m)tu)

 

Table 2. Selected EQ morphology vis-à-vis Quechuan and non-Quechuan (First two columns reproduced from Table 1 above)

 

3. Discussion

 

Muysken (2009: 98) identifies the following timeline for selected changes in EQ:[15]

 

all known sources

ca. 1700

ca. 1900

ongoing

loss of clusivity

loss of nmlz-{pers}

loss of -sun-ki ‘3→2sg’

V-ku > V-kuna in pl

loss of -puben’

loss of psr suffixes

loss of -yki ‘1→2’

-ku > -rirefl’

-yka > -kudur’

-paben’

loss of -wa1.obj’

 

His final comments on these changes is indicative of his koineization (non-substrate-influenced) vs. creolization (substrate-influenced) distinction mentioned at the outset of the present article:

 

The earliest changes […] may possibly be viewed as cases of morphological simplification independent of individual substrates, while later changes, such as the shift in meaning of the verbal suffixes -ku- and -ri-, must have a different explanation, and could be due to substratal influence. This would suggest a gradual restructuring towards the possible substrate languages (which have now disappeared). Nonetheless, at least some of the morphological simplification must have been gradual in nature as well. (Muysken 2009: 98)

 

The data presented in Section 2 above lead to the identification of possible substrate languages as schematically summarized in Table 3 below; they mostly point towards Barbacoan. Nevertheless, as Muysken himself says, some other, now extinct, substrate languages cannot be ruled out.

 

 

EQ

Possible substrate

Person/number

- verbal.1

- verbal.2

- nominal

 

no clusivity

no portmanteaus

none (exc. Pastaza)

 

Barbacoan, Jivaroan

Barbacoan

Barbacoan (+Jivaroan)

Voice

- ben applicative

 

valency-neutral

loss of -pu

 

(Barbacoan)

Subordination

- Adverbial clauses

- Purpose clauses

 

SR (no person)

SR (no person)

 

Barbacoan

Barbacoan

TAME

- PQ *-yku

- PQ *-yka

- PQ *-ku

- PQ *-ri

 

(lost)

(lost)

dur

refl

 

 

 

*!

*!

Table 3. Selected morphological phenomena in EQ

 

Note that Bruil (2008) and Muysken (2009) disagree as to what motivated some of the comparatively early changes in EQ; in particular, the restructuring of morphological make-up of subordinate clauses is said to be due to Barbacoan influence by the former author but are presented as possible cases of contact-independent innovation by the latter. Muysken says that later EQ innovations might have been contact-induced, and they may very well have been, but the extant evidence from Barbacoan seems to make the unknown-and-extinct-language(s) suggestion more attractive — hardly firm ground to continue reconstructing the exact details of EQ prehistory.

 

In fact, two important questions posed by Muysken seem to me to be still somewhat open. First, “[c]an we identify a single major dominant substrate language that may have influenced [EQ]?” (Muysken 2009: 85). Based on the present-day descriptions available to me, such a language would have to be Barbacoan (possibly Caranqui, as suggested by Bruil 2008, 2011), rather than Jivaroan — with a strong emphasis on may, however. The new switch-reference morphology in EQ seems to be the best evidence for contact-induced change. Second, “[w]hat features can be readily explained through autonomous simplification processes and what features would require a different, possibly substratist, explanation?” (Muysken 2009: 85). I see two problems with the quote at the beginning of Section 3 in this respect: Why is the earlier loss of morphology autonomous, rather than the later loss? Why is the shift of some TAME markers rather contact-induced? There does not seem to be a direct motivating or driving force in either Barbacoan or Jivaroan, let alone in Spanish, for such developments. I thus conclude that, intuitively appealing though Muysken’s koineization hypothesis is, a strong case in its favor is yet to be made, and the extant descriptions of possible substrate languages does not allow one to be more confident here.

 

Finally, let me zoom out and briefly comment on an interesting hypothesis formulated by Muysken when discussing the evolution of EQ. He suggests that there might be a correlation between the outcome of such contact situations and the morphological make-up of the participating languages:

 

[The fact that radical restructuring in EQ led to the loss of morphology in specific gram­matical contexts and regularization rather than a total loss] may be due to the agglutinative character of Quechua; in that case we would predict that in general agglutinative languages undergo different processes of restructuring than inflectional languages (Muysken 2009: 97).

 

In other words, the outcome of such an alleged koineization process would be decisively determined by the morphology-typological profile of the superstrate language. Nevertheless, since the case for the specific koineization-cum-creolization process in EQ proposed in Muysken (2009) is not particularly strong, and since, to my knowledge, no similar cases have been made for other Quechuan varieties, we still do not know whether this applies to Quechuan in general — even though, as mentioned by Pieter Muysken (p.c.), the evidence in favor of EQ as spreading comparatively late is indeed convincing. Within Indo-European, to judge from the reasonably documented morphological change documented in Germanic, Romance, Celtic, Slavic, and Indo-Aryan, it would seem that radical restructuring does not necessarily lead to total loss of morphology. Are there parallels from Turkic, Tungusic, Eskaleut, Uralic, Bantu, and Dravidian (all examples of “agglutinating” families) that support Muysken’s proposal regarding the resistance of agglutination to total morphology loss? His tentative prediction is both attractive and worth pursuing further, but more work needs to be done, not only within Quechuan but also within other language families, in order to obtain conclusive support for such a claim.

 

Abbreviations

A agentive argument of bivalent predicates, acc accusative, agt agentive, all allative, alloph allophoric, AP Awa Pit, asp aspect(ual), ben benefactive, decl declarative, dim diminituve, dir direct, dirct directional, DS different subject, dur durative, dyn dynamic, egoph egophoric, EQ Ecuadorian Quechua, evid evidential, excl exclusive, foc focus, fut future, gen genitive, GRs grammatical relations, hon honorific, imper imperative, inch inchoative, incl inclusive, inf infinitive, inter interrogative, intm intermediate, inv inverse, ipfv imperfective, irr irrealis, loc locative, mpass mediopassive, neg negative, nmlz nominalization, obj object(ive), P patientive argument of bivalent predicates, pers person, pfv perfective, pl plural, PQ Proto-Quechua, psr possessor, pst past, qappl quasi-applicative, rec recent, refl reflexive, S single argument of monovalent predicates, SAP speech act participant, sbj subject, seq sequential, ser serial, sg singular, sim simultaneous, SR switch-reference, SS same subject, TAME tense-aspect-modality-evidentiality, top topic

x→y x acting on y

 

References

Adelaar, Willem (with the collaboration of Pieter Muysken) (2004) The Languages of the Andes. Cambridge: Cambridge University Press.

Arends, Jacques (1993) Towards a gradualist model of creolization. Atlantic Meets Pacific. A Global View of Pidginization and Creolization, ed. by Francis Byrne & John Holm, 371-380. Amsterdam: Benjamins.

Arends, Jacques (ed.) (1996) The Early Stages of Creolization. Amsterdam: John Benjamins.

Bruil, Martine (2008) Innovations in the Ecuadorian converb systems: grammatical change in language contact situations. M.A. thesis, University of Leiden.

Bruil, Martine (2011) The emergence of future converbs in Imbabura Quichua: Pre-Hispanic language contact as a possible explanation. Language Contact in Times of Globalization, ed. by Cornelius Hasselblatt, Peter Houtzagers & Remco van Pareren, 27-44. Amsterdam: Rodopi.

Cardoso, Hugo (2009) Jacques Arends’ model of gradual creolization. Gradual Creolization: Studies Celebrating Jacques Arends, ed. by Rachel Selbach, Hugo Cardoso & Margot van den Berg, 13-23. Amsterdam: Benjamins.

Cerrón-Palomino, Rodolfo (1987) Lingüística quechua. Cuzco: Centro de Estudios Rurales Andinos Bartolomé de Las Casas.

Ciucci, Luca & Pieter Muysken (2011) Hernando de Alcocer y la Breve declaración del Arte y Bocabulario de la lengua del Ynga conforme al estilo y vso de la provincia de Quito. El más antiguo manuscrito de quichua del Ecuador. Indiana 28: 359-394.

Cole, Peter (1982) Imbabura Quechua. Amsterdam: North-Holland.

Cole, Peter (1983) Switch-reference in two Quechua languages. In Switch-Reference and Universal Grammar, ed. by John Haiman & Pamela Munro, 1-15. Amsterdam: Benjamins.

Curnow, Timothy (1997) A grammar of Awa Pit (Cuaiquer): An indigenous language of South-Western Colombia. Ph.D. thesis, Australian National University.

Dickinson, Connie (2002) Complex predicates in Tsafiki. Ph.D. dissertation, University of Oregon.

Gnerre, Maurizio (1999) Perfil descriptivo e histórico-comparativo de una lengua amazónica: el shuar (jíbaro). Ms., University of Naples Federico II.

Hartmann, Roswith (1994) Rimaykullayki: Unterrichtsmaterialien zum Quechua Ayachuchano - Peru. Zusammengestellt nach Clodoaldo Soto Ruiz’ Quechua – manual de enseñanza. Berlin: Dietrich Reimer Verlag.

Macario, Florens (2014) Influences of contact phenomena on Ecuadorian Quichua morphology, under consideration of Jivaroan and Barbacoan languages. B.A. thesis, University of Bern.

Moore, Bruce (1979) Método para aprender el idioma colorado. Gramática pedagógica, tomo I, lecciones 1-6. Quito: Instituto Lingüístico de Verano.

Muysken, Pieter (1977) Syntactic Developments in the Verb Phrase of Ecuadorian Quechua. Lisse: Peter de Ridder.

Muysken, Pieter (1980) Sources for the study of Amerindian contact vernaculars in Ecuador. Amsterdam Creole Studies 3: 66-82.

Muysken, Pieter (2000) The genesis of Lowland Ecuadorian Quechua and semantic transparency. Creoles, Pidgins, Sundry Languages: Essays in Honor of Pieter Seuren, ed. by Jacques Arends, 973-988. Berlin: Mouton de Gruyter.

Muysken, Pieter (2009) Gradual restructuring in Ecuadorian Quechua. Gradual Creolization: Studies Celebrating Jacques Arends, ed. by Rachel Selbach, Hugo Cardoso & Margot van den Berg, 77-100. Amsterdam: Benjamins.

Muysken, Pieter (2010) The copula in Ecuadorian Quechua. Linguistics and Archeology in the Americas: The Historization of Language and Society, ed. by Eithne Carlin & Simon van de Kerke, 191-206. Leiden: Brill.

Muysken, Pieter (2011a) Relative clauses in Ecuadorian Quechua. Subordination in Native South-American Languages, ed. by Rik van Gijn, Katharina Haude & Pieter Muysken, 251-266. Amsterdam: Benjamins.

Muysken, Pieter (2011b) Change, contact, and ethnogenesis in Northern Quechua: Structural phylogenetic approaches to clause embedding predicates. Ethnicity in Ancient Amazonia: Reconstructing Past Identities from Archeology, Linguistics, and Ethnohistory, ed. by Alf Hornborg & Jonathan Hill, 237-258. Boulder: University of Colorado Press.

Muysken, Pieter (2011c) Multi-verb constructions in Ecuadorian Quechua. Multi-Verb Constructions in the Americas, ed. by Alexandra Aikhenvald & Pieter Muysken, 133-156. Leiden: Brill.

Muysken, Pieter (2012) Contacts between indigenous languages in South America. The Indigenous Languages of South America. A Comprehensive Guide, ed. by Lyle Campbell & Verónica Grondona, 235-258. Berlin: De Gruyter Mouton.

Overall, Simon (2007) A Grammar of Aguaruna. Ph.D. thesis, La Trobe University.

Parker, Gary (1976) Gramática quechua: Ancash-Huailas. Lima: Ministerio de Educación.

Soto Ruiz, Clodoaldo (1976) Gramática quechua: Ayacucho-Chanca. Lima: Ministerio de Educación.

Turner, Glen (1992) Una breve gramática del shuar. Quito: Instituto Lingüístico de Verano.

Van de Kerke, Simon & Pieter Muysken (2014) The Andean matrix. The Native Languages of South America: Origins, Development, Typology, ed. by Loretta O’Connor & Pieter Muysken, 126-151. Cambridge: Cambridge University Press.



[1] The author is grateful to Rik van Gijn, but especially to Pieter Muysken and an anonymous reviewer for numerous and valuable comments on a previous version of this article. The usual disclaimers apply.

[2] See also Muysken (2010, 2011a, 2011b, 2011c) and Van de Kerke & Muysken (2014). 

[3] See Muysken (2011c) for a slightly different (but compatible) list of differences in morphosyntax between EQ and other Quechua varieties.

[4] It is worth noting that there is non-negligible variation within EQ; as pointed out by an anonymous reviewer, the Imbabura variety is geographically closer to the Barbacoan languages than to the Jivaroan ones.

[5] Bruil (2008) points out the relevance of the extinct Barbacoan language Caranqui as probable substrate of EQ, but not much is known about the grammatical structure of this language.

[6] Or some other case marker; in Ayacucho Quechua, e.g., it is the benefactive -paq.

[7] According to an anonymous reviewer, however, -wa is still in use in the Imbabura variety. It seems that there is more variation in this respect within EQ in general, and within Imbabura Quechua in particular, than the comparative literature on Quechua has acknowledged hitherto. 

[8] The plural suffix -ya becomes -yaa in open syllables.

[9] A different issue —not treated here— is the existence of an innovative benefactive compound predicate based on cara- ‘give’ in EQ. Such applicative or pseudo-applicative constructions are well-known from West Africa, as well as from southern and eastern Eurasia, but the existence of a close parallel in Cha’palaa (southern Barbacoan; Ecuador), mentioned by an anonymous reviewer, is certainly worth exploring in greater detail.

[10] Simple reflex of PQ *-špa (Cerrón-Palomino 1987: 196).

[11] Either from PQ *-pti or innovative from *-q-pi (-agt-loc) (Cerrón-Palomino 1987: 195). The latter could well be a calque from Jivaroan: Aguaruna reportedly has a subordinate deverbal form featuring nominalizing morphology and a locative case marker (anonymous reviewer, p.c.).

[12] Similar forms in other EQ varieties (Orr’s 1978 “ecuatoriano de la selva” as opposed to “ecuatoriano de la sierra” like Imbabura EQ): ‑ngawa in Pastaza, ‑ngaj in Tena, ‑ngapa in Napo (Cerrón-Palomino 1987: 240).

[13] See also Muysken (2011b) for an overview of verbal morphology used in complementation and subordination across Quechua varieties that pays special attention to some variation within EQ. This author explicitly mentions “possible substrate influence” when discussing the following EQ-specific complex-predicate constructions: (a) verb of saying + future V form to express desire/intention (Shuar); (b) verb of doing + V to express ‘going to’ (Barbacoan, Chibchan, Paezan and Tucanoan; for the EQ variety under scrutiny, the relevant contact language is probably the isolate Waorani); (c) ‑chun in subjunctive/purposive DS forms (Tsafiki); and (d) use of yalli ‘exceed’ in comparisons (perhaps from Barbacoan). 

[14] As Pieter Muysken aptly points out (p.c.), this raises the issue why the chun-forms are so widespread, extending to Quechuan varieties spoken both in Tungurahua and the jungle; Barbacoan could well have had a pervasive influence in all EQ varieties. See Bruil (2011) for additional discussion of -chun as DS marker in the context of Barbacoan influence.

[15] See also Ciucci & Muysken (2011) for additional supporting evidence regarding the early date of many of the EQ innovations.

[ Home | Current Issue | Browse the Archive | Search the Site | Submission Information | Register for Updates | About | Editorial Board | Site Map | Help ]

Published by the Dartmouth College Library.
Copyright © 2002 Trustees of Dartmouth College.
For comments or feedback E-mail the site editor.
ISSN 1537-0852

Linguistic Discovery HomeDartmouth College Home