fredag den 28. juli 2017

Reviewing Kaufman’s evidence for Mixe-Zoque, Wastekan and Totonakan borrowings in proto-Nahuan

In a 2001 paper, distributed on the internet through the website of the Project for theDocumentation of Languages of Mesoamerica (PDLMA) the eminent linguist and expert in Mesoamerican languages Terrence Kaufman analyzed the prehistory of Nahuan languages. He focused specifically on showing how influence from the languages of the Mesoamerican Language Area participated in shaping the Southern Uto-Aztecan dialect proto-Nahuan into the Mesoamerican language Nahuatl. The data used for the paper is very impressive, his conclusions well argued, and Kaufman’s writing style is as always very authoritative, and so the paper has been cited quite a few times (30 citations in google scholar).

In this post, I will take issue with some of the conclusions in Kaufman’s paper, specifically I will show that Kaufman significantly overstates his evidence for substantial lexical influence from Mesoamerican languages on proto-Nahuan, because he does not adequately take into account alternative, potential or probable etymologies from Uto-Aztcan sources. I show that most of his proposed borrowings into proto-Nahuan are in fact equally (or more) likely to have Uto-Aztecan etymologies, either from proto-Uto-Aztecan, from proto-Corachol-Nahuan or can be plausibly analyzed as originating as combinations of Nahuan roots.

My conclusion is that there are much fewer borrowings from Mixe-Zoquean, Wastekan and Totonakan in proto-Nahuan than often thought, and that we therefore cannot use this contact as evidence that proto-Nahuatl was spoken in the area of north-eastern Mesoamerica where Kaufman locates the speech community. Rather we should locate the proto-Nahuan speech community on the north-western periphery of Mesoamerica in close contact with Corachol and with Oto-Pamean languages. 

Proposed loans from Mixe-Zoque in all Nahuan

Kaufman’s source
Potential UA etymology
PUA *kawa “shell”
PZ *kɨ’ak
PCN *kakai
PMZ *kopak “head”
*PUA *kupa “top of head/hair”
PMi *pus
Huichol *purusi “stub, cut short”<PCN *puyusi “stub”
PZo *pata’
PCN *pɨta
Old man
PMZ *na’w
PCN *nawari “thief”  
*nawa “steal”
PMZ *(hah)-¢uku
PMZ *tu’nuk
Corachol *tutuvi “large parrot”, Nahua toto “bird”
PMZ *sam “heat”
PCN *sia “sand/clay” + mi “collective plural”
PMZ calque of *tɨ’k-ɨy “house enter”

Kaufman proposes 9 borrowings and a lexical calque from proto-Mixe-Zoque, proto-Zoque or proto-Mixe into proto-Nahuan. Of these borrowings, 7 have equally probable Uto-Aztecan etymologies, and 5 have definite cognates in Corachol, suggesting that if they are borrowings and not inherited then the borrowing would have been between proto-Mixe and proto-Corachol-Nahuan. The calque seems likely, and the word for ant seems possible. Also, I actually think the word for cacao is a likely borrowing from Mixe-Zoque, since the alternative “shell” etymology proposed by Dakin and Wichmann is somewhat weak, and given the fact that it is extremely unlikely that proto-Nahua was spoken by people who lived in a cacao-producing region whereas proto-Mixe-Zoque almost certainly was.  Nevertheless, the claim of Mixe-Zoque contact with proto-Nahuan seems to lack real support once the alternative etymologies are examined.  This is particularly significant because the words proposed as borrowings are highly culturally significant suggesting that Mixe-Zoque speakers had a profound culturalizing influence on proto-Nahua speakers, teaching them to use foot-wear, live in adobe houses with cultivated liverstock such as turkeys, and to use the culturally salient luxury good cacao, and that through them the Nahuas adopted the pan-Mesoamerican belief in shapeshifting sorcerers. With these borrowings, the role of Mixe-Zoque in this regard seems much less significant. Kaufman has been a major proponent of seeing Mixe-Zoque speaking Olmecs as the drivers of the development of the Mesoamerican cultural area, and they probably were – but it does not seem to me that there was any significant contact between Mixe-Zoque speakers and the proto-Nahuan speech community. This probably means that the Nahuas entered Mesoamerica after the decline of Olmec civilization in the centuries before the beginning of the first millennium.

Proposed loans from Wastekan in all Nahuan

Kaufman’s proposal
Potential UA etymology
Sp. of Parrot
Nahua oši-ƛ “sticky dirt”
Nahua ne:-te:č “reciprocal-together”

Kaufman proposes 5 loans from Wastek Maya into proto-Nahuan. Of these only pulque, and deer-foot seem likely loans. Kochotl is not a general Nahuan word, and there is no reason to reconstruct it for proto-Nahuan – likely be an exclusive eastern or Huasteca Nahua loan. Netech is morphologically analyzable as ne-te:ch. Ohoxihtli seems a likely reduplicated form of oxitl “dirt that comes of when you wash”. Nahuas in fact associated the origin of pulque with the Huastecs, so it seems likely that this is indeed a likely loan. In conclusion, there may have been contact between proto-Nahuan and Wastekan, but if there was it was quite limited – the only likely loan is the word for pulque, and in fact not all Nahuan varieties have this root, as many use the inherited word for “honey” nekwƛi instead.

Proposed loans from Totonac in all Nahuan

Kaufman’s Totonac source
Alternative source
PUA (Stubbs, 2011, #557)
-¢in diminutive
Otomi-Mazahua či-
Corachol ¢i-/-ši (š is a regular cognate of Nahuan ¢ in Corachol)

Corachol *siuri “tadpole” regularly becomes Nahuan *šoli-.
Nahua: wa(k/h)-kal-ƛi “drying house”.
Corachol ¢ɨ¢ɨ
Sp. Of fish
waapa “tilapia”
Brother in law
Older sister
-pi “sister” (not younger)
SUA *saka “grass”, Hopi tïïsaqa ”grass”, NUA *saka “willow” (Stubbs 2011 #1055)
Plate/flat bowl
Wild avocado
Avocado is yewka in Coracholan suggesting an origin as proto-Cora-Nahuan *pewaka
Proto-Corachol-Nahuan *siwi “sour/bitter”. *iw becomes Nahua o, but the question is where the -ko element then comes from.
Totonac *ƛ

Kaufman’s 14 proposed loans from Totonac fare a little better when checked for plausible alternative etymologies. The forms šolotl, wahkalli, chichi, pihtli, pawatl have viable UA etymologies. Šolotl and chichi are shared with Corachol. The diminutive -tzin could be borrowed from Totonac, but Otomi-Mazahua has a diminutive/honorific prefix či- and Coracholan has a diminutive prefix ¢i- and a honorific suffix -ši.  The Totonac form does match the Nahua form better than either of those sources. In any case there is basis for considering the -¢i diminutive morpheme to be an areal trait since it is shared between Mesoamerican languages of three different linguistic families (Totonakan, Oto-Pamean and Uto-Aztecan).
The words pochotl and xonotl, describe species with restricted distribution that likely arose as local borrowings in the Nahuatl varieties spoken where these species are found and only subsequently spread through inter-Nahua contact – I would not reconstruct these words to proto-Nahuan. Wapotl and čone are not found in all (or most?) Nahuan dialects, but are local (recent) borrowings.
That leaves the words for plate, brother in-law, tilapia and xonote, as well as the phoneme ƛ, as likely borrowings from Totonacan into proto-Nahuan. 


Out of 29 proposed borrowings, only 9 seem more likely to have been borrowed, than to have been inherited. So, having reviewed the evidence of borrowings from Mixe-Zoquean, Totonac and Wastekan, I must conclude that the extent of lexical borrowings from Mesoamerican languages into proto-Nahuan is greatly overstated by Kaufman.

Kaufman also shows a long list of borrowings from Wasteko into Huastecan Nahuatl – the Nahuatl variety that we know has been spoken in close contact with Wastekan Maya for centuries. Here, most of the proposed borrowings seem completely plausible, but a couple to me suggest the direction of borrowing to be the opposite of what is assumed by Kaufman.

For example the Wastekan word kw’itš’a “grind in mortar” which Kaufman proposes as the source of Huastecan Nahua tekwicha “pestle” seems likely to be related to the Nahuan word for grinding kwečoa “to grind” from PN kwe¢iwa, and related to Huichol rakwi¢i “nixtamal”, Cora kwei¢i “dough” – suggesting a loan from Nahuan into Wasteko.

The Wastek word molik “elbow” is suggestive, but it is not restricted to Huastecan Nahuatl as Kaufman implies, it is found also in western Nahua branch (and as molic in Molina’s dictionary). This suggests either borrowing into Wastek from Nahuan or an additional example of Wastek contact with PN. Given the otherwise unconvincing evidence for Wastek/proto-Nahuan contact, it is probably best to see the default hypothesis as a loan from Nahuan into Wastek. The proposed borrowing of Wastek či’im “maguey juice” as čiimiƛ “mothers milk” in Wastek Nahuan is unlikely, since Cora has ¢i’imé “mothers milk” suggesting again borrowing in the opposite direction.

 In the paper itself, Kaufman states that Mesoamerican languages are seemingly reluctant to borrow and that therefor any situation in which a language is permeated by borrowings shows very intense contact. I think the review of the paper suggests that proto-Nahuan was not permeated with borrowings from Wastekan, Totonakan and Mixe-Zoquean.

References Cited: 

*Dakin, K., & Wichmann, S. (2000). Cacao and chocolate. Ancient Mesoamerica11(1), 55-75.

*Kaufman, T. (2001). The history of the Nawa language group from the earliest times to the sixteenth century: Some initial results. Paper posted online at http://www. albany. edu/anthro/maldp/Nawa. pdf. University of Pittsburgh.

6 kommentarer:

  1. I've never seen ¢ in a linguistic transcription before. How is it thought to have been pronounced?

    1. I just use that instead of c for ts. It's a sub convention of APA that I've learned in Mesoamericanist historical linguistics - in the olden typewriting days they would write the slash through the c's by hand before publishing the manuscripts.

  2. Regarding the Wastekan loans:
    1. N. čočolli – W. čočob 'deer-foot': assuming the /b/ was borrowed as /p/ which was later lost, is that consistent with the absolutive -lli? Also, what is "deer-foot"?

    Regarding the Totonacan loans:
    2. N. wapo-ƛ – T. waapa 'fish sp.': Why wouldn't this be borrowed as something like wāpa-ƛ? What kind of fish is it in Nahuatl?
    3. N. tex-ƛi – T. tiiš 'brother-in-law'. Wouldn't it be borrowed as tīš-ƛi? Also, I see in Stubbs that this kind of term does not have a widespread etymon in UA. In other words, there are still a number of in-law terms in UA without much of an etymology. I would guess that they have internal etymologies, just not very easy ones to guess. With so few loans, I wouldn't think kin terminology is too likely to be one of them.
    4. And on the same topic, if there are just a couple of loans from Totonac to Nahuan, the level of linguistic acculturation between the two was very slight. Given that, how do you justify ƛ coming in from Totonac?

  3. If we are to accept both the deer-foot and the fish etymologies, are the environments of the *p similar enough that we need to set the relative chronology Wastekan contact, then p-loss, then Totonac contact?

  4. A deer foot is the foot of a deer. They are used as ceremonial objects I think. I think Kaufman considers that borrowings can be borrowed with phonological changes that are no necessarily rule-structured- for example leaving out segments or changing them semi-randomly. Whereas inherited words have to follow the sound changes.

    The idea of p loss being diagnostic for inherited versus borrowed words bothers me a lot, because it seems to be simply wrong. Nahuatl has literally dozens of inherited words with p in them, both word initially and in "protected" positions inside thre word. The reality is that only some p's dissappeared, some of them became y and some stayed p. I think that Langacker's idea of something like a fortis/lenis distinction in PUA is probably required to account for the different paths of p. In any case, as I see it the reduction of p to h in many words is shared with Coracholan and must have happened much before proto-Nahuan moved into contact with any Mesoamerican languages.