Notes on Ixhuatlancillo Nahuatl: An Exercise in Dialect Classification

Candelaria ceremony in Ixhuatlancillo,
from Spanish Wikipedia 

Ixhuatlancillo Nahuatl is spoken in the municipality of Ixhuatlancillo, Veracruz, just northwest of the city of Orizaba. There are some 8000 speakers of all ages in the municipality, most of them around the town center, whereas the outskirts are primarily Spanish speaking recent arrivals.  The last week I have been living in Ixhuatlancillo, carrying out interviews and studying the language, and here I give a little taste of the language, while trying to classify its position among other Nahuan varieties.

When trying to analyze and understand any new variety of Nahuatl, an important part of the analysis requires us to compare with other varieties that we know and to establish if the traits that make a given variety different and unique are evidence of its historical roots or evidence of innovation. There are no varieties that do not innovate, and there are also no varieties that do not contain retentions of ancient variation. However to group two varieties together as being descendants of a common ancestral variety we have to see that they share some trait that can best be understood as a shared innovation, rather than as independent innovations in the two varieties. Some innovations are extremely common in Nahuatl and in the world's languages in general, and they are not good evidence for shared history, whereas other types of innovations are uncommon and therefore suggest that the innovation happened in a shared ancestor of the varieties that have them. This is the kind of analysis I will undertake here.

Two of Ixhuatlancillo Nahuatl's characteristic traits.

To me the trait that stand out the most in Ixhuatlancillo Nahuatl relative to the neighboring varieties is that in Ixhuatlancillo the speakers pronounce a glottal stop [Ɂ] where the Zongolica varieties tend to use a glottal fricative [h]. The glottal stop is well known among Nahuatl scholars because it is the likely pronunciation of the so-called saltillo, in the colonial Nahuatl variety of Tenochtitlan/Mexico city - the so-called classical Nahuatl language. The presence of the glottal stop in Ixhuatlancillo, might therefore immediately be interpreted as a connection to the varieties of central Mexico. But analysis shows us that this is not the case, the saltillo in Ixhuatlancillo is an independent innovation that comes from a previous /h/.


In Ixhuatlancillo the glottal stop occurs in words such as [teɁwatsin] "you (with respect)" which in Zongolica dialects such as Tlaquilpa is [tehwatsin], but which was also [teɁwatsin]  in Colonial Tenochtitlan Nahuatl. But in Ixhuatlancillo the glottal stop also appears in words such as [koɁjoh] "forest" and [xiɁjoh] "leaf" in which the glottal stop corresponds to [kwafjoh] and [ʃifjoh] in Tlaquilpa, and to [kʷaʍjoɁ] and [ʃiʍjoɁ]. Here the glottal stop corresponds to an [f] or [ʍ] - both of which are allophonically devoiced versions of the phoneme /w/ in either variety. This correspondence may seem odd since there is no obvious reason a /w/ would turn into a glottal stop. The solution to this question however is that in other dialects, some in Zongolica, some in Puebla and some in Morelos (e.g. Hueyapan), the devoiced allophone of /w/ is [h]. Hueyapan for example has [kohyoh] and [ʃihjoh] in the corresponding words. The simplest explanation then is that in the ancestor language of Ixhuatlancillo Nahuatl, as in Hueyapan, /w/ was changed to [h] before unvoiced consonants and that [h] then subsequently changed to [Ɂ]. Furthermore the [h] pronunciation is in fact still found word finally after vowel, for example in the plural ending distinguishing /cho:ka/ "he cries" from /cho:kah/ "they cry". Interestingly, it seems that sometimes long vowels are glottalized in a way that make them sound almost as a short vowel plus glottal stop, so in fact /cho:kah/ often ends up pronounced [choɁkah], further complicating the picture. 

There is one more feature of Ixhuatlancillo Nahuatl that may trick us into positing a connection with a specific group of varieties.  An important feature that distinguishes two main groups of dialects, the eastern and western  dialects, is whether they have /i/ or /e/ in a small group of words including tesi/tisi "to grind", ihte/ihti "stomach",  iste/isti "fingernails" and sentli/sintli "corncob". Ixhuatlancillo Nahuatl has i in all these words suggesting a relation to the Eastern group of dialects. But again appearances may be tricky unless we take into account what else is going on in the language we're looking at. It turns out that Ixhuatlancillo also has /i/ in a number of other words in which most varieties, both Eastern and Western, have /e/. For example tikitl "work", titiki "cut",  tipe:tl "mountain",  nimi "walk", niki "want",  timoa "descend" and kitza "stand up" - all of which have /e/ in the first syllable in most varieties. This suggests that rather than being a retention of /i/ from the ancient proto-Eastern language, the /i/ in these words in Ixhuatlancilli may rather be an indepedendent innovation that has turned a previous /e/ into /i/. But to be sure whether this is the case we have to know the conditions under which this change happened - because obviously not all /e/'s have turned into /i/'s.  In this case all of the changes seem to affect the first vowel in a lexical root, since most of the roots are bisyllabic this is also the syllable that tends to be stressed (depending on the suffixes it may take). It is very common in the world's languages for stressed vowels to be raised like that. But if this is the case then we would also expect monosyllabic roots to be affected since they are likely to have stress on their only syllable - this also turns out to be the case. Monosyllabic roots such as tlitl "fire", titl "stone" and yitl "beans" also have /i/ instead of the /e/ expected in most dialects.  Two further arguments support the innovative analysis: 1. only short vowels are affected by the change, e:wa "to sit up", pe:wi "to begin" and ye:wayoɁ "skin/bark" all have the expected long /e:/; 2. in reduplicated forms the /e/ surfaces in the reduplicated syllable so that we get /yeɁyikoa / "to practice" and /te:titiki/ "to cut repeatedly". This means that in all likelihood at some point in its history Ixhuatlancillo Nahuatl has changed the vowel /e/ to /i/ in stressed (or root initial) syllables, and which means that we can't ascertain whether the language originally had /i/ or /e/ in the words that are diagnostic of the split between Eastern and Western dialects. However there is one exception, one of the diagnostic words, the word for stomach, has the diagnostic vowel in the second syllable - in Ixhuatlancillo /iɁti/. So maybe this single word could be a sign of the language having roots in the Eastern dialect area after all. But then again the other word that has the diagnostic vowel in the second syllable is iste "fingernail", which has /e/, grouping it with the Western dialects. So in this case the evidence turns out to be completely inconclusive as to the Ixhuatlancillo Nahuatl's membership in either the Eastern i-branch or the Western e-branch.

Other dialectally significant features:The reflexive subject prefix is -mo-  for all persons, suggesting that this area of the language hasn't been influenced by the central dialects of the Mexico valley that have no- for the first person singular "myself" and to- for the first person plural "ourselves". 

The pronoun system has forms with the root  /eɁwa/ suggesting a relation to the Western dialects instead of with the Eastern ones that tend to have the root /eha/ or /aha/.

 The negation is aɁmo in all cases, suggesting an affiliation with the Western branch rather than with the Eastern branch that has a number of different negations.