Nahuatl Scholar Interview: R. Joe Campbell

Believe it or not, this was Joe's first selfie.
It's Joe on the left.
(taken in Guadalajara July, 2018)

In this blogpost, I give you an interview with one of the great Nahuatl scholars alive today, R. Joe Campbell. I have had the good fortune to meet Joe a number of times at Nahuatl meetings, and in 2018 when we met at the Uto-Aztecan meetings in Guadalajara Joe and I hung out together for a couple of days after the meetings. I asked if I could do an interview with him for my blog, and he agreed. Then the interview sat there for awhile, so I asked Joe if he wanted to update it with some more details, which he did. And here it is, the first in a series of interviews with great Nahuatl scholars.

Joe is one of the popularizers of the Nahuatl orthography system sometimes known as "Andrews-Campbell-Karttunen" orthography, which was first developed by J. Richard Andrews, subsequently employed in Joe's and Fran Karttunen’s influential "Foundation Course in Nahuatl Grammar", and Frances Karttunen's "Analytical Dictionary". (His first major work, “A Morphological Dictionary of Classical Nahuatl” came out in 1985 and precedes his use of the Andrews orthography.) He is also a pioneer in computerized Nahuatl study, having begun working on Nahuatl computerized lexicography back when computers were the size of a building and used punch cards. He is still working on two major works: an integrated version of all of Alonso de Molina's three dictionaries, analyzed into individual morphemes and with the possibility of running all kinds of searches for combinations of different morphemes and a dictionary combining the vocabulary of both Molina and the Florentine Codex.

Joe has worked with Nahuatl for more than 50 years, and in this interview he tells a little about his career in Nahuatl studies and his different contributions to the field. 

MPH: So Joe, would you mind telling a little bit about how you got into Nahuatl in the first place and when?

RJC: OK, In 1962, I was a graduate student at the University of Illinois studying Spanish linguistics. And I took a couple of courses with Ken Hale, who was a young PhD, fresh out of Indiana University studying with Carl Voegelin, and in the spring of 1962, he said "Joe, some of the anthropology students are going to Mexico this summer and they're going to study Nahuatl. We're going to Tepoztlán and they're going to do fieldwork." And he said "you could go along, and would you be interested in going along and studying the Spanish dialect of Tepoztlán and see what the effects of Nahuatl were on the Spanish?" And I said "Oh, I'd really like to do that." And so, my then-wife and I and two daughters came to Mexico, and they went to spend the summer with her parents in Guadalajara and I went directly to Tepoztlán. And Ken and I went up to San Juan Tlacotenco at the top of the mountain, and while we were there we met a guy that spoke Nahuatl fluently -- one of the youngest people in town that did -- met him, made friends with him. And then later on Ken took me to see a rather elderly man who had fought during the revolution and sat me down to study with him, and as we worked along I found that he was a very nice guy, but he was very impatient. I didn't have a good ear at the time. I think it has gotten a lot better now. But when he said, I asked him how to say "camino" and he said "ohtli", and I said "otli", and he said "no! ohtli!", and I said "sí, otli". And with his reaction, I saw that he and I probably couldn't work together. We turned out to be very good friends all summer, and I found a much

Joe working with a friend from Tepoztlán

younger person to work with, with considerably more patience.

MPH: So what did you do when you came back from Tepoztlán? How did you continue with you Nahuatl studies after your first visit?

RJC:I wrote a paper for Ken, and then with the things I had to do, write the thesis at Illinois and search for a job at Indiana, Nahuatl sort of got put on the backburner. I got my first job at Indiana University and after I had been there well, like five years, one of the young professors came down the hall and he was doing legwork for the chairman, looking for a federal grant. Because the federal government was interested in doing unusual languages, languages that were not likely to be spoken by ...quote ..."the enemy". And he said, if we apply for a grant and say that we can offer Nahuatl, would you be willing to teach it? And I said, "well I am really not prepared to teach it on the basis of six weeks of experience", and also with the intervening years. And so I said "I would be willing to do it if I had an informant, if I had somebody to work with me in the class". And they came back to me and said "We're more than willing to do that; we'll pay for an informant and all the expenses and everything, but ... you would need to go find somebody". And to make a long story short, I was going to be in Mexico all the summer of 1970, and I went to Tepoztlan -- a friend of mine in Tepoztlan, Karen Dakin -- and she paired me up with a friend of hers, the anthropologist Judith Friedlander, and we went to Hueyapan and we found a young lady who was very happy to go to the United States and work for a year, that was Elvira [Hérnandez]. 

MPH: So you really learned Nahuatl when you were teaching it with Doña Elvira Hernández from Hueyapan?

Joe with Elvira Hernández in Hueyapan

RJC: That is exactly right. She and I, we started at the beginning of the school year. And she educated me in class, and then we had three hours a week out of class, and it was just like having an intensive teacher.

MPH: And when did you start getting into studying the colonial version of the language?

RJC: When Elvira came back to Mexico, I taught Nahuatl for two more years by myself and used sources like González Casanova's folktales and a lot of structural material. But during those two years I started realizing that if I really wanted to get more details of Nahuatl I would have to look at printed sources. So I started putting the Molina 1571 Nahuatl-Spanish half dictionary on punched cards. And it ended up that the whole second half of 1571 fit on something like 20,000 cards. That's ten boxes of cards. And then from there on, I worked in San Antonio with the whole card deck, inserting English translations into the boxes. So it swelled up, and so in one year -- from I think it was 75-76 -- I managed to translate all of the Spanish of the second half into English. And while I was doing that I was realizing that I should also enter the record by means of another punched card with the morphology codes typed out on that card.

MPH: So maybe you can explain a little, what is a perforated card and what is a morphology code?

Perforated card used for early digital computing.

RJC: A punched card was a card with eighty columns for punching holes in different combinations to represent letters. And people who used them actually started understanding what the perforations meant: There was a top row and a second row and the top row and second row then correlated with a set down below of one through ten and a combination of top and 1 was an A and top and 2 was a B and so on. 

MPH: Ok, so it was a very simple form of machine readable code? 

RJC: Yes, it weighed a lot more than a floppy disk.

MPH: So basically it would be fair to say that you were probably one of the first people to start using computational technology for working with Nahuatl?

RJC: I was the first one that I knew of, and I don't remember anybody else doing it while I was working on it in the seventies. But here you must remember that when we’re talking about punch cards, ‘computerized’ only means how we stored and re-arranges data. I have seen more than one suggestion that somehow the morphological analyses are done “by computer” and adds. They are NOT. Every morphological analysis is my considered opinion. The computer really has nothing to do with it.

MPH: So, the first work of yours that I became aware of was an article that you had in IJAL in 1976, about the phonology of Hueyapan Nahuatl, which I read for the first time before I went to do field work in Hueyapan myself in 2003. Could you summarize how that article came about and what is the interesting part of the argument?

RJC: I am not sure if I can remember the data very well, but I was working with paradigms that I learned from Elvira and I was trying to come up with a very logical description of the morphology. And it turns out that when you look at the representation of some words it is very straight forward: there is part one, two and three and they sound the same in either present tense or preterit. But then there was some verbs that looked like they had an underlying /w/ that was phonetically a [v] that in the preterit would turn into a velar [ŋ]. And, I thought well it's underlyingly a /w/ its phonetically a [v] and in another form it is an [n], what kind of segment could that possibly be? So because of the influence of an article by Sol Saporta in Spanish where he talked about the problem of palatal l and l and y in Spanish, in words like [mil] "thousand" and [miyón] "million" [this article], I came up with the idea that what that segment was could probably be a /ŋw/ a velar nasal with a labial release. So that would explain why Hueyapan Nahuatl had the present tense form [noga:va] "I stay", and the preterit form  [onoga:ŋ] "I stayed". 

MPH: Then another really important thing you've done has been the foundation course. How did that come about?

RJC: When I went back, well I went from San Antonio to Nebraska and stayed in Nebraska for two years, and then I went back to Indiana. And they had the Nahuatl course waiting for me when I got there and asked me to do that. And so, I didn't have a textbook, and I thought, rather than do a sort of struggling course the way I did in the early 70s, I thought what I ought to do would be to write simply a series of disconnected lessons, starting with the most simple things and doing lots of exercises where people had to do recognition and recall. And I started writing as many lessons as it took to get through the semester. And as I gave the course a couple of times more during the 80s, I wrote more and more lessons. -- Not a lot of explanation, because the explanations were mainly things that I did in class. But I gave them a lot of heavy take-home stuff to work on and come back and build up their memory on things. And as a matter of fact in 1987, I had already shown this rather big stack of exercises to Fran Karttunen who had been a friend of mine for a long long time. And at one point Fran said - that was in 1987 - she was going to have a government funded institute in the summer of ’89, and would I teach the modern dialect and she would do the classical track and could we use a selection from my set of exercises. And if we could, she would supply a denser set of explanations so it would be easier to work on at home; and she would also put in the vowel length which I had not taken the trouble to do. And that's what we did, we used it for a six week National Endowment for the Humanities Summer Institute in Austin Texas. 

MPH: And then the other thing is that when you finished the translation of the 1571 Molina in 1975, that was not at all the last thing you did on the Molinas. Can you tell us a little about what you have been doing with the Molina dictionary since then? 

Frontispiece of Molina's 1571 vocabulario

RJC:The translation of the second part of the 1571 dictionary was just the first step. I followed it up by preparing a morphological analysis of the Nahuatl vocabulary of the Nahuatl-Spanish dictionary and doing the programming to make it possible to sort all Nahuatl entries along with their morphological analyses and English translations under each of their component morphemes.

I happened to show the resulting print-out to professors Lloyd Kasten and John Nitti of the University of Wisconsin during a reception at our house following a lecture that they gave in Bloomington involving the use of computing technology in the study and preparation of old documents. This was probably in 1984. They expressed interest in having the Hispanic Seminary of Medieval Studies publish the morpheme index and gave me explicit instructions to make it possible for me to insert the printing codes into my data so that the final publication came directly from the computer tape that I submitted to them -- by this time I had moved past cards-- since re-setting something so complicated would have been impossible. Somewhere there is a picture of me standing at the Post Office window posing with the 12x12 inch package in my hand. The result was my A Morphological Dictionary of Classical Nahuatl: a morpheme index to the vocabulario en lengua mexicana y castellana of Fray Alonso de Molina (Madison, 1985). 

The morphology I did in the Morphological Dictionary is somewhat more rudimentary than what I am doing now. For instance if you have a word like <tlacatl>, that is a simple morpheme (well actually, it’s two, but people don’t think much about the -tl being separate), but if you have a word like <nitetlacatilia> there are actually five pieces. In the Morphological Dictionary I just coded this as tlacatl-verb-caus for the noun tlaca-tl, the verb-forming morpheme -ti- (become) and the causative -lia. I did not code the ni- (I) -te- (someone) or the other personal prefixes at all. 

Later I made up numerical codes for the prefixes that would set them up in ways that were reasonable instead of spelling them out and I made a number of separate categories for the verb-forming morphemes as well as for the causative and benefactive morphemes and the preterits. Today the morphology for nitetlacatilia is p11-p53-tla:catl-v01a-caus04. I still treat the absolutive suffixes as part of the spelling for the noun morphemes because it makes the nouns more recognizable for most people.

Since I had done the English translation of the 1571 dictionary...The 1571 dictionary is really two dictionaries and I call them Molina 1 and Molina 2 - and I had only worked on Molina 2 - the Nahuatl-Spanish. And so I thought that what I should do would be to put in the 1555 and the first half, the Molina 1, into machine readable form. By then I had graduated from punched cards to putting them in in electronic mode - just keying them in, which struck me as a rather strange thing to do rather than using cards, but I did that. And I by-passed translating the other two dictionaries in favor of spending all my time inserting morpheme codes after each entry in all three Molina dictionaries. My goal here is what I call The Integrated Molina Dictionary -- the gathered Nahuatl entries with their morphological analyses, their Spanish equivalents and, where I have it from M2, the English translations of the Spanish. This is actually very nearly finished, except that I keep setting it aside to work on the Florentine Codex vocabulary.

MPH: So basically, rather than just translating the Spanish translation of the Nahuatl word, what you did was to also analyze the Nahuatl word piece for piece and type it up as part of the entries for the three Molina dictionaries. So about how many entries is that for the entire Molina corpus?

RJC: Well, actually, it wasn’t “in addition” to translating. I only translated the Spanish of Molina2. Ok, I can't remember off the top of my head. But the Molina 2, the Nahuatl-Spanish dictionary has 23,623 words, and the 1555 has something over 15,000 and I don't remember what the Molina 1, the Spanish-Nahuatl, has - but it is somewhere in between those figures. And of course one of the kind of things I did was to write a program to sort Molina 1555 and Molina 1 and 2 together on lexical items, and then I had a cluster that showed what each dictionary said about any given item.

MPH: So how about many entries do you have in your morpheme database at this point?

RJC: So I am a little bit unclear on this, because I also have an extra little ambitious thing that I did, which was to also code items for various kinds of things; say I would code them for an error that Molina did in parsing things, for example when Molina stripped the prefixes of a verb he would sometimes he would cut off the beginning of the verb itself. So I put in a code so that I could gather all of those things. And I was also interested in the fact that there was a certain amount of spelling variation like the dropping of n, y and w, and so I coded for either dropping or retention. And if you simply add all of the codes for those different kinds of notes plus the morphemes, right now, my "morpheme" count is around one million one hundred thousand.

MPH: That's a lot of words. Can you tell us briefly some of the cool things that your program can do with that data? The type of searches you can run and the kind of information you can extract?

RJC: The very first thing I was aiming at was being able to give an alphabetical index of all the words in the three Molinas integrated, and that is pure vanilla dictionary construction. But the next thing I wanted to do was to compile a list of morphemes so that anybody could look up alphabetically in a list of morphemes and find all the words in any one of the three Molinas that had that morpheme. So from the very beginning I was constructing a morpheme index of the whole thing. And then because of some of the advantages I saw in looking at the whole vocabulary from the end of the word, I also wrote a program that could give you a backwards index. So for example everything that ended in -tilia, you'd find every single word that had that. Parallelly and not meaning to get distracted fom Molina, but yes I did get distracted from Molina, I also together with a friend of mine from France, Marc Eisinger, we put the Nahuatl of the Florentine Codex into machine readable form, we each did one book and another, and then he did the entire verification of all the Florentine Codex, just a monstrous job. (Verification is when you re-enter data so that a program can find where one copy differs from the other.) And then I spent three years in the late eighties integrating the Nahuatl and the Dibble and Anderson English translation - it took three years to get them to match sentences. And then I spent a long time adding a regularized spelling line to the Florentine entries so that the computer could bring together instances of a word that was spelled differently in different places. And then, I could correlate words from the Florentine with words from Molina, that's one thing, and then the other thing independently from Molina, since the Florentine was text it was sentences in contexts, it would be interesting for anyone to see how a given word was used, in context, so I wrote a concordance program that will allow you to interrogate the Florentine database for any string that you wanted. For instance... well, there were two different things: one would construct an index based on any alphabetical representation so you could look for "tlacat" or you could extend it and look for "tlacatil", whatever... But the other thing was, and certainly based on favors that people did me in enlightening me - Fred Householder at Indiana had a computer colleague that had been doing some work on texts and he had the idea of hanging codes on words in context… He wasn't working on Nahuatl but I turned it into working on Nahuatl. If you wanted to look at anything that was related to <tlacatl> then you could hang a morpheme under that and you could make an index of every morpheme and then you would get the word in-text, in context. So what I did was to label, I had to fill in for every word that I could, and that means a lot that I couldn't, I made every word in the sentence a key to that sentence and then every translation of those given Nahuatl words was a key to the Nahuatl word which was a key to the sentence. In this way I could retrieve a lot of interesting information.

MPH: So you have been working on this for 43 years. When are you going to be done, and what is the final product going to look like?

RJC: I originally hoped... Well I keep changing my mind about what the goal could possibly be. The original goal was to do Molina. But then I started realizing that it would be a lot more valuable to do a combined study of the Molina and the Florentine. And then I started realizing that it was very unlikely that I would have time in life to do all of it, so I said Ok, I will just do Molina. But while I have been working on Molina I never stopped analyzing sections of the Florentine. As a matter of fact it is not sections of the Florentine, but I look for a given thing in the Florentine, and I go through and find every token of it and mark it for whatever it needs - morphology, phonetic variation and things like that. And if there are errors in the Dibble and Anderson translation, I put in a note on every sentence where I find that going on. And every time I do, I always think, I am in awe of what they did, but we are all humans, we all make mistakes.

MPH: Yes, and we can keep improving on it. Thank you very much Joe, for telling us about your work.

RJC: You're welcome, I enjoyed it.

---

Here, we end the blog with a list of some of Joe's most significant publications (several coauthored with his wife Mary Clayton who is also a Nahuatl scholar):

• Campbell, R. J. (1976). Underlying/ŋw/in Hueyapan nahuatl. International Journal of American Linguistics, 42(1), pp. 46-50.

• Campbell, R. J. (1985). A Morphological Dictionary of Classical Nahuatl: A Morpheme Index to the Vocabulario en Lengua Mexicana Y Castellana of Fray Alonso de Molina. Hispanic Seminary of Medieval Studies. University of Texas.

• Campbell, R. J., & Clayton, M. L. (1988). Bernardino de Sahagún’s Contributions to the Lexicon of Classical Nahuatl. The Work of Bernardino de Sahagún: Pioneer Ethnographer of Sixteenth-Century Aztec Mexico, pp. 295-314.

• Campbell, R. J., & Karttunen, F. E. (1989). Foundation course in Nahuatl grammar (Vol. 1). Institute of Latin American Studies, The University of Texas at Austin.

• Clayton, Mary L., and R. Joe Campbell. (2002) "Alonso de Molina as Lexicographer." Making Dictionaries: Preserving Indigenous Languages of the Americas. pp. 336-90.

Tongues of Aztlan: The Nahua migrations and dialectology

In my last post I showed the basic split between eastern and western Nahuatl dialects, and how different dialects could be classified as belonging to either of the two branches. In this post I summarize a talk I gave at the Mesoamerican conference at CalState L.A. this past saturday, in which I gave my interpretation of the history of migrations that brought the Nahuas to their current locations from their ancestral homeland.

My narrative is based on three avenues of evidence:

1. Linguistic evidence. By analyzing how the structure of shared linguistic innovations between dialects reflect historical relationships and splits, give us a way to understand how ancestral dialect groups moved, split and moved again. 
2. Ethnohistorical evidence. By analyzing accounts told by the Nahuas about their own past we can find clues to how they themselves understood the relaiton between groups of speakers and their historical and geographic relations.
3. Archeological evidence. By analyzing material culture in different locations we can trace population movements and cultural innovations or changes. 

There are several different views of where the Nahuan languages originated:

Terrence Kaufman 2001 thinks the proto-Nahuan language must have been spoken somewhere North of the Huasteca region in Tamaulipas or San Luis Potosi before entereing the basin of Mexico where according to his interpretation they the Western group split off.  His main argument for this location is what he sees as loanwords in proto-Nahuan from Wastek/Huastec Mayan (Teenek) and the Otomanguean language Pame. 

Christensen and Beekman 2003, rather see the homeland in the Bajio region of Zacatecas, Jalisco, Aguacalientes and Queretaro. This view they base on the fact tha they have archeologically documented a large population movement out of this region into central Mexico starting in about 500AD. 

Other, less well supported suggestions for the location of the ancestral Proto-Nahuan speech community include the coast of Nayarit and even Durango/Sonora/Chihuahua. I follow Christensen and Beekman, both because I disagree with Kaufman regarding the Teenek influence in proto-Nahuan which I find negligible, and also because my own investigations have found more commonalities between Nahuan and Cora (Naayerite) and Huichol (Wixarika), suggesting to me a more westerly origin in contact with these languages. My interpretation of contact in Nahuan prehistory is thus more consonant with Christensen and Beekman than with Kaufman's account. 

This then, using the three kinds of evidence and the Bajio as a place of origin, provides us with the following narrative: 

Late Classic (500-600AD):

By the late classic Speakers of Eastern Nahuatl are entering Mesoamerica from the Bajio, drawn by the metropolis of Teotihuacan. They settle in a broad area across the central highlands from the Huastec region to Guererro. Possibly the Huastec and Guerrero branches already split from the other Eastern languages - or mor likely they formed a dialect continuum, which was only subsequently broken apart.

Epiclassic (600-800AD):

Then at the fall of Teotihuacan around 600AD the demographic weight of Central Mexico shifts towards the new centers of Cholula/Cacaxtla, and Xochicalco. Here Eastern Nahuas come into contact with Maya speakers - Maya influence in the Cholula region in this period is well documented. Perhaps, including the ancestors of the Teenek Maya people, if we assume that they had not yet arrived in their current location (this is in contrast to Kaufman's view of Huastecs arriving in their current position much earlier than that, but consonant with Robertson and Houston's argument for a much later Huastec migration). Cholula is one of the centers from which the religion based around the worship of Quetzalcoatl spreads, southeast in to the Maya region and west into the highlands.

Early Post-Classic (800-1000)

In this period Western Nahuas, who had stayed behind in contact with Naayerite, Wixarika and Otomi people move into Central Mexico settling at the site of Tula, Hidalgo. Previous scholars such as Una Canger 1988 and Terrence Kaufman have seen Tula as a center of dispersal of Eastern Nahuatl. I disagree, since I consider this date to be much too late, and because there are no Eastern varities spoken in the area of Tula, only Western ones and Otomi. Kaufman consider the Otomies to be a later intrusion in the area - I know of no convincing evidence for this. Rather I think that clearly the Western Nahuas were closely aligned with Otomies (e.g. the Acolhua of Texcoco were multiethnic Otomi-Nahuas even in the14th century), and probably Tula, was composed of both Otomi and Nahua populations. The many important calques between Otomi and Nahua described by David Wright Carr, date to this period, and I believe a number of loans in both directions as well.

Then after settling in Tula, the Western Nahuas who use /ahmo/ as their only negation word, move into the Central Mexico Valley and arrive in Cholula. Here they rout the Eastern Nahuas and Mayas who live there, and make them flee towards the gulf coast, towards the province of Olman-Xicallanco. This event is suggested by the account in the Historia Tolteca-Chichimeca which tells us about the invasion of Tolteca-Chichimeca who rout the Olmeca-Xoiclanca from Cholula. It is also supported by the work of Geoffrey McCafferty whose excavations at Cholula have demonstrated warfare and a change in the material culture of the site around the time of the entry of the Eastern Nahuas (his Olmeca-Xicallanca) in the 7th century and at the time of the entry of the Western Nahuas (his Tolteca-Chichimeca) in the 9th century. My argument here is that the Olmeca-Xicallanca only came into existence when they left Cholula, for Olman-Xicallanco - that is they were named after the destination where they went after leaving Cholula, and not for their origin before arriving there. The arrival of the Western Nahuas split the Eastern Nahuas into three groups - the Isthmian group who settled the gulf coast of southern Veracruz and well into Tabasco, and the North-Eastern Huastec-Nahuas who settled in the Sierra Huasteca.  The Sierra de Puebla varieties also split off at this time. The Isthmian and Sierra de Puebla groups subsequently changed /tl/ to /t/, developed new grammatical strategies for negation, and began dropping their /y/s wordinitially before /e/.  Perhaps the fact that one of the Eastern group of Nahuas the Isthmians, changed their /tl/ to  /t/ is due to Maya influence, since there is no /tl/ sound in Mayan, and Maya speakers must have found it hard to pronounce. The Isthmian Nahuas, also seem to have had the closest ties to the Chontal Mayas in the Tabasco region where both languages are still spoken in close contact.

Another group of Western Nahuas split from the central group going into Mexico state, backtracking towards the pacific coast and north into Michoacan, Colima, Jalisco, Nayarit and Durango. They became the Western Peiphery. They are characterized by a frequent if sporadic change of /tl/ to /l/, and the use of -lo for the plural subject, and a number of other innovations. 

Postclassic: 1000-1500AD

The Postclassic period sees the continued dominance of Western Nahuas centered in the Mexico Valley, the rise of Azcapotzalco, and later the triple alliance. During this period Western features extend into most of the historically Eastern varieties, creating a zone of diffusion in the central Nahua speaking area  (from the sierra Zongolica across Southern Puebla into Morelos and Central Guerrero) where Western and Eastern traits mingle freely. With the rise of Tenochtitlan, Western influence reaches its maximum extent and, except for Pipil which has by then reached in El Salvador and Guatemala, no Eastern varieties escape being influenced by Western Nahuatl.

Conclusions:

This narrative leads us to the current distribution with clearly demarcated Western and Eastern peripheries, and a large central diffusion area marked by dialect mixture and koineization (as suggested by Canger 2011).  My narrative builds on and is largely consonant with most other studies, and mostly differ in the timing and in the identification of Tula as the center of the Western dispersal and Cholula as the center of Eastern dispersal. 

Bibliography:

Canger, U. (1988). Nahuatl dialectology: A survey and some suggestions.International Journal of American Linguistics, 28-72.

Canger, U. (2011). El nauatl urbano de Tlatelolco/Tenochtitlan, resultado de convergencia entre dialectos: Con un esbozo brevísimo de la historia de los dialectos. Estudios de cultura Náhuatl, 42, 243-258.

Canger, U., & Dakin, K. (1985). An inconspicuous basic split in Nahuatl.International journal of American linguistics, 358-361.

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.

Kaufman, T., & Justeson, J. (2009). Historical linguistics and pre-Columbian Mesoamerica. Ancient Mesoamerica, 20(02), 221-231.

Robertson, J., & Houston, S. (2015). The Huastec Problem. in Faust, K. A., & Richter, K. N. (Eds.). (2015). The Huasteca: Culture, History, and Interregional Exchange. University of Oklahoma Press.

McCafferty, G. G. (1996). Reinterpreting the great pyramid of Cholula, Mexico.Ancient Mesoamerica, 7(01), 1-17.

McCafferty, G. G. (1996). The ceramics and chronology of Cholula, Mexico.Ancient Mesoamerica, 7(02), 299-323.

Carr, D. C. W. (2008). La Sociedad Prehispánica en las Lenguas Náhuatl y Otomí◊. Acta Universitaria, 18(Esp), 15-23.

Eastern and Western Nahuatl Dialects

As noted in previous blogs, I am working on analyzing and understanding Nahuan dialectology, dialectology of course being the study of patterns of regional variation.

The way this works is that one looks at the linguistic traits that vary between how Nahuatl is spoken in different regions, and then classify the dialect of each region into the varieties that they are most similar to. Classification can be based simply on superficial similarity, grouping everything similar together, or it can be based on historical development, where you classify dialects together that seem to share a single origin because they display shared innovations.

Another way is to classify by region. Here the tricky part is that the number of possible regions is almost infinite. One could potentially classify each of the many thousand local communities where the 1,5 million Nahuatl speakers live each as a separate dialect. Because every local community has certain ways of speaking that set them off from their neighbors. Dialect areas then are areas regions where certain traits a more frequent than in other regions, with the caveat that not every dialect in the area necessarily has all of the features characteristic of the area were it is spoken, and some dialects in other regions may have them too.

In the literature on Nahuatl dialectology the basic division is both regional and historical. It divides all the Nahuan languages into two basic groups: The Eastern and the Western dialects. This grouping was first proposed by Una Canger and Karen Dakin in 1985. They had realized that some dialects systematically had the vowel /i/ in certain words where other dialects had /e/. They also noted that all of the words where this correspondence was found corresponded to words that had oeiginally had the vowel *u in Proto-Uto-Aztecan (PUA).  The words they focused on were tesi/tisi "to grind" (from PUA *tusu), sentli/sintli "corn" (from PUA *sunu), iste/isti "fingernail" (from PUA *sutu) and ihte/ihti "stomach" (which Daking and Canger reconstruct as coming from PUA *patu, I am personally a little dubious of this form). The i forms were generally found in the dialects in the extreme eastern range of the area where Nahuatl is spoken, and the e forms were generally found in the western extreme.

Later Dakin (2000) and Canger (1988) each added some new traits to each of the two areas. Canger noted that in the Western area, the prefix o- is used to mark the past tense, and in the Eastern area it is absent, and that Eastern dialects tend to put the adverb ok after the predicate it modifies whereas in the West it precedes it. Dakin noted that the Western dialects tend to have ye- in a number of forms where the Eastern dialects have e- (e.g. yetl/etl (bean), yestli/estli (blood), yeyi/eyi (three) and yepatl/epatl "skunk"). Dakin saw that each of these words originally began with *p in PUA, and she surmised that the *p had changed into /y/ in the West but had disappeared entirely in the East.

In a recent paper of my own I argue that the two dialect areas also differ in the way that they negate sentences. Western dialects all use the adverb ahmo or ammo as the negator in all sentences. But in Eastern Nahuatl other words are used, either instead of ahmo or in addition to ammo. For example in the Huasteca region the negator is the prefix ax-, in the Puebla Highlands it is kanah, in Guerrero it is ka, and in the isthmus it is aya', ate, the prefix ah- or in Tabasco the word até and in El Salvador nite, inte ot tesu. In many Eastern varieties there are more than one negator and the use depends on the mood of the clause, one may be used for imperatives, subjunctives and conditionals and another for realis/indicative sentences for example.

The map below shows the locations of Eastern (red) and Western (blue) dialects. It also includes the extinct Pochutec language which seems to have split from proto-Nahuan before the split between Eastern and Western varieties. I have colored the varieties of Central Guerrero purple, that is because they seem to me to be conservative sharing few of the innovations of the other areas, and possibly have split also very early perhaps before the East-West split. (Note also that the varieties in Chiapas went extinct in the early 20th century, and that there are very few speakers in Jalisco).

"Classical" and Central Nahuatl are mixed dialects!

So what about "Classical Nahuatl". Well first of all I don't think that is a very good name for the colonial Nahuatl that we know from so many sources. I think it is better to call it colonial Nahuatl. The best known variety of colonial Nahuatl is the one from Mexico City, based on the dialect of Tetzcoco. Canger (2011) has argued very convincingly that this dialect which she calls Urban Nauatl, was a mixture of Eastern and Western varieties, that arose as people form different dialect areas mingled in the great city of Tenochtitlan.

Definitely many of the Central varieties seem to have traits from both Eastern and Wstern branches, and it is sometimes difficult to classify them. This is probably because the Eastern varieties must have been in the area longest and Western dialects then arrived later imposing their prestige language on the earlier speakers of Eastern Nahuatl. This would mean that many of the Central dialects (e.g. in Puebla, Morelos, Zongolica) are Western dialects with an Eastern substrate.

Works cited:

*Canger, Una. 1988. “Nahuatl Dialectology: A Survey and Some Suggestions” International Journal of American Linguistics 54:28-73.
*Canger, Una. 1988b [1978]. Subgrupos de los dialectos Nahuas. In Smoke and Mist, Mesoamerican Studies in memory of Thelma Sullivan. Kathryn Josserand & Karen Dakin (eds.) B.A.R. International Series 402.
*Canger, Una. 2011. "El nauatl urbano de Tlatelolco/Tenochtitlan, resultado de convergencia entredialectos, con un esbozo brevísimo de la historia de los dialectos". Estudios de Cultura Náhuatl (Mexico: UNAM): 243–258.
*Canger, Una and Karen Dakin. 1985. An inconspicuous basic split in Nahuatl. International Journal of American Linguistics 51: 358–361.
*Dakin, Karen. 2000. Proto-Uto-Aztecan *p and the e/ye isogloss in Nahuatl Dialectology. In UtoAztecan: Structural, Temporal, and Geographic Perspectives. Eugene Casad & Thomas Willett (eds.) Hermosillo, Mexico: University of Sonora. pp. 213-19

Tabasco Nawat: A *not* extinct Nahuan variety

According to The Ethnologue, the world's largest catalogue of languages and dialects, the Nahuatl language of Tabasco is extinct - it no longer has any living speakers. [UPDATE: As of today 2/22/2015, two days after I posted this blog, Ethnologue has updated their listing of Tabasco Nahuatl, no longer categorizing it as extinct, and noting a population of 30.]

The Maya ruins of Comalcalco.
 Twenty minutes from where Nawat is still spoken.
Maya and Nawat has long coexisted in Tabasco, and also Ayapa Zoque
 is spoken within a ten minute drive.
Fermin's neighbor in fact speaks Chontal Maya.

This is something that would probably concern Fermín Cruz Álvarez and some 30-40 of his neighbors and family members in his community in the municipality of Comalcalco, Tabasco, since this is the language they grew up speaking. It is also the language that Fermín has spent his free time trying to promote the past 7 years. He has been teaching the language to youths and adults, and with the help of several linguists he wrote and published a small dictionary through CDI (Comisión del Desarrollo de los Pueblos Indígenas).

So the Tabasco Nahuatl language is not extinct. It does have native speakers, although there are very few of them, and the language is in imminent risk of disappearing unless someone assists Fermín and the community. His efforts to revive his language are hampered by the fact that he has to work, cultivating cacao, black pepper and fish to sustain his family, and that he receives little or no remuneration for his work.

Fermin Cruz with his daughter,
who already knows the numbers
 up to ten in Tabasco Nawat

I visited Fermin in the summer of 2014, and the rest of this blog post is dedicated to giving a brief sketch of Tabasco Nawat, and to argue that this is an imminently interesting and important variety of Nawat that really deserves support and recognition. And if you read all the way to the end, you will get a chance to listen to what the language sounds like and read a small sample text of Fermin telling about how to make cacao.

Some Background:

Nahua people have a long history in the south-east of Mexico. It probably arrived in Tabasco sometime before 800 AD, as Nahua speakers migrated south from Central Mexico along the gulf coast, through Chiapas, and Guatemala and into Central America. At the time of the Spanish invasion, Nahuan languages were spoken all along the gulf coast of Veracruz and into Tabasco, where the language coexisted with Mixe-Zoquean and Mayan languages. The municipality of Comalcalco, Tabasco is the location of an important classical period Mayan archaeological site, also named Comalcalco, which was probably built by the Chontal maya people who still inhabit the area. The Ayapa Zoque language, which has become famous because of a media hyped story about two of its last fluent speakers, is spoken less than fifteen minutes from Fermín's house by car. Tabasco has always been an important area in Mesoamerica – it was here that the Olmec culture flourished in the pre-classic period, and in the classic and post-classic periods it was an important site of trade, because of the abundance of Cacao and other tropical luxury goods. Comalcalco is itself a Nahuatl name, meaning “In the house of comales”, the comal being the round ceramic griddle used by many Mesoamericans for baking tortillas on.

Tabasco Nawat:

The small family of Nahuan languages has two main branches: The Western branch includes most of the dialects spoken in the Center of Mexico and on the Pacfific coast in Michoacan and in Durango, and historically in Jalisco, Zacatecas and Colima. The Eastern branch includes the Huastecan dialects spoken in north-eastern Mexico, and the dialects of central Puebla, and Southern Veracruz, as well as the Pipil Nawat language spoken in El Salvador – and many varieties that are no longer spoken such as those of Chiapas and Guatemala. The Tabasco variety belongs to the Eastern branch. Some of the characteristics of the Eastern branch is that they use the pronouns neha/naha ”I”, teha/taha “you” and yeha/yaha “he/she”. They also do not use the word “ahmo” as the only negation word, and most Eastern varieties do not use ahmo at all (Pharao Hansen 2014). And they are also characterized by having the vowel /i/ in a number of words where Western Nahuan has /e/ (Canger & Dakin 1985), and by only using the plural morpheme -meh or -met and not the of the morpheme -tin which is used in most Western varieties. Tabasco Nawat, along with the Nahuan languages of Central Puebla, Southern Veracruz, and the Pipil Nawat language of El Salvador also do not use the tl-sound which is such a famous characteristic of the so-called Classical Nahuatl language and most of the Nahuan languages spoken in the Center of Mexico, but instead have changed all their previous tl's to t. That is why I write Tabasco Nawat and not Tabasco Nawatl. This is not a feature shared by all the Eastern varieties, as for example Huastecan Nahuatl still has the /tl/ sound.

In terms of phonology, apart from the lack of /tl/ another interesting feature is that Tabasco Nawat has five vowel qualities /i, e, a, o, u/. Most other Nahuan languages have only four vowel qualities, as they do not have the vowel /u/ or if they have /u/ then they don't have o. But Tabasco Nawat has both. This is because it has not changed all its o's to /u/, but only some of them following a rule that I have not yet been able to fully work out. It seems that particularly long /o:/'s have become /u/. The language has not changed its final stop consonants /t/ and /k/ to glottal stops as has happened in the nearby Nahuan languages of southern Veracruz, and it has /h/ where “Classical Nahuatl” has the saltillo. It also voices the velar stop /k/ to /g/ both between vowels and even frequently at the beginning of words. Tabasco Nawat also has the sound /f/ corresponding to the consonant cluster /hw/ in other dialects. For example the word “feather” which in some other dialects is /ihwitl/ in Tabasco Nawat is /ifit/ (this is not uncommon in other varieties either, and is found both in Morelos and Zongolica).

Interestingly Tabasco Nawat also has a morphophonological process that turns /w/ into /n/ whenever it appears at the end of a word. For example the past tense of the verb /chowa/ “to do” is [ogichin] “he did it” (also note that the vowel varies, the /w/ having colored the /i/ to [o], but since the n does not color the preceding vowel the past form retains the vowel /i/. This strongly suggests that the rule turning /w#/ to /n#/ is old, or at least older than the change of /i/ to /o/). And when nouns take the possessive suffix that is /w/ in most other dialects here it is /-n/, for example “my wife” is /no-soa-n/, and “my name” is /no-tu:ga-n/. This morpho-phonological rule is shared with the dialects of Morelos and (some of) the dialects of the Zongolica region in Veracruz. It is kind of weird that this would be shared between these two areas, that are otherwise both linguistically and geographically quite far removed, and changing /w/ to /n/ is not so common in the languages of the world that one would expect it to appear independently (changes such as the voicing of [k] to [g] and the fusion of [hw] to [f] are much more common changes, and very likely to occur independently of each other in different varieties).

But the really interesting differences between Tabasco Nawat and other varieties are in the grammar of the language – both in the areas of morphology and syntax.

One really interesting aspect, is how plural subjects are marked. In most Nahuan languages plural subjects are marked on verbs using a suffix /-h/ or /Ɂ/. But this is not the case in Tabasco Nawat. Here  plural subject is unmarked for the third person, but marked with -lo for the first person, e.g. ti-k-pia-lo "we are guarding it", but ti-k-pia "you are guarding it". 

The language also doesnt have a  future tense with the suffix -s as in most other varieties, rather future is constructed with the suffix -tiin which in other varieties means "to be going to do X". Negation is expressed with the word até. And the progressive present is expressed periphrastically with the verb nemi which is also used as a general copula.

Below are some examples of phrases in Tabasco Nawat demonstrating most of these interesting features:

chinechtemoli in nokwach porkeh nimopatatiin

find me my clothes because I am going to change

niyah nitamatiin chinechtemoli notegun iwan se morral

I am going fishing find me my fishing rod and a morral

nehpa nemi one tagat yun nigittak yalla

there is the man I saw yesterday

one tagat yun wallahka yahkiya 

the man who came has already left

nigan nemi takwali yun mokwatiin 

here is the food that were going to eat

iwa:n one tagat yun yahkiya mokwepatiin 

the man who has left will come back

ka niga tigochitigiweh

where will we sleep?

Nigan gan tigochiskiah

this is where we were going to sleep

onehpa gan mottak one taagat

this is where we saw that man

nigan gan tigochkeh yallah

this is where we slept yesterday

yallah tichuugakkeh iwan ami nemilo welia

yesterday we cried, but today we are well

até tun kichin

he didnt do anything

tikpialo

we're taking care of it

iiyomexti nemi chuuga

the two of them are crying

geeski yahwan nemiya?

how many people were there?

How Cacao is Made

A Cacao pod rotting on the tree:
Fermin and many other Cacao producers
are plagued by pests, such as disease and squirrels.

Fermin Cruz Alvarez, produces cacao as do many people in Tabasco, and here is a small text in which he explains in Nawat how it is done. The recording can be heard here: gagawat.mp3

Bueno pues nah nechittago nin noknin nigan nugal, ge wallah pan dinamarca iiwan negi tahtani tunu tikmati pan noaltepet pal nologar. Iiwan pos ne ginegi ma nigili lo que gen motuga gagawat, gen mochowa producir, gen mowaatza, iiwan cuando nemi, genon mochowa.

Well, I, this friend has come to see me here in my house, he comes from Denmark and wants to ask what we know here in my community, in my place. And well he wants me to tell him how we plant cacao and how it is made to produce, and how it is dried and when it is there how we do with it.

Bueno pues gagawat cuando motuuga, primero motuuga kwawit, kwachipili wan tzopankilit iwan ya despues de ume o eyi xiwit cuando wefeya kwawit, motuuga gagawat, gagawat cuando peewa taagi de eyi xiwit nawi xiwit peewa xuchowi iwan ya despues giisa ixuchigagawat op unpewa giisa imasorka iwan opun peewa tagit ya despues gichowa itaagilo.

Well then, cacao, when it is planted, first the tree is planted, the kwachipili and tzopankilit, and then after two or three years when the tree grows, the cacao is planted. Then when the cacao begins giving fruit at three or four years it starts flowering and then afterwards the cacao bloom comes out, and the cob starts coming out and it begins to give fruit, and then it makes its fruit.

Bueno gagawat despues de seligagawat wehkaawa eyi meztli cuando wehweiya despues yoksi, despues de yoksi motegi, motegit gagawat cuando yoksika, despues motapana iiwan luego mowaatza, mowaatza pan toonatin, iiwan luego ya de wahki pos monaamaga.

Well, the cacao, then the tender cacao lasts three months and when it grows then it ripens, after it is ripe it is cut, the cacao is cut when it has ripened, then it is cracked open and then it dries, it dries in the sun, and then after it is dry, well it is sold.

Mowaatza an despues monaamaga, despues de monaamaga, despues gigoowa, tataxtawilia a veces. A wehka teemagaya miyak gagawat, ahorita ne ain tiempo ke nemi, nemi enfermedad pal [onemoliasis]. An ya despues giiski motohtzin

It is dried and then it is sold, after it is sold, they buy it, sometimes they pay it. Long ago it gave a lot of cacao, but now in these times that are here, there is disease [?]. And then afterwards the squirrel came out.

Motohtzin gigwa gagawat, gigwa seligagawat, despues yoksi entero […] produccion de gagawat porke tami gigwa, iiwan aparte de aparte de enfermedad tamik migi, iiwan gigwa motohtzin, yiga yo gagawat até nemi, ate teemaga: Ume plaga nemi de enfermedad pal gogolisti iiwan aparte pal motohtzin gitamiya.

The squirrel eats the cacao, it eats the tender cacao, after it ripens the production of cacao [is lost], because it finishes eating it. So apart it finishes dying from the disease, and then the squirrels ear it. Now the cacao isnt there, it doesnt give anything, there are two pests, the disease and then apart the squirrels finishes it off.  

Studying the Nahuan family with Methods of Molecular Biology

In this blog post I will give a sneak-peek on some ongoing work that I am doing which involves learning the techniques biologists use to build family trees (called phylogenies) for species or other groups of organisms and applying those techniques to understanding the relations between Nahuan dialects.

As we know, languages form "families", which basically means that a group of languages can trace their lineages back to a shared ancestor. Language families appear because most languages split into daughter languages each of which diverge from the parent in a different way. This process is what we can observe when we look at dialects of a single language. Each dialect is a variety of the language that is in the process of breaking away from the parent language and form its own language. Divergence like this happens because all languages innovate, by introducing new ways of speaking, that didn't exist in the parent language.

The same is of course what happens in biological organisms where new mutations accumulate and spread in the DNA of a population causing daughter populations to diverge from the parent population. Biologists working with molecular data analyze how these mutations can be compared and traced so that we can reconstruct the genealogy of an entire family of related organisms by grouping those organisms that share specific mutations together.

Because of each new way of speaking that arises and spreads in a community can be compared to a mutation in the DNA that spreads in a population, we can use the same methods that biologists use to trace biological phylogenies to understanding the divergence of dialects.

Nahuan Dialects:

I am in the process of understanding how all the different regional varieties (dialects) of Nahuatl are related to each other. This would help us understand where the Nahuan parent language (called proto-Nahua by linguists) originated and how it spread. In order to do this I have to find out what the shared innovations that characterize each dialect area are, and then trace them back to build a family tree. I am well underways with this project, and I presented a preliminary version of this at the Meeting of the Friends of Uto-Aztecan in Nayarit this summer (Here is a link to the paper).

But what I am doing now is mapping this information into the same kind of model that biologists use, in order to apply statistical methods to better understand the relations. Doing this allows me to generate nice graphic trees using a program called Mesquite, and this is what I want to let you peek at. 

Here is an example of how each dialect (identified with a three letter code)
is plotted into a matrix of shared innovations.
These are only some of the characters I am using.

The data is entered in binary form: each innovation present in a dialect is entered as a 1, and the lack of a specific innovation is entered as a zero (i.e. when the original form is retained). Each position of 1/0 is called "a character". By listing all the characters that I am tracking, each dialect is identified with a string of ones and zeros (called a Markov chain). And by analyzing all the characters together a phylogeny is formed. For example the dialect area of Morelos is identified with the string: 011110101 - 100000100001 – 1000000000 Where each 1 is an innovation the dialects in Morelos have, and each 0 an innovation they don't have. 

I am working with 22 distinct dialects, most of them are regional, though a few are limited to only a single community. The communities and the codes I use to identify them are: Durango/Mexicanero (DUR). Michoacan (around Pomaro) (MCH), Mexico State (around Toluca) (MEX), North Guerrero (around Coatepec Costales), Federal District (mainly Texcoco)  (DFE), North Puebla (around Huauhchinango) (NPU), Morelos (MOR), Tetelcingo, Morelos (TET), SOuth Puebla (around Tehuacan) (SPU), Zongolica, Veracruz (ZON), Tlaxcala (TLX), Western Huasteca (Hidalgo/San Luis), Eastern Huasteca (Hidalgo/Veracruz), Sierra de Puebla (SDP), Isthmus (Area north of Coatzacoalcos, Veracruz) (IST),  Pajapan, Veracruz (Town in the Isthmus area), Tabasco Nawat (TAB), Chiapas Nahuat (extinct) (CHS), Pipil Nawat of El Salvador (PIP), Central Guerrero Nahuatl (GRO), Oapan (a town in Central Guerrero) and Southern Guerrero Nahuatl (CGR). Each of these have been assigned a specific binary chain, although I still need to double check some of them in the literature to be sure I am assigning the right values, and perhaps adding more characters to the chains.

The preliminary result of my analysis is a tree that looks like this, which is interesting because it has a much more treelike branching structure than most of the previous classifications. Otherwise it does not diverge in major ways from the recent classifications by Canger (1988) or Kaufman (2001). That is partly because I have used their analyses to identify the innovations that have relevance for classification.

The bushiness comes from the fact that I have been able to identify chains of innovations in the eastern branch, where particularly the development of the negation provides a signal that I believ can can be traced along several steps. Previous classifications have been shallow bushes because they have not to the same extent tried to trace independent innovations in different areas. Canger and Dakin introduced the basic split between eastern and western branches, based on a single innovation. I have been able to add a few further innovations to the basic split. But this is about the only deep isogloss in the standard classification - the rest of it is basically a synchronic classification of the distinct areas assigning them to one or the other of the two branches. By looking at the relations between the areas within the two branches I have been able to get a bushier tree.

The tree has two main branches an Eastern and a Western branch. The Western branch is divided into a Periphery (including the dialects of Durango, Michoacan, Mexico state and North Guerrero), and a Center (including the dialects of the D.F. Morelos, North Puebla, South Puebla, Zongolica and Tlaxcala). The Eastern branch is divided into the Eastern Periphery (including the dialects of the Huasteca, the Isthmus, and Central America), and Guerrero (Central and South). The Eastern periphery is subdivided into a Huastecan and an Isthmian branch, with the Sierra de Puebla dialects in a kind of intermediate position between the two.

The next step after double and triple checking the data for the tree is to map each dialect to its geographical correlates and use a phylogeographic mapping program to calculate the probable paths that each community took to arrive to their current locations.

Meanwhile, here is the pretty tree for you to take a look at:

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.