Nahuatl word classes

John Sullivan idiez at
Tue Jan 1 21:59:59 UTC 2013

Piyali Magnus huan nochin notequixpoyohuan,
	I've taken a long time to respond to this mail, but I assure you it is only because, on the one hand, I've had a lot of work, and on the other hand, it really helped me to clear up some of my ideas. The most important thing I got out of this was the realization that indeed I was confusing form with meaning. This isn't a negative thing: the fact that I came to the realization that noun structures had all of these functions in Nahuatl was very enlightening. I now realize that, as you pointed out, the word classes should be based on form, in other words, how they work, and not what they mean. And, also as you pointed out, this is the way all of the major Nahuatl linguists do it. So anyway, here are the categories for dictionary entries we will be working with:
1. tlatocaxtiliztli, "noun"
2. tlachihualiztli, "verb"
3. tlapantiliztli, "relational". And here, I am assuming that what people have called relational suffixes, etc., in the past, are actually the root of the word. Nouns can be incorporated into some of them, and some of them can take possessors and other affixes (reverential). I agree with you that these words are a form of noun. I think that they originated as nouns, but now work a bit differently. The main structural difference that I would used to separate them from the noun class is that they cannot take subjects (at least in the variants I work with).
4. piltlahtoltzin, "particles". These are words that are made of a single morpheme, or chains of single morpheme words, and perhaps, words that have become fossilized in some way. This is going to be the hardest class to work out.
5. tlatenmotzquiltilli, "affix"
6. tlamotzquiltihquetl, "ligature". There is only one, the "-t(i)-" ligature. "ca:-" can work as a ligature and in most cases is structurally a verb suffix, but sometimes it is used in cases where no verb is present (cualcan, macehualcayotl, etc.), and I have not made up my mind if these are actually two different morphemes.
7. piltlahcuiloltzin, "letter"
	Listeros, please continue to criticize these categories. Tweeking makes perfect.
	Second to last thing. I don't believe one has to go through the process of  linguistic academic preparation in order to have the right to work with language. For practicality's sake I have to say this, because if not I'd be screwed and would have to give up a job that I really love. I took an introductory class in linguistics when I was a freshman a long time ago, but I dropped out because I didn't like it. This is not to say that formal linguistics is boring. It is just boring to me. I deeply respect the work of linguists, profit from reading some of their work, and most immensely enjoy sitting down and talking about Nahuatl with them in person. It's just that I prefer to do this thing my own way. And I think the history of science is full of cases of outsiders making contributions to disciplines. And I'm not talking about re-inventing the wheel. It's just that it is very important to go over what scholars have done and look for ways to make things better. The idea of the academic "vaca sagrada" has always made me sick to my stomach: the idea that the purpose of getting a Ph.D. is to aspire to get into an academic club and wind up subordinating oneself to a big name (this is very common in Mexico, I don`t know if this is also the case in the US or in Europe). When you actually get know the big names who are the real thing, in the vast majority of cases, they are friendly, ego-less, accessible and generous with their time and research, and not at all interested in being king or queen of the hill. 
	And last of all, Magnus, in your list of eminent grammarians of Nahuatl, you forgot to include the greatest Nahuatl grammarian in history (so far). 

On Nov 6, 2012, at 6:27 PM, Magnus Pharao Hansen <magnuspharao at> wrote:

> Dear John and listeros
> I'm responding to the inquiry about Nahuatl word classes, I am a little
> worried that your approach to grammatical analysis is not the most useful
> for the project you are undertaking. I think the best thing you can do is
> to base those analytical choices on research done by the many excellent
> linguists who have worked on Nahuatl. Personally, I think you should adopt
> either Andrews or Launey's analysis - and I recommend Launey's because it
> is more compatible with standard linguistic terminology. If you don't want
> to do this I think you would need to go back a few steps to make some tough
> decisions about how to approach grammar at a theoretical level. And here
> for the sake of the utility of the database you want to build I think the
> best choice would be to assure that it is compatible with what is by now
> called "basic linguistic theory" which is used for all kinds of linguistic
> typology and almost all langauge documentation. A good example of this
> theoretical perspective is Thomas Payne's "describing Morphosyntax" which
> gives the basics of how to do a typologically based language description
> that can be used for cross-linguistic comparison. Subsequently some more
> typologically oriented literature such as the series of mongraphs  by Dixon
> and Aikhenvald might be a useful read.
> What makes me say this is that in your question you are unclear on several
> key grammatical distinctions which I think stems from a lack of a decision
> about what grammar is and how you want to describe it, this leads you to
> mix up formal (syntactic) and functional (grammatical and semantic)
> criteria of "wordness". For example you conflate the notions of "word",
> "root", "part of speech/word class", "morpheme", "semantic function" and
> "grammatical function". The way you use the concepts are not in synch with
> how they are used in descriptive linguistics, you can of course choose to
> adopt a new theoretical framework, but that would seem to require a good
> reason.
> In linguistics a word class, also called "part of speech" is traditionally
> syntactically defined. A group of words form a word class if they can be
> seen to have complementary distribution to other such classes and to be
> characterized by a shared underlying syntactic/grammatical function (e.g.
> that of forming predicates or arguments). In language's such as Nahuatl
> that have a very loose word orde and a complex morphology,  the main
> criteria for describing a word as belonging to one class or the other tends
> to be morphological. Verbs is any word that can take verbal morphology, and
> a noun is any root that can take nominal morphology. The criteria are not
> fully waterproof since certain morphologicaol categories are shared (e.g.
> the subject marking morphemes), but nonetheless with careful analysis it is
> almost always possible to discern differences. (e.g. verbs never take
> possessive morphemes and nouns never take object morphemes (except in Oapan
> Nahuatl where kinship nouns do!) or tense/aspect/mood related morphology).
> Now for adjectives and adverbs this is much more complicated, because there
> are no completely clear definitions of these categories, accepted by all
> linguists. I think that consensus in linguistics currently is that not all
> languages have adverbs and adjectives, and that only those languages have
> these word classes where these categories have specific morphological or
> syntactic patterns of distribution. In Nahuatl there is a small class of
> words that can be considered adjectives or adverbs, but it is a small and
> ambiguous class of words that are neither fully nouns nor fully verbs but
> which can form predicates (I consider them to be "statives" and some of
> them may be considered adjectives (e.g. hueyi, istac, yancuic, cualli) or
> adverbs (e.g. yolic, huilihui). because this class of words is small and
> closed instead the aspects of meaning that are carried out by adjectives
> and adverbs in English, in Nahuatl  are carried out by either nouns, verbs.
> But none of these classes correspond directly to what we would call
> adjectives or adverbs in English, since both nouns and verbs can carry out
> the functions carried out by adverbs and adjectives in English. In a
> conventional analysis this does not mean that these words become adjectives
> or adverbs, it just means that in this language those semantic functions
> are also fulfilled by other wordclasses.
> The confusion of these categories is evident for example in your examples
> of *cuauhtli*. I.e. /kwaw/ is a morpheme, not a word - it doesn't belong to
> any wordclass even though it clearly is nominal in its semantics and is
> clearly most often used to create nouns. When constructed with the
> absolutive, c*uauhtli *is a noun because it can function as an argument of
> a predicate, and stand as a free word in argument position in the sentence,
> and because it takes the absolutive ending, and because it can be possessed
> and pluralized. In *cuauhpillli *it is still a noun root, it has just been
> incorporated into another noun - which is what Nahuatl does most of the
> time when it wants to modify nouns. That does not make it an adjective
> though, because "adjective" is usually defined as a syntactic category with
> the main function of modifying nouns (in Nahuatl the only ones are kwalli,
> weyi and perhaps a few others).  I.e. /kwaw/ is a noun without regards to
> the semantic function it carries out in a given context, because in all the
> cases it functions exactly as all other nouns, and in opposition to either
> verbs, particles and adjectives. In the same way teopixcatequitl is also a
> noun that is made by combining two nouns one of which modifies the other -
> teopixquetl/teopixqui does not become an adjective because it is used in
> this way.
> It is simply not the case that in Nahuatl there is a category of words that
> can randomly function as nouns, adjectives or adverbs - this idea goes
> against everything we know about Nahuatl grammar. The fact is that Nahuatl
> has a class of nouns and that that class of nouns can be combined in ways
> that convey the meanings of English adjectives and adverbs - but which are
> still nouns syntactically and grammatically. You may wish to to take a look
> at my short article on the question of Nahuatl Adjectives in Kansas Working
> Papers in linguistics ) to see a little bit about how complicated it is to
> define wordclasses other than "verb", "noun" and "particle" in Nahuatl
> grammar (even your proposed "relational word"  I wouldn't consider a valid
> word class since they are all either nouns, affixes (i.e. morphemes not
> words) or particles - most of them are nouns marked for relationality with
> wifferent  combinations of possession and suffixes). I end the article with
> my analysis of wordclasses in Nahuatl, which is basically the same as
> Launey''s and Andrews'.
> It
> is not a great piece of work, but it is an exercise inthe kind of
> grammatical reasoning that must go before making any decision about
> analyzing word classes in Nahuatl.
> I think that the thing to do is to take the time to do a thorough survey
> comparing analyses in the major grammatical works and seeing how they
> divide up word classes and analyze their functions. This is a huge task
> that will take many hundred hours of study and a really good familiarity
> with linguistic theory, and how linguists make analytical choices based on
> different theoretical perspectives and on analysis of evidence. I don't
> think it is enough to be very good at Nahuatl, this tasks requires intimate
> familiarity with linguistic theory and Nahuatl scholarship.
> For this reason I don't see why anyone would undertake this endeavor from
> scratch since so many eminent grammarians of Nahuatl have already done it
> for us, e.g.  Carochi, Launey, Andrews, Canger, Lockhart, Lastra or Dakin.
> I don't understand why you'd want to reinvent the wheel on this, and if you
> go with an analysis that is too idiosyncratic you risk that the entire
> documentation project will be of little use to others in the discipline,
> especially if the the data format is not based on a full systemic analysis
> of the language but rather on scattered observations and gut feelings.
> best regards,
> -- 
> Magnus Pharao Hansen
> PhD. student
> Department of Anthropology
> Brown University
> 128 Hope St.
> Providence, RI 02906
> *magnus_pharao_hansen at*
> US: 001 401 651 8413
> _______________________________________________
> Nahuatl mailing list
> Nahuatl at

Nahuatl mailing list
Nahuatl at

More information about the Nahuat-l mailing list