On chantli and Pipil

Alan King alanrking at yahoo.com
Thu May 22 08:04:33 UTC 2008

Alan King here, returning after a long absence -
during which my work on Pipil (which we call Nawat)
has not ceased - with greetings for Joe Campbell if he
remembers me and the rest of the list. Ken annemit?

This is in response to Joe's "comment 2" where he
raises some issues about the status of chantli and
tzolli about which I would like to chip in. Since this
seems to have nothing at all to do with turquoise
diadems, I thought it best to move over to a new
thread with it.

I was struck by Joe's digression about how to
interpret the absence of a noun entry for chantli in
Molina. In Pipil, the conclusion I have drawn from the
entire text corpus at my disposal (written
transcriptions of originally oral materials; my
present corpus includes both previously published
texts, written but unpublished stuff, and new oral
material), as well as from direct observation of the
language in use is that either there simply is no
autonomous *noun* "chan" 'house' in Pipil, or else its
presence is rather marginal and involves many

For the ordinary noun 'house' Pipil has the word kal
(cf. calli). Note that noun stems ending in an 'l'
regularly take no suffix whether absolute or
possessed. So: se kal "a house", ne kal "the house",
(ne) nukal "my house". 

First a word about inalienability. If there were an
equivalent of chantli in Pipil and it had an absolute
form, the form we would expect is *chanti (cf. "sinti"
- "nusin" 'maize' - 'my maize'). However, *chanti is
not attested. Now there are many, many nouns with no
absolute form in Pipil because they are obligatorily
possessed (let's call such nouns inalienable), whereas
in classical Nahuatl apparently they have absolute
forms (at least in dictionaries). So for example
'mother' is "nan", it is inalienable so there is no
*nanti, only "nunan" etc., and in at least one dialect
also "tenan" is attested. 

Notice that this certainly does not prove that "nan"
is not a noun, only that as a noun it is
"inalienable". And of course relationals are also
"inalienable"; while often described as noun-like,
they likewise have no absolute form. My point,
therefore, is not merely that in Pipil "chan" is
inalienable (which would not be so surprising), but
rather that it seems to be a relational (rather than
just an inalienable noun) or, at best, something
intermediate between a true noun and a true

For the most part, in Pipil "chan" certainly seems to
function as a relational meaning 'chez', e.g. "Niaw
nuchan" 'I am going home', "Nemi (ka) ichan" 'He/She
is at home', "(i)chan ne palej" 'at the priest's
house'. (Note that omission of "i-" when the possessor
follows, as in "chan ne palej", is very widespread
both with relationals and inalienable nouns in Pipil.)
The corpus is chock-full of this kind of usage of

In the Santo Domingo de Guzmán variety of Pipil, I
have hardly encountered evidence in actual language
use to contradict the hypothesis that this is the
*only* way "chan" is used as an autonomous word in
this dialect. By autonomous word I mean to exclude
lexical derivatives or compounds such as "chanej"
'inhabitant', or "techan" lexicalized in the meaning
'village'. For another variety, that spoken in
Cuisnahuat, as far as I know the same can be said
except that here the word that has been lexicalized as
meaning 'village' is "tuchan". "Techan" (Santo
Domingo) and "tuchan" (Cuisnahuat) are synchronically
true nouns, but I suspect that "chan" as such is not -
at least in these dialects.

At least three tests of "nounness" might be suggested
here (see below for yet another): (1) can it take a
determiner such as a definite or indefinite article or
a demonstrative? (this test could be expanded to
included other determiners and quantifiers) (2) can it
be qualified by an adjective? Both of these tests
question whether "chan" really is a nucleus of a noun
phrase admitting normal noun phrase expansions. (3) a
semantic test: is "chan" only used to refer to a
location (as relationals typically do) or can it also
refer to an object (as "kal" certainly can).

By these tests, my assertion about "chan" is confirmed
for these dialects. (The only example with "ne" +
"chan" are "ne techan" in Santo Domingo and "ne
tuchan" in Cuisnahuat, both meaning "the village".)

This conflicts with a claim made to me by one
present-day speaker who thinks that "kal" and "chan"
are both nouns and "really" refer to different types
of construction. I have found nothing to support this
in actual usage and suspect it is a spurious
rationalization of the existence of two different
words which are patently not quite equivalent, by
someone lacking the linguistic sophistication needed
to comprehend the notion of a grammatical rather than
semantic differentiation between lexemes, especially
when without a parallel in Spanish. I only mention
this for the sake of completeness and in honour of his
opinion, whether correct or in error.

However, in one important component in the Pipil
corpus, the texts recorded by Leonhard Schultze Jena
in the 1920s in Izalco, where the language is now all
but extinct, there are some counterexamples regarding
test (1), given that the definite article "ne"
sometimes precedes "ichan", as in (with standardized
spelling): "Ne shulet mukwepki tik ne ichan kilia ne
isiwaw" 'The husband returned home to tell his wife'.
In Santo Domingo this would be: "Ne shulet mukwepki
(ka) ichan kilwia ne isiwaw", omitting both the
article and the locative preposition "tik", which also
seems to be incompatible with relationals. (This
suggests another syntactic test for nouns: can they be
preceded by a preposition other than "ka"? "Ka" acts
somewhat differently to other prepositions - long
story - and its optional presence doesn't prove
anything here.)

But even in the Izalco data, and even when the article
"ne" is present, only locative phrases are found. By
contrast, with "kal" there are some non-locative uses
in the corpus, such as: "se lamatzin ka kipiatuya ne
ikal tik se kujtan..." 'an old woman who had a (lit.
"her") house in the woods/country' etc. On the basis
of my findings on Pipil usage, I would expect *ichan
to be starred in a sentence such as this one.

As far as I know, this distinction between "kal" (as a
noun) and "chan" (as something other than a real or
full noun) has not previously been noted in linguistic
descriptions of El Salvador Nawat, hence my need to
argue the case as I have done. Having done that, I
thought this might just have a bearing on the absence
of a dictionary entry for a noun "chantli" in
classical Nahuatl, as noted by Joe Campbell. If
relevant, then this would not only tell us something
about classical Nahuatl but also about Pipil Nawat by
suggesting that the usage I have described for the
latter continues a pattern of considerable antiquity.

Sorry that had to be so long!



Nahuatl mailing list
Nahuatl at lists.famsi.org

More information about the Nahuat-l mailing list