Mechanical grammar

[Ricardo J. Salvador]

Hey Joe,

On Friday, January 10, 2003, at 05:25 PM, r. joe campbell wrote:

> I thought at first that my comments might not be of general enough
> interest for the list

I'm glad you wrote, because that recounting of your first experiences
with Nahuatl provided a fascinating glimpse into the workings of the
mind of a linguistics whiz ;-). I knew a little bit about that first
experience of yours in Tepoztlan from the moving paragraph you
contributed to the "In Memoriam" site that MIT put up for Kenneth Hale
(the part about his catching on immediately to the "ayotochtli"
compound was particularly insightful.) Your message of yesterday also
caused me to flash back a few years. I didn't realize when you asked me
in Bloomington for the name of my Tepozteco friend that it was because
you had great depth of experience with the town (and probably with all
its Mexicano speakers.)

I've often wondered (and envied) how language specialists such as Hale
(and you) can evidently just inhale a language (no pun intended ;-).
Conceptually, it is clear that ability must relate to having developed
very effective mental schemas for what a language is and the key
functions that it must serve. You must immediately focus on the
fundamental mechanisms that each language employs to solve its
universal functions, and then "layer" the secondary aspects. For
example, just your description:

> Since I thought the key to reasonable early progress was the
> aquisition of a clear image of the one dimensional intransitive verb
> matrix (ni-, ti-, --, ti--h, am--h, --h) and the two dimensional
> transitive verb matrix (the subject prefixes intersecting with the
> object prefixes (nech-, mitz-, c/qui-, tech-, amech-, quim-),

provides such an economical and visual metaphor that it is clear you
approach a language from a "10,000 meter overhead view," rather than
wandering from the side into a vast and dark forest, which as an
amateur is what I do. But OF COURSE the number of combinations becomes
"two-dimensional" when shifting to the transitive and what an obvious
and practical way of generating pragmatic "challenges" for conjugation
(for either a learner or an algorithm--although what is a learner but
an algorithm with nacatl on it ;-)). So instead of blindly stumbling
upon useful expressions for quotidian use over an indeterminate period
of time, here is a way of methodically building up skill and compacting
time. That must be so clear to linguists that I'm just confirming my
idiocy by noting that it took your example for this to dawn on me! (To
borrow one of my favorite expressions from Doña Luz: "Ican Ricardo coza
titicuintli amo quimatia tlen quichihua." :-/ ) Your casual comment has
given me some immediately useful ideas for both programming and
learning (am currently working on a forest preservation project in Los
Chimalapas and Selva Zoque where it would be very handy not to have to
force the local collaborators into Spanish. Many years ago I got a
decent start with Lacandón Maya, so I can entertain Maya of the isthmus
to distraction, but I'd like to actually communicate ;-). (There are
many refugees from Chiapas streaming into the forest, which is part of
the issue).

> I needed a large set of exercise items.  Was I going to laboriously
> write all these out in the present, future, and preterit?  ...Maybe on
> purple ditto masters?
...
> So I punched a large set of verb stems on those 80 column cards and
> wrote a program to generate all the possible verb forms in three > tenses
...
> certain "irregularities" had to be built into the program
...
> (There was no hard disk and no floppies...)

Ummm.... now that there ARE hard disks, are those algorithms still
lying about someplace? What a fabulous learning tool! I see from the IU
catalog that Nahuatl is no longer taught, so perhaps you've not had a
reason to maintain/update such methods, but they are so obviously
suited to the WWW that it would be handy for the Nahuatl Home Page to
sport such support for the many learners who are trolling
electronically for authoritative didactic material. The moment I type
that I want to hasten to say I'm not urging this, the last thing I
intend is to distract you from the fascinating integrative Molina
project you described the other day, the Florentine work, and the other
things you are doing, but if things are just lying around and could
readily be adapted... I see your department is currently searching for
a computational/natural language processing linguist, and these days
folks like that are probably legion and would find it child's play to
update and generate such a tool for the web (along the lines of what
Jonathan Amith started putting up at Yale a short while back.)

Incidentally, the irregularities you described were very interesting,
particularly "/k/ --> [g] intervocally." I see I've been unnecessarily
rough on some of the folks to whom I've passed along tips for Puebla
valley Nahuatl, whereas I should've instead just said "that's alright,
that's the way they say it in Hueyapan" (just go across the volcano!
:-).

> Now my cackle has subsided into a chuckle, but I am still enjoying the
> shared experience, even the situation later when I was speaking to
> people from Hueyapan in Nahuatl (partly because I became influenced by
> "classical" Nahuatl) and made a non-Hueyapan verb like 'ninokopa',
> they removed my embarrassment at the error by saying, "No, no, esta'
> bien, asi' hablan en Santa Cruz".

How utterly and typically gracious. Glad to have stimulated pleasant
memories and to have provided some amusement. OK, yotlan.

> Joe
>  (Cencah Xochichil)

Your story on this and your word list from yesterday have provided the
perfect epigram for me to wear on the back of MY belt:
Cuitlapilyahualtic.

Hasta moztla.

Ricardo J. Salvador          Voice: 515.294.9595
1126 Agronomy Hall         Telefax: 515.294.8146
Iowa State University        e-mail: salvador at iastate.edu
Ames, IA 50011-1010       WWW: http://www.public.iastate.edu/~rjsalvad