Posted by: ayac | December 23, 2015

-tl vs. -tli

Why do some Nahuatl nouns end in -tl and others end in -tli? The obvious answer is that -tl is used after vowels and -tli after consonants (with -l- + -tli assimilating to -lli). But why do some Nahuatl nouns have roots that end in consonants in the first place? The cognates in related languages have vowels, so Nahuatl must have deleted them. What were the conditions?

Looking at it statistically, you can see that noun roots rarely end in certain consonants, while other consonants are frequently found before -tli and rarely before -Vtl. For simplicity I’m only going to look at CVCV and CVC roots here; I’m also ignoring words with a long vowel in the second syllable, and nouns that are derived from verbs.

Given the above restrictions, I count 13 nouns of the shape CVlli, some of which have cognates showing the lost vowel:

calli ‘house’ : Tübatulabal kaniˑl
cālli ‘tongs’
cōlli ‘grandfather’
chīlli ‘chili pepper’
cualli ‘good’
ēlli ‘liver’ : Tohono O’odham ‘eḍa ‘inside; interior’
mīlli ‘field’
mōlli ‘sauce’
nelli ‘true’
olli ‘rubber ball’ : Hopi pölö ‘ball, lump (of any material)’
pilli ‘nobleman’
tlālli ‘land’
xālli ‘sand’

Not counting long vowels, I’ve found only two words of the shape CVlVtl:

quilitl ‘leaf vegetables’
talatl ‘a kind of ant’ (used only in Tetelcingo?)

Based on this, I would assume that vowels are regularly deleted after l, at least in this particular context. Quilitl and talatl might be loanwords. Talatl in particular seems likely to be a loanword, based on its restricted geographic range, its specific meaning, and the presence of the sequence ta-, which is rare in Nahuatl.

Doing the same with k (c/qu in traditional orthography), I find 11 nouns of the shape CVkVtl:

ācatl ‘reed’ : Mayo baaca
mecatl ‘cord’
nacatl ‘meat’ : possibly Southern Paiute naγa-ˢ ‘bighorn sheep’
ocotl ‘pine tree’ : Mayo guocco
tlācatl ‘person’ : Hopi taaqa ‘man’
tocatl ‘spider’
tzīcatl ‘a kind of ant’
xocotl ‘fruit’
yacatl ‘nose’ : Hopi yaqa
zacatl ‘grass, fodder’
zoquitl ‘mud’

But there’s almost as many CVctli nouns:

cactli ‘sandal’
huictli (vowel length unknown) ‘digging stick’ : Hopi wiikya
octli ‘pulque’
tlāctli ‘body’
tzictli ‘chewing gum’
xīctli ‘navel’
xoctli ‘pot’
yēctli ‘good’

However, I believe that tlāctli ‘body’ and xīctli ‘navel’ are back-formations from the equivalent possessed forms notlāc ‘my body’ and noxīc ‘my navel’. Vowels are deleted when word-final after all consonants, independently of the deletion in the absolutive form under discussion, and anatomical terms like these are used in the possessed form more frequently than in the absolutive. Other words affected by the same process include iztlactli ‘saliva; lies’ (which has combining forms iztlac- ‘saliva’ and iztlaca- ‘false’) and cantli ‘cheek’ (which exists alongside camatl ‘mouth’). Different varieties of Nahuatl vary as to which words are affected by this; while Classical Nahuatl has yacatl ‘nose’, Isthmus Nahuatl (Mecayapan and Tatahuicapan) has yakti. Furthermore, tlāctli ‘body’ is actually etymologically the same word as tlācatl ‘person’.

Something similar may apply to yēctli ‘good’. The usual word for ‘good’ is cualli; yēc- is mostly only found in compounds. If we exclude tlāctli ‘body’, xīctli ‘navel’ and yēctli ‘good’, we’re left with only 5 CVctli nouns, less than half the number of CVkVtl words, and I suspect that some or all of these are loanwords.

All of them refer to culture-specific things that are more likely to be loaned than words like ‘nose’ or ‘mud’. Cactli ‘sandal’ and octli ‘pulque’ have been claimed to be loanwords before, for unrelated reasons. Cactli does have related words in other Uto-Aztecan languages, but it’s geographically restricted and the correspondences aren’t regular, so it may be a wanderwort. Nahuatl huictli and Hopi wiikya could be cognate, but I’d rather assume that huictli is a loanword than assume that this word is the one exception to an otherwise regular phonetic development.

Doing the same sort of comparison with other consonants, my conclusion is that the final vowel of bisyllabic noun roots is lost if the preceding consonant is z, tz, x, ch, or l, and kept if it is k, m, p, tl, or y. I’m not sure about cu (=kw), h, hu (=w), n, or t. This is mostly due to limited data, except for w, which has contradictory data:

xihuitl ‘herb’ : possibly Hopi siwi ‘a kind of shrub’
ēhuatl ‘skin’ : Mayo beegua
teuhtli ‘dust’ : possibly Tohono O’odham ce:g ‘mesquite bean flour’
cihtli ‘hare’ : Hopi sowi
tlohtli ‘hawk’ : Mayo taahue

Something more complex may be going on there.



  1. Isn’t this the same process of penultimate vowel loss described by Canger 1980? The process that motivates the differences perfect forms on different verbs?

    • I don’t think so, the conditions are different.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s