etymologizing word roots

[Mech]

Na'vi etymology is quite simple. We have compound words composed of words stuck together.


In some compounds, we have some simplified forms of recognized words, usually in names

Cf. rumut "ball tree", recognizably rum + ut(ral) or Fwakiwll "mantis plant" from fwaki + ('e)wll

Other examples are 

za'ärìp where we recognize za('u) + 'ärìp
lì'fya where were recognize lì'(u) + fya('o) and generally all of the "way" words (pefya, tengfya etc) where -io is missing

Now, there are two possibilities

Common understanding
The first possibility and the most acceptable one, that full words are contracted in order to simplify pronunciation. So in "ancient Na'vi" we had plants called *rumutral and *fwaki'ewll but then they were simplified. *lì'ufya'o was an ancient name for "language" and was simplified to much more prettier lì'fya
This is supported by some small words with etymology given by PF. Fwa < fì'u a, furia < fì'uri a, futa < fì'ut a

My theory
The second possibility I just thought about (and you can see some suggestions at http://en.wiktionary.org/wiki/Appendix:Na'vi) is that these compounded elements are perhaps not shortened words, but old etymological roots describing abstract concepts which were fossilized in the compounds

For example, the ancient root for the "tree" concept was perhaps *ut. The word Rumut is a word that contains this root (later fossilized). Utral is a full noun that also evolved around the root *ut (perhaps with an ending -ral).

Similarly *wll is the generic word for "plant" (perhaps meaning "growth") which is found in many plant names; and out of which the full noun 'ewll later evolved. Fwakiwll contains the actual root *wll, and not a contraction of the noun 'ewll.

The link above suggests that *lì' and *za are roots meaning "spoken sound" and "direction" respectively, although they are not attested in the texts as independent words.

Why I thought so is that (for example) plant names insist on containing invariably the element -wll. Not in a single instance I found any trace of e or the '. No plant name ended in -ewll or -'ewll. So why this insistence on always keeping only the final letters?

What is your opinion? Can we guess some... "ancient Na'vi" roots by examining the compounds? Or should we stick to the theory that they are just contracted words?

[Taronyu]

Tutanìl nong oeyä txelanit...

I used to think the same (still do.) I don't think that they are compounds. There's a lot more to find - iknimaya, for instance, could be from the same root for ikran, where *ik is the sky, and ikran is a compound of *ik and *ran, from tìran, meaning walk - walks the sky. Lots of cool stuff.

I stopped thinking this way because it wasn't helpful to anyone, and since I made the most-used dictionary, people started using my roots as words in their own compounding. Just this morning I deleted a lì' definition in a helpful document for beginners, where someone stated, from my original deductions, that it meant "to speak". Dangerous, really.

The other problem is that this language, to be frank, doesn't have a diachronic history. Sad times, eh?

I've started to use compounding as an example of how to link them together, but I think your root idea is correct. Just look at all of the *kll words, for Eywa's sake.

[wm.annis]

What is your opinion? Can we guess some... "ancient Na'vi" roots by examining the compounds? Or should we stick to the theory that they are just contracted words?

Well, for now I'm inclined to stick with contraction.  If we take your ancient roots notion, we then have to explain why we never see the other parts of these compounds.  That is, if we had several other words with *'e or *ral elements your theory would seem more likely to me.

Another example, venzek toe, from venu toe and zekwä finger.  We have wä, but it's an adposition that doesn't suggest to me, at least, any etymology for zekwä.  Or the syllable sìl- is regularly used first in compounds for sìltsan, but we've never run into *tsan alone or at the head of a compound, but only at the end of a compound in txantsan.  That sure looks like a pattern of apocope ("cutting off") and apheresis at the compound juncture.

[Mech]

The other problem is that this language, to be frank, doesn't have a diachronic history. Sad times, eh?

Now that you mention it, I consider the reverential forms "ohe" and "ngenga" as ancient forms. They were simplified to "oe" (pronounced "weh") and "nga" in everyday language, but to show that you respect someone, you revert them back to their archaic forms.

So, first theory: ancient medial -h- was lost

If this is true, then I guess that PF thinks a bit like Tolkien, and certainly there is a lot up his sleeve. I expect we see sometime ancient na'vi or na'vi dialects (why not retcon illegal words like Beyda'amo and T'riti so jahmka as dialectal?)

[Taronyu]

Where is Beyda'amo from? You seem to have a fair amount of words in your database that I haven't seen before as well.

*is very curious*

[Mech]

He is from the game

http://james-camerons-avatar.wikia.com/wiki/Beyda%27amo