Na'vi Reference Grammar

Started by wm.annis, August 13, 2010, 08:46:50 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

wm.annis

So, here it is, at last.

  Horen Lì'fyayä leNa'vi (Version 1.33, March 9 2024)

A few months ago, I wrote up a very dense Na'vi grammar cheat sheet designed to fit into a single, double-sided sheet of paper.  I had to leave out a lot of useful information, especially syntax, so I said I'd work on a larger document.  I had no idea how much larger it would turn out.  On it's first release (August 13, 2010), the thing is 54 pages long.  Now, some of those pages are just for the index, but still.

About this Grammar
This document will not teach you Na'vi.  It's a reference, and I've tried to cover everything we know for sure about the language.  So, I hope it will be a good guide when you have questions about the details of some point of Na'vi grammar.

I use a lot of cross-referencing in the grammar.  When you see a section number, like 6.2.1, you can click on it to be taken to that section.  The index takes up several pages, and once again, you can click on any page number to be taken straight to that page.  For now, there's no table of contents, but I hope the index will be more useful for people.  Please let me know where I've missed obvious things to index.

If you see text in maroon, that means I'm uncertain about something.  Hopefully the maroon parts will disappear over time.

In the Syntax chapter especially, you will see "F" floating in the margin.  This means I'm using a Na'vi example straight from Frommer.  If you see any places where I've created an example, but for which you know there is a Frommerian example, please let me know.

Save the trees!  For now, at least, please don't print this.  It is 54 pages!  And it's going to be changing fairly rapidly for a few months, I'm sure.  I will probably not update this as quickly as Taronyu updates the dictionary, but any substantial release of information about Na'vi from Frommer is likely to lead to a new release in a day or two.

I hope everyone will find this useful.


You can see the LaTeX sources under the LearnNavi github space: https://github.com/LearnNavi/Horen

omängum fra'uti

#1
Lesar nìtxan nang!

Reading through it a few comments...

Quote from: §2.1.4.A syllable with a pseudovowel must start with a consonant or consonant cluster
Psuedovowel syllables are strictly a single consonant and never a cluster or no consonant kefyak?

Quote from: §2.1.4.1.muiä [mu.i.æ], ioang [i.o.aŋ].
Just a little nitpick that the IPA is being given to show syllable boundaries, but it is incomplete. (No stressed syllable indicated.) The impression I get though is that this is intentional to avoid cluttering it up?

Quote from: §2.1.4.2.tsenge is [tsɛ.ŋɛ] not *[tsɛŋ.ɛ]. Reduplication may override this, as in kangangang [kaŋ.aŋ.aŋ], where the echo effect is desired.
Again with the stressed syllable, but also...
It's kxangangang not kangakgang...
Also I'm not sure I agree with the assertation that it is reduplication overriding it.  I think that in that example, it is desired sound of the onomatopoeia which is setting the syllibafication, not reduplication.  And further, that is one of a small number of things which may override syllabification.

§2.2.2.
I saw a thread recently about ADP+, and what exactly the lenition is in - and I don't recall there being an example where an ADP+ and its noun was separated by another word which would lenit (IE - *mì feyä kelku or *mì peyä helku) - so unless you know otherwise, perhaps this is an unanswered question worthy of a red mark?

§2.3.6.
There's also many cases of the vowel in lu being elided, probably worth mentioning.  This appears both in the songs and the movie dialog, though the movie dialog is backed by an actual explanation of the dialog by Frommer.

Quote from: §2.4.3.1.The first person pronoun root oe, though pronounced we when taking a suffix, retains the original spelling.
Probably worth adding in that the first person dual inclusive oeng gets the "we" treatment all the time as the ng(a) gets the same treatment as suffixes.

§3.2.
It is probably important to mention that oeng and all its derived pronouns regain the "a" from nga when taking a suffix.  Comments on this should also be included in §3.2.2.5. where oeng becomes oengeyä

And other general comments...
Some places you note different registers, but other places you do not.  Notably we have an example of the participle used alone with no verb and no attribution - tìkan tawnatep.  There is also a specific section on register (§7.3.), but most of the more informal rules are spread out through the document, while the Register section is more about formality.

More comments as I get more time to read through it, but what I've seen so far is txantsan!
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

Kì'eyawn

Tewti, ma William, fì'u leiu txantsan!  Ngaru irayo seiyi oe!
eo Eywa oe 'ia

Fra'uri tìyawnur oe täpivìng nìwotx...

Plumps

Ma William,
this is truely amazing work!!! Thank you so much for the time and effort you put into this!

Thoughts:

I'm officially confused :P I thought, K. Pawl said when looking at your grammar cheat sheet that the case suffixes after diphthongs would be either -ìl, -it, -ti etc., not the one-letter endings? That's why I was confused seeing *keyeyt in his first blog post...

6.2.6. I think there are better examples for the gerund... I think the *tìtaron lu lehrrap example was given at a time when the gerund form with tì-‹us› was not yet developed(?). With this example it seems more correct to say »the hunt is dangerous«

6.7.9. minor mistake in example sentence; it's: tafral ke lìsyek oel ngeyä keye'ungit therefore I will not heed your insanity.

(?)6.15.3. Have you not marked a sample sentence as Frommarian when it didn't come originally from him? I would think that corrections by him can be considered attested... But that can be discussed, can't it?
(I'm aiming at fo ke lu 'ewan kaw'it, which (I think) was a sentence by Prrton but was corrected by K. Pawl; I was expecting an F in the margin...)

pp. 39 f. Don't know if you have an influence on that but the footnote 13 is on the wrong page...

So much at a first glance.
Nìmun, nìngay txantsana tìkangkem! *Karma* nefä! :D

Lance R. Casey

Tewti! Txantsana tìkangkemìri aftxavang ngeyä irayo seiyi ayoe nìwotx ma Wìlyìm!

After reading it through once, this stood out:

Quote from: p3, §1.2Notatation and Conventions
p4, §2.1.1: Is it correct to sort /ɾ/ under "liquids"?

p6, §2.2: No wavy underline to mark the lenitable consonant in 'eylan, although in this case it is hardly required. Also, perhaps pe+ should be mentioned here.

p6, §2.2.2: According to, well, you, pxisre is also ADP+.

p11, §3.2.3.1: From the movie we have ohengeyä as the genitive form of ohe ngengasì.

p12, §3.3.3: ay+ marked as non-leniting.

p15, §3.6.3: How about a quick mention of the s-futures, with a link to §6.7.9?

Quote from: p17, §4.1.1The independent forms of the numerals from one to eight are:
p18, §4.3.1: Pxelo has also been confirmed.

p19, §5.1.1: Bob ;)

p19, §5.1.1.2: And pronouns, e.g. nìayoeng.

Quote from: p21, footnoteMany human languages are more strict.
Quote from: p24, §6.1.2.2The subject may be either subjective in or agentive.
p28, §6.5.1 See §2.2.2 above.

p30, §6.5.17.2: We also have mì sìrey in (my) life.

p30, §6.5.28: Pxisre is ADP+ as per above (if that's correct, of course).

Quote from: p33, §6.7.4Because the imperfective presents an on-going state of affairs, it can be used in in complex sentences to indicate simultaneous action, [...]
p34, §6.8.3.1: Include sìlpey without tsnì?

p39, §6.17.2: "Relative clause" (and nothing else) mentioned thrice.

p43, §6.20.1: Should also mention that either san or sìk can be dropped when circumstances so permit.

p43, §6.20.2: Also thoughts, according to WB.

p48: No title?

// Lance R. Casey

bommel


wm.annis

Ah, so many excellent comments.  I've incorporated most already, but a few need some more careful thinking about than I'm capable of just now.  My schedule is going to be crazy today, so I'll probably not get an update out until tomorrow.

Taronyu

No rush.

I'm glad that this is here. :)

Nyx

Wou.. fì'u lu txantsan!

I still have to read through the whole thing, 54 pages is quite a bit. Thank you for making this :D

Kä'eng

Tewti, this is amazingly detailed. Nice work!

With everyone else listing corrections, I'll jump on the bandwagon as well:


  • p4 §2.1.1: missing right bracket: "n [n"
  • p5 §2.1.4.5: oìsss misspelled as oìss
  • p6 §2.2.3: "lention"
  • p10 §3.2.2: oeng, ayoeng imply impossible syllabifications
  • p11 §3.2.2.6: "gentive"
  • p14 §3.6.2: participles claimed to be pre-first. The Language Log post said:
    QuoteFirst-position infixes indicate tense, aspect, or mood; there are also participial and reflexive infixes in this position, the latter being in "pre-first" position so it can co-occur with other first-position infixes. [emphasis mine]
    Many people seem to have interpreted "the latter" as meaning both the participial and reflexive infixes, but note the singular pronoun "it". I think this comment was only meant to be about -äp-, not about -us-/-awn- too.
  • p17 §4: "forty-two" spelled as "fourty-two"
  • p17 §4: mrrvomun spelled as mrrvolmun
  • p17 §4 footnote: "couting"
  • p20 §5.1.2.1. Says that kelfpomtokx is from lefpomtokx with no accent shift, but the original word is actually lefpomtokx. (§5.3.1.2 has it correctly)
  • p20 §5.1.4.1: "speach"
  • p21 §5.3.1.1: (if I've understood what "head" means) the head of fpomtokx would be fpom, while tokx is a modifier.
  • p23 §6.1.1 footnote: "gibberish" spelled as "jibberish"
  • p33 §6.7.5: perfective referred to as "perfect" (not called this anywhere else)
  • p46 §7.1.2 footnote: "MaSempol.org"
  • p48 §7.3.1.3: "formaltiy"
Ma evi, ke'u ke lu prrte' to fwa sim tuteot ayawne.
Slä txo tuteo fmi 'ivampi ngat ro seng, fu nìfya'o, a 'eykefu ngati vä', tsakem ke lu sìltsan.
Tsaw lu ngeyä tokx! Kawtu ke tsun nìmuiä 'ivampi ngat txo ngal ke new tsakemit.
Ha kempe si nga? Nì'awve, nga plltxe san kehe. Tsakrr, ngal tsatsengti hum!

Taronyu

Ya'll are making my pre-release edits look bad. :P

wm.annis

#11
Quote from: omängum fra'uti on August 13, 2010, 09:20:21 PM
Quote from: §2.1.4.A syllable with a pseudovowel must start with a consonant or consonant cluster
Psuedovowel syllables are strictly a single consonant and never a cluster or no consonant kefyak?

Though I can find no examples of a cluster + pseudovowel combo, I can think of no good reason to claim that's not possible.  Does Frommer say that somewhere?

QuoteJust a little nitpick that the IPA is being given to show syllable boundaries, but it is incomplete. (No stressed syllable indicated.) The impression I get though is that this is intentional to avoid cluttering it up?

Exactly.

QuoteAlso I'm not sure I agree with the assertation that it is reduplication overriding it.  I think that in that example, it is desired sound of the onomatopoeia which is setting the syllibafication, not reduplication.

That's clearer, done.

Quote§2.2.2.I saw a thread recently about ADP+, and what exactly the lenition is in - and I don't recall there being an example where an ADP+ and its noun was separated by another word which would lenit (IE - *mì feyä kelku or *mì peyä helku) - so unless you know otherwise, perhaps this is an unanswered question worthy of a red mark?

Lenition is primarily a phonetic phenomenon, not a grammatical one.  I would be stunned if *mì peyä helku turned out to be correct.


QuoteIt is probably important to mention that oeng and all its derived pronouns regain the "a" from nga when taking a suffix.  Comments on this should also be included in §3.2.2.5. where oeng becomes oengeyä

3.2.2.2.  :)

QuoteSome places you note different registers, but other places you do not.

Argh.  You're right.  I used to have three different register sections, but I obliterated the poetry one until I had more of certainty I could say.  I'll add a section on informal and military registers.

Quote from: Plumps on August 14, 2010, 03:50:07 AM
I'm officially confused :P I thought, K. Pawl said when looking at your grammar cheat sheet that the case suffixes after diphthongs would be either -ìl, -it, -ti etc., not the one-letter endings? That's why I was confused seeing *keyeyt in his first blog post...

I am also confused, which is why I added note 3.1.1.3.

Quote6.2.6. I think there are better examples for the gerund...

It's not intended as a gerund example, but of adjective predication (the only simple one from Frommer I could find).  I've altered the translation as you suggest, however.

Quote(?)6.15.3. Have you not marked a sample sentence as Frommarian when it didn't come originally from him? I would think that corrections by him can be considered attested... But that can be discussed, can't it?
(I'm aiming at fo ke lu 'ewan kaw'it, which (I think) was a sentence by Prrton but was corrected by K. Pawl; I was expecting an F in the margin...)

Sentences that made it past Frommer without him correcting them get to be added to the grammar, but I still don't put the "F" on them unless he himself reworked them quite a lot.

Quotepp. 39 f. Don't know if you have an influence on that but the footnote 13 is on the wrong page...

I have moderate control over that.  Fixed.

Quote from: Lance R. Casey on August 14, 2010, 07:54:48 AM
p4, §2.1.1: Is it correct to sort /ɾ/ under "liquids"?

That is a traditional place for rhotics, and Frommer classifies it so.

Quotep6, §2.2: No wavy underline to mark the lenitable consonant in 'eylan, although in this case it is hardly required. Also, perhaps pe+ should be mentioned here.
Note about pe+ added.

Quotep6, §2.2.2: According to, well, you, pxisre is also ADP+.

Pxisre issue clarified throughout the document.

Quotep11, §3.2.3.1: From the movie we have ohengeyä as the genitive form of ohe ngengasì.

Eew.  Where?

Quotep15, §3.6.3: How about a quick mention of the s-futures, with a link to §6.7.9?

Good idea.  Done.

Quote
Quote from: p17, §4.1.1The independent forms of the numerals from one to eight are:
p18, §4.3.1: Pxelo has also been confirmed.

p19, §5.1.1: Bob ;)

p19, §5.1.1.2: And pronouns, e.g. nìayoeng.

Quote from: p21, footnoteMany human languages are more strict.
Quote from: p24, §6.1.2.2The subject may be either subjective in or agentive.
p28, §6.5.1 See §2.2.2 above.

All fixed.

Quotep39, §6.17.2: "Relative clause" (and nothing else) mentioned thrice.

Hmmm.  Not the clearest sentence I've ever produced.

Quotep43, §6.20.1: Should also mention that either san or sìk can be dropped when circumstances so permit.

Good point!

Quotep43, §6.20.2: Also thoughts, according to WB.

Without sourcing, I'm hesitant about that.

Quotep48: No title?

There's a bug I haven't been able to track down, where if I give the index page a title, it inserts a blank page between the title and the actual index.

Quote from: Kä'eng on August 14, 2010, 03:12:23 PM
With everyone else listing corrections, I'll jump on the bandwagon as well:

All of your spelling, punctuation, etc. corrections have been rolled in.

Quote
  • p10 §3.2.2: oeng, ayoeng imply impossible syllabifications
Explained in 3.2.2.1

Quote
  • p14 §3.6.2: participles claimed to be pre-first. The Language Log post said:
    QuoteFirst-position infixes indicate tense, aspect, or mood; there are also participial and reflexive infixes in this position, the latter being in "pre-first" position so it can co-occur with other first-position infixes. [emphasis mine]
    Many people seem to have interpreted "the latter" as meaning both the participial and reflexive infixes, but note the singular pronoun "it". I think this comment was only meant to be about -äp-, not about -us-/-awn- too.
Nope.  Frommer looked at and commented on my cheat-sheet summary of Na'vi grammar.  The layout I use here is taken directly from that.  If it were incorrect, Frommer could have said so.

Quote
  • p21 §5.3.1.1: (if I've understood what "head" means) the head of fpomtokx would be fpom, while tokx is a modifier.
Hrmm.  That's an interesting word.  I've picked a different example where the head is indisputable.

omängum fra'uti

Quote from: wm.annis on August 15, 2010, 12:53:39 PM
Quote from: omängum fra'uti on August 13, 2010, 09:20:21 PM
Quote from: §2.1.4.A syllable with a pseudovowel must start with a consonant or consonant cluster
Psuedovowel syllables are strictly a single consonant and never a cluster or no consonant kefyak?

Though I can find no examples of a cluster + pseudovowel combo, I can think of no good reason to claim that's not possible.  Does Frommer say that somewhere?

Not directly, but from ye olde Language Log...

Quote from: Paul FrommerThat is, the syllable structure is (C)(C)V(C), where V represents a vowel or a diphthong. [...] In CV syllables, the liquids l and r can replace the vowel.

Perhaps I'm reading too much into it, but when he spelled out that syllables are (C)(C)V(C) rather than just saying (C)V(C) and declaring that the first C can be a cluster, the impression I got is that for pseudovowel syllables, it must start with a consonant, not a cluster.  I could be wrong though.

Quote
Quote§2.2.2.I saw a thread recently about ADP+, and what exactly the lenition is in - and I don't recall there being an example where an ADP+ and its noun was separated by another word which would lenit (IE - *mì feyä kelku or *mì peyä helku) - so unless you know otherwise, perhaps this is an unanswered question worthy of a red mark?

Lenition is primarily a phonetic phenomenon, not a grammatical one.  I would be stunned if *mì peyä helku turned out to be correct.
I agree, but I suspect most people who are just learning how to use lenition and not WHY it is might try to use *mì peyä helku, and that doesn't answer the question of if *mì feyä kelku (Or even *mì sawla kelku) would be correct.

Quote
Quote from: Lance R. Casey on August 14, 2010, 07:54:48 AMp11, §3.2.3.1: From the movie we have ohengeyä as the genitive form of ohe ngengasì.

Eew.  Where?

Tsun oe fì'uteri 'iveyng!

This is Norm's 'awvea ultxari ohengeyä, nawma sa'nok lrrtok siveiyi...  And given that he is only Human and Frommer has given us counter-factual information about the use there, I think we can dismiss it here.
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

Kä'eng

Quote from: omängum fra'uti on August 15, 2010, 02:26:57 PM
Perhaps I'm reading too much into it, but when he spelled out that syllables are (C)(C)V(C) rather than just saying (C)V(C) and declaring that the first C can be a cluster, the impression I got is that for pseudovowel syllables, it must start with a consonant, not a cluster.

Clusters are allowed, for example: lefkrr. Prrton said:
QuotePS: There are touches of unblessed vocab in here that are currently under consideration by K. Pawl. If I get answers, I'll post them here and in Lang. Updates (or he may unveil them via his blog). The only thing that's blessed is the adjective «lefkrr» for "current". (Le.F(ì).Krr). He decided a while back to allow CCrr/ll, but I don't think every announced it publicly.
Ma evi, ke'u ke lu prrte' to fwa sim tuteot ayawne.
Slä txo tuteo fmi 'ivampi ngat ro seng, fu nìfya'o, a 'eykefu ngati vä', tsakem ke lu sìltsan.
Tsaw lu ngeyä tokx! Kawtu ke tsun nìmuiä 'ivampi ngat txo ngal ke new tsakemit.
Ha kempe si nga? Nì'awve, nga plltxe san kehe. Tsakrr, ngal tsatsengti hum!

omängum fra'uti

Eh, then I would say I was right but he revisited the idea and opted to change his mind.
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

Kì'eyawn

Ma William, i just got a chance to sit down and read the whole thing, and took notes.  I tried to remove any corrections from my list that other people already caught; apologies for any repeats.

2.3.1.4. — I e-mailed Frommer about this—specifically, the case of eo oe -> *oeeo?  If i hear back before you get an answer, i will of course let you know.

2.3.2. — "syallble"

2.3.4. — "positition"

3. 3. 4. — Fra- + ay+ = fray+?
   I had been using this like "every" in English, and so thought it always attached to nouns in the singular.  But i've been wrong before.

4.1.1. — "The independent forms of the [?] from one to eight..."

6.1.2.2. — "The subject may be either subjective in agentive."

6.7.9. — "keyeug'it"

6.14.2 — "...a yes-no question can [?] made with an idiom.."

6.16.3 — "...may be joined without an conjunction connecting them."

6.17.3.1 — The more i look at this example, the more surprised/confused i am that it doesn't use pum—but that's my issue, not yours.

6.19.5.1. — "Though sì is most often found..., but [it?] can join verbs that are closely related."

7.3.1.3. — "formaltiy"

Zene oe pivlltxe nìmun, san Fìtìkangkem txantsan lu nìngay sìk!  Fìkemìri ngaru irayo seiyi oe, ma William.
eo Eywa oe 'ia

Fra'uri tìyawnur oe täpivìng nìwotx...

Zeykoyu Aean

Aloha William:

Thank you for a most significant piece of work!

I did read it through last night but I am so new at this that my mind was spinning by the time I was about 1/2 way through it.

Zaphod
[ Zeykoyu Aean ]

wm.annis

Quote from: tigermind on August 15, 2010, 04:06:30 PMMa William, i just got a chance to sit down and read the whole thing, and took notes. 

Fìtìkangkemìri oe irayo seiyi ngar.  These changes are all worked in.

Dreamlight

Quote4.1.2. Powers of Eight. Rather than "tens," Na'vi has "eights:"
This is true, but the text counts up to only the second power of eight (64 = "zam").  Powers of eight in the language go up to 84 "zazam" or 4096 in decimal (and of course the language counts up to 32,767 or 85-1).  Maybe a little more completeness?  (See Taronyu's dictionary for the Na'vi octal counting system.)
http://www.reverbnation.com/inkubussukkubus
"Peace on Earth" was all it said.

wm.annis

Quote from: Dreamlight on August 16, 2010, 07:16:07 PM
This is true, but the text counts up to only the second power of eight (64 = "zam").  Powers of eight in the language go up to 84 "zazam" or 4096 in decimal (and of course the language counts up to 32,767 or 85-1).  Maybe a little more completeness? 

I have enough questions about the proper use of zam and above — accenting, composition — that I've left them out for now.