A newly discovered rule of Na'vi word order?

[wm.annis]

I haven't sent this yet to KP for verification, but I do think I have discovered a rule of Na'vi word order (at about 95% confidence). Below is the text I'm planning to send to him fairly soon, but I was wondering if people had other comments or examples.

_____
I have recently had two things on my mind.  First, I want to use a few
more lines from the films as examples in the Horen where possible.
Second, the word order section is old and could use an update, and you
yourself have suggested the discussion of the Topical could stand a
review.  I was reading through the LearnNavi.org wiki which has a page with
all the Na'vi lines from the first film, which has been filled out and
refined over the years.  I was looking through it to find examples of
different core word orders, and I noticed a pattern.

  - pot zamunge
  - pot lonu
  - fìswiräti ngal pelun molunge fìtsenge
  - tsaswiräti lonu
  - fìketuwongti oel stìyeftxaw
  - pot tsun oe tspivang nìftue
  - fìtaronyur apxan teswotivìng äiet angay
  - tsampongut Tsu'teyìl iveyk
  - foti awngal tìyakuk mì te'lan  (txe'lan?)
  - fra'ut fkol skera'a
  - mefoti yìm
  - ikranti makto
  - (ma sempul,) ngati oel kin
  - (ma 'ite), tskoti munge
  - fìpoti oel tspìyang fte tìkenong lìyevu aylaru

Some recent things by you from your blog (I assume many vocab example
sentences are generated not by you but by the community, so I leave
those out of the survey):

  - Ayngeyä aysìralpengit ngop nì'o'!
  - Tsat oel zosleykolu.

There is a strong tendency here, I think, though not a hard-and-fast
rule:

   Korentsyìp amip: definite constituents are often fronted.

By definition, personal pronouns, most uses of demonstratives, and any
phrase with a demonstrative prefix are definite.  Many languages treat
anything possessed, or possessed by a personal pronoun, as definite,
too.  The topical is of course also definite.  So, the fronting of
phrases in the topical *case* is just the grammaticalization of a
general Na'vi tendency.

My guess is that when you have several definite constituents, if
fronting happens at all, the most salient one is chosen.  I do note,
though, that definite nouns, such as with fì-, seem likely to oust
pronoun, which are background information in most sentences (to the
degree that in plenty of languages pronoun forms in core argument
cases (S/A, O roles) are often clitics).  Any remaining definites then
go pretty much anywhere, with the proviso that sometimes last position
is used for focused constituents.

[Vawmataw]

Interesting! I'm looking forward to his reply.

[Toliman]

Eltur tìtxen si nìtxan

[Alyara Arati]

Hmmm.  That would make sense in light of what I recall about sentence stress, namely that the most important position is the last word, but the second most emphasized location is the first word.

[Toliman]

Hmmm.  That would make sense in light of what I recall about sentence stress, namely that the most important position is the last word, but the second most emphasized location is the first word.

Srane, oe mllte ngahu

[Blue Elf]

Uh, can you explain by simple words for us, who are not linguists? Does it mean, that most important part of sentence goes first? AFAIR, originally was said that "punch" goes to last part of sentence, but as time change, importance of this rule decreased. And little correction:

pot longu

[`Eylan Ayfalulukanä]

Although I don't have a lot of trouble understanding most linguistic stuff, this went right over my head  

[wm.annis]

Uh, can you explain by simple words for us, who are not linguists? Does it mean, that most important part of sentence goes first?

Nope. It means almost the opposite. I used some linguistics jargon that will be familiar to Paul in this, but I should have given a bit more preparatory detail here.

AFAIR, originally was said that "punch" goes to last part of sentence, but as time change, importance of this rule decreased.

What Paul called "punch" is in linguistics terms usually called "focus." That is, it is the newest or most important part of the sentence. It's usually something new, unexpected, or having attention drawn to it in particular for some reason.

Definiteness, on the other hand, is all about old information, or at least information you can be expected to know. It's only being mentioned at all to help listeners orient themselves in relation to new information. In English, we mark definite information with the article "the." That just says "this word is referring something we've already talked about, or which I expect you to know about from the context." However, "the" is not the only way something can be definite. Certain kinds of phrases are definite by definition, even if they don't have "the." For example, personal pronouns (oe, nga, etc.), demonstrative pronouns (fì'u, tsapo, etc.), nouns with the demonstrative prefixes (fì-, tsa-). Possessed nouns are also very often considered definite, especially if the possessor is a personal pronoun (my hand). Finally, topics are definite.

In very many examples from the first movie, Paul moves definite phrases to the start of the clause, way more often than I'd expect to happen by stylistic chance. I went through the blog, looking for things he's said nìNa'vi there, and I can find further examples of this in his non-movie Na'vi, as well.

Putting definite and topical information at the start of a clause is quite common in the world's languages. It's used to help a listener orient themselves to the difference between old background information and new incoming information. Paul does speak a language which fronts definite and topical information — Malay.

He does this so much that I'd be confident to include this in the Horen even without his approval (as we had to do early, before he was answering our emails regularly and posting on his blog). I also have a more radical interpretation of this, but that might not get his approval, so I've not included that part here.


____
There are actually some circumstances where demonstratives might not be definite in some languages, but I don't think we've seen anything like that for Na'vi.

[wm.annis]

And little correction:

pot longu

Irayo!

[Vawmataw]

Something just popped into my mind: you seem to list only direct objects, but this could also apply to somes indirect ones.
For example, here are several conversational phrases (that - I hope - come from Karyu Pawl):

oeru syaw fko...
oeru sunu...
Xru tìyawr
Xru tìkxey
eltur tìtxen si
ngaru tsulfä
oeru meuia
oeru txoa livu

It's not happening all the time, but there is surely some pattern that Karyu Pawl can elucidate.

[wm.annis]

Something just popped into my mind: you seem to list only direct objects, but this could also apply to somes indirect ones.

Absolutely. I include one dative example in my list, but there are certainly others I could use.

The one confusion here is that in some sense, some of these datives are acting in a very subject-like way. This is another thing that occurs in natural languages reasonably often. They are sometimes called "dative experiencers," and I left them out to really focus on less subject-like things being moved to the front.

[Toliman]

Uh, can you explain by simple words for us, who are not linguists? Does it mean, that most important part of sentence goes first?

Nope. It means almost the opposite. I used some linguistics jargon that will be familiar to Paul in this, but I should have given a bit more preparatory detail here.

AFAIR, originally was said that "punch" goes to last part of sentence, but as time change, importance of this rule decreased.

Spoiler

What Paul called "punch" is in linguistics terms usually called "focus." That is, it is the newest or most important part of the sentence. It's usually something new, unexpected, or having attention drawn to it in particular for some reason.

Definiteness, on the other hand, is all about old information, or at least information you can be expected to know. It's only being mentioned at all to help listeners orient themselves in relation to new information. In English, we mark definite information with the article "the." That just says "this word is referring something we've already talked about, or which I expect you to know about from the context." However, "the" is not the only way something can be definite. Certain kinds of phrases are definite by definition, even if they don't have "the." For example, personal pronouns (oe, nga, etc.), demonstrative pronouns (fì'u, tsapo, etc.), nouns with the demonstrative prefixes (fì-, tsa-). Possessed nouns are also very often considered definite, especially if the possessor is a personal pronoun (my hand). Finally, topics are definite.

In very many examples from the first movie, Paul moves definite phrases to the start of the clause, way more often than I'd expect to happen by stylistic chance. I went through the blog, looking for things he's said nìNa'vi there, and I can find further examples of this in his non-movie Na'vi, as well.

Putting definite and topical information at the start of a clause is quite common in the world's languages. It's used to help a listener orient themselves to the difference between old background information and new incoming information. Paul does speak a language which fronts definite and topical information — Malay.

He does this so much that I'd be confident to include this in the Horen even without his approval (as we had to do early, before he was answering our emails regularly and posting on his blog). I also have a more radical interpretation of this, but that might not get his approval, so I've not included that part here.


____
There are actually some circumstances where demonstratives might not be definite in some languages, but I don't think we've seen anything like that for Na'vi.

That is very good explanation, irayo!

[Toliman]

It's not happening all the time, but there is surely some pattern that Karyu Pawl can elucidate.

I hope that It's really interesting!

[`Eylan Ayfalulukanä]

Could this also be an unconscious stylistic choice on Pawl's part?

[Toliman]

Could this also be an unconscious stylistic choice on Pawl's part?

Hmm... it could make sense too 

[Vawmataw]

It's possible too, but it would be more interesting as a grammatical feature. hrh

[Toliman]

It's possible too, but it would be more interesting as a grammatical feature. hrh

hrh... but yeah, you are right

[wm.annis]

Could this also be an unconscious stylistic choice on Pawl's part?

I'm guessing it's unconscious or he would have told us about it. However, because of how the topical also works, I think there's a more significant principle than just style at work.

[wm.annis]

Could this also be an unconscious stylistic choice on Pawl's part?

I guess I will add, there are places where I make special note of Paul's practice that are probably stylistic. He's the lì'fyayä sempul, so copying his style is meritorious, too.    One example we copy is the [ADJ lu nìtxan/nì'it/etc.] pattern, which is less about grammar than feel.

[Wllìm]

Hmm... It is an interesting observation, but I don't know if it is true. I looked at a handful "randomly" selected sentences from Plumps' corpus document and while some confirm this hypothesis, there were also plenty counterexamples. In particular, sentences such as tseri oel futa ..., fpìl oel futa ... sl., in which according to the hypothesis the verb probably shouldn't come first, are very common. Fpìl oel futa happens 8 times in the corpus, while oel fpìl futa does not occur at all. (The other option consistent with the hypothesis, futa ... (fpìl oel / oel fpìl) is hard to specifically search for in the corpus, but does not happen often in my experience.)

One also wonders whether, even if Pawl has a tendency in Na'vi he writes himself to do X, that means that he intends X to be a rule/tendency in Na'vi in general. Maybe this becomes a bit philosophical, but the way I see it, there may be a difference between what Pawl intends for the Na'vi language, and what he expresses himself in the Na'vi he writes. I think a descriptivist position is hard to maintain if the person generating the text we're describing is the same person that prescribes the rules in the first place (That being said, I would be interested in descriptive studies on Na'vi produced by the community, where there some word order patterns are clearly preferred over others, but that's a whole different discussion.)

In any case, I am very interested in Pawl's opinion, so please do let us know if/when you receive a reply

I assume many vocab example sentences are generated not by you but by the community, so I leave those out of the survey

I don't think that is true, at least not in the more recent posts, as the LEP has been dead for years at this point. Also as Pawl generally does not accept LEP submissions without considerably changing them to his taste, I would assume most example sentences are his own. This may be an interesting thing to ask him.