clause partitioning

[`Eylan Ayfalulukanä]

In an excellent and instructive post yesterday, Kewnya Txamew`itan writes

A clause is formed around a verb and its arguments and, in general, there is a bubble around each clause so that elements outside it cannot affect those inside (for example, the ergative case is used for subjects if there is an accusative. If that accusative is in a different clause however, the 'bubble effect' stops the accusative from affecting the subject and so it remains in the nominative e.g. in "new oe tivaron yerikit" there is a subject and accusative so you might think that oe should be oel, but because of the bubble affect the accusative on yerik stops oe needing to be ergative).

This answers some important questions involving complex sentences (which I struggle with more than anything else), but also generates others.

First of all, could some examples of this be offered that show how these clauses 'partition themselves' and how to clearly recognize them?

Second, if a clause consists of a verb and its arguments, how can the argument of a given verb end up outside its clause in such a way that it loses (or perhaps gains) case marking?

Clear and confusing at the same time!

[kewnya txamew'itan]

There is never a perfect way of knowing how a sentence is partitioned into clauses, {new oe {yerikit tivaron}} is correct and, at the same time, {new {oel yerikit tivaron}}.

Whenever there is an "a" involved, a clause has either just finished or is just being introduced. In "tsun oe ngahu pivängkxo a fì'u prrte' lu oer" it is showing that an attributive clause has just finished but in the sentence "{kelku {a oe tsamì kelku si} teyr lu}" it marks the start of a new clause ("a" doesn't fit nicely in either clause, it's effectively a border so I may be inconsistent with where I draw the border when I partition sentences).

Conjunctions like fte, ulte and fu also all introduce new clauses although in this case there isn't always one matrix clause. Examples: {oel yerikit taron} slä {pol 'ängtsìkit taron}; {oel tsat tspamang} fte {hawnuyu sivi olo'ur} (in this case the second clause is arguably an argument of the first which would be the matrix clause); {ayuniltìranyul awngat zivong} fu {skiva'a} or lastly {oe yom} ulte {po taron}.

It's also easier to understand some case markings if we treat modal verbs as always introducing a new clause for the verb following them.

Generally it's easy to tell where clauses begin, where they end is often harder to see and is mainly practice.

[`Eylan Ayfalulukanä]

I need to read this a couple dozen times and study more examples. But this is really helpful.

I am hoping someone eventually will put together a guide that will help people figure out how to build the more complex sentences. (Your example of tsun oe ngahu pivängkxo a fì'u prrte' lu oer is a great example of a sentence that at first face, does not have a  really clear meaning.)

[Muzer]

I don't quite know what you mean - although it doesn't seem very common in Frommerian writing to me (prove me wrong), I instantly recognised the "a fì'u" and knew it was referring to the whole of the clauses before it. Maybe I'm just naturally good at breaking up clauses, or it was a fluke, but it didn't seem hard to me.

[kewnya txamew'itan]

Ma Muzer, the entire sentence (except possibly the position and the choice of -r over -ru) is Frommerian.

He does seem to use fì'u a more commonly though and this seems to be the trend in na'vi as a whole (hence the contractions only occur when they end with a instead of starting with it.

[`Eylan Ayfalulukanä]

I guess it is the challenge of overcoming my English preconceptions, and this is probably worse because I did not previously know any other language but English. Thus, the very different constructions are taking a serious getting used to, and I can't pretend to fully grasp it.

That said, I instantly recognized this sentence, as it had been thoroughly picked apart some time in the past. I also know that a fi`u is a 'backwards' form of futa. No matter how hard I have tried, I have never been able to build a sentence that emzawna`u that uses a fi`u or futa.

The only reason I have not been submitting more sentences is a severe lack of time to work on stuff like that.

[Lance R. Casey]

I also know that a fi`u is a 'backwards' form of futa.

Of fwa, actually. Mind the cases.

[`Eylan Ayfalulukanä]

I also know that a fi`u is a 'backwards' form of futa.

Of fwa, actually. Mind the cases.

Thanks for catching the missing a. I edited that into my post.

I (think I) see what you mean now. a fi`u is kind of a 'root', and indicates that the words behind the a essentially make up a nominative case noun. Since futa indicates the clause that follows is accusative case, the 'backwards' version would be a fi`ut.

So if you wanted to indicate that a preceding clause makes up an ergative case noun, would that be a fi`ul?

If that is correct (and I think there is also a fi`uru and a fi`uä for dative and genitive cases, respectively), why isn't there a futa-like contraction for the ergative, like fula (or furua or fuäya)?

[kewnya txamew'itan]

The lack of a contraction reflects how much less commonly they are used. The most common fì'u[CASE] a formation is accusative followed by nominative and topical so they'd contract before the less commonly used ones like fì'uä a.