First sentence, correction required!

[Shazzle]

Kaltxí, ma frapo!

I bet you experts are all tired of correcting people, so I'll be a bad person and put up another task!
After doing the excellent activity book twice, I now decided to try and make a sentence. (Deja vu?) I'd be delighted for a correction!

Ha, txo utralìri zup, utral pamit si, srak? Ngian, txo kawtu tok tsatseng stawm utralìl kllte takuk,-
pam mi lu srak?
So, if a tree falls, the tree makes a sound*, right? However, if no one's there to hear the tree hit the ground, does the sound still exist?

*A "BAM!" sound.

I'm very unsure about this. From what I can see, the tree is the subject of the story, as it "creates" a sound, which is the "second object", or what?

Thanks in advance! 

[kewnya txamew'itan]

Hej, jeg lærer dansk, så den oprindelige (på dansk) ville være nyttigt for mig. (Som man rettelser til denne)

Anyway, back to English and na'vi. The first things is that the "so" in the English isn't the same so as the one that "ha" describes which is closer to "therefore"/"which implies that" (using the mathematical terms e.g. ∴ or ⇒ signs) but instead is a general conversation/discourse starting which, in na'vi is "tse".

Then you've made the very common mistake of unnecessarily using the topical. Unfortunately there are no clear guidelines about its use which is far more nuanced than the other cases, and even the best of us are rarely 100% sure when to use (although here it is wrong) and so the general advice is to avoid it where possible. Anyway, fall is intransitive so "utral" just takes the usual, plain nominative case which is just "utral".

Then, you've made the reasonable (albeit wrong) assumption that si behaves much the English verb "to make" or, probably more accurately, like the Danish "at gøre" alas, it does not. Instead, si is only used in compound verbs, that is to say, in words where it and a noun or adjective are used together as a single unit which is intransitive (takes no direct object). From that, you would be forgiven for thinking that "pam si" is correct, but unfortunately "si" is not productive which is just a fancy way of saying that, if we do go around sticking on things all over the place we don't know for certain whether it's correct or whether a new verb will come along that says exactly the same thing but is said (and written) completely differently. So instead, although it sounds rather clunky, I'd use the verb "ngop" meaning "creates" which is transitive.

Then there's a slight confusion on the  case marking in the clause "utral pamit ngop" (I've swapped my ngop in for you si), because there is a direct object in the accusative case here, utral has to take the ergative case which has the ending -ìl for words that end in a consonant. That said, because in the previous clause "txo utral zup" you established the tree as the subject, you can drop it in the next clause and it will be assumed to be the subject, so I'd just say "pamit ngop".

Then, as it is there's nothing linking the two clauses which isn't wrong, but might make it confusing, so I'd add in a "tsakrr" between the two clauses and put the second one in the subjunctive mood (which is used for the effect of a cause in constructions like this) with the infix <iv> which goes in the verb "ngop".

Lastly (for the first bit) srak implies a question whereas you're asking more for confirmation. English doesn't really distinguish the two (and I don't  think Danish does either), but na'vi does so I'd use "kefyak" instead at the end of this sentence.

So the first sentence would be:

tse, txo utral zup tsakrr pamit ngivop kefyak?

Then it gets a bit counter-intuitive, first of all, in na'vi double negatives are required when you use a noun or adverb like "kawtu", "kawkrr" or "ke'u" etc. so you need to put a ke before the "tok". To make matters even worse, unlike English and Danish where "to be somewhere" is intransitive (fancy linguistic words for it not having a direct object marked as such), in na'vi it is transitive so you need to put kawtu in the ergative (using the suffix -l) and tsatseng in the accusative (either as tsatsengit or tsatsenget/tsatsengeti using the long form of tsatseng).

Then, even more complicated (I'm sorry, there's not really any good way to explain this), but there's nothing linking the act of hearing the "nobody" (in English this is done with the "to" infinitive), in na'vi we'd do this by attributing the entire clause to kawtu using "a", what this means is that the entire "stawm ..." bit is placed straight after the kawtu and then an "a" is put in between the kawtu and stawm.

Then, the object of the hearing is the sound (pam) which, being object should take the accusative case as pamit) but we still haven't said what the sound is. To do that we have to attribute another clause, this time to the pamit, again using a, so we stick the a straight after the "pamit" and then put the "takuk" clause straight after that "a".

Again, we need to link the cause and effect (I'd use tsakrr and <iv> again) Unfortunately there's a little but of work to be done on the "takuk" clause. First of all, think about what is striking what. You should realise that the tree is being struck and so takes the accusative and becomes klltet or kllteti and you've already got utral in the right case (the ergative), we can't imply this utral like we did earlier, because we've now established "kawtu" as the subject. So that clause becomes "utralìl klltet takuk".

For the last clause, whilst it's not wrong, I'd probably say "pamit ngop mi srak" to put some symmetry in the sentence.

So in total I'd say:

TSe, txo utral zup tsakrr pamit ngivop kefyak? Ngian, txo kawtul a stawm pamit a utralìl klltet takuk tok tsatsenget tsakrr pamit ngivop mi srak?

1 thing to note, I'm not happy with ngop, pam si might be better, I'm just not using it because it isn't productive whereas ngop is.

Anyway, it was a very complicated sentence for a first try, I suggest that you try a few simpler things and work your way up because it can get quite hideously complicated with long things like this. You did very well considering what you chose though.

[Shazzle]

Many thanks for the help, really! *Gives karma cookie*. When I first started on the sentence, I did think of using "ngop", but I later decided to make it into si. - Big mistake. 
Now I do realize how complicated that sentence was going to be, so I'd better study some easier sentences.

Again, thanks! 

[kewnya txamew'itan]

Si is more elegant and I prefer it as a solution, the only reason I used ngop is that we know it works, and I'm struggling to think of any ambiguity that might be caused by using si so it should be fine even if a better way of saying it does come along in a month or two. (although then you'd replace "pamit ngivop" with "pam si" because si verbs are intransitive so "pam" would no longer have to be accusative).

[`Eylan Ayfalulukanä]

Txantsan ma Kewnya! Nga siltsana karyu lu

And you are getting better all the time.

Your post brings up a couple questions about subjects:

1. If you establish a subject in nominative case, and then your next verb is transitive, do you need to reestablish the subject?

2. You showed that s subject 'remains in force' until the subject changes. Can you do this with objects as well?

3. Is it incorrect to not imply subjects, as you did here?

[kewnya txamew'itan]

1. If you establish a subject in nominative case, and then your next verb is transitive, do you need to reestablish the subject?

2. You showed that s subject 'remains in force' until the subject changes. Can you do this with objects as well?

3. Is it incorrect to not imply subjects, as you did here?

1. No, weirdly (given the distinction made between transitive and intransitive subjects in the case system) it is the subject that's implied not a nominative.

2. You can certainly do so with accusative/direct objects, I don't know about dative/indirect objects, it's possible you can, but a verb in between which takes no dative object might prevent it from being implied after that. So yes with accusative objects and possibly with dative ones.

3. No, although it might sound strange, or, like Spanish emphasise that noun (for example in Spanish, because verbs are conjugated for the person performing them pronouns are dropped when they are the subject so "quiero" = "I want" but "yo quiero" = "I want"), we don't know what effect it would have (if any), but it shouldn't be wrong.