Genitive case refinement; declension of tsaw

[roger]

Since it has been established that diphthong-final words get case endings like consonant-final ones (i.e one that starts with a vowel), I don't understand why these aren't tsawit and tsawìri...

Except that it hasn't been established. That was just the simplest predictive explanation that fit the forms known at the time. But now we also have kemri as an alt to kemìri.

[tsrräfkxätu]

We've never actually seen "tsa" on it's own from a canon source.  Only as a prefix, or with case endings.  Not sure on tsaw vs tsa'u.

We kinda have.

For inanimate "it" you shouldn't use po but rather tsa: a tsane po karmä.

Here, he refers to tsa as a standalone word (not a prefix), and we see it used with a postpositional affix (and not a case ending.)

I just figured it was -ä after high & back vowels (which would cover diphthongs and rr, ll as well) and -yä elsewhere.

[ɪ] is also high, so this might not work, unless there are no nouns ending in an ì, which would explain why F didn't mention it.

Oh, sorry, my bad. Well then your rule seems to be hard to beat, maybe with:

Katot taftxu oel nìean nìrim (Weaving Song)

You managed neatly, though. Had I been right, this would be *nìyean. I could still work around it, but creating complicated phonotactics to explain an few simple exceptions just doesn't seem to be worth it.

[wm.annis]

Except that it hasn't been established. That was just the simplest predictive explanation that fit the forms known at the time. But now we also have kemri as an alt to kemìri.

And it's really not clear what motivates different endings for words ending in consonants anyway — no rules of phonotactics are broken if you cram any ol' ending onto any ol' word.  This puzzled me way back when we had only a few Na'vi sentences.  In particular, fìskxawngìri.  The word fìskxawngri is also legal (according to the sound rules).  So why do different sorts of nouns get different sorts of endings?  I assume some prosodic consideration is in play, but I don't know for sure.

[roger]

ì isn't high, but only almost high. But we have a problem here, because -yä is also attested after i, for example in slu Na'viyä hapxì in Jake's dialog. Of course, it is Jake, and there are other discrepancies there, such as eu for ew, but there's also Utral Aymokriyä in Paul's current vocab spreadsheet.

[roger]

Except that it hasn't been established. That was just the simplest predictive explanation that fit the forms known at the time. But now we also have kemri as an alt to kemìri.

And it's really not clear what motivates different endings for words ending in consonants anyway — no rules of phonotactics are broken if you cram any ol' ending onto any ol' word.  This puzzled me way back when we had only a few Na'vi sentences.  In particular, fìskxawngìri.  The word fìskxawngri is also legal (according to the sound rules).  So why do different sorts of nouns get different sorts of endings?  I assume some prosodic consideration is in play, but I don't know for sure.

I need to take some of that back. As for kem(ì)ri, consider the context: a variant of the Golden Rule. The loss of the ì may be due to meter rather than to a general rule; we see similar things in the songs, where for example sì reduces to s. Whenever I've seen Paul put a letter in parentheses, it's either a known variation (such as -r(u)), or meter is involved.

And what are the only other cases of case-suffix vowel loss after anything other than a simple vowel? Tsawt and tsawri may have these forms because they are colloquial contractions of tsa'ut, tsa'uri, which are perfectly regular with those forms. So that may be the reason for the exception, rather than any free variation.

Do we have any other cases where we cannot predict the case forms based on the final phoneme of the word, orthographic C vs V? Other than the "strong" accusative in invariable -ti ?

As for why they behave this way, grammatical affixes tend to have a much simpler phonology than the lexicon. A strong tendency toward a CVCV structure would not at all be unusual.

[omängum fra'uti]

The interesting thing about the golden rule is that if the (ì) were marked because of meter...  It would need to be in there to KEEP the meter.

Kem-(ì)-ri  a nga-ru prr-te' ke lu
Tsa-kem rä-'ä si -vi ay -la -he-ru

But the running hypothesis is that kem would require the ìri form, so why would that need to be marked as such?  Even odder is that if he DID want to drop it....  He could easily have used the contracted "aylaru" on the second line to keep the meter, admittedly at the loss of a rhyming syllable at the end.

[roger]

The interesting thing about the golden rule is that if the (ì) were marked because of meter...  It would need to be in there to KEEP the meter.

That's not how I see it:

    Kem-ri-a-nga-ru-prr-te'-ke-lu
  Tsa-kem-rä-'ä-si-vi-ay-la-he-ru

  = o , , o , o , ? ,
    o , , o , o , o , ,

Which works if we stress ke and contract to aylaru as you suggested. It's not very satisfying as verse (I prefer the other version, which is rather elegant), but kemìri would add an extra syllable, making it o , , , o in the first line.

[Taronyu]

So, do we have any cases of o -> e \ _yä?

[omängum fra'uti]

You mean besides "Oeri ta peyä fahew akewong ontu teya längu"?

[Na'rìghawnu]

So, do we have any cases of o -> e \ _yä?

holpxay ayzekwäyä feyä "the number of their fingers"

http://wiki.learnnavi.org/index.php?title=Canon#Some_Conjunctions_and_Adverbs

[Swoka Swizaw]

ì isn't high, but only almost high. But we have a problem here, because -yä is also attested after i, for example in slu Na'viyä hapxì in Jake's dialog. Of course, it is Jake, and there are other discrepancies there, such as eu for ew, but there's also Utral Aymokriyä in Paul's current vocab spreadsheet.

I assume that you posting this means that you've sat upon your seat and, hopefully with no one else around to wonder if you're "having a moment," rendered these sounds to yourself. If so, I also assume that you think the glide from {i} to [{] is distinct in sound from {i} to [j{]. Personally, I don't hear it, if I don't try to...

Essentially, I feel that the canon CAN be debated as to whether the glide ([j]) is actually there or not. We've not had any examples, as far as I know, that really hindge on whether of not the /y/ is there in situations like you've listed.

[Kì'eyawn]

ì isn't high, but only almost high. But we have a problem here, because -yä is also attested after i, for example in slu Na'viyä hapxì in Jake's dialog. Of course, it is Jake, and there are other discrepancies there, such as eu for ew, but there's also Utral Aymokriyä in Paul's current vocab spreadsheet.

I assume that you posting this means that you've sat upon your seat and, hopefully with no one else around to wonder if you're "having a moment," rendered these sounds to yourself. If so, I also assume that you think the glide from {i} to [{] is distinct in sound from {i} to [j{]. Personally, I don't hear it, if I don't try to...

Essentially, I feel that the canon CAN be debated as to whether the glide ([j]) is actually there or not. We've not had any examples, as far as I know, that really hindge on whether of not the /y/ is there in situations like you've listed.

I don't know if this makes any difference, but maybe we see this "long" genitive attaching to Na'vi and Aymokri because these are parts of "proper nouns"?

[roger]

Hey, I got an answer on this one!

As you've said, there's not much difference between -iä and -iyä (perhaps no difference at all), so it's essentially a spelling question.

Here's what's out there:

-iä

1. muiä (LEXICON)
2. niä (LEXICON)
3. menariä (VIDEO GAME)

(Fmawnit menariä ke tsun oe spivaw. 'I can't believe what I'm seeing with my eyes.')

-iyä

4. Na'viyä (SCRIPT)
5. Aymokriyä (LEXICON, SCRIPT)
6. ritiyä (VIDEO GAME)

1 and 2 are roots, so they shouldn't be a problem. 3 through 6 are genitive forms. So if the official spelling is -iyä, the only outlier is 3, which I'll change.

Bottom line: the genitive for bases ending in -i is -yä. Apologies for going back and forth on this.

One more inconsistency:

For bases in -a, the gen. was -yä in 3 cases (tompayä, Eywayä, tìska'ayä), but there was also Omatikayaä. I feel comfortable leaving that last word alone, as an exceptional form: *Omatikayayä doesn't seem stable--I can see yayä > yaä.

[Swoka Swizaw]

I don't know if this makes any difference, but maybe we see this "long" genitive attaching to Na'vi and Aymokri because these are parts of "proper nouns"?

Sure, but not all proper nouns will end in a vowel... /yä/ can't attach to consonants.

[Kì'eyawn]

I don't know if this makes any difference, but maybe we see this "long" genitive attaching to Na'vi and Aymokri because these are parts of "proper nouns"?

Sure, but not all proper nouns will end in a vowel... /yä/ can't attach to consonants.

Of course, ma smuk.  I only meant that maybe there's a rule that proper nouns ending in a vowel always take the long form, instead of sometimes taking the short.  But it seems we have an answer from Karyu Pawl, in any case.

[Swoka Swizaw]

Sure, but not all proper nouns will end in a vowel... /yä/ can't attach to consonants.

Of course, ma smuk.  I only meant that maybe there's a rule that proper nouns ending in a vowel always take the long form, instead of sometimes taking the short.  But it seems we have an answer from Karyu Pawl, in any case.

Sorry. I figure you knew that...

[tsrräfkxätu]

As you've said, there's not much difference between -iä and -iyä (perhaps no difference at all), so it's essentially a spelling question.

Here's what's out there:

-iä

1. muiä (LEXICON)
2. niä (LEXICON)
3. menariä (VIDEO GAME)

(Fmawnit menariä ke tsun oe spivaw. 'I can't believe what I'm seeing with my eyes.')

-iyä

4. Na'viyä (SCRIPT)
5. Aymokriyä (LEXICON, SCRIPT)
6. ritiyä (VIDEO GAME)

1 and 2 are roots, so they shouldn't be a problem. 3 through 6 are genitive forms. So if the official spelling is -iyä, the only outlier is 3, which I'll change.

Bottom line: the genitive for bases ending in -i is -yä. Apologies for going back and forth on this.

One more inconsistency:

For bases in -a, the gen. was -yä in 3 cases (tompayä, Eywayä, tìska'ayä), but there was also Omatikayaä. I feel comfortable leaving that last word alone, as an exceptional form: *Omatikayayä doesn't seem stable--I can see yayä > yaä.

So is it -yä again for all V-ending stems? Or...?

[NeotrekkerZ]

Bottom line: the genitive for bases ending in -i is -yä. Apologies for going back and forth on this.

I take that to mean the o and u still take only -ä, not -yä

[roger]

Bottom line: the genitive for bases ending in -i is -yä. Apologies for going back and forth on this.

I take that to mean the o and u still take only -ä, not -yä

Correct. And after i it's pretty much just an orthographic convention.