Gì, ma eylan! 🙂
The previous post on Reef Na’vi phonetics and phonology elicited some really perceptive questions in the comments section, so let me respond to those here. Then I’ll mention a few other things you might be interested in.
FAQ about Reef Na’vi Phonetics and Phonology
Q: Do the phonetic rules of Reef Na’vi extend to the numeral prefix pxe-?
A: Yes! In RN, ‘three guests’ is befrrtu, ‘three rivers’ is behilvan, and ‘three fires’ is betep. (Do you see why it’s not *bedep?)
Q: What happens to words like atxkxe or ekxtxu? Would they be changed into atxge and ekxdu, or would they become adge and egdu?
A: It’s adge and egdu. This is an exception to the rule that the changes from ejectives to voiced stops only occur at the onset of a syllable. In these cases, when ejectives are in contact, the change in the second ejective influences a change in the first. It’s an example of what linguists call “regressive assimilation,” where a sound reaches backwards, so to speak, to influence a preceding sound.
Q: If a noun ends in an ejective (e.g., ’awkx), what happens if we append a case ending? Is the ejective reanalyzed to become the onset of the next syllable? So, is ’awkx + -ìl pronounced as ’aw-gìl or as ’awkx-ìl?
A: It’s ’awgìl. You’re right about the reanalysis.
Q: Is the glottal stop also lost between two words, e.g. oe ’ia, or does that rule only apply to glottal stops in a single word?
A: Only within a single word. Initial glottal stops are generally retained.
Q: We know that in FN, if a word has a closed syllable containing u, it can be pronounced [ u ] or [ ʊ ] (and apparently now, anywhere in between). But if u appears in an open syllable, like in tìfnu, it is always pronounced [ u ]. What is the situation for RN?
A: In RN, ù is perfectly possible in non-final open syllables. So for example, txula, the word for ‘build’ in FN, is dùla in RN. [Edit 15 Jan.: Example changed. See comments.] (That’s a word that will be on our list of FN-RN correspondences where RN has a ù.) In final open syllables, however, ù, although theoretically possible, is very rare. I doubt we’ll be seeing examples of that. This situation is different from the i-ì contrast, where both vowels are frequently found in final position.
Q: Would RN speakers insert an optional tìftang in sequences where FN has identical vowels together like spono-o and zekwä-äo?
A: No. The tendency in RN is to smooth over the glottal stop, so they wouldn’t go in the opposite direction and insert it.
Q: Would RN allow sequences of identical vowels to remain where FN cancels them, such as apxaa (apxa), meeveng (meveng), seii (seiyi)?
A: Given that RN speakers are used to hearing sequences of identical vowels, as in rää, you’re right: forms like apxaa, meeveng, and seii remain as they are in RN.
Really great questions!
And now, two announcements:
First, I’ll be giving a little online talk this Sunday, 15 January, as part of OmatiCon. It’ll be about the Reef Na’vi dialect, with some additional exploration of dialects in general. There won’t be much new about RN phonetics and phonology, but I’ll also be previewing the morphological and syntactic differences between FN and RN, which I haven’t yet gotten to here on the blog.
The talk will be at the following times (I hope I’ve gotten this right!):
US Pacific Time: 11:00 am
US Central Time: 1:00 pm
US Eastern Time: 2:00 pm
UK Time: 7:00 pm
European Continental Time: 8:00 pm
Here’s a link you can use to join the Zoom meeting:
https://us05web.zoom.us/j/85404913238?pwd=bEhzK1dseG9SUlVJWU9HQ2xJQzliZz09Meeting ID: 854 0491 3238Passcode: T994RJ
Secondly, if you’re not tired of articles about Na’vi where Pawl say things that have become all too familiar to you, there’s a new one you might like to take a look at. It’s on Salon.com, an online magazine with quite a large readership. I think the author did a very nice job, and I’m quite pleased with how it turned out. As you’ll see, the genre is one where the interviewer transcribes what the interviewee said almost verbatim. It’s not polished prose, but it does give you a sense that someone is speaking spontaneously rather than reading prepared text.
Hayalovay!
Tìyusuneri ngaru fìmuntrrkrrka oe so’heia nìtxan nang, ma Karyu Pawl! Layu trray ‘awvea skxom a tsun oe tivok tsatsenget! Keyeyri oeru txoa livu 🙂
Keyey srak? Kea keyeyti ke tse’a oel kaw’it, ma tsmuk! 😊 Sìlpey oe, ngal fpìyevìl futa tsasäftxulì’utsyìp oeyä eltur tìtxen si.
Furia tsaysìpawmur kangay si irayo ngaru nìtxan! || Thanks for addressing these!
Just for clarification though on the glottal stop/vowel front (I thought it was in the comments last time). We are including diphthongs in this rule as well, correct? So it’s nìaw in RN. (btw. ’a’aw could get really weird 🙂 )
Speaking of adjectives: you said initial glottal stop is usually retained but since the attributive a attaches to the adjective, it should disappear, right? In short, tute aipu or tute a’ipu for a ‘humorous person’?
Ngeyä säftxulì’uri trray srefereiey nìprrte’!
Yes, the rule for glottal stops with vowels applies to diphthongs as well: nìaw is correct in RN.
Good point about ’a’aw! In a phrase like ’eylan a’a’aw ‘several friends,’ the rule would seem to imply that this could become *’eylan aaaw’! Weird indeed. That’s where the “optional” part of the rule kicks in. In a case like this, RN speakers would not drop both glottal stops. The most likely pronunciation, I think, would be ’eylan aa’aw.
‘Humorous person’ would be tute aipu, since the original glottal stop is now between two different vowels and no longer initial.
how it would work for stressed tuté word? would it become *tutee in RN?
Stress is not the same as two identical vowels. the é only implies that that syllable is stressed (as opposed to tute where the first syllable is stressed) and should therefore not become tutee in RN afaik 🙂
Well, yes, that too, but seeing how some dual vowels are no longer merged into one by rules, like FN does, I thought that (just in theory) ethymologically things like “gender” root additions would make sense, since original word is just tute (with e unstressed) and most of gender-like nouns appear to have just those endings added.
Does the deletion of the tìftang apply to both sides of a diphthong or only the “vowel side”? For example, does kaw’it become kawit?
And what about pseudovowels: does mll’an turn into mllan, or even tsa’llngo into tsallngo?
Sì’eyngìri irayo nìtxan nìmun ma KP!
Another question just came up. -yu is productive for verbs, including si-verbs, and as far as we know, “…siyu” would be spoken as “…syu” in colloquial speech, kefyak.
Would LW turn that into “shu” as well?
Sorry if this question is redundant, I just want to make sure 😀
Good observation. Of course, -siyu doesn’t have to turn into -syu in colloquial speech, but it can. When it does, you’re right: it would be pronounced “shu” in RN. So, for example, tsamsiyu would come out as tsamshu. Hmm. Does that sound like a fiercer warrior than tsamsiyu or a less fierce one? I’m not sure! 😊
Haha, this is great, thanks for your reply, as always!
In my experience, as is often the case with (German, for example) dialects, it’s a very subjective thing… for an “outsider” it may sound endearing or funny in a way, maybe even ridiculous, but to one who grew up with that dialect, it would just be perceived without any of those connotations, I would guess 😀
So to a Na’vi from the forest, tsamshu would maybe sound funny or endearing, but to a Na’vi born and raised at the reef, it would just sound normal and fierce as it should? Fun to think about!
This is awesome! I have a question tho, regarding all the u vowels, is there a rule when it changes (apart from that final position role you just told us about) we can use to update our dictionaries or will we just have to wait until we get a more complete list released?
Unfortunately there’s no rule. We’ll just have to wait for the list. 😊
Does this mean that kxu was originally kxù and LW preserves the ù in derived forms like kxùtu but not in kxù itself? What about regular inflected forms like kxùl or kxùä?
[Edit 15 Jan.] Looking at this with “fresh eyes,” as they say, I’ve changed my thinking on kxutu. Although “frozen forms” are certainly possible, the explanation for why gùtu appears alongside gu in LW seemed forced and implausible. So I changed the example to one in which the problem didn’t occur. And it’ll be gutu for ‘enemy’ in LW, not *gùtu. Irayo for bringing this my attention!
We came across a discussion: You mention in your clarification on the u vs ù topic that kxutu becomes gùtu and that syllables which are final and open usually contain a u not an ù. However, the word you used as an example, kxutu, is deprived from the word “kxu” for harm, isn’t it? Which now rises the question if kxu becomes gu, following the more or less rule you just told us about or does it become gù following the model of the word related to it?
Never mind, I should read all comments, before sending my own
No worries! It’s a good question, and I appreciate your pointing out the problem. As you’ll see, I’ve changed the example. Irayo!
So’ha! I love seeing the subtle implications of the dialect handled here. One question, how about glottal stops between a vowel and a consonant like in the character Ao’nung vs Aonung? I’ve seen both variants floating around the net, but is Ao’nung the FN pronunciation and Aonung the RN one?
That’s a good question. To be honest, I’m not sure why there are those two variants. Even though Reef Na’vi doesn’t normally delete glottal stops between a vowel and a consonant, it’s possible, as you’ve suggested, that Aonung is the RN version and Ao’nung the FN one. But until we get confirmation of that, or until we find out that one version is correct and the other a misspelling, it’s safest to assume that they’re simply two alternate versions of the same name.
You said in the Zoom meeting something interesting about the word tìkankxan which becomes tìkangan and how we’re able to distinct the g sound so it doesn’t look like a Ng sound. You proposed a – (tìkan-gan) but also said if someone has other (more aesthetically pleasing) ideas to let you know. I would maybe propose to write the g as a capital letter if it is its own sound (like tìkanGan) it’s still weird to look at, but it looks more like it’s still the same word
I have also heard this statement from Pawl and suggest the use of an interpunct. See here: https://en.wikipedia.org/wiki/Interpunct
Especially in the section about the Catalan language.
I love this! Thanks for bringing it to my attention. It’s easy to type on my Mac–OPTION + SHIFT + 9. And it’s less obtrusive than the other solutions. This situation will occur only rarely, but when it does, I think the interpunct is the best solution.
I also fielded the interpunct idea in a Discord server. It’s quite original and aesthetic!
Txantsan!
I found more support for the interpunct today, when I found out Occitan uses it in almost exactly the same way that’s been proposed for Na’vi. Per Wikipedia:
In Occitan, especially in the Gascon dialect, the interpunct (punt interior, literally, “inner dot”, or ponch naut for “high / upper point”) is used to distinguish the following graphemes:
s·h, pronounced [s.h], versus sh, pronounced [ʃ], for example, in des·har ‘to undo’ vs deishar ‘to leave’
n·h, pronounced [n.h], versus nh, pronounced [ɲ], for example in in·hèrn ‘hell’ vs vinha ‘vineyard’
I’m liking it even more now.
Säfpìl asteng, tìkan ateng! Great minds think alike! 😊 I had considered cap-G as well. That or the hyphen would work–it’s a question of aesthetics. But I really like Viewing guy’s suggestion about the interpunct, which I was unaware of! I think that’s the best solution to the problem.
It’s probably too late to say about but an idea came to me to use ñ symbol just in places of ambiguity, on -nkx- combinations.
So it would be tìkañgan just for tìkankxan, and the usual ng would not be affected, making -ngkx- into -ngg- in RN.
Thank you so much for your presentation during Omaticon, I really enjoyed listening/watching and am grateful for the opportunity to see it live!
That presentation raised another question, hrh. According to what you said, in unstressed syllables, Ä often gets changed to E.
Okay, two questions, ngaytxoa.
1) What does “often” mean, are there any rules or details to know when exactly that sound is changed? I would guess you will probably touch on this in the 2nd part about this new dialect, but I wanted to ask anyway.
2) According to this, ätxäle (ä-TXÄ-le) should result in edäle, kefyak, would that be correct?
Also, not sure if this is intentional or a typo, just wanted to point it out or rather ask about it: Is apxaa supposed to be abaa?
And one last question… Oeru txoa livu, ke new fpomit ngey txivung! Newomum nìtxan nì’aw, hrh.
I gather that in RN it’s seii, because RN simply doesn’t implement Y, as FN does, to avoid two identical vowels being right next to each other in a verb.
If this is correct: What about tìng? In FN it would be teiyìng, kefyak, so would RN just say teiìng?
I was discussing this very question with Wm. Annis, author of our Horen. Let me get back to you on this.
This seems related to the question of when the genitive ending is -yä (-ye) rather than just -ä (-e). It always seemed to me that the purpose of the y there is to separate vowels that are too similar, but then again for the genitive (unlike the infix ei) LN extends that to all front and all open vowels even though normally sequences like iä would be entirely acceptable… So I guess the question is, how old is this y and to what extent is it actually governed by phonological considerations like LN’s aversion to sequences of identical vowels?
In any event, forms like veiyll would still need the y, wouldn’t they?
Would change of genitive ending -yä (-ä) to -ye (-e) mean also that infixes like and change into and in RN? (since infixes don’t change stress of a verb)
I meant, if äp and äng change into ep and eng respectively, ngaytxoa.
Ma elongater, I believe it is very likely that the infixes äp and äng change to ep and eng, considering that not only are they always unstressed but also äng can under certain circumstances change to eng even in Forest Na’vi.
Irayo, ma tsmuk. Glad you enjoyed the presentation!
Yes, “often” is pretty vague. I’m afraid I’m going to have to leave it that way for a while until I hear more of the Reef people’s speech (which I will at some point!) and form some conclusions about when they implement this sound change.
But you’re right, ätxäle corresponds to edäle when the RN changes are implemented.
As for apxaa, that was how the word appeared in the original question (in the last post’s comments), and I just went along with that. If you’re going to spell out the b, d, g sounds in RN, however, then of course you’re right–it would be abaa.
Mesì’eyngìri… ìì, kolan, pxesì’eyngìri, hrh, irayo nìtxan! 🙂
One example of the unstressed ä not changing is rää. But for me this is completely understandable and natural. With the deletion of the tìftang, we have sort of a similar situation to adge in which the latter sound makes the former conform to it. It would be more effort for me to say reä, because that stressed ä—specifically now without any separation from the unstressed—puts me into ‘ä mode’ in anticipation.
Ngaru tìyawr. Mllte oe nìwotx.
Kaldì ma KP,
Mipa aysäomumìri irayo nìdan!
I’ll have to catch up on the presentation you gave today, irayo for having taken the time for us. 🙂 I gathered some information from the presentation, and I’m looking forward to exploring the topic more in depth (for example, the application of the phonetic rules and the grammar of RN).
Nìvingkap, who is the friend you’re greeting at the beginning of the post? 😉
Hayalovay!
HRH! Ah, those stray apostrophes! I could say it’s a secret friend who must remain nameless . . . but I think I’ll just remove the offending tìftang instead. 😊 Irayo!
It is an unrelated to this post question, but I had that for a long time: why is no *pxoeng there, but straight away ayoeng after moeng?
Na’vi has pxoeng like Pxoeng uvan seri – “We three (both linsteners, inclusive) are playing”. 🙂
*moeng on the other hand would make not much sense because oeng already includes two persons which is even visible in the word itself oe (I) + nga (you), and of course oeng is dual already.
Do not confuse it with the exclusive pronoun of moe, which is the dual exclusive version of oe, like Moe uvan seri. – “We both (but not the listener, exclusive) are playing.”.
Oh, wait, how to say then “we both and you”? Or is it “pxoeng” itself? Is there no way to say “we three and you” as a separate pronoun then?
“we both and you” would be pxoeng because there are three people total: you (who say this words), someone else who is actually with you, and the person you talk to.
“we three and you” would be ayoeng since there is no prefix for “four” or even a pronoun that includes four people total only. I am not sure if one can say pxoe sì nga to get that what you want to have separatly.
Hi, I wonder, are there words for grandmother/grandfather terms in Na’vi? How was Mo’at referred to by her grandchildren?
Hi,
as we practiced reef speech we had a little discussions about what happens when one word ends with an ejective but the next starts with a vowel (like in this sentence: ” ‘efu ohakx oe”), is it smoothed out (like “ohag oe”) or not?
You state here that the glottal stop rule doesn’t breach borders of one word to the next, but when one word ends usually with an unreleased plosive, it becomes a “real one” when the next letter is a vowel.
Come on, Paul, update us!
Kxì!
I know, old topic, but one question arose today: Is the -‘ite/-‘itan in full names affected by the dissappearing tìftang rule if the mother’s/father’s name ends with a vowel? So for example: My mother’s name is Claudia, would I be Kìlawdia’ite or Kìlawdiaite in Reef Na’vi? I assumed second, but after others questioned it I was wondering if that is really seen as one word, or if it is seen as two words actually, which then would lead to the tìftang being retained.
ta Tekre