Na'vi Linguistics: the Syllable

Started by wm.annis, January 22, 2010, 05:55:22 PM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

wm.annis

I assume everyone reading this already knows a bit about Na'vi sounds and spelling, that, for example, px is a single consonant, even if written with two letters, and that ts though composed of two sounds is considered a single unit in Na'vi.

Another fun thing Na'vi forces us to think about in more detail than most of us are used to, is the syllable.  If we cannot dissect a syllable correctly, we're likely to misplace an infix.  Now, I know some people here wail about "walls of text" and linguistic jargon, but linguists came up with this terminology for a reason — it's useful.  So I'm going to give a quick overview here of a more technical way of thinking about the syllable.  When I'm done, hopefully we'll all have a way to very concisely and accurately say where an infix should go.

Every syllable has three parts: an onset, a nucleus and a coda.

Let's start in the middle, with the nucleus.  In every language on the planet [suomichris — can we ignore Nuxálk for now?] the nucleus is the only required part of a syllable.  It is the peak, central part of the syllable, and is nearly always a vowel sound (simple vowel, diphthong).  Here are some simple words in English with the nucleus underlined: see, at, flock.  In Na'vi two sounds most of us think of as consonants can be the nucleus of a syllable, rr and ll.  English actually has the ll sound, again underlining the nucleus, people (or pipll in Na'vi spelling).

Going back to the front, the onset of a syllable is the consonant or consonants the syllable starts with.   Different languages have different rules about what are allowable onsets.  English onsets can be very complex.  Na'vi can have either (1) no onset, sometimes called a null onset, (2) a single consonant onset or (3) a consonant cluster with f, ts or s followed by another consonant (see the Language Log post for the legal elements of the cluster).  Note that because Na'vi does not require an onset, you can get words like äie vision, a three syllable word with each syllable having a null onset.  Going back to my English examples, here are the words again, with the onset marked: see, flock, Øat (I've used a Ø to mark a null onset).

And now to the end of the syallble, where we have the coda.  This is the conclusion to the syallble.  English can have very complex codas, but in Na'vi you have two choices, (1) no coda at all, as you might expect also called a null coda, or (2) a single consonant (again with restrictions laid out in the Language Log post).  Back to my three English words: flock, at, seeØ.

The fundamental rule of Na'vi infixes is that they come before the nucleus.  If there's a null onset, that doesn't change anything.  The word eyk, lead has one syllable with a null onset, the nucleus ey and the coda k: Øeyk.  To add everyone's favorite infix, ‹iv›, just insert it before the nucleus: iveyk.  Similarly, if the onset is a consonant cluster, that also doesn't change anything: tspang has a tsp onset, a a nucleus and a ng coda.  After you add the near future infix ìy, you get tspìyang.

Extra mumbo-jumbo:  A syllable that has no coda is sometimes called an open syllable, and one with a coda is called closed.  This matters in Na'vi only because there's a restriction on the vocalic rr and ll.  They may not occur in closed syllables.  This constrains the shapes words can take in Na'vi, and can even intrude on grammar.  To add the ergative ending to the word trr day, you have to use the form you'd use with a word ending in a consonant, because *trrl would be illegal, since it puts rr into a closed syllable.


Edit: clean up closed syllable matter.

suomichris

Nice post!  And very helpful for those new to linguistics/Na'vi, I imagine!

Quote from: wm.annis on January 22, 2010, 05:55:22 PM[suomichris — can we ignore Nuxálk for now?]
Well, I guess... But can we talk about Berber instead??? ;)

QuoteEnglish actually has the ll sound, again underlining the nucleus, people (or pipll in Na'vi spelling).
It might also be worth noting that English also has an /r/ sound (although not the same /rr/ as in Na'vi) that can stand as the nucleus of the syllable, at least in some dialects: e.g., "pepper."

QuoteGoing back to the front, ...
!!!

QuoteThe fundamental rule of Na'vi infixes is that they come before the nucleus.
This is a far more important point than it may appear at first, and I'm going to try and explain why, hopefully without making this confusing.

In Na'vi, we have all of the vowels that can be the nucleus of a syllable, along with the sounds /rr/ and /ll/, but note that /rr/ and /ll/ are still consonants, even though they stand in as the the nucleus.  The Wikipedia page talks about (or used to) putting infixes before the first vowel.  But what about the (deduced) root /lrr/ "smile"?  It doesn't have a vowel, and if we try to apply this "before the vowel" rule, we're going to have some trouble.  But, by viewing it as the nucleus, we're now able to deduce that the infixes should go l<here>rr.

suomichris

Quote from: wm.annis on January 22, 2010, 05:55:22 PM
Extra mumbo-jumbo:  A syllable that has no coda is sometimes called an open syllable, and one with a coda is called closed.  This matters in Na'vi only because there's a restriction on the vocalic rr and ll.  They may not occur in closed syllables.  At the moment I can't think of a grammatical process which puts you in danger of running into that problem.  But Frommer made a special point to remind people of it, so I'll mention it here.
I think this is important because of the case allomorphs.  So, if we have a word like /trr/ "day," and we want to make it ergative, we can't just put an /l/ on the end, because this would create */trrl/ a form that is not allowed in Na'vi.  So, we have to do something else, and use the allomorph of the ergative marker /-ìl/ instead, giving us /trrìl/, which is fine as far as Na'vi phonology goes.

omängum fra'uti

For the point of people learning, if you are going to call it (Properly) a nuclease of the syllable then that means you have to first teach people what the nuclease of a syllable is.  Which brings you right back to the explanation that caused confusion in the first place.

I think Frommer's approach of calling /ll/ and /rr/ psuedo-vowels, and calling the whole set of vowels, diphthongs and psuedo-vowels as "syllabic vowels" may be the least confusing approach for just teaching the language (vs teaching linguistics).
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

Eaite Randjam

All of this is very important and informative to a new student to the Na'Vi language, at least it is with me.  Thanks for this all important lesson!

suomichris

Quote from: omängum fra'uti on January 22, 2010, 07:10:01 PMI think Frommer's approach of calling /ll/ and /rr/ psuedo-vowels, and calling the whole set of vowels, diphthongs and psuedo-vowels as "syllabic vowels" may be the least confusing approach for just teaching the language (vs teaching linguistics).
I haven't see this, but it is confusing, as diphthongs can't stand as the nucleus of a syllable...

wm.annis

Quote from: suomichris on January 22, 2010, 07:06:40 PMSo, we have to do something else, and use the allomorph of the ergative marker /-ìl/ instead, giving us /trrìl/, which is fine as far as Na'vi phonology goes.

Ahh!  I hadn't thought of that.

QuoteBut can we talk about Berber instead???

No.  ;)

omängum fra'uti

#7
Quote from: suomichris on January 22, 2010, 07:35:26 PM
Quote from: omängum fra'uti on January 22, 2010, 07:10:01 PMI think Frommer's approach of calling /ll/ and /rr/ psuedo-vowels, and calling the whole set of vowels, diphthongs and psuedo-vowels as "syllabic vowels" may be the least confusing approach for just teaching the language (vs teaching linguistics).
I haven't see this, but it is confusing, as diphthongs can't stand as the nucleus of a syllable...
They can in Na'vi.  It's spelled out on his explanation of phonetics.

In that way, "tspayk" (Just threw together some sounds there, didn't feel like searching the corpus for an example) is a perfectly acceptable syllable.

In Frommer's words, from the language log,
QuoteEvery syllable has a single vowel or diphthong at its center. Each vowel or diphthong in a word corresponds to a separate syllable. A single vowel or diphthong may be a syllable by itself.

Within syllables, Na'vi vowels and diphthongs can be preceded by either one or two consonants. They can also be followed by one consonant. That is, the syllable structure is (C)(C)V(C), where V represents a vowel or a diphthong. Restrictions on which consonants can occur in which positions are given below.
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

Nìwotxkrr Tìyawn

hehe ma suomichris you just got buuurned.
Naruto Shippuden Episode 166: Confession
                                    Watch it, Love it, Live it

suomichris

Quote from: omängum fra'uti on January 22, 2010, 07:43:24 PM
Quote from: suomichris on January 22, 2010, 07:35:26 PM
Quote from: omängum fra'uti on January 22, 2010, 07:10:01 PMI think Frommer's approach of calling /ll/ and /rr/ psuedo-vowels, and calling the whole set of vowels, diphthongs and psuedo-vowels as "syllabic vowels" may be the least confusing approach for just teaching the language (vs teaching linguistics).
I haven't see this, but it is confusing, as diphthongs can't stand as the nucleus of a syllable...
They can in Na'vi.  It's spelled out on his explanation of phonetics.

In that way, "tspayk" (Just threw together some sounds there, didn't feel like searching the corpus for an example) is a perfectly acceptable syllable.

In Frommer's words, from the language log,
QuoteEvery syllable has a single vowel or diphthong at its center. Each vowel or diphthong in a word corresponds to a separate syllable. A single vowel or diphthong may be a syllable by itself.

Within syllables, Na'vi vowels and diphthongs can be preceded by either one or two consonants. They can also be followed by one consonant. That is, the syllable structure is (C)(C)V(C), where V represents a vowel or a diphthong. Restrictions on which consonants can occur in which positions are given below.
Ah, sorry, I misunderstood you; "diphthongs" is, to me, always included in the category of vowels.  I thought, from what you were saying, that you meant that "y" and "w" by themselves could stand in as a nucleus.

That said, though, we still see differences between vowels (included diphthongs) and rr/ll: namely, as your example "tspayk" shows, you can still have a consonant following a diphthong, but not rr/ll, so it is worth distinguishing them...

suomichris

Quote from: wm.annis on January 22, 2010, 07:40:07 PM
QuoteBut can we talk about Berber instead???

No.  ;)
Pfft. Fine.  :p

I actually have only read about the so-called "strident vowels" (I think?) that show up in Berber, and don't know much about them.  I'd sure love to HEAR them, tho!

omängum fra'uti

That's a good point about /rr/ and /ll/ vs real vowels...  But again looking at this from a lessons point of view, you're grouping them with vowels under one term, I'm just doing it under a different term that my thinking is would be a little less intimidating to people.

But they are still distinct.  I never call them vowels, and nether does Frommer.  He calls them psuedovowels, and I think that is a good descriptive term for them because they are standing in place of a vowel without actually being them.
Ftxey lu nga tokx ftxey lu nga tirea? Lu oe tìkeftxo.
Listen to my Na'vi Lessons podcast!

suomichris

Quote from: omängum fra'uti on January 22, 2010, 08:18:30 PMBut they are still distinct.  I never call them vowels, and nether does Frommer.  He calls them psuedovowels, and I think that is a good descriptive term for them because they are standing in place of a vowel without actually being them.
Yes, they are distinct in some ways, but not in others.  This is what I meant about the Wikipedia page (or maybe it was the guide?)--despite the fact that rr/ll are NOT vowels, in some of the materials that I've seen, they are treated as if they were, which is a) wrong and b) confusing.

roger

#13
We have a cline from V to C. The simple vowels are the most vocoid. The diphthongs are less vowel-like, since they trigger the consonant allomorphs of the case endings. The syllabic liquids are even less vowel-like, as they do that and in addition disallow coda consonants. Then of the consonants, the frics (incl. /ts/) are the most consonant-like, in that they cannot form codas, and perhaps we could argue the voiceless frics are even more consonant-like, because they can form clusters w all the rest.

As for the "pseudo-vowels" not being vowels, is the nucleus of "church" a vowel in rhotic dialects of English? There's some disagreement among linguists as to whether it's a rhotic vowel or a syllabic consonant. I wonder if there's any real difference between the two. (For me, English syllabic ar is sesqui-syllabic, the way diphthongs are; none of them really accept /l/ as a coda: peel "pee-uhl", pearl "purr-uhl".)

suomichris

Quote from: roger on January 23, 2010, 06:34:46 PM
We have a cline from V to C. The simple vowels are the most vocoid. The diphthongs are less vowel-like, since they trigger the consonant allomorphs of the case endings. The syllabic liquids are even less vowel-like, as they do that and in addition disallow coda consonants. Then of the consonants, the frics (incl. /ts/) are the most consonant-like, in that they cannot form codas, and perhaps we could argue the voiceless frics are even more consonant-like, because they can form clusters w all the rest.
Yeah, I think there is something like that going on here with conditioning the case allomorphs....  Would that we had more examples!

QuoteAs for the "pseudo-vowels" not being vowels, is the nucleus of "church" a vowel in rhotic dialects of English? There's some disagreement among linguists as to whether it's a rhotic vowel or a syllabic consonant. I wonder if there's any real difference between the two. (For me, English syllabic ar is sesqui-syllabic, the way diphthongs are; none of them really accept /l/ as a coda: peel "pee-uhl", pearl "purr-uhl".)
Well, in my dialect, words like "pearl" and "rural" definitely don't have a schwa, and pearl is... Well, probably only one syllable, but if it is two, it's a syllable /r/ followed by a syllabic /l/, which not schwa... I've always thought some of the disagreement here comes from people who speak different English dialects :p

Nìwotxkrr Tìyawn

It is my understanding that in english for something to be a syllable it has to have at least one vowel so I don't know why anyone would say "pee uhl".
Naruto Shippuden Episode 166: Confession
                                    Watch it, Love it, Live it

donjoe

Quote from: wm.annis on January 22, 2010, 05:55:22 PMWhen I'm done, hopefully we'll all have a way to very concisely and accurately say where an infix should go.
Then wouldn't this topic fit the "Prefixes, Infixes and Suffixes" subforum better? ;)