Mrra tìpängkxotsyìp    Five little discussions

Kaltxì, ma frapo.

From time to time I receive emails from members of the lì’fyaolo’, asking for clarification about certain aspects of Na’vi. The questions often demonstrate a lot of insight into the language and help me clarify and deepen my own understanding. I’d like to share a few of those more recent inquiries with you here, along with my responses. Sìlpey oe, ayngari faysìpängkxotsyìp eltur tìtxen sìyevi!

’Awvea Tìpängkxotsyìp: Does the modal verb new ‘want’ have a causative form neykew?

(Note: The original version of this discussion was posted to LearnNa’vi on Nov. 29.)

New is a modal, but it’s also a vtr, a transitive verb. There are six such “dual function” verbs in our current dictionaries, labeled either vtrm or (a) vtr and (b) modal:

fmi                  ‘try’
kan                 ‘aim, intend’
may’               ‘try, taste, sample’
new                ‘want’
nulnew          ‘prefer’
sto                  ‘refuse’

In the case of new, how would you say, for example, ‘I want to dance’? The dual nature of this verb means you have a choice:

1. 1. Oe new srivew. (new used as a modal)
   2. Oel new futa srew. (new used as a vtr)

A is more common, but B is certainly possible. (By the way, B is also the only way to have the “wanter” and the dancer be different: To say, ‘I want you to dance,’ it’s got to be Oel new futa nga srew.) Note that after futa in such constructions, we don’t need <iv> on the verb, although it’s not wrong to have it. (See the next section below!)

So far so good. Now . . . how do we convert this to a causative construction? How would we say, for example, ‘This music makes me want to dance?’

Well first of all, the causative infix <eyk> doesn’t go with modals. So it’s the B version that gets “causativized,” not the A version.

Second, the causer—the one making something happen—is always in the agentive case. Our wonderful Horen Lì’fyayä leNa’vi states the rule clearly:

6.11.2. Causative of Transitive Verb. When a transitive verb is made into a causative, the causee, which had been in the agentive case, goes into the dative. This leaves the original accusative in place.

Applying this rule to B, and realizing that “the original accusative” here is futa, we get:

Fìpamtseol oeru neykew futa srew.
‘This music makes me want to dance.’

The new agent here, with the l case marker, is fìpamtseol ‘this music’: it’s the music that’s making something happen!

 

Muvea Tìpängkxotsyìp: Is the <iv> infix used with the complements of modal verbs?

This question is related to the previous one.

Recall that for ‘I want to dance,’ we have two equivalent versions:

A. Oe new srivew.                                  (new used as a modal)
B. Oel new futa srew.                             (new used as a vtr)

But is there a third version as well? What about:

C. Oel new futa srivew.              (new used as a vtr)

C is indeed possible, but it merits some explanation.

There was a time in the early days of Na’vi when I would have used the C version exclusively. As my feeling for Na’vi evolved over the years, however, I realized that with fwa and futa, the bare verb will do just fine. For example:

Sunu oer fwa srew.
‘I like to dance.’

That is, literally, ‘The dance-thing brings me enjoyment.’ Today I would judge *Sunu oer fwa srivew as ungrammatical, since it would be saying the equivalent of ‘The might-dance-thing brings me enjoyment.’

With the dual-function verbs, however, the situation is a little different. The A and B versions of our ‘want to dance’ sentence are the most expected versions in Na’vi. But given that the simple version A is much more common than B, there’s “analogical pressure” on B for the verb to conform, and so we get C, a pattern which, for these verbs, is also considered correct.

By the same token, we have both:

D. Oel new futa nga srew.
‘I want you to dance.’

and

E. Oel new futa nga srivew.
‘I want you to dance.’

 

[EDIT 28 Feb: THIS SECTION NEEDS TO BE REVISED. The Koren stated towards the end is incomplete. Further discussion will appear in a subsequent post.]
Pxeyvea Tìpängkxotsyìp: What can the pronouns po and sno refer to?

Let’s begin with an English example:

F. John thinks that Bill likes his car.

The question is, whose car is it that Bill likes—John’s car or his own (= Bill’s) car? I think most English speakers would say that without any context, the referent of “his” is ambiguous: it could be either one. But what about this slightly modified version:

G. John thinks that Bill likes his own car.

I think most people would say that G is no longer ambiguous: it has to be Bill’s own car.

In Na’vi, sno, in all its forms, works somewhat like ‘his/her own’ in English. Take a look at these examples:

H. Ateyol fpìl futa Ralul peyä tsmuket ve’kì.
‘Ateyo thinks that Ralu hates his sister.’

I. Ateyol fpìl futa Ralul sneyä tsmuket ve’kì.
‘Ateyo thinks that Ralu hates his sister.’

Although the English translation is ambiguous, the Na’vi sentences are not: In H, it’s Ateyo’s sister. In I, it’s Ralu’s own sister.

For those who like technical linguistic rules, here’s a Koren a teri tsalì’u alu sno, a rule about the word sno:

Sno, in all its forms, can only refer back to a noun phrase within the same clause.

In particular, sno in a subordinate clause can’t refer to a noun phrase in the main clause. This means that in I, sneyä, being in a subordinate clause, can only refer to Ralu, a noun phrase in that clause. It can’t refer to Ateyo, which is in the main clause.

 

Tsìvea Tìpängkxotsyìp: When are final stops unreleased?

Here’s an interesting pronunciation question:

We know that stops are liable to be unreleased under certain conditions but are wondering about the exact scope of this rule. In particular:

1. 1. Are they unreleased only at the end of a word, or at the end of every syllable (as seems to be more common in human languages that do this)?
   2. Does this rule include the glottal stop?

First, a little linguistics. 😊

What’s the difference between a consonant sound (C) and a vowel sound (V)?

The answer is that with V’s, the air flow through the vocal tract is not blocked; the air flows freely. With C’s, there’s some blockage that restricts the flow of air. Sometimes the blockage is only partial, as in the case of s, z, f, v, etc. With those sounds, the passage for the airflow is narrowed, creating friction and a characteristic sound. (The sounds I just listed are in fact called fricatives. 😄) Sometimes, however, the blockage is complete, and the airflow is momentarily stopped. And guess what: the C’s that do this are called stops! As you might imagine, there’s more to this story, but that’s the basic idea.

In Na’vi, the stops are k, p, t, kx, px, tx, and ’, the glottal stop.

Now as you know, when three of these stops—namely, k, p, and t—occur at the end of a word, they’re “unreleased.” As I think I’ve mentioned before, this phenomenon occurs optionally in English. If I say, “What’s up?” I can either “explode” the p, releasing the air that’s been trapped, or not release it, keeping my mouth closed. The sound is a bit different in each case. In proper Na’vi, these final stops are unreleased.

With that background, what about question 1? If k, p, and t are non-final (that is, not at the end of a word), can they still be at the end of a syllable? Yes they can, but only if they’re followed by another consonant. For example, k, p, and t are syllable-final in nik.re, txep.mì, and ’ok.trr. They’re not syllable final in a.kum, tsa.po, and nì.teng. When they are syllable-final, as in the first group, they’re unreleased as if they’re word-final.

As for the second question, in my experience I’ve never heard “released” and “unreleased” applied to the glottal stop. Take the word olo’. Do you hear a difference between a released and an unreleased tìftang? I’m not sure what a released glottal stop would sound like. But if there is a difference, it would follow the same rules as for k, p, and t.

 

Mrrvea Tìpängkxotsyìp: How is fpap used?

Finally, a correction:

In the previous post, I gave this example for the vtr fpap ‘pound’:

J. *Krra sti nìtxan, pol mesyokxit fpap sìn fyanyo.
‘When he’s angry, he pounds his hands on the table.’

Some astute readers asked if it shouldn’t be:

K. Krra sti nìtxan, pol fyanyot fpap fa mesyokx.
‘When he’s angry, he pounds the table with his hands.’

The question is, when you pound something, what exactly are you pounding—your hands, or some external object?

I see in retrospect that that example I gave, J, was clearly influenced by English, since in English we can say both “He pounded the table with his hands” and “He pounded his hands on the table.” But that seems to be unusual; I don’t know of other languages where that happens. (If anyone does, please let me know!) Although I’m not positive, I assume that at some point in the history of English, some kind of semantic shift occurred, where the grammatical object of “pound” could be either the external object that gets pounded or the instrument of pounding. But that’s English, and there’s no reason to think such a shift occurred in Na’vi as well. So K represents the correct use of fpap, and I’ve corrected the example in the previous post.

For this unusual holiday season, ma eylan, I wish you all the best celebrations you can manage lefkrra tìfkeytokmì.

Please stay safe, everyone . . . ulte makto zong.