Kxì, ma frapo!
In the previous post, some questions that came up in the comments were left hanging. Rather than answering them there, I thought I’d do that here, since I don’t won’t anyone to miss them.
RN U VS. Ù
Thank you to Zángtsuva, Txonpay, and the others who pointed out that I haven’t been consistent in marking new vocabulary for u vs. ù in Reef Na’vi. Going forward, I’ll try to remember to do that. As before, if a word containing u is not otherwise noted, you can assume the RN version is u. Only words that have ù in RN will be called out.
New or recent words with ù in RN:
lemùngwrr
parwùn
fleyùl
Mangkuan, ’ulte, and <uy> are u in RN, not ù. And to confirm Záng’s assumptions, yes, derived words like wätùm and utùrtu have the expected pronunciations based on their etymologies, and we still have no exceptions to the tendency that ù does not occur at the end of a word.
THE ORDERING OF FINAL -PE AND CASE ENDINGS
This deserves some discussion.
The question came up in this example sentence:
Ayioangìlpe fìsponot parwun?
’What animals inhabit this island?’
As you know, ayioangìlpe ‘what animals’ is in the Agentive case. The breakdown is:
ayioang (animals: noun) + ìl (Agentive case ending) + pe (interrogative particle)
Several people felt the ordering of ìl and pe should have been the other way around: ayioangpel.
It’s remarkable to me that after all this time, there are still basic questions to be asked about grammar and structure! In this case, the question is:
When interrogative <pe> is a suffix, does it precede or follow case endings?
I don’t think there’s an immediately obvious answer to this question. Na’vi could go in either direction. Although I’m leaning one way, before I make a ruling I’d like to get your input—whether you feel strongly it should be this way or that, and why.
As some people noted, we have a possibly parallel situation with the -o suffix. In Wiliam Annis’s authoritative Horen, the rule is stated clearly:
3.1.3. Indefinite -o. A noun may take the indefinite suffix -o, one, some. Case endings follow the -o, such as puk-o-t.
So if -pe is analogous to -o, we’re done: it’s noun + pe + case endings. But are -o and -pe really parallel? Or perhaps the question is, do they need to be treated as parallel?
Before anything else, if anyone knows of a canonical example where one or the other of these orderings is found, please let me know! We should be consistent with what’s already been established. But I haven’t been able to find such examples, so at this point it’s an open question.
To facilitate your thinking, consider these examples:
1. Which teacher gave you that book?
A. Peharyul tolìng ngar tsapukit?
B. Karyulpe tolìng ngar tsapukit?
C. Karyupel tolìng ngar tsapukit?
2. Which book did the teacher give you?
A. Karyul tolìng ngar pefukit?
B. Karyul tolìng ngar pukitpe?
C. Karyul tolìng ngar pukpet?
3. To which student did the teacher give that book?
A. Karyul tolìng tsapukit penumeyur?
B. Karyul tolìng tsapukit numeyurpe?
C. Karyul tolìng tsapukit numeyuper?
4. To which teacher’s student did you give that book?
A. Ngal tolìng tsapukit numeyur peharyuä?
B. Ngal tolìng tsapukit numeyur karyuäpe?
C. Ngal tolìng tsapukit numeyur karyupeyä?
Items 1 through 4 illustrate <pe> with the Agentive, Patientive, Dative, and Genitive case endings respectively. In each item, A has <pe> as a lenition-triggering prefix, where there’s no controversy. B and C have <pe> as a suffix. In B, the ordering is case ending + pe. In C, the ordering is pe + case ending.
Looking at the B and C versions, which seem more natural to you? Which do you think would be easier to understand and process in conversation? Do you have a gut feeling about which one is preferable?
Let me know in the comments!
ta P.
ADDENDUM 18 April
Tewti! Tìpängkxo akosman nìtxan nang!! Irayo seiyi oe ayngaru nìwotx, ma eylan!
What great comments! Thanks so much to each one of you who contributed. I can hardly imagine a more intelligent, informed, and helpful discussion. Lì’fyaolo’ri awngeyä leiu oeru nrra nìtxan!
I’ve now read through all the comments (well, TJ-K, I’m still working on yours, but I’ll get there! 🙂 ). I’ll try to respond to all of them individually, although it may take a little time. But for the present, let me tell you my conclusions.
Although a few of you preferred the B versions, where pe was in final position, and provided well-considered arguments for that choice, most preferred the C versions, with the case endings final. Since it became clear that (1) the C-version is the ordering that’s been taught up to now, (2) most people feel more comfortable with C, and (3) most persuasively, we have canonical examples of C, which I had forgotten (’upet, tupel, yìpet), C will be the standard ordering. That is,
When interrogative <pe> is a suffix, it precedes case endings.
Accordingly, I’ve changed the example in the previous post to ayioangpel.
That said, I believe there are situations in which some speakers may come up with B-type orderings spontaneously. (See Neytiri’s interesting example in the previous post.) But those will be considered non-standard.
A bit more, if you haven’t already had your fill:
I suspect one of the reasons *ayioangìlpe sounded right to me when I wrote the previous post was influence from another language. As research for a language project I’ve been working on, I’ve been looking closely at Finnish, a non-Indo-European language I’ve found extremely interesting but also quite a challenge. (It has 15 cases!) In Finnish, you have forms like this:
autossani ‘in my car’
The breakdown is:
auto ‘car’
autossa ‘in a/the car’
autossani ‘in my car’
Importantly, <ssa> is not a postposition but rather a case ending (the Inessive case). So in this language, case endings are not necessarily word-final. That fact may have unconsciously influenced my thinking about Na’vi.
There’s more to say about how <pe> interacts with adpositions compared to how it interacts with case endings, but I think that’s enough for now. 🙂
We have a few examples of ‘upet (‘u + -pe):
Nga new yivom ’upet fìtxon? from https://naviteri.org/2013/03/whoever-whatever-whenever/
Fol fte ayspe’etut livonu fngo’ ’upet? from https://naviteri.org/2012/01/mipa-zisit-ayliu-amip-new-words-for-the-new-year/
Kaltxì ma Karyu Pawl,
I found a few instances where you seem to confirm C:
https://naviteri.org/2011/07/number-in-na%e2%80%99vi/#comment-985
Ngeyä tsafkxilet tupel rengolop?
https://naviteri.org/2013/01/awvea-posti-zisita-amip-first-post-of-the-new-year/
Furia täftxu ngal tok yìpet?
https://naviteri.org/2013/01/lifyari-po-peyi-how-good-is-her-navi/
Nga new yivom ’upet fìtxon?
https://naviteri.org/2013/03/whoever-whatever-whenever/
(Lì’fyaolo’tu Tsyili made a search engine called Eywa, it’s very practical: https://tsyili.learnnavi.org/)
For consistency reasons I would vote in favour of the case ending coming after -pe. The same rule applies with other suffixes such as -tsyìp, and adpositions (when acting as postpositions), which are like semi-case endings for me, also follow the same order. That being say, could there be any cases where it’s preferrable to have the case ending before -pe?
Eywa ngahu
I think -pe followed by the case ending makes the most sense to me. I’m used to it always being word-final, and in my opinion, I think that’s the proper way to do it because you can think of the case endings as “[noun]-[role in the sentence]”. The [noun] part can and should be any affix to make the noun whatever it needs to be.
– “those three ikran” – tsapxeikran
– “which school of fish” – snapayoangpe
– “every herd of direhorses” – snapa’lipe
– “some type of rock” – fnetskxeo
Then whatever that noun is, it receives a case ending to indicate its role. I think that the Na’vi would always expect to hear the case ending at the end of the word, and never somewhere within it.
Ope, I was switching my example words around and forgot to update the third one. I meant frasnapa’li, ngaytxoa hrh
Kxì,
My previous comment is waiting for approval, but in the meanwhile here’s a spam filter-friendly version for everyone:
If you search the posts in January 2013, you’ll find
– Ngeyä tsafkxilet tupel rengolop?
– Furia täftxu ngal tok yìpet?
In March 2013, we also have this example: Nga new yivom ’upet fìtxon?
You’ll also find in a comment under the post “Numbers in Na’vi” where “tupel tok tsatsengit?” is considered as correct.
Ta Vawmataw
Wion ftxozä sayeii fìdon! Speaking of the reef, the fact that –uy– doesn’t have ù means that all the –uyu homonyms work in reef dialect! Such as ayewkuyu, fyawìnduyu, täftxuyu and zìmauyu.
“and we still have no exceptions to the tendency that ù does not occur at the end of a word.”
So that means the datives in reef are -r, -ru and -ur, kefyak?
I once thought when looking at the noun affix order in Na’vi that putting –pe before a case ending felt like the only unepected part of the affix order. But I also thought that was my own bias talking (Na’vi is the first language with noun cases that I learned).
By the way, no matter the outcome of ioangpel vs. ioangìlpe, adpositions as suffixes would go in the same slot as case endings, kefyak?
Also, I had a thought. If le’awnì’aw means “unique”, what is “uniquely”? 😁 Is there some general rule around this which hasn’t been written yet?
-pe to me feels a bit more belonging to the stem/root of word than being a situational modifier as case endings or ending adpositions. And I am kinda more used to -pe then case/adp than to case/adp then -pe. Still, it would be quite an imteresting gimmick, if case ending then -pe would be allowed. I would not exclude a possibility of seeing this more often, should people prefer this over -pe then case ending.
Well, I see that now the question regarding ordering -pe and ending adp is a separate one, which would be interesting to hear about too.
There are also such fancy thingies as -sì and -to, which I’d still prefer to see past everything else, as it implies some consistency for me. In case -pe stays after case ending, then how would -sì/to behave then?
In any way, if there would be several of such words with -pe&co, I think it’s rule that every and each one of them would follow the one pattern within a given sentence/clause.
I had more of questions in regards of ordering all the noun suffixes. Should I ever find my backlogs or texts (some are not quite accessible due to reasons), I will let know my thoughts about here.
As others have pointed out, we have a few official examples for -pe + case ending in that ordering. That ordering always felt more natural to me, and in fact, I wasn’t even aware that this is an open question! If learners asked, I (and other teachers from what I have seen) have always answered that it would be A. or C..
It honestly just feels weird to not have a case ending being word-final. Other affixes change the word themselves (and therefore it feels natural to me to apply them first, they are part of the word building process), case endings however, instead of changing the core meaning of the word, place the word in the sentence, which feels like the “final step”. Of course this is just a feeling that I apply to Na’vi here – To my knowledge, there is no universal rule that case needs to be “the furthest away” from the word core in language. Maybe if I have some time later today I’ll spend the afternoon researching 😀
Did a quick check on Wals.org immediately, and as always, they do not disappoint and have a chapter on case marking – look like there are even languages (albeit very few, 7 out of more than 1000 investigated languages) that use infixes! I’ll read the chapter later, maybe look at some languages in detail and if I find something intersting I’ll report back.
https://wals.info/feature/51A#2/10.9/151.0 – if someone is interested 😀
I think there is no inherent grammatical reason why either B or C should be easier to understand or more natural.
If I try to think about the question with a “fresh brain”, I think I actually somewhat prefer B iover C. The reason is that pe+, when used as a prefix, comes before the plural prefix (pemeikran, not *mefeikran). In this case the question marker is further “outside” than the plural market – so it would make sense that the question marker also goes further “outside” than the case marker. Besides this, there’s also a parallel with -sì, which also goes behind the case ending (ikranìlsì, not *ikransìl, kefyak?)
However, as mentioned by others, there are examples for style C and correspondingly, in the lì’fyaolo’ C has been taught as correct for as long as I remember. So I’d say version C feels more natural to me by exposure.
The question di, by the way, immediately remind me of your discussion of language universals from your talk last year, and the conclusion that Na’vi supposedly violates one of them 🙂
Addendum: for words like tupe I don’t know how style B would work. A declension like tupe, *tulpe, *tutipe, … seems very wild to me. For pronouns in general I strongly lean towards C. There’s also a parallel with the question of tsayu vs. aysa’u where the “pronoun use” may be distinct from the “noun use”.
Kaltxì!
Personally, I would prefer to stick with the existing rules (see examples that were provided by Vawmataw sl.); however, one idea might be to introduce this new variation or rule as a dialectal feature, for example for the Mangkuan. 😉
Eywa ayngahu 🙂
Kaltxì ma Karyu Pawl,
Irayo seiyi ngaru, for specifying words using ù. It may interest you to know that me and Zángtsuva maintain a list of ù words based on your original 2023 list.
https://docs.google.com/file/d/1UzPcnMUQ1fOGr-ch0OUDZzgtGKl9U4_S/edit
Irayo adan!! Nìia tseiùn zeykivo aylìupùkti sì mesestarsìmti aylìue a hu u fu ù! 🙂 🌊 I also appreciate that words that have u in lì’fya wione are indicated explicitly in this post and in previous posts (such as when «nun» was introduced) and of course the original u/ù list; it’s good to have the confirmation that it’s an intentional choice and not just an instance of forgetting! hrh
Concerning «ayioangìlpe», I see that the lì’fyaolo’ has not failed to list the reasons to think it should rather be «ayioangpel», so I will merely mention the weaker arguments that I see suggesting that «ayioangìlpe» might also be allowed:
1) Although, as you say, the order «puk-o-t(i)» is preferred, you also commented about «puk-it-o» in an old blog post: “Let’s say it’s a non-standard form you hear among some speakers, that inadvertently crept into my example. I’ve now changed it to the standard pukot. [True confessions: This is a place where the rule changed. In the language document I submitted to Lightstorm at the end of 2009, the rule was that the indefinite/durative -o could either precede or follow the case endings. I later tightened that up so that -o only precedes the case endings. But I guess the old form is still in my head. Fìtxeleri zene oe nari sivi!]” So the analogy with «-o» would imply that «ayioangìlpe» is a form that some speakers would use sometimes alongside the more universal «ayioangpel».
2) Of course, because «pe-» exists as a prefix and not only as a suffix «-pe», there is already a sort of precedent for flexibility in its position.
3) As a consequence of 2) and the fact that nouns do not always have a (non-zero) case suffix, speakers might end up analogizing like this: “if I can change «payioang» to «ayioangpe», moving «pe» to the end of the word, then changing «payioangìl» to «ayioangìlpe» seems like the same thing, doesn’t it?”
4) Arguably, having the case suffix reduced to a single consonant at the end of the word with no possibility to change it to a “fuller” form is undesirably inflexible (at least by Na’vi standards hrh) and in real speech liable to result in a loss of information. (Then again, this would apply even more so to other nouns ending in vowels, and there’s been no indication that they might have any alternative.)
ta Zang
To me, -pe + case ending feels much more natural than the other way around (as Tekre pointed out, it’s weird not to have the case ending at the last position). However, I’d like to add that the genitive form would always be -peyä, which could easily be analysed as the possessive pronoun “peyä” in spoken language, possibly making understanding harder.
Kaltxì ma karyu Pawl!
Thank you for such a quick update – I still have a lot of words to learn from the previous one, but for now, I’d like to join in on the discussion of -PE+.
As only few people here, I too incline towards having -pe as final suffix. My reasoning is very similar to Wllìm’s, as he explained in his comment (that is, when pe+ is used as a prefix, it precedes other markers, and its relation to “sì” and other markers used as suffixes).
-PE+ has always felt like it kind of “closes” the word (from either side) and gives it a final “touch” – I’m not sure if this explanation makes a lot of sense, but to me, putting -PE+ as either starting or final affix feels much more natural than finishing it with other markers (including adpositions). Since in English, many of the question markers resebling -pe+ are one separate word that can function on its own (who, what, where, when), it feels more natural to not have -PE+ completely included in the word.
I acknowledge this can be affected by being used to talk and often think in an English way, but as someone who hasn’t used, nor been exposed to the sentences type C (and therefore doesn’t have needed knowledge to be grammatically correct), I must say my instinct would be to compose the word as is shown in sentences type B.
Thank you for the new post!
As others have stated, we already have examples and have taught this always like in examples A and C, so maybe we just feel those more natural because of usage. However, if this finally changed I’d see it pretty na’vi and somehow coherent (like Wllìm said, pemeikran, not *mefeikran). I recognize that it would be quite weird to get used to it but makes sense due to how the language is made itself and how do we always mix parts due to flexible word order.
So, I wanted to tell about my other questions about suffixes of nouns. Some years ago, I made this kinda mind map. https://i.imgur.com/loKFmXL.png
It’s outdated, since peme+/pepe+ question was closed iirc and to some of my past speculations I’d approach differently today. All that mind map is a speculation, but that red/pink-ish texts are questions/speculations of speculations. I’ll mark my speculations as (spec.N) as in further.
I won’t tell much beyond suffixes (endings) questions in this one, but I should tell of some core assumptions. There, I assumed fra-/pe+/tsa-/fì- as interchangeable and used as “only one option”. In fact, only tsa-/fì- do that, while there might exist fra- with tsa-/fì- in any order (spec.1), plus pe+ might take it with tsa-/fì- OR with fra- alone or in any spec.1 combo. (spec.2)
Next go number prefixes, which are clearly interchangeable. While pe+/tsa-/fì- and fray+ aren’t under any questions, other fra- and combined cases from spec.1 and spec.2 quite much are. (spec.3)
Assuming pe+/fra-/tsa-/fì- plus numbers go first, or rather most outward, there are also suffixes of munsna-/sna-/fne-, which, I assumed go after ones mentioned above due to “more belonging to the stem” factor. I know of sna- limited productivity, it’s here for sake of completeness, and it might be assumed that sna- and munsna- are interchangeable, but I’d probably see that *munsnasna-, “a pair of sets” or *snamunsna-, “a set of pairs” would exist, albeit in rare/unlikely case. (spec.4)
As for combining fne- with munsna-/sna-, it’s a bit complicated, but for a rule of thumb, I’d take this as a postulate: it depends on what meaning it has. It can be a pair of types (munsnafne-) or a type of pairs (fnemunsna-), the both are entirely possible imo. Same goes for a set of types (snafne-) or a type of sets (fnesna-). (spec.5) And here I assumed the most outward as the core one (spec.6), but in fact how to decipher the meaning correctly is an open question, as the most inward might be the core.
Then, finally goes the root noun. (Yay!)
And then, I assumed that a group of -o/-fkeyk/-tsyìp/-pe goes most inward, then cases/adps go, and then -sì/-to. (spec.7) That -to can be used with and after case endings is kinda not under qiestion for me, tho not sure about the other people.
I also assumed that -o/-fkeyk/-tsyìp/-pe are allowed to be used together, though it depends on the cases, and some I consider as clearly mutually exclusive. (spec.8) That is, pe+/-pe vs. -o are mutually exclusive. One makes noun into a question, the other makes noun into “an undefined noun”, “some noun”. (spec.9)
Pe+ and -pe specifically with -fkeyk and -tsyìp: if not count that there are lexicalized nouns (i.e. having separate entries in dictionaries), I’d be more prone to using -pe after -fkeyk or -tsyìp, because of seeing them as “more close” to the root, and of seeing pe+/-pe as more widely used affix. (spec.10)
-O and -fkeyk/-tsyìp: I’d see this much like as -pe and -fkeyk/-tsyìp, almost for the same reasons. (spec.11)
On other side, if to take spec.6 in consideration, in case of -o and -fkeyk specifically, it might kinda depend on the intended meaning. “An undefined state of noun?” (*-fkeyko) Or “state of an undefined noun?” (*-ofkeyk) (spec.12)
Then, -tsyìp and -fkeyk together remain. Their usage I’d consider much based on spec.6 and spec.11 respectively. I.e. concerning to what exactly by meaning the diminitive is appled, either to a noun or to its state. (spec.13)
These assumptions on “binary” interactions of these suffixes, I’d take as the base bricks for their interactions, should three of them meet ever together. Never four, I’d think, because of thinking of -pe and -o as mutually exclusive. (spec.14)
Spec.14 leads me to assuming two possibilities: -fkeyk -tsyìp -pe and -fkeyk -tsyìp -o.
In first one, there is -fkeyk and -tsyìp interact with each other based on spec.13, with -pe much likely going to the most outward, based on spec.10 (spec.15)
Second one… Since I consider both -tsyìp and -o more like “modifiers of meaning” than “bearers of meaning”, it again brings me to spec.6, namely that there would certainly be *-tsyìpo, but it would be placed around -fkeyk as the meaning would see fit. So to say, uhm… “a state of undefined diminitive noun” (*-tsyìpofkeyk) or “an undefined diminitive state of a noun” (*-fkeyktsyìpo). (spec.16) I lean to *-tsyìpo mostly due to basing on spec.11, plus not forgetting that spec.6 is still up to debate.
Now… If to assume denying spec.9 and to consider it’d be possible to ask “what some thing?”, it comes to considering how -pe and -o would interact. On one side, both -ope and -peo would feel equally possible to exist. (spec.17) On the other side, there is already kinda rule that case endings go after -o. It might be a base for assuming that it is *-peo only. (spec.18)
So, considering possibility of spec.17, it leads also to possibility of seeing all 4 together. So, with specs 6, 10-13 & 16 in mind, it comes to *-tsyìpo (first), then to *-tsyìpo and -fkeyk in order which depends on meaning (second), and then to -pe leaning the most outward of all, unless spec.18 comes into the action. “Kinda breakdown” of this would look like: [*-tsyìpo | -fkeyk] -pe. Or *-tsyìpofkeykpe (“what state of a diminitive undefined thing?”) and *-fkeyktsyìpope (what diminitive undefined state of a thing?”), under specs 6, 10-13, 16 and 17. (spec.19) If to replace spec.17 with spec.18, the latter changes to the following: *-fkeyktsyìppeo. (spec.20)
What about possibility of other unmentioned combinations of these 4? They’d require denial of -pe being the most outward with/out -o, plus the brick of *-tsyìpo would either revert to *-otsyìp or even -o would be unbound from -tsyìp. While I do feel like I exhausted my logical possibilities of combinations onto these 4, the other people may have well the other reasoning of things. I, for one, will move onto the next from here.
The next, there are case endings, ending adpositions and -sì/-to. I saw pretty much that case endings and adpositions are mutually exclusive, but I’m not sure to what refer. In light of question with -pe and case endings placement, it might begin feeling like placement of “more core” suffixes and case endings might come to redefining.
In any combo, -o seems to come as always “most outward of core 4” unless it redefines the meaning. So, in cases when -o is the most outward, it implies case endings would always come after “core 4”. In presense of -fkeyk and where -o belongs to the stem (family of *-ofkeyk), I’d still prefer to see case endings after “core 4 suffixes”, for the matter of consistency and because of complexity of the construction. (spec.21)
I’d see spec.21 also applied to -pe, in case of more complex construction with more than just -pe present. In spite I might not mind having case endings + -pe as some option in simpler combos, where only -pe and case endings are present. (spec.22)
Case endings earlier than -fkeyk or -tsyìp? I guess, it’d be “no please?” Much due to seeing both as something “closer to the stem”. (spec.23)
Now to question about ending adpositons coming before “core 4” in general and -pe in particular. I guess, it doesn’t work with -o, because of its rule of case endings? I guess, the same for -tsyìp and -fkeyk.
-Pe: well… in naive and simpler manner and matter it comes to “just changing the positions of bricks within the word”. In matter of meaning and perception, it comes to seeing it something like a mistake or something which break the habits, or “set in stone” things quite a bit, which would not come as easy. And since ending adpositions is a separate entity anyway, I guess, it doesn’t into question that they go after so-called “core 4”. (spec.24)
Kinda last one remains, -sì/-to: I’d place them AFTER everything else, even if “case endings before -pe” would be possible. Even if “case/adp endings before core 4 suffixes” would be possible. Because… probably, because some things should stay as they always are. Out of order, consistency, preferrences… You name it. (spec.25)
->——————————————————–
Ok, now I think I ran through my views on suffixes placements in the full scale I’m/I’d be ever capable of. Please, still consider all of these as the speculations and the opinion of just one person. Should someone else bring another breakdown or chime in more views or opinions, it would be much, much appreciated.
Plus, should someone more favored ponder, contemplate and inquisite the matters above, please refer to their views as more authoritative than these.
I won’t go more into the arguments/existing examples already stated here. Those are the ones I would have named as well.
After 16 years of exposure (great choice of words, ma Wllìm 😉 ) A and C feel most comfortable to me.
Another point we could take into account are adpositions as suffixes. I could find at least one example: Nga zola’u tsengpeftu? ‘Where did you come from?’
Here *tsengftupe would sound weird to me because I parse it as [place-from] < which whereas tsengpeftu = [place-which] < from is more in line with the other way around as ftu peseng/tsengpe Does this make sense?
mllte oe nìwotx!
In pure concept it could go either way, but intuitively B feels strange for me for the same reasons Zik mentions, although there are languages that do something similar. I don’t think the “pemeikran” case is the same, that can be analyzed as which-(two ikran) so it makes sense for the interrogative to be “outside” too.
Given that there are available examples of case endings after -pe and it’s what the community has been using for years, I think it should stay as A/C. However, I agree that B could be a “non-standard” form like “pukito”, or a dialectal difference like EanaUnil said.
(ouch, please, kindly feel free to disregard my previous unformatted bulk)
So, I wanted to tell about my other questions about suffixes of nouns. Some years ago, I made this kinda mind map. https://i.imgur.com/loKFmXL.png
It’s outdated, since peme+/pepe+ question was closed iirc and to some of my past speculations I’d approach differently today. All that mind map is a speculation, but that red/pink-ish texts are questions/speculations of speculations. I’ll mark my speculations as (spec.N) as in further.
I won’t tell much beyond suffixes (endings) questions in this one, but I should tell of some core assumptions. There, I assumed fra-/pe+/tsa-/fì- as interchangeable and used as “only one option”. In fact, only tsa-/fì- do that, while there might exist fra- with tsa-/fì- in any order (spec.1), plus pe+ might take it with tsa-/fì- OR with fra- alone or in any spec.1 combo. (spec.2)
Next go number prefixes, which are clearly interchangeable. While pe+/tsa-/fì- and fray+ aren’t under any questions, other fra- and combined cases from spec.1 and spec.2 quite much are. (spec.3)
Assuming pe+/fra-/tsa-/fì- plus numbers go first, or rather most outward, there are also suffixes of munsna-/sna-/fne-, which, I assumed go after ones mentioned above due to “more belonging to the stem” factor. I know of sna- limited productivity, it’s here for sake of completeness, and it might be assumed that sna- and munsna- are interchangeable, but I’d probably see that *munsnasna-, “a pair of sets” or *snamunsna-, “a set of pairs” would exist, albeit in rare/unlikely case. (spec.4)
As for combining fne- with munsna-/sna-, it’s a bit complicated, but for a rule of thumb, I’d take this as a postulate: it depends on what meaning it has. It can be a pair of types (munsnafne-) or a type of pairs (fnemunsna-), the both are entirely possible imo. Same goes for a set of types (snafne-) or a type of sets (fnesna-). (spec.5) And here I assumed the most outward as the core one (spec.6), but in fact how to decipher the meaning correctly is an open question, as the most inward might be the core.
Then, finally goes the root noun. (Yay!)
And then, I assumed that a group of -o/-fkeyk/-tsyìp/-pe goes most inward, then cases/adps go, and then -sì/-to. (spec.7) That -to can be used with and after case endings is kinda not under qiestion for me, tho not sure about the other people.
I also assumed that -o/-fkeyk/-tsyìp/-pe are allowed to be used together, though it depends on the cases, and some I consider as clearly mutually exclusive. (spec.8) That is, pe+/-pe vs. -o are mutually exclusive. One makes noun into a question, the other makes noun into “an undefined noun”, “some noun”. (spec.9)
Pe+ and -pe specifically with -fkeyk and -tsyìp: if not count that there are lexicalized nouns (i.e. having separate entries in dictionaries), I’d be more prone to using -pe after -fkeyk or -tsyìp, because of seeing them as “more close” to the root, and of seeing pe+/-pe as more widely used affix. (spec.10)
-O and -fkeyk/-tsyìp: I’d see this much like as -pe and -fkeyk/-tsyìp, almost for the same reasons. (spec.11)
On other side, if to take spec.6 in consideration, in case of -o and -fkeyk specifically, it might kinda depend on the intended meaning. “An undefined state of noun?” (*-fkeyko) Or “state of an undefined noun?” (*-ofkeyk) (spec.12)
Then, -tsyìp and -fkeyk together remain. Their usage I’d consider much based on spec.6 and spec.11 respectively. I.e. concerning to what exactly by meaning the diminitive is appled, either to a noun or to its state. (spec.13)
These assumptions on “binary” interactions of these suffixes, I’d take as the base bricks for their interactions, should three of them meet ever together. Never four, I’d think, because of thinking of -pe and -o as mutually exclusive. (spec.14)
Spec.14 leads me to assuming two possibilities: -fkeyk -tsyìp -pe and -fkeyk -tsyìp -o.
In first one, there is -fkeyk and -tsyìp interact with each other based on spec.13, with -pe much likely going to the most outward, based on spec.10 (spec.15)
Second one… Since I consider both -tsyìp and -o more like “modifiers of meaning” than “bearers of meaning”, it again brings me to spec.6, namely that there would certainly be *-tsyìpo, but it would be placed around -fkeyk as the meaning would see fit. So to say, uhm… “a state of undefined diminitive noun” (*-tsyìpofkeyk) or “an undefined diminitive state of a noun” (*-fkeyktsyìpo). (spec.16) I lean to *-tsyìpo mostly due to basing on spec.11, plus not forgetting that spec.6 is still up to debate.
Now… If to assume denying spec.9 and to consider it’d be possible to ask “what some thing?”, it comes to considering how -pe and -o would interact. On one side, both -ope and -peo would feel equally possible to exist. (spec.17) On the other side, there is already kinda rule that case endings go after -o. It might be a base for assuming that it is *-peo only. (spec.18)
So, considering possibility of spec.17, it leads also to possibility of seeing all 4 together. So, with specs 6, 10-13 & 16 in mind, it comes to *-tsyìpo (first), then to *-tsyìpo and -fkeyk in order which depends on meaning (second), and then to -pe leaning the most outward of all, unless spec.18 comes into the action. “Kinda breakdown” of this would look like: [*-tsyìpo | -fkeyk] -pe. Or *-tsyìpofkeykpe (“what state of a diminitive undefined thing?”) and *-fkeyktsyìpope (what diminitive undefined state of a thing?”), under specs 6, 10-13, 16 and 17. (spec.19) If to replace spec.17 with spec.18, the latter changes to the following: *-fkeyktsyìppeo. (spec.20)
What about possibility of other unmentioned combinations of these 4? They’d require denial of -pe being the most outward with/out -o, plus the brick of *-tsyìpo would either revert to *-otsyìp or even -o would be unbound from -tsyìp. While I do feel like I exhausted my logical possibilities of combinations onto these 4, the other people may have well the other reasoning of things. I, for one, will move onto the next from here.
The next, there are case endings, ending adpositions and -sì/-to. I saw pretty much that case endings and adpositions are mutually exclusive, but I’m not sure to what refer. In light of question with -pe and case endings placement, it might begin feeling like placement of “more core” suffixes and case endings might come to redefining.
In any combo, -o seems to come as always “most outward of core 4” unless it redefines the meaning. So, in cases when -o is the most outward, it implies case endings would always come after “core 4”. In presense of -fkeyk and where -o belongs to the stem (family of *-ofkeyk), I’d still prefer to see case endings after “core 4 suffixes”, for the matter of consistency and because of complexity of the construction. (spec.21)
I’d see spec.21 also applied to -pe, in case of more complex construction with more than just -pe present. In spite I might not mind having case endings + -pe as some option in simpler combos, where only -pe and case endings are present. (spec.22)
Case endings earlier than -fkeyk or -tsyìp? I guess, it’d be “no please?” Much due to seeing both as something “closer to the stem”. (spec.23)
Now to question about ending adpositons coming before “core 4” in general and -pe in particular. I guess, it doesn’t work with -o, because of its rule of case endings? I guess, the same for -tsyìp and -fkeyk.
-Pe: well… in naive and simpler manner and matter it comes to “just changing the positions of bricks within the word”. In matter of meaning and perception, it comes to seeing it something like a mistake or something which break the habits, or “set in stone” things quite a bit, which would not come as easy. And since ending adpositions is a separate entity anyway, I guess, it doesn’t into question that they go after so-called “core 4”. (spec.24)
Kinda last one remains, -sì/-to: I’d place them AFTER everything else, even if “case endings before -pe” would be possible. Even if “case/adp endings before core 4 suffixes” would be possible. Because… probably, because some things should stay as they always are. Out of order, consistency, preferrences… You name it. (spec.25)
————————————————————-
Ok, now I think I ran through my views on suffixes placements in the full scale I’m/I’d be ever capable of. Please, still consider all of these as the speculations and the opinion of just one person. Should someone else bring another breakdown or chime in more views or opinions, it would be much, much appreciated.
Plus, should someone more favored ponder, contemplate and inquisite the matters above, please refer to their views as more authoritative than these.
Another argument occurs to me that I think is actually a fairly strong reason to forbid forms like «ayioangìlpe»: (unlike «-o») if «-pe» is allowed after case suffixes, then this will greatly increase the potential for ambiguous sentences where it’s not obvious whether the «pe» belongs to the preceding noun or the following noun, such as «Ayioangìl( )pe( )sponoti parwùn?» “Which animals inhabit the island?” or “Which island do the animals inhabit?”. If we allow it after enclitic adpositions then it gets even worse as for example «Tùlkun( )äo( )pe( )ilu slele?» could mean not only “Which ilu swims under the tùlkun?” and “Under which ilu does the tùlkun swim?” but also “Under which tùlkun does the ilu swim?”. (And if I had more time now I would consider what monstrosities could result if we additionally combined these with «to» and «sì»!) I know prosody would help at least with the clitics if not for «pe» itself, but it does seem like something that could get confusing fast even without going out of one’s way to set up such ambiguity.
I’ve also posted a comment two hours before the long one. Now it looks it was lost. Sorry in case it is too much, but I hope I’m allowed to restore it from backup, or longer one makes less sense:
-pe to me feels a bit more belonging to the stem/root of word than being a situational modifier as case endings or ending adpositions. And I am kinda more used to -pe then case/adp than to case/adp then -pe. Still, it would be quite an interesting gimmick, if case ending then -pe would be allowed. I would not exclude a possibility of seeing this more often, should people prefer this over -pe then case ending.
Well, I see that now the question regarding ordering -pe and ending adp is a separate one, which would be interesting to hear about too.
There are also such fancy thingies as -sì and -to, which I’d still prefer to see past everything else, as it implies some consistency for me. In case -pe stays after case ending, then how would -sì/to behave then?
In any way, if there would be several of such words with -pe&co, I think it’s rule that every and each one of them would follow the one pattern within a given sentence/clause.
I had more of questions in regards of ordering all the noun suffixes. Should I find my backlogs or texts (some I lost as time went), I will let know my thoughts about here.
Oeyä säplltxevi a ro fostì aham fìtsengmì matsan nì’ul.
In to second Plumps when he said
> After 16 years of exposure (great choice of words, ma Wllìm 😉 ) A and C feel most comfortable to me.
Regarding the “the case ending should always be the very last thing, it assigns a role to the whole noun built up to that point” argument, I agree with that, that’s how I’ve seen it too, I suppose. That, and just everyone teaching and using A and C style exclusively for nearly 20 years, sometimes even saying B is incorrect and should be A or C.
In also to agree with the sentiment of B making sense as well. It wouldn’t bother me to have yet more valid forms of words by affix permutation like this. If all 3 are valid, I just would probably be stuck in my A/C ways, hardly saying B but here and there. Some old habits die hard HRH
As for the “no prefix comes before pe+ prefix, so no suffix should come after -pe suffix, like a nice neat sandwich or bookends” argument, I can see that argument too, along with the others mentioned in favor of B. I’ll be using this mindset to settle in, if B does get ruled correct.
Basically, A and C are etched for me at this point. If it comes to pass after today that “A or B, but NOT C” instead of “A or C, but NOT B”, then it would be like “Nume is NOT Transitive 2.0” for me ;D Minor nuisance to retrain, ultimately fairly surmountable, good fun history lore for later.
Interesting discussion! Both viewpoints make strong cases, and there are existing examples that validate either usage. However, allowing the interogative marker to ‘float’ between B and C would not help with the practical useability of the language. Restricting the use of the interogative to case A on nouns would probably not work well at this late point in time, either.
When you apply an interogative to a noun with a case role, is is actually functioning entirely in that case role, or does it become something different? I seem to think it becomes something different, although it certainly may also still remain an agent or a patient. Because of its role-altering nature, it makes more sense, as Tirea Aena points out for -pe to be the final element on a word, just as it is the initial element when used as pe+
However, although only a few, we do have an established pattern of the case endings being the final element of a word. One one hand this establishes -pe as becoming before a case ending, but then this might become one of those ‘special rules’ or ‘exceptions to the norm’ that don’t make logical sense when we see pe+ preceeding all other prefixes at the beginning of a word.
Although this can be OK either way, I am slightly inclined to accept use case B as being the more correct one. We then need to deal with the few examples that are out there where -pe comes before a case ending. It might be easier to go back and alter these examples than create a special exception. Although Na’vi does have a few special exceptions, one of the beauties of the language is consistent use of grammar.
Tìeyngìri sì aylìuri atìtstunwinga’ irayo! 🙂
Now I’m wondering about the form «lì’fyaolo’ri» hrh: from our past knowledge of lì’fya na’rìnge I would have expected «lì’fyaolo’ìri», but we know that «olo’ru» is a valid alternative to «olo’ur» (and perhaps the only option in lì’fya wione because of the awkwardness of preserving the ending that is allowed only after consonants when the consonant in question is deleted?), so is this another variation of the same sort (and likewise the solution to *«oloìri» in lì’fya wione)?
Sätseri asìltsan.
The standard dative and topical case endings for bases with final consonants are ur and ìri respectively. Since the tìftang is a consonant, forms like olo’ur and olo’ìri are expected and correct. But as you’ve noted, there’s an exception for the dative following a tìftang: the ru variant is also permitted. So alongside olo’ur we also have olo’ru. (This goes back a long way. In A1, Tsu’tey, in translating Jake’s summoning of the clans, says, “Ayolo’ru alahe peng ziva’u.”)
Given forms like olo’ru, it’s virtually inevitable that olo’ri for the topical would have developed by analogy. So yes, for bases ending in a tìftang, the topical case can be either ìri or just ri. I don’t think I ever stated that explicitly, so thanks for pointing it out.
I agree that in Lì’fya Wione, with the intervocalic dropping of the tìftang, oloru and olori are the forms you’d be most likely hear—but not exclusively, since olour and oloìri are quite pronounceable!
Tsayuti set shawm a fìu dantsan! Only, I’m a little bit confused by «oloru» and «olori»: as you say, in Lì’fya Wione we have intervocalic dropping of the tìftang, but in «olo’ru» and «olo’ri» it’s not intervocalic, is it? Or are you saying that the tìftang disappears first and then the vowel: olo’ìri > oloìri > olori, *olo’uru > *olouru > oloru ?
Unrelated to the ongoing discussion, but should the title be “Hì’ia tìsung postä aham”? (instead of “postìyä”)
Given we learned that the added -ì disappears if through case endings the loan becomes pronouncable (kelnì -> kelnä). Or am I missing something here?
No, you’re not missing anything. 🙂 Ngaru tìyawr! It should have been postä. Thanks for noticing that. Tìkxey zoslolu.
This raises some interesting questions, though. We don’t have a lot of common nouns (i.e., things, not names of people, places, or organizations which in English would be capitalized) that end in unstressed ì. A little quick research yielded these root nouns: fwäkì ‘mantis,’ herwì ‘snow,’ lìlì ‘breast,’ pätsì ‘badge,’ postì ‘post,’ and tsopì ‘lung.’ (We also have the compounds with yì: fyanyì, heynyì, kxamyì, kxaylyì, tìmyì.) Of the root nouns, the borrowed terms are pätsì and postì; the others are native roots.
Comparing postì and tsopì is particularly interesting. Ironically, they’re anagrams of each other! However, they don’t behave the same. The genitive of tsopì, as a native root, is tsopìyä, while the genitive of postì, as a borrowed term with ì added at the end to make it conform to the rules of Na’vi, is postä.
How would native speakers of Na’vi on Pandora come up with the right forms for these words? It seems to me they’d have to be aware of the etymologies! That is, they’d have to realize that postì was borrowed while tsopì is native. I wonder if at some point the information that postì was borrowed from the Sky People’s language would be lost, and it would simply be thought of as a native word, in which case the genitive would indeed become postìyä. Eltur tìtxen si, kefyak?
Eltur tìtxen si!. This reminds me about the infamous situation of the plural for octopus. Octopuses just applies the more common English rules, octopi falsely assumes it’s a Latin word (it’s actually Greek) while octopodes is the rarest of the three English.
The other non-loaned nouns that end in unstressed Ì that I found are sätsìsyì (whisper), swayì (generation in a family), tomperwì (sleet), txanfwerì (blizzard) and Txewì (name of a male Na’vi).
By the way, postì (post) and tsopì (lung) are not only anagrams of each other, but also tìsop (journey). Out of all root words and non-productive compounds, the only other triple anagram is tawsìp (skyship), tìspaw (belief (not countable)) and tìpsaw (clumsiness). Of course I had a computer check this. 😁
In German language, we often respect the etymology of a word, e.g. when the plural of a borrowed word is used:
Atlas – Atlanten or Atlasse
Matrix – Matrizes, Matrices or Matrizen
Ananas – Ananas or Ananasse
Lexikon – Lexiken or Lexika
I’m sure Na’vi would be aware of the etymology of a word.