Mónate paináute he Pai le Koa 24si

Héite lami lakaméhepo nuisai le Koa hetia, kapaináute lúaku-néisi kacimi!

Hekei kavoa kesi ticímiki kémeni i palili oasi pameti mono lacimi ela eso. Tamieko kasavu oma, kapatune kaímita oma. Hetikehe kavoa vime ninasicutekuvi kepi kacimi kocusunu, kocupamuta... konicutehienu ve le púhuma eso i hala aaika iuluhu.

Paináute iolo, nii, Koa ni paloha. Kavóase sii i vimuniko taa i vela, e ninateota komiilo kaseculilo.

-Iúliki, Váhumaa, 2023-09-13


Lílite la le Níkili / English Translation:

Warm greetings to the worldwide Koa community on this, the 24th birthday of the language!

During the last year this little language of mine has transformed from a mere considered idea into a real living language. It took on its own style, its own sense of self. At this time last year I couldn't have imagined how much the language would grow and change... that I would be able to call myself a real speaker after such a short time.

Happy birthday, then, my beloved Koa. May your next year be even more astonishing, and I can't wait to discover what you will become.

-Julie, Portland, 9/13/23

Embedded questions

This is a much overdue followup to the seminal piece on embedded questions and theta clauses from a couple years ago. These expansions and adjustments quickly developed in the weeks after.

As a reminder, the standard way to say things like "I don't know what you want" or "I wonder where my clock came from" is apparently via nominalization of the sub-clause, thus

ni-na-ilo ka-se-halu
1SG-NEG-know DEF-2SG-want
"I don't know what you want," lit. "I don't know the thing you want"

toko keo ka-sáti-ni (ko-tule)
wonder origin DEF-clock-1SG (NOM-come)
"I wonder where my clock came from," "I wonder about the origin of my clock('s coming)"

Figuring out these structures was a huge relief from my previous anxiety that I might be thoughtlessly calquing IE strategies, but once I had them in hand I realized right away that that fear wasn't really cross-linguistically motivated. A more familiar variety of finite embedded clauses is well represented in languages from many other families, and as such I really don't see a reason to ban them from Koa:

ni-na-ilo [ kea sa se-halu ]
1SG-NEG-know what FOC 2SG-want
"I don't know what you want," "I don't know what it is you want"

toko [ o-kea sa ka-sáti-ni i tule ]
wonder ABL-what FOC DEF-clock-1SG FIN come
"I wonder where my clock comes from," "I wonder where it is that my clock comes from"

As glossed above, these kinds of forms might see slightly different usage pragmatically, feeling a little wordier, possibly softer, possibly less formal.

Entirely new to this discussion are embedded yes/no questions. Here too a finite structure is possible, using ai "whether, or":

ni-na-ilo ai ta-cu-tasi
1SG-NEG-know whether 3SG-IRR-repeat
"I don't know whether it will happen again"

...and again this option exists against another possibility with a nominalized clause headed by one of the infamous ke-compounds: kema "current, ongoing," kesi "past," kecu "future, what would/will be," kete "possibility," kelu "desire," keki "necessity." Thus

ni-na-ilo kecu ko-ta-tasi
1SG-NEG-know future NOM-3SG-repeat
"I don't know whether it will happen again," lit. "I don't know the future of its happening again"

Similarly, with the other types:

kema kotatasi "whether it is/was happening again"
kesi kotatasi "whether it had happened again"
kete kotatasi "whether it could happen again"
kelu kotatasi "whether there was a desire for it to happen again"
keki kotatasi "whether it has to happen again

Any of these could be rephrased with ai and the particle in question within the clause rather than outside it: kete kotatasi = ai tatetasi "whether it could happen again, whether there's a possibility of it happening again." Here I have the sense -- which as usual will have to be confirmed in time by usage -- that the ai-type clauses may be more comfortable and vernacular, the ke-clauses more literary.

Obiter: in the original theta clause post, I was omitting the nominalizer ko in the embedded clauses with ke-compounds (i.e. just kete tatasi rather than kete kotatasi). I'm unclear on why I did this, and at this point I think it was either just a mistake...or I was getting a little ahead of myself with an experimental idea that ko might be optional in clauses lacking a nominal subject. In this event, we could also have e.g. niilo selóhani "I know you love me" alongside niilo koselóhani "I know that you love me." I'm not sure yet what the implications of this would be for parseability, so an official verdict is pending more rigorous exploration.

Bride of the nominalized clause

What should happen to a Koa clause with a nominal subject when the clause is nominalized? This has become a bit of an albatross. The last eleven years have seen numerous well-thought-out, well-intentioned, sincere attempts to answer this question; moments of satisfied relief have followed, but a seed of anxiety has always remained that we haven't got it quite right. As, apparently, now.

My last bid for a final answer, in which the nominalizer ko actually replaces the finiteness marker i, is odd to my Indo-European sensibilities but quite elegant. The problem -- if indeed it is a problem -- is that the literal translation of such clauses according to the usual Koa syntactic rules ends up looking anomalous at best and like word salad at worst. For example, a phrase from the previous post:

na-vi-ana po-mítitinu ko-cita
NEG-IMP-give GEN-bedbug NOM-bite
"don't let the bedbugs bite"

Here the nominalized clause is pomítitinu kocita "for bedbugs to bite," "bedbugs' biting." If I parse that phrase according to ordinary principles, though, it seems to mean something like "bedbugs of biting," parallel to e.g. poétete kocini "acts of kindness." So..."don't allow bedbugs of biting!"

Maybe that's fine, because outside of Loglan and friends the mechanism of language itself was never "supposed" to be logical: merely parseable. Which I think this entirely is, and as I mentioned in the post in which I introduced this, it's not so dissimilar from what Latin does.

I guess the point where I started to fret again was when I realized that another structure -- which I have also proposed for this purpose in the past, though perhaps for not entirely the right reasons -- might address this in an even more parseable form. The core concept here is that any verb phrase can follow a head noun as a modifier, just like any other adjective: thus

mítitinu veta
bedbug giant
"giant bedbugs"

mítitinu ma-cita
bedbug CONT-bite
"biting bedbugs, bedbugs that are biting"

This being so, it would appear that we could create a nominalized clause simply by -- on the surface -- deleteing the finite i entirely, and marking the whole thing with ko at the beginning:

po-mítitinu i cita
GEN-bedbugs FIN bite
"bedbugs bite"

po-mítitinu cita
GEN-bedbugs bite
"biting bedbugs"

ko po-mítitinu cita
NOM GEN-bedbug bite
"bedbugs' biting," "that bedbugs bite," "for bedbugs to bite"

This would give us an alternate, less poetic, more easily parseable injunction,

na-vi-ana ko po-mítitinu cita
NEG-IMP-give NOM GEN-bedbug bite
"don't let the bedbugs bite"

...which happens to be identical to the most recent proposed syntax other than the position of ko! Thus again

na-vi-ana po-mítitinu ko-cita
NEG-IMP-give GEN-bedbug NOM-bite
"don't let the bedbugs bite," "don't allow the bedbugs biting"

Maybe there's no decision to be made and the two structures could coexist, just like in ditransitive VP's like

ana ka-nosu pe-vii
give DEF-elephant OBL-mango
"give the elephant a mango"

ana po-vii la-ka-nosu
give GEN-mango DAT-DEF-elephant
"give a mango to the elephant"

I'm honestly not sure. I don't see a strong reason to disallow either structure, but the muse is also not sending me a resounding chime of rightness in either case. One thing is certain, which is that I am very, very tired of going around in circles about this year after year, and on that basis I'm inclined to leave both options in circulation and allow usage rather than theory to cast light on the question. I would be relieved for this whole area to stop feeling like a crisis after 11 years of ceaseless gear-grinding.

Of course, this discussion has completely omitted to mention the fact that another structure also exists for this same meaning, a finite one with ve:

na-vi-ana ve po-mítitinu i cita
NEG-IMP-give as/like GEN-bedbug FIN bite"
"don't let the bedbugs bite"

I'd previously confidently affirmed that these ve-structures are identical in meaning to the the nominalized ones with ko, but Nahuatl has been making some gentle suggestions recently that this might not quite be so. That's potentially a really big revelation that's still taking shape; hopefully more on this soon, once I'm confident I understand it myself.

Worthy of brief mention, summer 2023 edition

In the spirit of a belated spring cleaning, I hereby present the following list of topics with which my Koa process document is beginning slightly to overflow. These seem like they probably ought to be mentioned, but have been patiently waiting because they may not merit an entire post all to themselves.

1) Words for "something" and "nothing" have been all over the place through the years. Most recently we were fretting about aha "something in particular" vs apparently hua "something in general, something unknown," but at some point late last year I realized that (A) this is needlessly complicated, and (B) aha/aka for "something/someone" didn't really make any sense because a- is not a formative anywhere else in the language. This leaves us with a much simpler hua/huka for "something/someone."

"Nothing" and "no one" have traditionally been nahua/nahuka, literally "not anything/anyone," and those forms still exist as options. Completely unique within the grammar of Koa, though, there is now another standard allomorph for "no one," naka, and two others for "nothing:" naha and naa. Naha is the preferred form where naa might create ambiguity with na "not," in practice essentially whenever it constitutes the whole NP by itself: naha i (na)cuapu "nothing will help." Naa shows up elsewhere, in particular when preceded by adjunct particles: nenaa "nowhere," henaa "never."

2) As the careful reader may have noticed from the foregoing exposition, we're still nowhere near a decision about how to construct clauses with negative pronouns. Should "I don't want anything" be nihalu naha "I want nothing," ninahalu hua "I don't want anything," or ninahalu naha "I don't want nothing?" Or are these all acceptable? I suppose I'm saying this out loud as incentive to finally get my head on straight about it sometime this year.

3) Subject pronouns are only required in complement clauses when they differ from that of the matrix clause. Thus nihalu komene "I want to go," not *nihalu konimene -- cf. nihalu kotamene "I want her to go." Similarly,

ni-si-lule ko-io-na-cu-náe-se
1SG-ANT-think NOM-STAT-NEG-IRR-see-2SG
"I thought I wasn't going to see you anymore"

...compared to

ni-si-lule ko-ta-io-na-cu-náe-se
1SG-ANT-think NOM-3SG-STAT-NEG-IRR-see-2SG
"I thought she wasn't going to see you anymore"

4) I've alluded to the fact but never said outright that "before" and "after" are verbs, not prepositions: koe "precede in time" and hala "follow in time." Thus

ka-élate ni-nánaka i koe ka-kémeni
DEF-life 1SG-grandfather FIN precede DEF-mine
"my grandfather's life preceded mine" or "my grandfather's life was before mine."

5) It appears that Koa may have an emergent future/nonfuture distinction, in contrast to the more familiar past/nonpast orientation of e.g. Old English or Finnish: bare verbs can be interpreted to have either present or past meaning, but the future -- I think -- must be marked with the irrealis cu. Thus nináese "I see you" (usually) or "I saw you," but nicunáese "I will see you"; nisúsota "I kissed her" (usually) or "I kiss her," but nicusúsota "I will kiss her." I think this is pretty cool! Now that I think about it this is also what accidentally happened in Seadi, which raises interesting questions about the structure of my brain...

6) I've struggled with presentative structures in Koa. For a very long time I used the focalizer sa to try to do this work, before finally realizing a few years ago that I was confusing different kinds of fronting. I really wanted to be able to achieve something like the Polish szła dzieweczka do laseczka, lit. "went a girl into the forest," when the whole scene is introduced and no topic has been defined yet. It was such a problem that I couldn't even translate the first sentence of Are You My Mother?, "A mother bird sat on her egg."

Boarding the plane home from Dallas once, though, I suddenly thought of a very idiomatic Welsh structure for this kind of situation -- dyma fi'n sefyll ar yr awyren, lit. "here I am sitting on the plane" -- and it occurred to me that Koa could do something parallel. Rather than just

ni-eki ne-léhukone ne-tie lai o le Tálasi
1SG-sit LOC-airplane LOC-way return ABL NAME Dallas
"I'm sitting on the plane on the way back from Dallas"

...which is semantically correct but pragmatically totally neutral, I can use the presentative particle vo "ecco, voilà, вот, jen" to do this:

vo ni-eki ne-léhukone...
PRES 1SG-sit LOC-airplane
"here I am sitting on the plane..." or "so I'm sitting here on the plane..."

I'll still need to work out the specifics of usage, but I think this might become an important pragmatic device for Koa style.

7) I've been reconsidering "7." For more than 15 years it's been sapi from Basque zazpi, but with the quantity particle pi we frequently end up with the unhappy sequence sapi pi... "7 of..." This is especially grating given my particular fondness for the number 7. Hitu from Samoan has long been an understudy, and for the moment I'm trying it out to see if it'll stick. I'm not sure I can bear reassigning sapi, though.

8) The particle io "change of state, already" can follow a predicate as an intensifier: tule io! "come ON!", ika io "all right already!" This is a little reminiscent of the Japanese particle yo, which is some sort of whimsical good luck: 美味しいよ oishii yo "this is delicious!" (The Koa particle ho is actually often closer in meaning to Japanese yo, but I'll take my wins where they come.) Similarly, ca "steady state, still" can have a kind of conciliatory force: ika ca "there there, it's okay," tule ca "come along then."

9) In the original version of Aika Konuku I had translated "so tired" as toa pi kiuni. By the time I was writing about it on the blog I'd realized this was wrong, but for the wrong reason! I thought it meant "that specific referential amount of tired," but pi governs a nominal...which means that this phrase would actually translate as "so many tired people." Oops. This is a cautionary tale to remind us that qualifiers like vaha "slightly, a little" ano "sort of," nai "somewhat," aiva "quite" and poli "very" must always follow their predicates, never precede with pi as I'd previously thought possible: kuma vaha "a little hot," not vaha pikuma "few hot people!"

10) I've translated "at least" as mocekie, literally "as the merest thing," and "et cetera" as e tei motoa, etm. for short, lit. "and onward like that." I think I'm pretty happy with "etc.," but for "at least" I've gone for a calque and fear that I haven't really taken the time to understand the semantics like I would wish. I have an intuition that it should maybe have its own morpheme, because there's this whole concessive thing going on that's way beyond the literal meaning of the underlying words in any IE language. The structure would be something like

X ko-ta-ia-holo ko-tule
X NOM-3SG-AFF-decide NOM-come
"at least he did decide to come"

or

ta-ia-holo ko-tule i X
3SG-AFF-decide NOM-come FIN X
"he did decide to come at least"

Actually I quite like that. Okay, an independent morpheme it is.

11) À propos de bottes, this sentence, clearly a critical necessity for fluency in any language:

na-vi-ana po-míti-tinu ko-cita
NEG-IMP-give GEN-bed-parasite NOM-bite
"don't let the bedbugs bite"

12) I just wanted to say out loud that this post brings me level with 2010 for the highest number of posts in a calendar year...and it's only June. I am so happy and excited that Koa has found its way to such an explosion of growth after so many years of slow simmering; as a life's work, I feel pretty proud of it.

Purpose clauses à la lettone

At some point in the past several years the allative la plus nominalization sort of unexaminedly became the standard way to form purpose clauses:

ni-lai lakoto la-ko-núkuki
1SG-return home ALL-NOM-nap
"I went back home to take a nap"

...or with a different subject in the subordinate clause,

ni-lai lakoto la-ko-ta-núkuki
1SG-return home ALL-NOM-3SG-nap
"I went back home for him to take a nap"

ni-lai lakoto la le Keoni ko-núkuki
1SG-return home ALL NAME John NOM-nap
"I went back home for John to take a nap"

I suppose this is plausible as an intuitive strategy, but I'm always suspicious of unexamined IE calques. This has been on my list to research properly for a long time. I still need to, but in the mean time I was recently reading a Latvian grammar and happened across a different structure that I thought was pretty cool. It uses essentially an imperative formation (Esperanto might call it volitive), and the purpose clause remains finite:

lai drusku atpūs-tos, mēs aizbrauk-s-im uz Jūrmalu
IMP a.little rest-SUB.REFL, we travel-FUT-1PL to Jurmala
"in order to rest a little, we shall go to Jurmala"

I wonder if the parallel Koa structure might work too, as an alternative to nominalization. Something like

ni-lai lakoto i vi-núkuki
1SG-return home FIN IMP-nap

Maybe? It's certainly nice to have options, and this new way feels a little more fluid and poetic somehow. One thing about this, though, is that it looks like a serial verb, but then the second verb seems to have different TAM marking than the first...which I think isn't typical for things called serial verbs? I'm also not sure whether we could have a different subject in the purpose clause:

?ni-lai lakoto i ta-vi-núkuki
1SG-return home FIN 3SG-IMP-nap
"I went back home for him to take a nap"

??ni-lai lakoto le Néliki i vi-núkuki
1SG-return home NAME Nellie FIN IMP-nap
"I went back home for Nellie to take a nap"

Actually I do seem to remember serial verbs sometimes having different argument structures from one verb to another, but TAM? It doesn't help a ton that my examples tend to come from pretty isolating languages without much tense marking. I need to review Describing Morphosyntax and probably Finnish, Turkish and Nahuatl while I'm at it...doesn't Finnish use some kind of translative suffix? Anyway, I'm clearly not prepared yet to offer an intelligent assessment here, but it's certainly interesting.

More about agglutination

After a lot of excited scuffling earlier this year, I confess I'm still no closer to any kind of formal typological assessment of Koa. At base, this comes down to the fact that, despite the fact that I can speak this language with ever-increasing fluency and it feels completely intuitive, I still can't articulate what the particles actually are: that is, how to classify them in any standard system of linguistic description.

If Koa is agglutinative and all these particles are best described as prefixes and suffixes, I thought, there ought to be some kind of template dictating how they combine. I spent a good amount of time diagramming all this out, for verbal forms, at least; it turns out that there is in fact a pretty firm prescriptive ordering of particles/affixes around their predicate. I made charts like this one, showing the 15 slots that verbs apparently admit:

The most complex "word" I've been able to come up with so far is kotapunacusivatesutásiluketu, "the fact that he apparently won't regularly have been able to finish rereading them," containing 12 morphemes (2 predicates and 10 particles):

ko-ta-pu-na-cu-si-va-te-su-tási-luke-tu
NOM-3SG-HEAR-NEG-IRR-ANT-HAB-ABIL-CES-repeat-read-3PL
"the fact that he apparently won't regularly have been able to finish rereading them"

Quite a bit of recursion is possible as well, which means that there's no theoretical limit on the length of a "word": for example kotaiasivamimuminúkutu "the fact that he really used to start putting them to sleep" containing two instances of the inchoative -mi- in different places.

Of course, the fact that there are ordering rules doesn't say anything about typological classification. Catalan and Bulgarian famously have very complex obligatory ordering rules for pronominal clitics, but I've never seen anyone propose that these must therefore combine with their verb into a single "word."

Perhaps these particles could be described as clitics rather than affixes, then? Unfortunately I have too little experience with complex clitic systems to have any intuitive sense, but I do notice that Koa permits pauses for thought between particles in a way that I could produce between verbs and clitic pronouns in e.g. Polish (ja się...staram "I'm...trying"), but absolutely never between verbs and affixes (*co chce-...-my? "what do...we...want?").

Can we just call them particles and be done with it? I have the frustrating sense that that word may be akin to "emphasis" in being a placeholder for actual understanding, and I'm not sure why this is turning out to be so hard for me to think through. As a person who loves to put everything in neat little labeled boxes, it drives me a little crazy...especially since the answer might affect my judgment on how Koa should ideally be written. I'll just have to soldier on as I have been for the moment.

One note: just because absurdly long strings of particles can be well-formed doesn't mean that they're necessarily desirable, clear, or constitute good style. That 12-morpheme word up there might be better expressed in more manageable chunks, something like

koputai ve nacusivatai ve tatesulúketu i tasi
NOM-HEAR-be CP NEG-IRR-ANT-HAB-be CP 3SG-ABIL-CES-read-3PL FIN repeat
"the fact that it's apparently the case that it won't regularly have been the case for him to be able to finish reading them again"

Even that's pretty ungainly. The world of Koa style has clearly only just begun to burst into bloom.

Stress-determined minimal pairs

It's been clear from the beginning that stress was going to be contrastive in Koa -- indeed, it is in fact the feature that allows predicates to be parsed reliably in a phonological, morphological and syntactic system like this -- but until the language was significantly more developed and I began to use it more regularly it hadn't quite hit home just how necessarily ubiquitous the minimal pairs would be.

I've studied many languages in which stress is phonemic, but never one in which many or most stress shifts in fact result in a well-formed word with a new, often totally unrelated meaning. I'll just dance right past the question of whether this is quite proper in an IAL -- shhhhhh -- and accept it as an interesting poetic boon. What puns and other delightful turns of phrase might be created by a skillful Koa speaker to play off of this characteristic, or what conventions might arise to minimize its effect? Here are just a few of a million possible examples of the phenomenon, marking all stressed syllables and separating morphemes for clarity.

First, stress distinguishes predicates from clusters of particles:

hána "work"
ha-ná "unless"

áina "land, earth"
ai-ná "isn't it?"

láni "sky"
la-ní "to me"

ia-pu-té "they do certainly say that that may be so!"
iápu-te "an expectoration"

Stress often serves to differentiate particle-predicate sequences from predicate-suffix structures:

nána-e "grandmother"
na-náe "doesn't see"

úso-e "sister"
u-sóe "the rivers"

líe-si "purplish"
li-ési "probably the moon"

káka-va "move one's bowels"
ka-káva "the coffee"

Lastly, stress disambiguates predicate-predicate compounds, verbs with object incorporation, and predicates with both an initial particle and a suffix:

suo-kítu "eat liver"
su-óki-tu "finish pouring them"

mai-túni = "feel calm"
ma-ítu-ni = "using me"

kuce-túpo "ride a horse"
kúce-tupo "a riding horse, a horse for riding"
ku-cétu-po "obviously a bunch of deplorables"

No doubt context would also be instrumental in parsing streams of syllables, or resolving ambiguity where stress is misplaced or insufficiently distinct; I really wonder just how much sense could be made of a paragraph of Koa in the absence of stress marking or word breaks, and whether I'm a fluent-enough speaker even to make that assessment. Might be an interesting project for a rainy afternoon someday.

Dependent clauses re-re-reenvisioned

Whenever I hear myself using the phrases "dependent clause" and "final decision" in the same sentence, it really ought to set alarm bells ringing. The last such occasion was about two years ago, in which I said the following:

• Clauses used as predicates are syntactically and morphologically identical to independent clauses
• Clauses used as adjectives are preceded by ve when the head is not the subject of the dependent clause, optionally otherwise
• Clauses used as nouns are preceded by either ko or ve

Over the past several months I've been entertaining a different idea. It initially seemed like a wild flight of fancy, but I've stuck with it and it's been surprising me by feeling more and more legitimate, or maybe even unavoidable. I'm still skeptical, so don't hold me to any of this, but I want to give it a genuine chance.

In this new dependent clause universe, the rules above can still be considered correct regarding clauses with ve. Here the dependent clause is being set off by a subordinating conjunction, and as such remains finite and modular with syntax identical to its use as a main clause. I think I've realized, though, that ko isn't just in free alternation with ve for structures like these as I initially proposed. It seems obvious now -- given the fact that ko is in fact a specifier -- that its function here is that of a nominalizer.

I see the source of my confusion: with clauses with a pronominal subject, subordination with ko and ve does accidentally look the same...

ni-kulu ko-ta-koke poli
1SG-hear NOM-3SG-tall very
"I hear he's very tall"

ni-kulu ve ta-koke poli
1SG-hear CJ 3SG-tall very
"I hear he's very tall"

The apparent parallelism was more striking before I started grouping particles in writing, i.e. previously

ni kulu ko ta koke poli
ni kulu ve ta koke poli

...but however they're spelled, in speech the structures still sound identical. The wild difference appears when the clause has a nominal subject: in this situation, rather than occupying the same clause-initial position as ve, ko replaces the preverbal i that marks finite verbs. I actually identified this possibility in the post referenced above, and initially ten years earlier, but couldn't make sense yet of why this would actually be reasonable beyond an aesthetic inclination for its elegance.

There's a straight line to this new structure from a clause like this:

A. ana ko-ni-lahe
give NOM-1SG-leave
"let me go"

We might translate this literally into English as "allow my going," or "allow that I might go." Another equally valid way of phrasing this in Koa, though, would be

B. ána-ni ko-lahe
give-1SG NOM-leave
"let me go"

Here the literal translation would be more like "allow me [the act of] leaving," just like ánani kovapa "give me liberty." If we replace ni "me" in structure B above with a noun, we end up with this:

ana le Keoni ko-lahe
give NAME John NOM-leave
"let John go"

For contrast, here's the same meaning with ve and a finite clause. Notice that the finite verb marker i is restored in the position previously occupied by ko:

ana ve le Keoni i lahe
give CJ NAME John FIN leave
"let John go"

So it's clear that something pretty different is happening syntactically in subordinate clauses depending on which of these strategies we use. As a further example of how nominalization works, here's the above nikulu kotakoke poli "I hear he's very tall" with a nominal subject in the dependent clause:

ni-kulu le Keoni ko-koke poli
1SG-hear NAME John NOM-tall very
"I hear John is very tall"

For those following the plot exceptionally closely, this does have implications for the vaunted tetai ko structure we've tried so hard to preserve through all these years. Again with a nominal subject, we have two possibilities:

te-tai ve le Keoni i cu-lai he-tana
ABIL-be CJ NAME John FIN IRR-return TIME-today
"John may be coming back today"

te-tai le Keoni ko-cu-lai he-tana
ABIL-be NAME John NOM-IRR-return TIME-today
"John may be coming back today"

I find myself narrowing my eyes as I read this back to myself...I'll sit with it for a while. This is a pretty big change, so I know it make take some time to get used to. I am cheered, I will just say, by some parallels between this new structure with ko and complement clauses in Latin (not to mention English phrasings like "I know him to be trustworthy"):

audiō Ioannem longissimum esse (Latin)
hear-1SG John-ACC tall-SUPER-ACC be.INF
"I hear that John is very tall," lit. "I hear John to be very tall"

The Salishan languages have similar structures as well:

d-s-x̆aƛ' kwi s-tə̀xʷ-s tsulč (Lushootseed)
1SG.POSS-NOM-desire ART NOM-buy-3.POSS drum
"I want him to buy a drum"

...cf Koa's

ni-halu ko-ta-kou toe
1SG-want NOM-3SG-buy drum
"I want him to buy a drum"

One last example of a more complex sentence, from actual usage:

Ni-io-te-nae kemo ko-te-tai u-tie neva
1SG-TRANS-ABIL-see manner NOM-ABIL-be DEF.PL-road unstraight
"I can already see how the winding roads...

ne-múnu-nu ko-na-cu-luta a-cólute.
LOC-middle-1PL NOM-NEG-IRR-find INDEF-connection
...between us might not find a connection."

Let's see how this goes! It's a bit of a brazen move in a long history of syntactical conservatism, but I'm excited about it.

Color terms

Despite boundless enthusiasm for color as a topic in itself, I've dragged my feet on this for a long time, for reasons I'll get to in a moment. The progression of color term assignment in Koa has been interesting, reminiscent of Berlin and Kay's study: initially red and blue, then later green, then later yellow, then much later orange and purple.

It has been a very difficult decision which color terms to classify as basic, and I've been extremely cautious to move forward because of a sharp awareness that my English-based instincts were certain to influence my judgment. In the end, realizing that to some extent any choice will have arbitrary components, I decided on a base favoring symmetry on the color wheel from the standpoint of a classical RYB paint-based model. A couple hopefully justifiable additions at the end round things out. Here's the system as it currently stands.

kona "black"
liko "white"

puna "red"
kinu "orange"
mele "yellow"
vihe "green"
sini "blue"
lie "purple"

teta "brown"
lusu "gray"

I dithered on orange and purple for ages -- ten years longer than the others -- and still feel uncomfortable with them knowing just how many languages they're not basic in. Nonetheless I really can't see a way forward without them, so I'll deal with it!

For the rest, despite technically being a kind of dark orange I considered brown to be reasonable given its cross-linguistic distribution, and gray for its usefulness in describing other intermediate colors. In a way I wish that the system looked less like English, but on the other hand at least I didn't include pink, and as a language for the modern world I feel like it'll be capable of handling color theory without a lot of excesses.

Blends between primary and secondary colors are formed as compounds. Theoretically the head might be the base color, with the modifying predicate indicating a small amount of an additional color:

mélevihe "yellow-green" (a hue of green with a little yellow added)
víhemele "green-yellow" (a hue of yellow with a little green added)

In practice, though, it may be more reasonable to conceive of a single tertiary color between each primary and secondary, permitting either root ordering but preferring primary-secondary, thus

púnakinu "vermillion"
mélekinu "amber/marigold"
mélevihe "chartreuse"
sínivihe "teal/aqua"
sínilie "violet"
púnalie "magenta"

A wild impulse is urging me to create single roots for these, which would allow delicious quaternary color compounds like scarlet (red-vermillion) and indigo (blue-violet). I'll hold off for the moment (sigh) but this whole area is difficult because of how accustomed I've become to beautiful single-word descriptions of so very many colors these days. How will we express sage? Olive? Taupe? Most likely it'll be necessary to come up with more or less standardized forms based on mia "color" and a noun: mialumo "sand," maybe? And additions of nominal descriptors to the colors themselves: híkivihe "grass green."

Moving on to modifications of a starting hue, we can express saturation and tint/tone/shade as follows:

válosini "bright blue (high saturation of blue)"
kícasini "light blue (low saturation of blue, or blue mixed with white)"
pímesini "dark blue (blue mixed with black)"
lúsusini "gray blue (blue mixed with gray)"

There's also a possibility of compounds with lahu "murky, obscure, overcast" to denote, probably, combinations with brown (or with an opposing primary color, if you prefer).

Lastly, a small amount of pigmentation added to something else can be expressed with -si meaning "-ish":

púnasi "reddish"

There's clearly a lot more to come in this area, as usual no doubt to be determined (or at least necessitated) by actual usage. Hopefully this will at least be a solid starting point.

Obligation and necessity

Koa has developed quite a bit of nuance this year in its ability to express shadings of force acting on the will. Much of this started coming together just a few months ago as I suddenly became aware of just how much linguistic machinery we have at our disposal in natural languages for making fine distinctions in this area, and for veiling (or not) the context of force or control behind a communication. Unsurprisingly given the brain of its creator, Koa is rather more transparent than English in its basic structures.

Starting out with morphological strategies (if indeed Koa has inflectional morphology, on which utterly undecided question see here), we can get at an initial simple range of semantics using modal particles, with evidentials and viridicals adding subtlety:

nitelahe "I can go"
nilulahe "I want to go"

nikilahe "I have to go, I must go"

nivukilahe "I guess I should go"
nilikilahe "Seems like I should probably go"

nipukilahe "Apparently I've gotta go (based on what I hear/perceive)"

Note that ki is neutral as to the source of the necessity: nikilahe implies nothing about whether I have to go for internal or external reasons, or what those reasons might be -- see below for greater specificity here.

The particle vi, formally an imperative or optative marker, can have volitive force particularly in the 1st person. Here it's largely synonymous with the hortative ea: ea lahe "let's get going" can be regarded as semantically equivalent to nuvilahe.

nivilahe "I'd better go, I've gotta get going"

Deserving of mention at this point is the particle oe which has existed in something of an alpha version since 2003. Theoretically it might mean "should," as in "I should go, I ought to go," but though I adore that word for that meaning I've never managed to decide definitively whether it in fact really should exist and if so how it would be used. Nioelahe? Oe nilahe? Oe konilahe? Bookmark for sometime in the next decade...

Moving on to more complex structures, we can describe internal volition or need with the referent in question as the subject of a main clause:

nivoi kolahe "I can go, I'm capable of going"
nilahi kolahe "I know how to go'"
nihalu kolahe "I want to go" (equivalent to nilulahe above)
nisaa kolahe "I get to go, I'm allowed to go"
nitau kolahe "I need to go, I have a need to go (internal necessity)"

nivati kolahe "I demand/insist on going (internal requirement)"

To express external will or control, the referent becomes the subject of a nominalized subordinate clause governed by a predicate explaining the relevant semantics:

tau konilahe "I have to go (external necessity), there's a need for me to go"
pono konilahe "I should go, it's fitting/proper that I go"
halu konilahe "there's a desire for me to go (external volition)"

vati konilahe "I'm required to go (external requirement)"

noi konilahe "I've been asked to go, I have to go (external request)"

milo konilahe "I have to go (according to the rules/law)"
kase konilahe "I'm ordered to go"
pako konilahe "I'm forced/compelled to go"
i ana konilahe "I'm allowed to go, they're letting me go"

...and one with negative force:

levi konilahe "I'm not to go, I'm forbidden to go"

A few notes here:

* In this syntactic context, as is usual with predicates with presentative force, all the above forms can optionally be preceded by i -- i tau konilahe, etc. With ana it's required, as ana konilahe alone sounds like "let me go!"
* The expressions with tau and pono above are pretty common, neutral ways to translate "have to" and "should" respectively when the deontic source is external. The others are less frequent and/or have more marked or complex semantics.

* Evidentials and viridicals can of course combine with these structures as well for additional nuance: puvati konilahe "I'm required to go, so I'm informed," and so on.

These presentative forms downplay the actual agents of control: we may be talking about actual, specific people, or institutions, or just cultural forces. If we want to bring the presence of those agents further onto the stage, we can use hi- with the predicates that have transitive verbal valence (i.e. not pono, milo or levi):

hivati konilahe "'someone' is insisting that I go...(as in, we all know that 'someone' is my wife)"

Finally, just to spell this out, with a non-pronominal subject in those subordinate clauses, we'd have the usual syntax like the following:

pono ko/ve le Keoni i lahe "John should go"

...or maybe pono le Keoni kolahe? It will surprise no one that I've quietly replanted the nominalized clause briar patch in the past several months. A controversial proposal is currently in formulation for an upcoming post...

From the vaults: Quantifiers and specifiers

The following post was originally written on March 13, 2021, but never published because I was so frustrated not to be able to come to grips with the problem that became very manifest as I started trying to produce example sentences. I never deleted the draft, though, because I sort of enjoyed the exasperated sarcasm into which I rapidly deteriorated. Here it is for posterity.

I think I may thankfully have solved this problem eventually, at least in theory. Actual usage recommendations are still developing via...actual usage, and I still sort of cringe everytime I need an existential verb. Truly, resolving this overlap between the existential quantifier and the indefinite article may be Koa's most vexing, longest-standing problem.

----------

Ever since a Ling department party in 2002 I've been nervous about something. I've put it off and assured myself it was okay, I think because my Indo-European intuition was so strong, but today it has officially stopped making sense and by hook or by crook we're going to have to take another inventory of our articles and determine what they actually really mean.

The areas that cause nervousness are basically where quantifiers and specifiers intersect, for example:

* What is the difference between the existential quantifier and the indefinite article -- "a book is on the table" versus "there is a book on the table?"
* What is the difference between the universal quantifier and...well, actual statements of quantity -- "cats are evil" vs "all cats are evil?"

I've tried to tackle this before, first in this specifier flowchart from 2010 and then in this revision in 2012. I'm increasingly feeling that tickle, though, that tells me that I was making some assumptions based on misunderstandings or unwarranted conflation of levels of description. I'm seeing three levels at this point that these questions interact with, and we need to keep them separate: logical quantification, the discourse stage, and other pragmatic concerns.

Some specifiers are straightforward (I hope) and we can probably safely leave them out of this discussion: ka the definite article, signifying that the referent is already on the discourse stage and recognizable to everyone taking part in the discourse; ko abstraction of the predicate; and ti/to deictic markers ("this"/"that") when used to refer to physical proximity. What we're going to have to really rigorously investigate are a, hu, po, and ti/to in discursive rather than physical function.

What is actually the difference between a and hu? Let's explore in different syntactic positions and AFF/INT/NEG.

SUBJECT
hu mala i ne masa "there's a fly on the table, an (unspecific) fly is on the table"
a mala i ne masa "there's a (certain?!) fly on the table" uh what

ai hu mala i ne masa "is there a fly on the table? is an (unspecific) fly on the table?"
ai a mala i ne masa "is there a (certain) fly on the table?" yeah that's nonsense

na hu mala i ne masa, hu mala i na ne masa "there's no fly on the table"
na a mala i ne masa "okay this is stupid"

OBJECT
Where are you going? I need to buy "a" book today. Does that book already exist in the physical universe recognizable to the speaker, as in "a certain?" Traditional Koa would give you:

ni ki ala (po) pama he leo "I want to engage in book-buying today" (this surprised me)
ni ki ala a pama he leo "I want to buy a particular book today"
ni ki ala hu pama he leo "There's a book I want to buy today"

Shouldn't this really just be:

ni ki ala pama he leo "I want to buy a (some nonspecific) book today"
ni ki ala a/hu pama he leo "I want to buy a (specific but not on the discourse stage) book today" or "there's a book I want to buy today" depending on context
ni ki ala ti pama he leo "I want to buy a certain book today that's not on the discourse stage but will potentially be relevant to subsequent discourse FYI"

I don't want to do this anymore, this is idiotic. It's really just a question of which particle we want to keep in the a/hu role and which we want to free up for other meanings.

Experiential structures

Koa has a series of verbs representing sensory, perceptual or otherwise subjective scenarios, for which the "experiencer" theta role is encoded as the subject. Experiencer marking is all over the place cross-linguistically, obviously, but this is the default syntactic strategy for experiencers in Koa on the grounds that most or all other statives share that structure.

In making this decision there was an initial challenge around how to encode the feature of volition: "see" versus "look," "hear" versus "listen," etc. I sorted much of this out in 2016 with the realization that nae as "see" could be rephrased almost synonymously as tenae "can see," whereas the progressive form manae would be understood as "looking":

ni-na-ma-nae ala ni-si-náe-ta i meno
1SG-NEG-PROG-see but 1SG-ANT-see-3SG FIN regardless
"I wasn't looking, but I saw it anyway!"

Another common usage is the passive with mo as in panae mo "look like," literally "be seen as." There are many other parallel structures among these experience verbs, some of which correspond to lexical differences in English where others are harder to translate directly. Since I've never spelled this out fully, here are the roots I've identified so far as belonging to this class and their meanings in all of these structures (omitting those usages which seem anomalous or meaningless):

nae "see"
manae "look at, watch"
vinae! "look!"
konae "vision"
náete "a glance, a look"
panae "appearance, one's 'look'"
panae mo "look like"

kulu "hear"
makulu "listen to"
vikulu! "listen!"
kokulu "hearing"
kúlute "a 'listen'"
pakulu "a sound, a noise"
pakulu mo "sound like"

olo "smell"
maolo "sniff, smell (intentionally)"
violo! "smell!"
koolo "smell, smelling"
ólote "a sniff, a smell of something"
paolo "scent, fragrance"
paolo mo "smell like"

kihe "feel (physical, tactile)"
makihe "touch, feel"
vikike! "touch! feel!"
kokihe "touch, sensation"
kíhete "a feel, a touch"
pakihe "feeling, sensation, texture"
pakihe mo "feel like (physically)"

maku "taste"
mamaku "taste, try"
vimaku! "taste! try!"
komaku "(sense of) taste"
mákute "a taste"
pamaku "flavor, taste"
pamaku mo "taste like"

tune "perceive"
matune "focus on"
vitune! "focus! pay attention!"
kotune "perception"
túnete "an occasion of focusing attention/perception on something"
patune "a perception"
patune mo "seem like"

kue "experience"
makue "invite/seek out an experience"
vikue! "(intentionally) experience!"
kokue "experience"
kúete "an experience (the act of experiencing)"
pakue "a feeling, an experience (the thing experienced)"
pakue mo "feel like (experientially)"

huo "notice"
mahuo "take note of, observe, pay attention to"
vihuo! "take note! observe, pay attention!"
kohuo "notice, attention"
húote "an observation (the act of taking note)"
pahuo "a note, an observation (the thing noticed)"

kuvi "imagine"
makuvi "imagine, picture (intentionally)"
vikuvi! "imagine! picture!"
kokuvi "imagination"
kúvite "(an occasion of) imagining, a daydream"
pakuvi "an imagining, fancy, daydream, idea (the thing imagined)"

mai "feel (emotionally, subjectively)"
mamai "hold onto, seek a feeling"
vimai! "try to feel X!"
komai "feeling emotionally"
máite "(an occasion of) feeling something"
pamai "a feeling, an emotion"
pamai mo "feel like (emotionally)"

...and to some extent, though the +volition set of meanings is kind of marginal (I guess maybe specific to lucid dreaming?):

moe "dream"
komoe "dreaming"
móete "an occasion of dreaming"
komoe "a dream (the thing dreamed)"

There is almost certainly some unavoidable ambiguity around volition in certain syntactic contexts, though I'm pretty sure that semantic or narrative context, TAM marking, and/or definitiness of arguments would resolve it in most cases...and maybe in other cases the distinction isn't really that important.

Compounds and branching direction

This is a translation of yesterday's post, which I'd been meaning to write since last December and it suddenly occurred to me I might actually be able to express in Koa. It was challenging in places and not quite as free-flowing and expressive as it would have been in English, but still it's the first-ever meta-article in Koa which feels like a milestone!

There were lots of vocabulary and usages being invented on the fly, not all of which I'm sure about: nóava "name (tr)," since I don't have a word for "express" or "mention"; pamulíkema "compound," i.e. "put-together-thing"; páama "head" and nálama "dependent," and so on. (Should "head" be étuma or something? Either way étuvike would be "matrix clause" and nálavike "subordinate clause.") I also don't yet have a word for "order," which irritatingly necessitated the circumlocution nácate "arrangement." On the other hand I'm pretty pleased with VM as the acronym for vike méama "noun phrase," to which we could also add VE = vike étema "verb phrase" and VI = vike ílama "modifier phrase." I've been slowly working on other Koa-specific linguistic vocabulary on the side which I'll eventually want to unveil; some examples are étema aivu "passive verb" (e.g. nipaloha), méama lala (oétema) "(deverbal) instance noun" (e.g. súote), and méama litu eli "definite plural noun" (e.g. ukunu).

I might mention that the problem described in this post is getting kind of galling given how many compound nouns have found their way into regular usage, with more appearing all the time. Every single time I say or write ésipai "Monday" I ask myself whether it really ought to be paiesi; my excitement at coining koláimiena "reunification" is tempered by anxiety that komienalai may be a more native structure somehow.

One thing about the head-final ordering is that it can sometimes generate worryingly long words with initial stress, which feels alien to the usual Koa cadence for long words. Nisilumutepanáeta "I wanted to make it visible," on the other hand, has nine syllables but is still natural to pronounce and clear to understand. Left-branching compounds can also end up being tricky to parse when they contain prefixes, suffixes and compound nouns: munálavikema "complementizer," grouped like [[mu[nála[vike]]]ma] "cause-subordinate-clause-thing"; or sometimes seem to need two written accents -- is that even allowed? -- like símipúhute "telephone conversation." Again, I have no answers, only vague anxiety.

A last note: it came out in the course of writing the below that the adverbial "often" translates into Koa as mouse, a discovery which caused me almost to break a rib laughing.

--

For a few months I've been worrying about the fact that we apparently have two different ways of forming compounds in Koa, without any guidance or even discussion about the fact, really. In both methods the words are written together and the dependent bears the stress (I'm creating these grammatical words as I write, so they may not be permanent), but in one strategy the dependent precedes the head (címihale "grammar, e.g. language-structure") and in the other it follows (halecimi "structure-language").

In the beginning I thought that the second type was the only one possible, in accordance with the normal ordering of dependents and heads in Koa. But later on, maybe due to a bunch of time spent with Esperanto (how do we use foreign words, anyway?) my practice changed almost entirely. I might have been thinking that there wouldn't be enough distinction between ordinary adjective phrases and compounds if dependents were to follow in the same way.

After some research -- mainly an article by Laurie Bauer (great thanks to Allison for extremely kindly getting hold of it for me as usual) from a book I unfortunately haven't been able to identify -- I learned:

* Some languages use compounding more than others; some don't use them at all (?!)
* It seems that both structures are found cross-linguistically, even in head-initial languages
* There's a small tendency for compounds to be left-branching even in languages in which NP's are right-branching
* In general the ordering of heads and dependents in compounds is the same as that of possessed and possessor nouns
* Interestingly, in languages in which noun-modifies phrases and compounds use the same ordering, there is often no structural difference between compounds and ordinary NP's!

All this seems to mean...that we're not getting any help with this decision, aggravatingly. This leaves me in some uncomfortable confusion, because it seems like I should decide something and I have no idea which I like more. It could be that both strategies will have their own uses...maybe? Or maybe I just need to make a real decision, unfortunately. Anyway, I wanted to make the problem visible, at least, even if nothing at all is clear at this point!

Pamulíkema e taho koósava

Hekei akuu tele nimaholi pekonupumekemo lua neo lakomumuo popamulíkema cile Koa, nokoohe coa e nokomova coa petá i vela moeso. Ciukemo lua hivakile uupu molike e ponálama i vakana célite (nimateke tiupu címihale henikile, laa tetai kotunacucali), ala cikakemo énasi ponálama i vakoe popáama (címihale) e cikamutu tavahala (halecimi).

Heámate nilule ko kalei lúasi mono i sitai, mavonu kanácate eco popáama e nálama cile Koa. Ala helohi taa, tetai lopólipi aika momaluna mele Esipelánito (ai motoa? Le Esperanto? Le Tóvoma? Mokea sa nuvamova poupu uto?) kakeva i simuta i aci mosai. Nilimameti ko i nacumenéote pepovike ílama eco e pamulíkema haponálama i vahala mosama.

Hala náipi kosue -- moetu apakile cile Laurie Bauer (kito suli lale Áliso pekokíita mocínino laní moeco) otusi ninasitemuimi i papaho -- niopi:

* Hucimi i vaitu popamulíkema i ene pemutu; huka i navaítutu i hetu (?!)
* Patune moko uhale lua i hovapaluta nemunu kacímipo kanui, i vela necimi páaenasi
* I mekeva píkuki ve popamulíkema i vaósava lavase i vela necimi ve povike méama i vaósava laaka
* Molaca kanácate kapáama e kanálama nepamulíkema i sama ve kaoma e kapaoma
* Mukino, necimi ve mekanácate sama pevike ílama e pamulíkema, mouse i hononéote pepamulíkema e VM (=vike méama) eco!

Tipoa i lisema...veninamasaa cóapi ápute metihólote, momuhulu. Toa i iátini nekosopu sune, lokopatune mokopono konicuholo hua e ninailo i hetu kaniaima i taa. Tetai ko ukemo lua i cumeítute oma, neo...aité? Au tetai koietau koniioholo i eso, nipaho. Meno, nisilumutepanae kahákate movaha, i vela ha naka hetu i kica hetitia!

Yes and no

I've traditionally given ia and na as the Koa translations of "yes" and "no," but the situation is actually quite a bit more complex with some fairly neat opportunities for nuance. Na is a decent all-purpose "no," being the general negative particle, but ia is actually a veridical often used in verum focus: as such its true meaning is "definitely, certainly, absolutely." There's a negative corollary ianá that means "definitely not, absolutely not," and there are many other modal particles that can be used analogously for different shades of meaning.

For example, ku classifies the information within its scope as already known, whether it really factually is or the speaker deems that it should be so. The following sentence might be uttered if the speaker had just returned from out of town, and been asked if they'd done something, or heard something, that implied more time at home than they had yet had:

ni-ku-lai he-ele i mono
1SG-KNOWN-return TIME-yesterday FIN only
"I mean, I only came back yesterday, after all"

As an interjection, then, ku means something like "obviously," "clearly" or "of course," and the negative form kuná would be "obviously not," "of course not," "clearly not."

For those particles that do not have veridical or evidential force -- that is, other than ia, li, pu, vu -- the positive form can also be combined with ia for emphasis, thus kuiá "clearly yes," "well of course it is." All of these particles are modular and can be combined freely in whatever ways their meanings would allow: so iaté "yes, that very well may be," "it is definitely true that that is possible" vs teiá "yes, perhaps," "it's possible that that's true." Or even stacked with other modals, like kuté, kuteiá "yes, that may indeed be possible."

Below are the particles that can be used as interjections and their approximate English meanings. One note for the following examples: "I" in the translations is only illustrative, and could be replaced with any other referent relevant to the discourse (e.g. oená could also mean "no, she shouldn't" in the appropriate context).

ea "yes, let's"
     eaná "no, let's not"

hoiá "yes, actually, counter to expectations" (ho alone indicates surprise)
     honá "no, actually not, counter to expectations"

ia "yes, definitely"
     ianá "no, definitely not"

io "that's right, there it is, aha!, bingo, there we go"
     ioná "no, no way, no longer, no more of that"

ki "yes, I have to; yes, it must be"
     kiná "no, I mustn't; no, it must not be"
     nakí "no, I don't have to; no, there's no need"

ku "yes, obviously; yes, of course"
     kuná "no, obviously not, of course not"

li "yes, I suppose so, might be, probably"
     liná "no, I suppose not, probably not"

lu "yes, I want to"
     luná "no, count me out"
     nalú "no, I don't want to"

oe "yes, I should"
     oená "no, I shouldn't"

pu "yes, apparently; yes, so I hear"
     puná "no, apparently not; no, so I hear"

te "yes, I can; yes, maybe"
     tená "no, maybe not"
     naté "no, I can't; no, it can't be"

vi "yes, do it! yes, let it be so!"
     navi, naví "no, don't do it! ojalá que no"

vu "yeah, I guess so"
     vuná "no, I guess not"

I'm really pretty pleased with the amount of subtlety and precision that Koa has developed in this area sort of by accident, just by virtue of its system of modal particles. I know I keep saying this but it's been continually fascinating to keep discovering the language's emergent complexity as it's come into greater and greater use this year.

Word break conventions and emergent typology

I've been doing a lot of free-form writing in Koa this year and it's been a pretty revealing experience. There's nothing that exposes semantic gaps and structural shortcomings like trying to write complex, expressive prose; initially all my writing felt unbelievably clumsy, with none of the grace, sophistication or subtlety that I try to embody when I write in other languages I know well. After a month or two, though, I feel like I'm starting to find my voice in Koa -- or maybe more accurately, Koa is finding its own nascent voice.

This is really the first constructed language in which I've navigated this process and it's fascinating (and intimidating): coming from a place of only having written single, unconnected example sentences, how does the language in question construct, say, a whole paragraph? How does it flow structurally? I feel so practiced in other areas of language design, but here I'm just doing my best to move through it all in an intuitive way without getting hung up on my own anxiety. Someday I'll have to try to actually articulate some of these emergent principles, but I think they need time to emerge further first.

In the mean time, another thing that came up as I began keeping a regular journal in Koa was a discovery I only made when I tried to read what I'd written later on. For one thing, I knew theoretically that production and comprehension were different disciplines, but I wasn't quite prepared for just how unpracticed I was at understanding my own language. It makes sense: I'd really never had the opportunity to try to interpret speech or writing coming at me before! In response I added a word recognition module to my vocab learning program; previously it had only been testing me on production in the target language.

More surprisingly, though, it turned out that the way I've always represented Koa is kind of hard to parse. Here's an example block of text written in the traditional style:

Ta lai la ka ásulo ta la ko vúakupu e ko mivami, sii, ta mene la ko kóuva e tule lai la ni. Ni si vima poli lo kopato ve hua i cu misucu, ala he lopu poka i pea pono e ka lila ni sai i si kali. E ka tana i kali i koe ka sena. Hala kehe nu lu nike la ko mova ka kecu, ka nu lu ete la ko mupea ka háote nu ne kene koa.

As soon as I start to read it my eyes sort of go out of focus; with such a rapid stream of little words it's hard for me to keep track of where I am in the text, let alone where I am in the syntactic tree. As a result, over the past month I've been experimenting with writing roots with their particles attached to them. The precise rules about what should be attached and what should be left separate are still developing, but the essence of the system has come together nicely. Here's what that previous paragraph looks like with the new conventions:

Talai lakaásulota lakovúakupu e komivami, sii, tamene lakokóuva e tule lai laní. Nisivima poli lokopato ve hua i cumisucu, ala helopu poka i pea pono e kalílani sai i sikali. E katana i kali i koe kasena. Hala kehe nulunike lakomova kakecu, kanuluete lakomupea kaháotenu nekene koa.

Even though this was unfamiliar, I instantly found it massively easier to parse. Allison said that made sense to the extent that there were many more word shapes now for the brain to grab onto; it's also entirely clear which particles belong to which roots, and morpheme clusters mirror natural intonation groups. Here's an attempt to articulate the principles of the system.

1. Particles whose scope is a predicate -- regardless of how complex it is is -- are written together with that predicate. This may require the use of additional accentuation where possessive pronouns and directionals are suffixed to the root.

ninasitemuláheta = "I couldn't make him leave"


2. Particles whose scope is a clause with a pronominal subject are joined joined to that clause (but see point 6)

nisánota lakomutulu kakúmumani  = "I said it to make my teacher angry"


3. Particles whose scope is a clause with a full subject NP are separated from surrounding words

nitovo ko le Kéoni i cutule = "I hope that John will come"


4. Predicate clusters -- compounds and incorporated objects -- are written together, but plain adjectival phrases are not joined to their head nouns

kalopuviko = "the weekend," but
kapasano vime = "the last statement"


5. Pronominal particles follow the same rules as predicates when used as the head of an NP, but must be marked with an accent.

laní = to me
nahunú = none of us


6. Certain particles, principally with clause-level scope, are always written separately: i, e when it means "and," au, ai, ha when it means "if," ve when used as a complementizer, and ko when used alone as a complementizer (this list may not be exhaustive). Le is also separated from its head to avoid muddlement with capitalization and foreign words.


One point of uncertainty: when a particle is written separately from its head but is itself within the scope of other particles, are those particles also separated or should they be attached to the "frontmost" one? For example, which of the following should be the convention?

nisánota lakole Kéoni i cutule
nisánota lako le Kéoni i cutule
nisánota la ko le Kéoni i cutule
"I said it so that John would come"

I'm not sure yet; I'll get back to you after more experimentation. I suspect a standard will shape itself over time.

A bunch of this, incidentally, may actually be an artifact of trying to smoosh Koa into an alphabetic writing system. If the language could be written with a syllabary rather than an alphabet, and if there were some marking that identified the stressed syllable of predicates -- in other words, if predicates were instantly differentiated visually from particles -- then there would be a much closer match between writing and Koa's native structure.

But what, then, is Koa's native structure? I had always thought of it as a basically isolating language, but one thing that really surprised me when I first saw text written with these new conventions is how...agglutinating it looks. I'm sort of shocked that I've never asked this question before, but...where does the structure of Koa really fit, typologically?

The language is certainly about as close as you can get to monoexponential in that each morpheme is (theoretically) encoding one and only one semantic, and since I've been thinking of all particles and predicates as individual "words," my unconsidered classification of isolating seemed justified. But looking at forms like this one from above...

ni-na-si-te-mu-láhe-ta
1SG-NEG-ANT-ABIL-CAUS-leave-3SG
"I couldn't make him leave"

...I really wonder on what grounds I would not call that a "word." A word constituting a complete sentence, with seven morphemes, which a Turkish speaker could feel right at home with. And if that resemblance isn't just incidental but in fact diagnostic, then classifying those first five morphemes as "particles" is obscuring something important: they're actually prefixes. Occupying slots, in a specific order. Like an agglutinative language.

I'm actually not sure how to make a ruling on this, and more thought and research may be required. Some of those particles certainly can stand on their own in certain contexts -- nate "no, I can't," or keka sa? ni "who is it? me" -- and maybe more revealingly, the pronouns can appear to be gapped: 

ka-Ø-na-si-te-mu-láhe-ta
DEFᵢ-Øᵢ-NEG-ANT-ABIL-CAUS-leave-3SG
"the one who couldn't make him leave"

ka-ni-na-si-te-mu-lahe-Ø
DEFᵢ-1SG-NEG-ANT-ABIL-CAUS-leave-Øᵢ
"the one I couldn't make leave"

On the other hand I've vigorously maintained previously that gapping is in fact not the best explanation for these structures despite the fact that it's possible to draw the trees that way. It may be that this new word break convention and the kinds of apparent agglutinative "words" it produces is itself also obscuring some of the true nature of the base structures. Ultimately this is not a question of graphical representation -- whether we write ni na si te mu lahe ta or ninasitemuláheta -- but what's really happening below the surface. And I'm starting to tie my brain in knots which is a pretty clear sign that I need to put down this problem for a bit.

More to come, clearly.

A parting of the ways

My decision last month to remove all predicate roots beginning with /j/ from the Koa lexicon sparked a significant artistic crisis. I tried to accept the replacements, but as time passed I was confronted with a growing feeling that this change was not okay. I loved those proscribed roots, loved the variation in syllable structure that they provided, and realized that I would like Koa less without them; worse than that, that it would feel like it had lost some of the essence of itself. It would feel like it was no longer mine.

I was clearly right when I said that this phoneme had no place given Koa's charter, but it just doesn't matter: apparently at this point the language has developed such a strong sense of itself, especially after all the vocabulary creation and writing that's been going on this year, that honoring that personality is actually more important. The charter was supposed to be an inspiration, not a prison, and the fact is that I love what Koa has become so much that I would rather change the limits than stifle the language to fit within them.

This may seem like a lot of fuss over 20 roots and a marginal phoneme, but this is the first time I've ever consciously and intentionally prioritized aesthetics over the language's ease or clarity. It's uncomfortable, but also unquestionably the right decision.

Emboldened by this I've found myself thinking crazy thoughts, like considering adding another consonant phoneme. I experimented with [ŋ] and was shocked to discover that I actually loved it, and that it "felt" like Koa despite the fact that it would be completely off the deep end charter-wise. I don't know that I'll really go down that path, but it's sort of a wonderful thing that after 23 years there is something that Koa "feels like" to such a clear extent that it can begin to direct its own course into territory I'd never imagined.

Over the weekend I reinstated all my exiled vocabulary. It was a tremendous relief. Honestly I think I would have died on that hill for iolo alone.

Lament for the vanished on-glides

Although /j/ as a formal phoneme had been officially nixed a few years previous, beginning in 2011 a number of words appeared in Koa with an initial [j] sound: iolo "joyful," iuna "train," iune "steal," and so on. This was possible because of the adoption of a series with this sound among the particles: ia "yes, definitely," io "already," etc., for which I could think of no objection.

As of this morning the language contained around 20 predicates with this on-glide. I still think the particles are fine, but it suddenly crystallized for me that there was a serious problem with the predicates: the accidental adoption of this phoneme created a growing functional load on the distinction between prevocalic [j], [i] and [ij]:

ka kane iolo
[ka kane jolo]
"the joyful man"

ka kane i olo
[ka kane i olo]
"the man smells (something)"

ka kane i iolo
[ka kane i jolo]
"the man is joyful"

As I've been experimenting with writing Koa without spaces between bound morphemes (post eventually forthcoming) the problem became very stark: the phrases above come out kakane iolo, kakane iolo, and kakane iiolo! It's even worse when the preceding predicate also ends with /i/:

ka hapi iolo
[ka hapi jolo]
"the joyful ant"

ka hapi i olo
[ka hapi i olo]
"the ant smells (something)"

ka hapi i iolo
[ka hapi i jolo]
"the ant is joyful"

...so here we're making a distinction between [ijo], [iio] and [iijo]. Heavens above. As much as it -- truly, sincerely, kind of agonizingly -- grieves me, these just couldn't stand: in an artlang, sure, but not with Koa's charter. They just weren't meant to be.

And so, glumly, this afternoon I went through and reassigned all of these predicates. Some of them feel okay, others may take some getting used to or find themselves replaced eventually. The hardest one by far was iolo: there is just no other sequence of sounds that more clearly communicates joy to me after having it as a core predicate for more than ten years. I feel like I want to keep it around as an archaic alternative usable in poetry.

Anyway, for posterity, here are the lost on-glide roots; farewell, and I'll remember you always.

iaho -> auho "flour"
iali -> ali "put away"
iane -> ane "cord"
iapu -> epu "spit"
iehi -> ehi "hate"
iela -> sela "whole, unbroken"
ietu -> cetu "dishonor"
ieva -> teva "gradual"
ioco -> oco "copper"
iolo -> elo "joyful"
iomu -> omu "meat"
ioni -> coni "yoni"
ioti -> toti "perseverate"
iotu -> enu "curious"
iovi -> kovi "wise"
iule -> ulu "apart"
iuna -> vona "train"
iune -> lune "steal"
iuve -> uve "fall short"

Ooh, some of these are still not feeling great...I can tell I'm going to have to give myself time.

Consonant use statistics

This morning I focused for the first time on the fact that my little random word program -- the database that suggests Koa roots in need of meanings -- is suggesting roots containing /c/ a disproportionate amount of the time. This in itself wasn't surprising: /c/ only returned to active use about 15 months ago, so it would make sense that more roots containing it would be available. It made me wonder, though, just how much variation there is in consonant phoneme frequency in Koa. I ran some numbers...

This was not quite what I expected! It turns out that as of this morning at 11:30am, of the 840 roots assigned meanings so far, the average number of words containing a given consonant phoneme is 135.5. That puts /h m n s t/ right in the middle with approximately equal frequency. My expectation about /c/ was correct, with roots containing it only representing 46% of average...but who knew that /p/ is way down there too at only 69%? I knew I had a bit of anti-bilabial-stop bias -- Seadi didn't even have those phonemes originally, explaining them away via some extremely convenient historical change -- but I certainly was not aware of its having been working so effectively in the background of Koa word creation.

On the other end of things, /k/ and /l/ are significantly overrepresented at nearly 150% of average! ...Which also kind of makes sense because they're also favorites of mine.

I guess it just hadn't occurred to me that my own personal aesthetics would have figured so prominently in root choice with respect to phoneme frequency! I must have expected that each consonant would appear approximately equally, as odd as that would have been cross-linguistically?

That raises a really interesting point, though, which I also had never considered: the particular character of Koa as it has always existed manifests these frequency biases. Like any language, the phonemes are represented unequally, and that gives it an important part of its unique phonological character. As such, moving towards greater uniformity -- as my random picker would automatically tend to do -- would, over time, actually alter the feel of Koa.

And if I like the phonological aesthetics as they've been up to this point -- which it turns out I do -- I may actually not want to continue generating words this way! I'm not sure yet exactly how I'll do this, but what we really want is for the randomness to be weighted -- towards words with Koa's favorite phonemes, and away from words with those it prefers less -- such that a random sample of suggested words would tend to show the same frequency distribution as the language as a whole.

I almost wonder if I should go back to an earlier version of the file, run these numbers again, and use those statistics; the program potentially had a noticeable impact on the frequencies with those 200+ words in the past couple months. Though...on the other hand I was still vetting the choices so my aesthetics were still probably in force, even if being nudged. I could figure out the statistics of the recent additions on their own just to be sure.

Anyway this is certainly an interesting little surprise for me to ponder.

Etymology statistics

Just a point of interest as I continue to organize my lexicon...

Of the 790 predicate roots assigned so far:

* 163 (21%) are derived from Finnish
* 57 (7%) are derived from Hawai'ian, Sāmoan, Tongan or Māori
* 75 (9%) are derived from other languages (Arabic, Basque, Bislama, Chinese, Doraja, Esperanto, French, Greek, Icelandic, Irish, Japanese, Latin, Lapine, Latvian, Malay, Nahuatl, Polish, Proto-World [ha ha], Quechua, Quenya, Russian, Seadi, Spanish, Swahili, Tagalog, Turkish, Swahili, or broad international usage)
* 34 (4%) are internally-derived

This means 295 (37%) of the current Koa root stock was derived in some way from other languages, compared with 495 (63%) that was either randomly generated, internally derived, or selected/created in some way (unfortunately there's no good way to distinguish randomness from intention reliably at this point). I find these figures a little surprising: it was my impression that the significant majority of Koa words was based in something -- to the point that I was stymied for a long time in creating more vocabulary when I couldn't find enough existing linguistic inspiration. Also, again, let's just pause for a second to acknowledge that Finnish has provided a fifth of Koa vocabulary.

Worthy of special mention are 6 roots (1%) that were created by friends or family members -- I'd love to swell that number moving forward!

A first Koa publication

In response to my children's repeated requests, I decided late last year that I would do my best to assemble a printed, kid-oriented, concise Koa dictionary in time for the Solstice. As the project took shape it grew beyond my original intention, eventually including a phrasebook and mini-grammar as well, and in the end I was pretty pleased with it as a snapshot in time of the development of this language.

It was also an opportunity to buckle down to some serious vocabulary creation, which had been languishing a bit in recent years; I'm pretty happy to finally have words like siki "particle," mohi "predicate," lelo "sentence," cóepo "alphabet" and címihale (or halecimi...more on that soon) "grammar."

In fact the process of creating needed vocabulary for the Úputusi Énasi sort of unblocked me and I've been on a bit of a rampage since then, coining around 200 new words over the past two months. What's been amazing is discovering that all that toiling in the syntactical, pragmatic and morphological mud I was doing in 2021 moved the structure of Koa to a place where now vocabulary is its primary need. Suddenly having all these words available, I'm finding that the language is much more speakable than I had previously expected, and with surprising expressive power.

As of today at noon, 774 of Koa's 3330 possible predicates have been defined. Emotional vocabulary has been my focus of late, but I'm starting to wonder what other thematic categories deserve some attention. Materials? Science? Botany? Civics? Geography? I've always been so intensely focus on word-worthiness and concerned about running out, but after 23 years I've still only used up a quarter of my possible roots!

Anyway, this was such a fun project that really jump-started some major progress after a pretty slow year. Unfortunately the dictionary doesn't seem to have inspired my girls toward total Koa fluency yet, but surely it's only a matter of time...

...And if you'd like to download the whole thing, a PDF is available here.

Ni Ceso

This is a difficult moment in my life. It's not the first such moment that's passed since I started this blog, but it is the first time that I was actively working on Koa enough to have something to say. What this means, of course, is that I'm now going to make you read sad love poetry.

Seriously, though, this is the first native-Koa artistic composition since Aika Konuku in 2012, and the first non-translated work of poetry of any kind. It feels significant! It features Koa’s growing collection of emotive vocabulary, particularly "ceso" which means something like "incurious" but less highbrow: the opposite of "curious," desiring not to know, feeling pulled away from rather than towards understanding of something; it also makes heavy use of modal particles and clause nominalization.

As with the previous poem, I find myself really liking how compact, elegant and balanced Koa can be for poetry. I didn't expect this but it makes me happy! My shot at an English translation definitely loses some or all of that particular aesthetic sense of the Koa original.

Ni Ceso

Ni ceso
Noia na vi sano ni
Ka se cu nike he tana
Ka so cu ete mo kune
Ka ne se simo he ko meti pe to níkete
Ka ne se simo he ko meti pe to mehe
Ka ne se simo he ko meti pe ka kecu.

Ni ceso
Kelo se na te lu tai me ni
Kemo sisu ve se ca ma tala ko halu ni
Ka ma lolo se simo mo iule o ni
Ni na te koma ka natepakoma.

Ni na lu koma
Ni pavasu lo ko tala
Ni cu te hitui la hete to lise mo cali.
Ni na lu koma ka natepakoma
Ni na lu koma kelo se na te halu ni
Noia na vi sano ni
Ni ceso.

-Váhumaa, 2023-01-22


Translation:

I'm Not Curious

I'm not curious
Please don't tell me
Who you're seeing today
What you're doing together
What's in your heart when you think about that meeting
What's in your heart when you think about that person
What's in your heart when you think about the future.

I'm not curious
Why you can't want to be with me
How hard you're still trying to want me
What's holding your heart back from me
I can't understand the incomprehensible.

I don't want to understand
I'm worn out with trying
I could smash myself against that wall forever.
I don't want to understand the incomprehensible
I don't want to understand why you can't want me
Please don't tell me
I'm not curious.

-Portland, 1/22/23