My Koa enthusiasm is starting to run out of gas for this cycle, so I want to make sure I quickly document the ideas still bouncing around in my brain for next time. A lot of these are already mentioned in the lexicon but not developed anywhere yet.
1) In proper Welsh spirit, it occurs to me that "bring" and "take" should be translated as tule me and mene me. For example, ta si tule me ka sahi la nu talo "he brought the wine to our house."
2) Currently we have two motion verb roots: tule, which means motion towards the given deictic center, and mene any other kind of motion: either away, or within, or along, or just unspecified. It happens that this is the same arrangement we have in English, which makes me nervous. Do we need another morpheme for movement away from the center? It seems like, logically, either the answer should be yes, or we should dismiss tule and just make do with the allative and ablative particles.
3) I've chosen he as the temporal particle. This gives us he kea "when," he toto "when I was a child," etc.
4) Now that the 3rd-person pronoun has once again been freed from use as a demonstrative, this reopens the pre-predicate slot for pronominal possession. I propose to continue to allow it for the present: thus "my mother" is either ka mama ni or ni mama, with no difference in sense.
5) Speaking of demonstratives, it seems that I've pretty much decided that one is not going to cut it anymore. The problem is that deictic specification is bound to get pretty wordy if I'm determined to make it periphrastic: like, if "here" is aki and "there" is ila, we'd have to say things like ti kunu ne aki and ti kunu ne ila for "this/that dog." In other words, we end up having to use roots to get at these deictic concepts: here/there, this/that, now/then, etc. Who's to say which are important enough to get their own morpheme? In Hungarian there's ilyen and olyan for "in this way" and "in that way," but English and Polish don't distinguish.
Having ti as the near demonstrative and to as the far demonstrative clears this all up neatly, because now ti kunu can be contrasted with to kunu, and we have he tia "now" versus he toa "then," etc. "Here" is ne tia, "there" is ne toa. I think this is better, in the sense that it's far simpler and easier, than any periphrastic alternative; for this reason, even though it makes me a little nervous, it seems like the way to go.
6) I've made ho available as an optional vocative marker. It doesn't have to be used, but there are definitely times when it helps to clear up ambiguity, or disentangle a predicate from a clause that it's not supposed to be an argument in. The idea is that it means "hey!" minus the potential pejorative associations that this can carry in English. Another option would be to use oi, which has a more "hey"-like feel and is less like Latin/Esperanto.
7) Lo is the new causal particle, giving us lo kea "why?", lo tia "because," ni si mene lo ko ni na loha se "I left because I don't love you," etc. This will be used in a lot of situations that "because" wouldn't immediately spring to the mind/lips of an English speaker: a flag flapping in the wind, for example. We'll need to come up with lots of further ones.
8) I seem to be in a fit of particle assignment this time. I'm slightly uneasy because I haven't spent years agonizing about whether these semantics really warrant a particle of their own, etc., but I guess I'll just lay them out, and then I've got something to do for the next year or two.
The next particle is pe, which I haven't found a good Latin name for yet, but which means "about/concerning/with respect to." This can be used in the English sense of "about" in ta ma puhu pe po kunu "he's talking about dogs," but I think there will be a slew of other uses that will come up when I start trying to translate more complex ideas. Li vetas je ĉevaloj, that sort of thing.
If not, though, o might be able to serve this purpose as in Finnish, so let's keep our eye on it. For that matter, o might also be able to take the place of lo as causal, so let's definitely not etch this into marble yet.
Yes, o definitely makes sense with causal meaning; o ko is great for "because" as a conjunction. It's a little weird that o kea would mean both "why" and "from where," though. Either way, a thought: if pe means nothing more than "concerning," what's wrong with ta ma puhu ne *tema po kunu for "he's talking about dogs?" We're going to need some more motivation for this.
9) Unless I can come up with any reasons why it's not a good idea, I'm planning to use la to mean both "to" and "for." Thus, ni ma ana tia la se "I'm giving this to you," whence tia i la se "this is for you." Also, though, this could join clauses with ko: ni si tule la ko ni ana tia la se "I came (in order) to give this to you."
Note that the above sentence is different from ni si tule lo ko ni ana tia la se in that this would mean "I came because I give you these things," a meaning sort of in the same neighborhood but not at all the same. Is it a problem that lo and la are so similar?
10) I think I'm actually okay without having a distinct privative particle: na me is fine. After all, that's what's going on in English without too, right? Ni me sene ala ni na me kunu "I have a cat but I don't have a dog." Maybe na me could even be spelled name sometimes.
11) Indeed: should particles sometimes be joined in writing, or joined to predicates in set constructions? That is to say, should lako "in order to" be one word? Hetia "now?" Tipai or hetipai "today?"
12) As of now, I judge that mo as an adverbializer (or maybe more accurately: similative particle) is still necessary. Its exact uses still need to be figured out.
13) How do you say "how?" It would be so simple to make it mo kea. We'll have to come back to this.
14) I've got a new comparative particle, so. This can also function as in Latin, so ta so mehe could either mean "he's more masculine/male/etc.," or "he's quite a man." The actual comparative construction uses o as "than," and at present can function with or without the comparative particle: ta so iso o ni or ta iso o ni "he's bigger than me." I may prescribe something eventually, or just leave them both as open options: probably better typologically, especially given the actual meaning of so. In that case, there could be a slight difference between ta iso o ni "he's bigger than me," and ta so iso o ni "he's a bunch bigger than me."
By the way, this also means that so is one of three new ways of saying "very." See below for the others.
15) I realized with happiness in the shower last week that we've got a ready-made affirmative particle to contrast with na in exactly the same syntactic positions: ia! This would function to emphasize/topicalize/whatever the truth value of the proposition, thus ni ia lu tule! "I am coming!"
16) Another new particle, but a "verbal" one this time: abilitative li, terminology borrowed from Turkish. My feeling at present is that this encodes innate ability, possibility and permission, but not knowledge (i.e. not umieć). So then, we can say ai se (ma) li nae? "can you see?" as in a movie theater. This also means that the translation of the English -able suffix becomes li pa in Koa: li pa nae "visible," na li pa suo "inedible," etc. (Or should it be nalipasúo, vis-à-vis #11 above?)
17) I alluded to this possibility in my original post on causativity, but I want to formalize that there are (at least) two strategies for this: the "synthetic" causative using the particle mu, and the "analytical" causative using mei "cause." The former denies more direct, possibly physical causation; the latter implies that the causation is more remote. For example: ni si mu mua ta "I killed him," with my bare hands or whatever, vs. ni si mei ko ta mua "I caused him to die," like, by accidentally knocking his heart attack medicine onto the floor.
Note that we are, of course, going to have a "kill" root other than mu mua. This is not Esperanto.
18) Way #2 of saying "very": poli. I think the translation of this on its own should be something like "great," "intense," etc., but I'll firm that up later. The point is that it can follow an adjective to intensify it: koa poli "very good."
Note that this doesn't mean "a lot, much." I don't have a word for this yet...hey, unless it's poli pi X. I guess that bears thinking about: ta si *toma poli pi sahi "he drank a lot of wine." Well, well, well. (Although: does poli need an article here? We need to map this out through various environments.)
Also, why the hell do we still not have a work for "drink?"
19) Reduplication: we've always known we wanted to use it for something, but never settled on exactly what; I think I've more or less defined the semantics I want at this point. With stative predicates it's intensity: pipo puna puna "very red butterfly"; with process verbs it's perseverative: ta ma talu talu "he keeps pushing, pushes on and on, etc."; with punctual verbs it's iterative, but we don't have any roots yet for this -- coughing, jumping, etc. I think that's a good, useful, typologically sensible semantic for this.
I still don't know, though, what reduplication of the pronouns does, if anything: nini, sese, tata, etc.
20) And speaking of pronouns, I have a provisional decision. For a little while there I was doing singular/plural for all persons (ni/nu, se/so, ta/tu), but I don't think that's the way to go. Instead, I've kept nu for "we," possibly exclusive, on the grounds that it really does mean something different than "I," along with seni/senu for dual/plural inclusive. Also possible are ponu "we, all of us (exclusive)," poseni "y'all and me" and posenu "y'all and us." This then becomes the strategy for the other persons: "y'all" is pose, and "they" is pota.
What's uncertain is whether I really want to maintain an inclusive/exclusive distinction (my gut says probably not in an IAL, though seni is still a useful thing to have), and also what happens when these longer forms are used with verbs: i or no i?
Also, if there's seni, why not tani? A matter for further thought. Maybe I'm not as sure about the above after all.
21) We still need inchoative and cessative particles. I'm just saying this out loud so we don't forget. The possibilities are dwindling, by the way: at the moment what we've got left is au eo hi ie (io) iu ki ku mi no oa oe oi pu su te tu ua ue ui. I don't want to recklessly use them all up since I'm sure a whole bunch of heretofore unexpected needs are going to come up when I start translating, but these two I'm pretty sure we're going to want.
22) We might also want a "must" particle -- what is that, obligatory? Gerundive? Anyway, that lets us have a *mupi *su pa nae "movie you've got to see," etc.
23) I've had this problem forever of what I was going to do once I started wanting to join clauses together. In the old days I had put aside the CVV roots for this purpose, thinking (for example) that noo could be "but," and all that sort of thing. My later wisdom thinks this is nuts, because the last thing an IAL needs is a vowel length distinction, especially with the ridiculously unnatural distribution restraint of occurring only in conjunctions.
Later on I started thinking that maybe I could do this with the normal particles, and I have a couple example sentences using no for "but" (not to mention e for clause-joining pretty much forever). On further reflection, though, it started seeming like I wasn't going to have enough material to work with, and conceived the idea of using roots instead: ala for "but," for example, which could also potentially have a nominal meaning of "objection," for example. This would be the final rejection of monosignificance, since these conjunctions wouldn't have particles around them to clearly identify their role in the sentence.
Now I'm really not sure. The fact is that I've (purposely) given so little thought to this kind of higher-level stuff that I really don't even know the range of functions I'm talking about. Some should be coordinating and some subordinating, probably, but which? I shouldn't just be using IE logic here. For example, e for "and" is clearly coordinating, but theoretically I could do the same thing with me ko, thereby nominalizing the following clause. Is this better? Worse? By what criteria should I even be judging? Otherwise how am I to decide between no and ala for "but," or even a "conjunctive phrase" like me ala ko... or something?
Here are some examples of conjunctions, assisted by my Intermediate Esperanto reader: before, after, while, if, and, but, or, because, whether, so that, in order to, although, as if. That's quite a jumble: we're definitely going to have to tease all these semantics apart before proceeding. This is, of course, begging the perhaps more important question, that of how clauses are going to be combined in general, since we can't assume it'll be as in IE languages. Clearly we'll be returning to this one.
24) There has been a serious omission in my basic structure: that of referential instances of general ideas. For example, a polo is "a runner," and ko polo is "running in general," but what about "a run?" Like, "that was a great run?" Or, "I recommend that run over this one?" How do we distinguish between "theft (i.e. stealing)" and "a theft (a particular event)?" Because they're not the same thing, and this is important!
The best I can come up with is doubled specifiers. Thus, if ko polo is "running," then ka ko polo is "the instance of running up on the stage right now," and I can translate "that was a great run" as something like tika i si ko polo so koa, or ti ko polo i si so koa, etc. I don't really see anything wrong with this, except that I've generally tried to avoid stacking specifiers for aesthetic reasons. Note that this means ko polo is sort of shorthand for ko ko polo which I assume would never appear.
25) I had this idea that ha "if" should operate at the predicate rather than clause level, both to extend its usefulness and reduce Koa's reliance on IE structures. As such, it could be called conditional or irrealis or something (though clearly not marking every irrealis concept). So we'd have se ha teke ta, pota lu tule "if you build it, they will come," etc. All kinds of stuff about TAM marking needs to be worked out for both the protasis and apodosis, and of the latter, whether it needs to be introduced by some kind of conjunction or what. But anyway, this lets us say things like ka ha loha ni "my would-be lover," which I think is neat.
26) This post has grown at least 200% beyond my original intention during the days I've spent composing it, as more and more thoughts crystallized in my brain. This is the point where I cut myself off for the moment, as it's getting kind of ridiculous. This last point, though, is one of potentially the hardest hitting of all of them, and one which will require the most soul searching.
Koa ought, it seems, to use serial verbs. It's consistent with the typology of a language with this kind of morphology, and enables the expression of complex ideas in a straightforward (and totally non-IE) way; and what's more, Koa is totally set up to use it right out of the box (and, to my delight, exactly like Bislama). For example, take the sentence "the frog killed the bad man with magic." Right now, this would go something like ka iki i si mu mua ka mehe pua me eme. Lots of isolating languages, though, use the "use" verb to do this sort of thing, and there's no reason Koa couldn't as well! Thus: ka iki i si mu mua ka mehe pua i *usa (k)a eme. Or, turning the emphasis around, ka iki i si *usa (k)a eme i mu mua ka mehe pua, "the frog used magic to kill the bad man." Without serialization, this would be something like ka iki i si *usa (k)a eme la ko mu mua ka mehe pua.
It's the same thing for lots and lots of concepts. We could use i ana instead of la for benefactive expressions. And beyond fossilized semi-prepositional usages, it would work great for chains of events, rather than assuming we should an IE narrative framework: why not the "he drive car hit buffalo die" kind of structure, rather than "he crashed his car into a buffalo and died?"
I don't think there's anything wrong with any of these. I suspect IE-language speakers would tend to regard serial verb constructions as unsophisticated, and speakers of languages making frequent use of serialization would consider our systems as needlessly and possibly incomprehensibly complex. I'd really like to make all of these ways available, and let people choose what works best; but I have this feeling that serialization is going to end up being the more frequent strategy for a variety of reasons, and I want to make sure I give it its due in my descriptions and in eventual didactic texts.
As such, I really need to do some serious research on serial verbs cross-linguistically. Maybe Johanna Nichols could direct me to some resources; the web is failing me completely at this point.
And that's all for today. This is a big month for Koa! I think this may be the most it's grown at one time since about the year 2002.
