Monday, December 9, 2002

First thoughts on syntax

I see the man = ni nae ka mehe

The man is seen = ka mehe pa nae

I give a cucumber = ni ana sihi

I give the cucumber to the man = ni ana ka sihi ka mehe

The cucumber is given (away) = Ka sihi pa ana

The man is given the cucumber = Ka mehe lo ana ka sihi

The man is given the cucumber by the dog = Ka mehe lo ana ka sihi ka keli


I saw the man to whom the dog gave a cucumber =

     Ni si nae ka mehe lo ana sihi ka keli. or

     Ni si nae ka mehe o ka keli ana sihi ti.


I saw the man who was given to the dog =

      Ni si nae ka mehe pa ana ka keli. or

      Ni si nae ka mehe o i ana ti ka keli.


Note in this last example that in phrase type (1) we have an original sentence:

X i ana ka mehe ka keli, inverted with "pa" to:

Ka mehe i pa ana X ka keli.

The question is how we tell whether "ka keli" here is the agent or the beneficiary! We need rules or something about what can be deleted.


We need the indefinite article. "u"?:

      Ni nae u mehe "I see some unspecified man", as opposed to

      Ni nae mehe "I see men in general," "I men-see."

In this example, though, we see that there needs to be a better definition of what the Y in the XY verbal position actually means--if "ni kulo lika" means "I milk-drink," meaning drink milk partitively, what is actually the difference between this and the "u" article? It's possible that the articled form would be preferable because it both clearly marks the noun and leaves that position open for adverbial usage: "Kaca i colo koa" = "Karen runs well." Not sure what'd be the best here.


I hopi a colo = He likes running

I hopi se colo koa = He likes that you run well, he likes your running well


An important question is whether, as in Loglan, using the passivity prefixes actually switches the arguments (which would be nice because it would be possible to specify the agent of a "passive" verb) or simply removes the original agent/subject, so that passivity is not allowed simply for emphasis; of course, in "relative clauses" this switching is highly desirable as there is no other way to classify the syntactic role of the head noun as anything other than subject.


My dog likes cucumbers = Ka keli o ni i hopi (u) sihi / O ni ka keli i hopi (u) sihi


Is a "counting word" necessary? Perhaps so, to disambiguate what would otherwise be long strings of perhaps unidentifiable monosyllables:


I see two dogs = Ni nae he (pi) keli

I see 21 dogs = Ni nae he-ca (pi) keli

I see 215 dogs = Ni nae he-ca-pu (pi) keli


Yes, definitely--without that marker the number morphemes are way too hard to isolate.


Are the correlatives to be classed as individual lexemes? That is, besides "ke a" = "which thing?" should there be "kea" = "grasshopper"?


Tia i kea? = What is that? or Is that a grasshopper?

Kea sa tia? = What is that? or Is that a grasshopper?


I think that pretty much resolves the issue. :) Correlatives must be set aside as separate words as soon as we define the morphemes.


Real question: what is the difference between these sentences?:


1. ka cumo koa "the good cucumber"

2. ka cumo o koa "the cucumber of goodness" / "the good one's cucumber"

3. ka cumo o a koa "the cucumber of goodness"

4. ka cumo o ka koa "the good one's cucumber"

5. ka cumo o u koa "a good one's cucumber"


Clearly (2) needs to be better defined. Maybe it's a shortening of (4)? We'll think about it.


What did you see = Se si nae kea? Kea sa se si nae?

Whom did you see = Se si nae keka? Keka sa se si nae?


TOPICALIZATION


I gave the man the squash = Ni sa si ana ka cumo ka mehe

It was the squash that I gave to the man =

      Ka cumo sa ni si ana ka mehe (?), Ka cumo sa i si pa ana ni ka mehe

It was the man that I gave the squash to =

      Ka mehe sa i si lo ana ni ka cumo.

I gave the squash to the man = *Ana sa ni ka cumo ka mehe (???)


VERB AFFIXATION


subject pronoun negation tense passivity locative stem question adverb arguments

Kaca i na si lo ana ko ka sihi? = Karen wasn't given the cucumber?

Friday, November 29, 2002

In the beginning

Okay, so here's the question: I've been feeling kinda sorta quite excited about the prospect of Koa once again, since a few instances have come up recently where it would have been really helpful for Karen and me to have our own language. But, I have that ubiquitous problem of just not knowing where the hell to go from the phonology—I don't want Japanese, I don't want Loglan, I don't want English; I have this idea that what I really want is Hawaiian, but I don't think that's actually the case. Hawaiian just happens to do very well with a very limited phonology, which is exactly what I'm going for.


Well, here's what we've got so far:


Ten consonant phonemes, {p t k s c h m n l ?}

Five vowel phonemes, {a E i O u}


/S/ can be realised as [S] or [tS] or whatever. /?/ > Ø / V_V



Syllable structure is uniformly CV. All root words are CVCV according to the oldest plan; newer plans propose a stressed CV addition that either carries no semantic content but produces fifty additional forms for every original root, or that has some kind of defined meaning (derivational). The problem with the second option is that one then has to decide which words are "worthy" of plain roots, and which ones need a prefix...and then, also, running out of prefixes and words to attach them to. That is: if well is "water-hole" or something, what do you call an actual hole in the ground full of water, such as what a gazelle might drink out of?


And then I have the somewhat conflicting needs of making this something Karen can learn easily, something I find elegant and interesting, and something I can reasonable put forth as an IAL (since that *is* why I came up with this in the first place). I think *some* logic to the prefix system is probably a good idea; it would help intelligibility, though, if some "basic" words were also formed with prefixes for more diversity in word shape. A principle of compounding needs to be invented to cope with the "water hole" problem.


But! But! What if you want to add two prefixes? That would make it look like a noun, which could mean something entirely different. I guess this means that the prefix-adding is non-productive—only compounding is available as a means for new-word formation. Example (random words):


hole: kosi

great,  big: pu

water, drink: li

oil: puli


well: líkosi

ravine: púkosi

reservoir/oil well: puli kosi


This is always the problem with oligosynthetic languages—the number of available "units" for a meaning and the number of words needing a particular "unit" are not equally distributed. I suppose I could do it like that Nahuatl example Dirk Elzinga posted, although that was more of a joke than anything else...but really this is silly. I want something entirely isolating; why have such involved derivation?


Word order: SVO

Topicalisation: As in Yoruba—thus:


I see the well: Ni táse líkosi

It is I who see the well: Ni ha táse líkosi

It is the well that I see: Líkosi ha ni táse


I like the way Yoruba uses serial verbs and compounds "have," so "to be beautiful" is "to have beauty," and "bring the book" is "grab the book come."


I see that it's far more practical not to do all that case stuff—and indeed, with rigidish word order, why use it anyway, especially in an IAL? So then.


As an aside, I'm extremely uncomfortable and my stomach hurts. I think I need to eat. Yes. Will be back.