What to Do, Part Two
What to Do, Part Two
In honor of Koa Day 2026, a better thought-out continuation of a post from about five years ago: how does one say things like “I don’t know what to do” in Koa?
The whole “what to X” structure does not lend itself easily to translation via any established strategy. It is easy enough to say “what I do/did”:
ka ni ete
SPEC 1SG do
“what I do/did”
Using this structure, the closest we can get to an exact rendering of the target phrase would be something like
ni na ilo ka ni vi ete
1SG NEG know SPEC 1SG OPT do
lit. “I don’t know what I am to do”
…but this is not at all the right vibe for most situations in which the English phrase would be uttered. It’s too wordy, too syntactically heavy. We also cannot say
*ni na ilo kea sa ete
1SG NEG know what FOC do
If this meant anything at all, it would have to be something like “I don’t know what did (it),” because the absence of a first argument in the embedded clause causes ka to refer to that argument role:
keaα sa _α ete?
whatα FOC _α do
“what did (it)?”
The reality is that Koa just doesn’t seem to hang with this exact embedded-infinitival-interrogative-clause concept, as simple as it is in IE languages. We need to go back to the drawing board and consider the set of emotions and considerations that might animate the individual saying this phrase.
On reflection, the literal words “I don’t know” are not necessarily needed to get at this. The form of a question is much more straightforward, and gives us a key to expressing this kind of idea in Koa:
ete kea?
do what
“what to do?” lit. “do what?”
This is a shortening of i ete kea, with an omitted first argument deemed recoverable in the discourse context. We have to phrase it this way because, as above, the more usual kea sa ete structure would focus the wrong frame role. Ete kea?, with a less focused interrogative, is actually pragmatically appropriate to the sense here anyway.† We can dress it up with modals:
cu ete kea?
IRR do what
“what will/would I do?” lit. “would do what?”
vi ete kea?
OPT do what
“what am I to do?”
On the other hand, if we really do feel we need an “I don’t know” in there, there’s no reason it can’t be included…but as a separate stance-framing clause, rather than a matrix into which the following clause is syntactically embedded:
ni na ilo, ete kea?
1SG NEG know, do what
lit. “I don’t know, what to do?”
Or ni na ilo, cu ete kea?, ni na ilo, vi ete kea?, etc. This much is completely uncontroversial, solid and satisfying. I will not be able to leave Koa’s 27th birthday, however, without muddying the waters with some wild speculation…
If, as we suggested earlier, we can say things like this via a serial nucleus construction:
ni halu i mene
1SG want SN go
“I want to go”
…then might ilo i ete also potentially be understood in this same way: something like “know to do”? In which case, in addition to the above, we might see
ni na ilo i ete kea
1SG NEG know SN do what
“I don’t know what to do”
lit. “I don’t know to do what”
A small difference in intonation, a big difference in syntactic structure, and zero difference in actual meaning. …but no, tantalizing though it may be, actually that doesn’t work at all anyway. I confused the syntax: if anything, it would have to mean the same thing as kea sa ni na ilo i ete?, something like “what do I not know I should do?” Oops.
So we can leave the waters unmuddied, and keep the comma after all: ni na ilo, ete kea? Fabulous!
Pai le koa iolo, poka! I believe adjunctives are coming up next.
†Bookmarking for further discussion later: the pragmatics of in-situ interrogatives (se sano kea? lit. “you said what?”) as against the standard focused structure (kea sa se sano? “what did you say?”). I have an intuitive sense of this but it needs to be better defined.
