Some, all and none

I feel a little grumpy about this post: this is one of those embarrassing core areas that I've somehow managed to sneak by without ever actually fully working out. Even now I'm not positive that some of the structures I'm about to document are quite right, though I also can't quite think what I could feasibly replace them with. The question is how we express ideas like "some," "all" and "none."

To begin with, we have the particles hu "∃" and po "∀". The former in particular has been awkward to understand how to use and to explain because Koa is not, in fact, a logical language like Loglan; they needed to be pragmatically useful to humans, not just computers.

A post about hu is forthcoming so I'll set the complicated history aside for now. Hu essentially identifies a referent which exists in the world, but which is not identifiable within the discourse; sometimes that can be translated as "some," but often the better English translation uses the existential verb: "there is..." Po, on the other hand, means "all" in the sense that it identifies an entire category, but does not imply any internal structure (i.e. individuals) within that category. Thus the simplest of sentences with these quantifiers look like this:

hu nosu i lusu
EXIST elephant MAIN.CL gray
"there are some elephants that are gray; there are gray elephants"

ni-loha hu nosu
1SG-love EXIST elephant
"there are some elephants I love; I love some elephants"

po nosu i lusu
UNIV elephant MAIN.CL gray
"elephants are gray"

ni-loha po nosu
1SG-love UNIV elephant
"I love elephants"

Hu and po can be combined with the specifiers ka "the" and a "a" to form pronouns: hua "(there is) something," huka "(there is) someone, some of them," poa "everything," poka "everyone, all of them." These words have existed primarily in that usage -- pronouns -- through almost the entirety of Koa's history. In extremely early Koa, though -- like 2002 early -- it was clear to me that these could also be preposed to predicates to indicate "some" and "all," but in later years this certainty quickly became muddled.

The quantifiers kind of look and behave in some ways like specifiers, and since specifiers can't be stacked before predicates anywhere else in the language, these started to feel ungrammatical. For example, ti + a = tia "this" as a pronoun, but we can't say *tia nosu "this elephant": the correct form is ti nosu. Poa and friends, though they look parallel and I've even called them "correlatives" in unguarded moments, are actually made of a quantifier and a specifier; their syntax is quite different. As such, we can say:

hu-a nosu i lusu
some-INDEF elephant MAIN.CL gray
"some elephants are gray"

hu-ka nosu i lusu
some-DEF elephant MAIN.CL gray
"some of the elephants are gray"

po-a nosu i lusu
all-INDEF elephant MAIN.CL gray
"all elephants are gray"

po-ka nosu i lusu
all-DEF elephant MAIN.CL gray
"all of the elephants are gray"

Of course, there is no logical distinction between e.g. "elephants are gray" and "all elephants are gray," but there are clearly important pragmatic differences; the same is true of "there are some gray elephants" and "some elephants are gray." The nature of these distinctions would be good for me to try to spell out at some point, but I will need to be at my sharpest and tonight is not the time for that attempt!

The above structures with hu can also be negated (nahu, nahua, nahuka), in which case they come to mean "no" or "none"; on their own, of course, nahua and nahuka are pronous, "nothing" and "no one/none of them."

na-hu nosu i lusu
NEG-EXIST elephant MAIN.CL gray
"there are no elephants that are gray"

na-hu-a nosu i lusu
NEG-EXIST-INDEF elephant MAIN.CL gray
"no elephants are gray"

na-hu-ka nosu i lusu
NEG-EXIST-DEF elephant MAIN.CL gray
"none of the elephants are gray"

Aside: If you're holding your breath in hopes I'm about to issue a judgment on whether negation is doubled in clauses like "I don't love any elephants," I'm afraid today will be disappointing on that point too. Just to lay the question out there again, should the Koa translation of this be...

ni-loha na-hu-a nosu
1SG-love NEG-EXIST-INDEF elephant

...or...

ni-na-loha hu-a nosu
1SG-NEG-love EXIST-INDEF elephant

...or yet again...

ni-na-loha na-hu-a nosu
1SG-NEG-love NEG-EXIST-INDEF elephant

Irritatingly, I have not the slightest idea how to choose between Slavic and Germanic logic on this question, and continue to hedge by allowing both. Someday I'll need find some criteria by which to make a decision.

Anyway, back to quantification, there are a few other structures available within this general category worth mentioning here. With nai "some/rather," we can say

nái-pi nosu i lusu
some-QUANT elephant MAIN.CL gray
"some elephants are gray"

nái-pi ka nosu i lusu
some-QUANT DEF elephant MAIN.CL gray
"some of the elphants are gray"

If there's any real distinction between hua nosu and náipi for "some elephants," I think it might be that the latter would be drawing attention to the fact that the number is indefinite or unknown, rather that to the truth value of the overall proposition. In practice, I wonder if it might often be a question of register?

With visi "each, every," we have

po nosu visi i lusu
UNIV elephant every MAIN.CL gray
"every elephant is gray"

ka nosu visi i lusu
DEF elephant every MAIN.CL gray
"every one / each of the elephants is gray"

I've asked myself whether vísipi could possibly be a synonym for poa in the way that náipi is slightly synonymous with hua, e.g.

?vísi-pi nosu i lusu
every-QUANT elephant MAIN.CL gray
"all elephants are gray"

...and I'm not sure if that feels more marginal because it's new, or because it's wrong. Theoretically it could emphasize number like nai, as in "every single elephant..." That might work! I'll let it sit.

Lastly, we have tele "several," as in

téle-pi nosu i lusu
several-QUANT elephant MAIN.CL gray
"several elephants are gray"

or alternatively

hu nosu tele i lusu
EXIST elephant several MAIN.CL gray
"several elephants are gray; there are several gray elephants"

Coming soon will be that post about hu, and particularly the difference between hu and a! I'm hoping I can finally, once and for all, set that matter to rest.