Monday, March 15, 2021

Return of the Pronominal Predicate

Over the years there's been a lot of hemming and hawing about predicate-length (i.e. bisyllabic) versions of the personal pronouns, which I've tended to refer to as "emphatic." I'm not sure where the idea came from originally; maybe the full versus clitic forms of the pronouns in Polish planted the seed? It occurs to me to wonder how well-represented a strategy this is cross-linguistically. The other clear examples I can think of come from Welsh (i vs innau), Ancient Greek (με vs ἐμέ) and Nahuatl ( vs nèhuātl): not much ground for generalizations.

Since the idea first occurred to me some version of it has never left at least the theoretical lexicon, despite occasional deep skepticism. Most recently, in this post -- since which more than two years inexplicably seem to have passed -- I think I pretty firmly established that there does need to be a form of the personal pronouns that makes it possible to use them as predicates. It's important to note that as such they may have nothing whatsoever to do with emphasis, though, except for the fact that these predicates have a tendency to show up in pragmatically marked syntactic structures. For this reason I'm thinking we should ditch the "emphatic" nomenclature that probably has its origins in IE languages anyway, and choose a better-motivated term like "extended" or "predicative."

As to form, we've seen three options:

A. Reduplicating the pronoun (the longest standing): nini, sese, etc.
B. Lengthening the pronoun to produce a bisyllable of analogous form: nii, sei, etc.
C. Adding -a on the analogy of ke > kea, to > toa, and so on.: nia, sea, etc.

I've thought of a couple others recently:

D. Adding -imi "self," currently used also as a true emphatic: niimi, seimi, etc.
E. Adding some other formative, like -pe: nipe, sepe, etc.

Option A was the original idea long ago and I'm still a big fan, other than the fact that this would necessitate giving up tata "dad" in favor of papa which I just do not like. Aside: this has been a bigger problem than you might think, because I really truly actually seriously detest papa and this has held up the whole thought process for probably almost 15 years. I honestly don't know what to do about this, because I apparently can't let go of either and I obviously can't have both. Could there be some completely different "dad" word? I feel like I might die on this hill while the whole of civilization grinds to a halt around me.

ANYWAY, back to the subject: I've never much cared for the aesthetics of option B, or the fact that the pronouns with mid vowels end up looking less like their simplex forms than the others, so I think we might as well just toss that one. Option C is totally reasonable and logical, though let's discuss the semantic implications below, and I'd just like to take a moment to bemoan the loss of taa for "surpass," as I've increasingly been thinking of it. Option D I apparently brought up just to instantly dismiss, because using a form with this literal semantic meaning in this pragmatic way feels very ad hoc and un-Koa. Option E is possible if we really need it, so maybe let's leave it in the back pocket for the moment.

As to option C: the idea here is that we already use -a to make pronouns out of specifiers, which initially makes this seem like a pretty solid plan. An important thing to notice, though, is that the Koa "personal particles" actually live a double life -- they do act as pronouns, but in other contexts they're specifiers! And as specifiers their value is possessive. If, then, -a makes a specifier into a pronoun -- or to look at it another way, into a predicate -- then nia should mean not "I" but "mine!"

I actually don't have a ready argument for why this shouldn't immediately be declared canon. A structure like this gives us possibilities like:

ti talo i nia
this house FIN mine
"this house is mine"

ka nia i pavasu
DEF mine FIN PASS-wear.out
"mine is worn out"

Currently, without these forms, we're forced to co-opt other words in what is unavoidably an arbitrary way, or repeat the coindexed referents redundantly:

ti talo i keme ni
this house FIN attribute/possession 1SG
"this house is my possession"

ka keme ni i pavasu
DEF possession 1SG FIN PASS-wear.out
"[this possession of] mine is worn out"

ti talo i ni talo
this house FIN 1SG house
"this house is my house"

Okay, so honestly, to my surprise I don't hate the structures with keme ni. I actually don't think they're arbitrary at all: this seems like a completely logical, consistent extension, and maybe even the most consistent way of expressing this concept. The real question, then, I guess, ends up coming down to aesthetics and word-worthiness. Before digging in here, though, note what happens when we use predicates like nia in an adjectival position:

ka lina nia
DEF city mine
"my city"

This gives us a genuinely emphatic meaning, as against the unmarked ka lina ni "my city." Important! And without these forms we'd be limited to

ka lina keme ni
DEF city possession 1SG
"the city of my possession" = "my city"

There's no question that the form with nia is more elegant. What I'm not sure of is (A) whether giving up 6 predicates to this cause is being unnecessarily improvident, and (B) what's the cross-linguistic word on possessive pronouns? Would these predicates be typologically motivated?

Indo-European has a variety of strategies, often the genitive of the personal pronoun or a separate possessive form. Modern Greek uses an adjective meaning "own," e.g. δικός μου "my own," "mine," which is kind of cool and a little like keme ni (and as to that, does Koa have a way of expressing that meaning of "own?" Oh yes of course, oma). Finnish and Turkish both use the genitive of the pronoun: tämä talo on minun, bu ev benim. Hungarian interestingly has distinct possessive pronouns, which I'd forgotten about: ez a ház az enyém, etc. Basque -- if I'm interpreting this forms correctly -- seems to put a definite ending on a genitive pronoun, so ni > nire > nirea...fascinating, a bit like el mío in Spanish, I guess? That about wraps it up for Europe, I think.

For a minute I got excited about Hebrew because the modern vernacular language seemed to use a structure 100% analogous to ka talo keme ni with its הבית שלי ha bait sheli but then I learned from Marisa that sheli is just the 1SG form of the preposition "of" and I was disappointed. But then I discovered that shel comes originally from something like "...which is to me," "to me" being the way Arabic still expresses "mine": هذا البيت لي hadha al-beyt li "this is my house," lit. "this the house to me." So a prepositional phrase -- that's a cool different strategy. (Thanks, Liorr, for these translations!)

Mandarin just uses the genitive of the pronoun again: 这房子是我 的 zhè fángzi shì wǒ de "this house COP 1SG GEN." Japanese says "my thing": この家は私のものです kono ka wa watashi no mono desu "this house TOPIC 1SG GEN thing COP." Interesting. Malay has "my possession"!!! rumah ini milik saya, "house this possession me"; apparently milik can also be a verb "belong."

Most of the rest of human languagedom is still untouched, but maybe this is enough. Clearly lots of very unrelated languages are happy with "my X," and I don't think I want discrete lexemes for possessive pronouns anyway: it just doesn't feel right. I'm intrigued by the semantics at the perhaps unlikely intersection of Greek and Malay, though, and I wonder whether we might use oma in this context:

ka talo ni oma
DEF house 1SG own
"my own house"

...wait, or is it

ka talo oma ni
DEF house own 1SG
"my own house"

Oh yikes. Let's not even get into the fact that the official lexicon gives oma for "self," when I thought it was imi! Okay, okay, no, actually we'd better. Deep breath: oma could be "one's own," whereas imi could be "the self" in, like, consciousness terms. So ni imi oma "my own self." Apparently reflexive pronouns do tend to come from the word for the self, so it's also not arbitrary to potentially say things like ni nae ni imi "I see myself."

That now being settled, which version above is correct? If we accept ni imi oma that would suggest that oma is applied to an existing possessive phrase, in which case it should be the first. Then:

ka ni oma
DEF 1SG own
"my own"

...and potentially

ti talo i ni oma
this house FIN 1SG own
"this house is mine"

Ugh, though, why is an alienable object ("house") being described with inalienable possession, then? Either we say that that's just how oma rolls, which would be realistic to natural languages but not exactly in line with Koa's goals of being reducible to first principles, or maybe the noun modified by oma "groups" together before possession happens? So ni [imi oma] "myself," obviously inalienable, but then [ka talo oma] ni "my own house," alienable. And since utterances like "this mother is mine" are pretty pragmatically anomalous and therefore unlikely outside of bad reference grammars, oma ni is a better form to be the default predicate.

WHICH MEANS that either of the following could mean "this house is mine":

ti talo i oma ni
this house FIN own 1SG
"this house is my own" = "this house is mine"

ti talo i keme ni
this house FIN possession 1SG
"this house is my possession" = "this house is mine"

However, there's a distinct semantic in the oma sentence worth pointing out: it's essentially saying "this is my very own house." In other words, there's no way besides ti talo i oma ni to say "this house is my very own." That's worth being able to say, but not really what we were trying to get at with this now absurdly long odyssey into the subject. Ti talo i keme ni seems much more neutral: this is just the house that I have.

This was extremely productive -- although it left me with the very same conclusions I'd previously reached here, albeit with much less exhaustive motivation -- but not even slightly the matter I actually meant to spend the apparently last 24 hours exploring! Will try again tomorrow...

No comments: