Friday, February 2, 2018

Pronouns into predicates

I was just reading my Nahuatl grammar and serendipitously reached the chapter in which they introduce the emphatic pronouns and describe them as basically a slightly anomalous kind of predicate. Nahuatl being one of only a handful of languages I'm aware of that handle lexical classes like Koa, I feel like it's worth taking note of how it uses these kinds of forms.

From what I just skimmed and what I recall from previous readings, the main uses to which Nahuatl puts its emphatic pronouns are topic/focus constructions and, sort of formally overlappingly, when the pronoun needs to be used as a predicate. For example:

ca nèhuātl
DECL 1SG
"It's me" (pronoun as predicate)

(ca) nèhuātl in ni.qu.i in ātl
(DECL) 1SG DEF 1SG.3SG.drink DEF water
"I'm the one drinking the water" (focus...hopefully I got this right)

I'm wondering whether these could be directly calqued to Koa and what that would look like! Remaining agnostic on the form of Koa emphatic pronouns but using the reduplicated ones just for the moment, the first could be put as i nini or more likely nini sa and the second as nini sa ka ma ipo ka anu.

This isn't why I started writing just now, but I can't resist comparing the syntax of that last form (1A) with the usual way we handle focus (2A):

1A. nini sa ka ma ipo ka anu
1SG FOC DEF IMPF drink DEF water
"I'm the one drinking the water"

2A. ni sa ma ipo ka anu
1SG FOC IMPF drink DEF water
"It's me who's drinking the water"

I glossed these differently but I think they're semantically equivalent. I also think there's no reason type 1 above could logically be disallowed, which means we need to figure out (as usual) the kinds of conditions that would determine its use. Most worthy of note here, though, I think, is that what's going on between these seemingly extremely similar sentences is in fact surprisingly different formally.

In type 1, nini is actually the predicate: without focalization, the clause could be rearranged as

1B. ka ma ipo ka anu i nini
DEF IMPF drink DEF water FIN 1SG
"The one drinking the water is me."

Therefore, if we were to say that in Koa, as in Nahuatl, an emphatic pronoun is required in order for it to play a predicative role, then type 1 clauses would have to appear as above and not, for example, as *ni sa ka ma ipo ka anu.

In type 2, what's going on is simple movement of the focus into initial position followed by sa, without structural change from the theoretical matrix clause: that is to say, the pragmatically neutral clause would be

2B. ni ma ipo ka anu
1SG IMPF drink DEF water
"I'm drinking the water"

Um...wait just a second, though. I previously said that ni sa ka ma ipo ka anu would be incorrect, but check out these three sentences:

1B. ka ma ipo ka anu i nini
DEF IMPF drink DEF water FIN 1SG
"The one drinking the water is me."

1C. nini i ka ma ipo ka anu
1SG FIN DEF IMPF drink DEF water
"I am the one drinking the water."

1D. ni ka ma ipo ka anu
1SG DEF IMPF drink DEF water
"I am the one drinking the water."

If focalizing nini, 1B and 1C would both yield the identically same clause, nini sa ka ma ipo ka anu! And focalizing ni in 1D would give us ni sa ka ma ipo ka anu...which means it's not incorrect after all. However, though, what on earth is the pragmatic difference between 1A, 1B, 1C, 1D and 2A, five different ways of expressing (seemingly) the exact same thing? Just for clarity, with apologies for what I've just realized is the total lameness of the example sentence I'm using here, here's that lineup all together, all saying something like "I'm the one who's drinking the water":

1A. nini sa ka ma ipo ka anu
1B. ka ma ipo ka anu i nini
1C. nini i ka ma ipo ka anu
1D. ni ka ma ipo ka anu
2A. ni sa ma ipo ka anu

First of all, I have no clue how to differentiate pragmatically between 2A and 1ACD. 1B is, I think, slightly different from the others in a way that might make its use a little clearer. Okay, like...maybe the choice between all the 1's and 2A is whether "the one drinking the water" is already an identifiable entity on the discourse stage. In Nahuatl, of course, this kind of structural strategy is the only way they can pull off focus, but Koa can introduce this kind of subtlety because it also has fronting à la Yoruba. This can be filed in that general folder of Advanced Koa Pragmatics...as to which, whatever happened to that document where I was trying to list every possible way of expressing the same transitive clause so we could try to determine how they were different? I think I might have burned out after the 25th permutation.

ANYWAY, none of the above is what I intended to write about here! What I wanted to point out is that, if a primary purpose for emphatic pronouns is providing a form to use as a predicate where required, we do actually already have an entirely different and extremely well-established way of doing this with a different set of pronouns: ti/to/ke. Here we have

na ipo to sahi!
NEG drink that wine
"don't drink that wine!"

but, I now realize

na ipo to.a!
NEG drink that.PRON
"don't drink that!"

This raises two questions for me: (1) could/should emphatic personal pronouns be done like this as well, i.e. nia, sea, taa, nua, soa, tua? and (2) going the other direction, could/should ti/to/ke be used as pronouns independently, e.g.:

ke sa se halu?
what FOC 2SG want
"what do you want?"

na ipo to!
NEG drink that
"don't drink that!"

...alongside the traditional

ke.a sa se halu?
what.PRON FOC 2SG want
"what do you want?"

na ipo to.a!
NEG drink that.PRON
"don't drink that!"

I'm not sure. This is a pretty big potential change, so we need to take the time to make sure we're clear on what this really means. Both of these sets of particles can be used directly with a predicate without an article -- i.e. they essentially replace the article -- as in

ti tako
this octopus
"this octopus"

ke tako?
what octopus
"which octopus?"

ni tako
1SG octopus
"my [inalienable] octopus" (incidentally also "I am an octopus")

I'm really into this strange little conversation I've just accidentally created. But the point is that the two sets have different meanings when used in this way: the demonstratives have deictic force, whereas the pronouns are possessive. As such I'm not whether what the predicates in -a when applied to pronouns should actually mean: should nia be emphatic "I," or just "mine?"

Maybe a way to think about this is that the personal pronouns actually -- at least superficially -- have two entirely different meanings when prefixed to a predicate, as in ni tako above, so there needs to be a way to create a predicative form for each of those meanings. This is getting kind of crazy, but what if we had nia "mine" AND nini "I"?

Although...this makes me think that titi and toto (uh-oh) should also exist, meaning...um...what, exactly?

In summary, I've settled absolutely nothing, but these are some interesting questions...


Tuesday, January 30, 2018

Nominalized pronominal possessives rescued

In this post I made the suggestion that ka ni must mean "mine" (i.e. "my one") in the same way that ka ka kunu means "the dog's (one)." I was uncomfortable with this for reasons I still can't articulate but which remain emotionally compelling and suggested ka asi ni -- literally "my thing" -- as a possible substitute. It turns out that we actually do already have a solution that I had forgotten about by 2016 when I was typing up that entry!

First of all, I want to note that because ni is a particle, structures with it are not necessarily going to be the same as with predicates: ka ni doesn't actually have to be admissible. I think. We'll just accept this for now for the sake of argument.

Anyway, what we do in fact have that we can use instead is keme "attribute, possession," which gives us a neat keme ni for "mine." For example:

ti kunu i keme ni
this dog FIN possession 1SG
"this dog is mine"

ka keme ni i puna
DEF possession 1SG FIN red
"mine is red"

kunu keme ni
dog possession 1SG
"my dog (?)", "a dog which is mine, my property"

This does raise the question of what exactly the difference is between ka kunu ni and ka kunu keme ni, but I think I have an intuitive sense of it...in vague terms, the salience of the possession, or the relationship of possession, is being emphasized.

Similarly, though, what's the difference between ka ka kunu and ka keme ka kunu? Both seem to mean "the dog's one." Could they be entirely semantically equivalent, just different structural options that one might choose based on considerations of aesthetics or clarity?

WAIT.

I had already posted this but had to come back. I was just practicing vocabulary and came across nini, which I apparently had still never changed to nii for the reason that I never actually liked that decision, when it occurred to me: if we really did have predicate-form emphatic pronouns, which as I'm sitting here I'm beginning to feel overwhelmingly that we should, then those would be used in any structural situations in which a predicate would ordinarily be called for...like nominalized possessors. If we did this, we could have:

ka nini i puna
DEF 1SG FIN red
"mine is red"

kunu nini
dog possession 1SG
"my dog"

ti kunu i nini
this dog FIN possession 1SG
"this dog is mine"

Whoa. That last one really took me by surprise. If we did this, would that then mean the emphatic pronouns would actually sort of have possessive value? Or am I messing something up here? Let's run through this again using ka kunu instead of ni as the possessor.

ka ka kunu i puna
DEF DEF dog FIN red
"the dog's is red"

lelu ka kunu
toy DEF dog
"the dog's toy"

ti lelu i ka kunu
this toy FIN DEF dog
"this toy is the dog's"

Ah. So the issue is that, if nini just means "I," then I'm apparently allowing that one predicate to be used without particles! If we want to be consistent then we have the choice of either having nini mean "mine," or saying ka nini as the emphatic form rather than just nini: so like... Loha ka nini! "Love me!" ...i.e., "love the me one!" That is really incredibly weird and now I'm unsure again about the entire thing.

Obviously this is making a mess of my Indo-European intuition, but let's sit with it a bit longer. Maybe it's not as bizarre as it initially seemed...like, we could write it as one word, kaníni, which might help a bit. Plus I was just realizing, if we did this we could have words like ko nini that would be so wonderful for philosophical conversations, meaning something like "the quality/experience of being me."

Another option: perhaps these are predicates but with slightly different rules -- since they are pronouns, after all -- and "assume" the presence of ka unless a different particle precedes. I'm not sure how I feel about this yet.

Issue #17: I was just thinking we could also have mo nini "my way, à la me." But then, we use particles with the particulate (?!) form of pronouns all the the time, like tule me ni "come with me." Are we suggesting this should be tule me nini? Or that "my way" could equally be stated as mo ni? Why do these not seem equivalent to me?

My thoughts are thoroughly tangled up now so I'm going to just let this percolate for a while. One note in the mean time: if we do want full-form pronouns like this, maybe we could have tata mean "he/she/it (emphatic)" and let go of tata "dad" in favor of papa. I really really really really didn't want to do this in past years, but...maybe it's time.

Monday, January 22, 2018

Reduplication types

I've discussed (at least) four different types of reduplication in Koa over the years, but never written them down in one place or really thoroughly explored/documented what they should mean. Here are the types I've talked about so far:

1) Whole-word reduplication, normal stress on first element

kuma > kúmakuma
loe > lóeloe
ake > ákeake

Meaning: intensifier (very hot, very cold, very sharp)
Notes: We talked about this here near the bottom, only written as two separate words without clear intentions about stress...though I have to admit that when I read Ni loha loha se aloud I stress the second. Hm.

2) Reduplication minus onset, stress on second element

kuma > kumaúma
loe > loeóe
ake > akeláke (note L-insertion)

Meaning: affective/minimizing diminutive, "ish." (warmish, chilly, ...pointy? dunno). Minaína "gal."
Questions: (1) How is loeóe different from lóeki? (2) Is this too similar to the first type given the fact that they kind of have the opposite meaning? (3) Is that L-insertion -- the only thing saving the opposite of ákeake from basically and laughably being akeáke -- too ad-hoc?

3) First syllable, stress on full word

kuma > kukúma
loe > lolóe
ake > aáke (aháke?)

Meaning: unknown

4) Reduplicate medial consonant with -a, stress on reduplicand

kuma > kumáma
loe > loáe
ake > akáke

Meaning: From here, "I was thinking this might serve to make the noun more euphemistic, more gentle, less objectionable, or something. So a pragmatic rather than semantic value."

Just for reference, here are the three sample words with each reduplication type in turn for comparison:

kúmakuma, kumaúma, kukúma, kumáma
lóeloe, loeóe, lolóe, loáe
ákeake, akeláke, aháke, akáke

This is going to require some thought, and maybe some study of what reduplication tends to be used for cross-linguistically. Just impressionistically, though:

* I don't like type 1 at all; as I alluded to before, my instinct is that these should sort of be separate words with the stress on the second one. Then, maybe weirdly, I feel like type 3 might be synonymous with type 1.

* If type 1 becomes separate words with stress on the second as I just suggested, type 2 kind of gets sabotaged. I always liked this one, though! Loeóe for "chilly" seems perfect...but on the other hand maybe lóeki really is enough here.

* Type 4 is really great for certain things (like luái) and I think it's valuable for a living language, though at the moment I'm not thinking of a lot of specific uses for it other than euphemisms. Could it also take on the meanings of type 2? I'm not sure. Thinking just of getting it on, I feel like there's an important difference between lúiki and luái!

Another thought I'm having here is that, just because a particular reduplication type technically has to be available for every word doesn't necessarily mean that it's going to be equally common with every word...so for example if mina mina means "a real woman" or whatever, and minaína means "gal," that doesn't have to entail that that type-2 strategy is now going to be used ALL the time and confuse everything.

Tuesday, December 6, 2016

Experiencer vs agent in verbs of perception

Just a note: I had said in a previous entry that the difference between "I see" and "I look at" is essentially aspectual, i.e. ni nae vs ni ma nae. I've realized since that the core semantics of a statement like "I see" require some reanalysis of the corresponding Koa structures.

A primary thing that makes "I see you" and "I'm looking at you" different is agency, of course, and therefore the semantic role of the subject. Koa doesn't mark agents and experiencers differently, but one could easily reframe this in terms of ability, which Koa does mark. In other words, "I see you" has approximately the same semantic content as "I can see you." The emphasis may be different, but that's the core idea.

I would now therefore list the basic and embellished possibilities for this phrase like this:

ni nae se
1SG see 2SG
"I see you" or "I looked at you" or (weakly) "I can see you"

ni ma nae se
1SG IMPF see 2SG
"I am/was looking at you"

ni te nae se
1SG ABIL see 2SG
"I can see you" or just emphasizing the experiencer interpretation "I see you"

ni voi nae se
1SG can see 2SG
"I can see you," emphasizing the abilitative interpretation

And to clarify, this isn't actually a change: it's just me getting clear on what my Koa words must actually mean.

On a somewhat different topic, I just want to mention that I made an error in that same entry with my glosses involving ipo sahi. I failed to notice that, because the object is incorporated in these VP's, they've just become intransitive and more like states in terms of Aktionsart category. Therefore:

ni na ipo sahi
1SG NEG drink wine
trad "I don't drink wine"
wrong "I didn't drink wine" or "I'm not drinking wine," depending
right "I don't drink wine" or "I am not a wine drinker"

ni na si ipo sahi
1SG NEG ANT drink wine
trad "I didn't drink wine"
wrong "I hadn't drunk wine"
right "I didn't drink wine" or "I wasn't a wine drinker"

To get the incorrect interpretations from above, you would need an article on "wine," thus:

ni na ipo hu sahi
1SG NEG drink EXIST wine
"I didn't drink (any) wine"

ni na si ipo hu sahi
1SG NEG ANT drink EXIST wine
"I hadn't drunk (any) wine"

Thursday, August 11, 2016

Comparison again

I'd honestly really like a comparative particle. I'm just not pleased with taha despite the typological motivation:

iso taha
big surpass
"bigger"

iso i taha ni
big FIN surpass 1SG
"bigger than me"

What about:

iu iso
COMP big
"bigger"

iu iso ____ ni
COMP big [THAN] 1SG
"bigger than me"

The question is what kind of mechanism for "than" can be derived from first principles. Iu iso o ni looks like an IE calque and is an arbitrary convention that has to be memorized: not what we want, even though (or especially because) it's familiar to me as an IE person.

iu iso taha ni
COMP big surpass 1SG
"bigger than me"

What if iu means "to such a degree?" iu iso taha, iu iso kase -- "big to such a degree it surpasses/meets" > "bigger, as big"

Do we have a way of saying "less?" Hm.

ai se halu ___ pi lele?
Q 2SG want [more] QUANT milk
"do you want more milk?"

Oh my...should there actually be a "more" word? That certainly shouldn't be taha in a sentence like that. Vese pi lele -- much better than taha pi lele. Quantity and comparison are not the same thing, clearly don't need the same structures.

asi iu vaka
asi vaka taha
asi vaha i taha
asi iu vaha i taha
"a more important matter" *sigh*

I don't know what to do here, but more work is needed.

Thursday, October 30, 2014

Discourse deixis

Ti/to for discourse deixis: ti is cataphoric (subsequent discourse), to is anaphoric (prior discourse).

[Neat, though I don't know where this came from...maybe Nahuatl?]

Monday, September 22, 2014

Exploring the possible scope of quoting

I need to look up a bunch of Bislama examples to do this properly, but basically, the idea is that le could expand its usefulness beyond just marking proper names. I think we can generalize to say that it could be used to mark anything that’s being mentioned rather than used. So maybe

kea sa le iuna?
"what's a train?"

But more than that, it could potentially also be used as an alternate strategy to mark clauses that are currently just being nominalized. So

ta sano le ni loha le Keoni
"she said that she loves John"
literally "she said, 'I love John'"

Apparently the above is (amazingly) the normal strategy for this kind of sentence in Quechua. Now I’m getting confused, though, because I was starting to think of it like Bislama se, in which case the sentence above should appear as

ta sano le ta loha le Keoni
i.e. hem i talem se hem i lavem John
“She said that she loves John”

More work is needed here. Probably for this usage, either ko or ve ought to be adequate, since the meaning would be identical [and the difference between ko and ve in this usage is...what?]. If we really are mentioning rather than using, I suspect direct quotation is the use we most need.

One final thought: trying to translate “mean,” as in “what does that mean?” Which of these is correct?

1. Kea sa le toa?
2. Le kea sa toa?

Ugh...I’m having trouble mustering an argument either way. Let’s think about it.