Re: Participles in ergative languages

On Sun, 15 Oct 2006 22:47:24 -0500, Eric Christopherson <rakko@...>
wrote:
>I'm wondering if anyone has some insight into how participles are
>used in ergative languages that have participles. (I am working on
>making a daughter language of an ergative protoconlang created by
>someone else.) I have a few issues with understanding them:
>
>1. Voice. AFAIK, in accusative languages there are different
>participles for different voices, e.g. active and passive -- thus
>"loving" is an active participle, because it describes someone who is
>an agent of loving; and "loved" is a passive participle, since it
>describes someone who is a patient of loving. Do ergative languages
>have e.g. an active participle, an antipassive participle, etc.?
>(Actually, I'm not even sure if it is correct to call the usual voice
>"active".) Do they have a passive participle, even if the actual
>verbs don't recognize a passive category (which, AFAIK, some ergative
>languages don't)?
The important distinction is whether the participle denotes the agent of
the verb or denotes the patient of the verb.  While this reasonably could
be labeled a distinction of voice, I don't see why the voice-labels applied
to participles necessarily have to be exactly the same as the voice-labels
applied to finite verbs in the same language.
If you use "agentive participles" and "patientive participles" instead
of "active participles" and "passive participles" (or "antipassive
participles" and "absolutive(?) participles"), you'll avoid any confusion.
That's only one idea, but it's the best one I've come up with so far.

>2. If I want to say "I eat an apple", I would put "I" in the ergative
>and "apple" in the absolutive. But what do I do if I want to say "I
>am eating an apple", using a participle?
Isn't this a distinction of aspect?
Why must such distinctions involve participles?
In English "to be" + "participle" gives progressive aspect (a sub-aspect of
imperfective aspect),
and "to have" + "participle" gives perfect (which could be considered a sub-
aspect of perfective aspect).
But in some languages  -- for instance, if I am not mistaken, some Slavic
languages? -- aspect is purely a morphological feature of the verb, and is
not periphrastically constructed via "light verb" + "participle or other
non-finite verb-form".

If your language has morphological progressives rather than analytic or
periphrastic ones, you'd likely want to keep the agent ergative and the
patient absolutive.  Of course if that's how your language handles
progressives, the question of progressives wouldn't involve the question of
participles.

>On the one hand, I can imagine using those same cases, but on the other
>hand, it seems like you could conceive of the whole phrase "eating an
>apple" as adjectival,
I think I see what you mean.

>and I think ergative languages generally use the absolutive for the
>subject of an adjectival predicate. I.e. both "I" and "apple" would be in
>the absolutive
>(unless of course the participle would require its objects to be specified
>in an oblique case, like the genitive).
That solution appeals to me.

>My intuition is that the choice of cases would depend on how
>grammaticalized the copula+participle construction is -- whether it's
>considered a periphrastic verb or simply the conjunction of a copula and a
>participle, parallel to conjunctions of copula and adjective or noun.
Obviously I think that would at least have something to do with it.
In the former case (copula+participle = periphrastic verb) you'd want the
agent "I" ergative and the patient "apple" absolutive.
In the latter case (copula+participle = copula + adjective) you'd want the
agent "I" absolutive; and I would recommend the patient "apple" be in some
third case.
A choice you didn't mention is "progressive is not expressed as
copula+participle".

>=========================================================================
Interesting question!

Thanks.

-----
eldin

Replies