Re: C'ali update: Split-S cross-referencing, agentive pivot

Sorry for the lateness of this reply; I've had lots of
things on my hands.

Quoting Pablo David Flores <pablo-flores@...>:

> Thomas R. Wier <trwier@...> wrote:
> >   (4) The man saw the animal and [X=the man] ran off.
> >   (5) *The man saw the animal and [X=the animal] ran off.
>
> As in C'ali and unlike English, (4) would be ungrammatical
> and (5) grammatical.
Actually, C'ali is like English is this respect, since "run off"
is an agentive intransitive verb.

> Stálág however has six voice operators,
> and there's one that could solve the problem, the antipassive,
> which makes a vi-P (intransitive verb with patientive subject)
> from a vt (transitive verb).
What you're describing here is the standard definition of a
passive, actually, not an antipassive at all.

> In this case it would apply to
> "ran off". The new meaning could be translated as "The man saw
> the animal and was/became a runner-off", but the antipassive
> is more of a pure syntactic, rather than semantic, device.
I'm a little confused here:  "run off" appears to be a patientive
intransitive verb which your operation turns into an agentive
intransitive verb.  As such, it's not what you described in the
immediately preceding sentence:  it's not a transitive verb becoming
a patientive intransitive verb (i.e., a prototypical passive
construction).

> >    *"The woman died and [she] shot at the small game."
>
> Same here on "shot". The antipassive cannot take a vi-A verb,
> but it's easy to make vi-A > vt using another voice, the
> applicative, with a particle that indicates allativity or
> goal.
Hmm.  I'm not sure if I'm following you:   with "vi-A > vt"
do you mean it would make an agentive intransitive *into* a
transitive verb or *from* a transitive verb? What you wrote
implies "into", but antipassives are normally considered to
be "from".

> I'm not sure whether "shoot" would be a vi-A or a vt
> in Stálág, but you can also have the opposite of the
> applicative voice applied, which is called "unergative"
> and uses the same particles.
>
> I've found it difficult to choose whether verbs with a
> goal or destination (like "shoot", "give" and discourse
> verbs) should be transitive or intransitive. That is,
> should one prefer to have
>
>     a) "speak" + Applicative voice (Goal) = "tell"
> or  b) "tell" + Unergative voice (Goal) = "speak" ?
Remember: syntax is autonomous from semantics.  Thus, it's
simply up to you to decide which is more basic in your language,
the intransitive, or the transitive, version of that verb. But
my first instinct would be (a).

> I'd like to hear your thoughts on this. I made up a structure
> with three basic applicative particles which I'll call APP:OBJ,
> APP:DAT and APP:ABL, and which works as follows:
>
>     1. kapos 'A-speaks' (basic form)
>     2. kappos 'A-says-P' (APP:OBJ = discourse topic = zero mark)
>     3. gye kappos 'A-tells-P' (APP:DAT = hearer = |gye|)
>     4. har kappos 'A-quotes-P' (APP:ABL = source = |har|)
>
> The root is |kap-|, with gemination marking the applicative
> voice, and APP:OBJ is a zero morpheme.
So, these are not bound to the head? That is somewhat unusual for
an applicative construction in my experience, but I don't see in
principle why you couldn't work it this way. It's an interesting
system though, and seems to me typologically reasonable.

> > I plan to work on voice-operations, such as antipassivization
> > and passivization, which would get around this problem. Any questions,
> > comments or criticisms?
>
> Just one idea, though you probably figured it out already. Split-S
> languages need *two* antipassives (one of them is the traditional,
> ergative-language antipassive, and the other is my "unergative"),
> or else you'll have to live with those unwieldy pivots. You can
> avoid the unergative by making all verbs with goals intransitive,
> but then you need something (the applicative voice) to add an
> argument.
Actually, you don't in fact *have* to have two *anti*passives --
impersonal passives like German "Es wird getanzt" accomplish the same
thing. Crosslinguistically, such impersonal passive constructions appear
always to demote the *agent* NP of unergative verbs ("dance", "sing",
etc.).  Another alternative to avoid the pivot is to have derivational
(not inflectional) pairs of intransitives, which differ only in how
they assign case to their single NP, agentive or patientive.  Indeed,
they need not even be morphologically related; they could simply be
suppletive.

(I say this hoping I've understood what you mean by "unergative
voice".  You seem to be using terminology in ways that are not
always canonical or widely used.)

> Another idea: an "inversive" voice that exchanges the patient
> and an oblique complement of a vi-P, using applicative particles.
Yes, this is an interesting feature of some Bantu languages.

> How do you manage perception verbs?
Perception verbs are all in a class of inverse verbs which function
differently from the voice of the same name in Stalag.  See my recent
post on them for details.

 =========================================================================
Thomas Wier            "I find it useful to meet my subjects personally,
Dept. of Linguistics    because our secret police don't get it right
University of Chicago   half the time." -- octogenarian Sheikh Zayed of
1010 E. 59th Street     Abu Dhabi, to a French reporter.
Chicago, IL 60637

Replies