Re: A few questions about linguistics concerning my new project

--- In conlang@yahoogroups.com, Nick Scholten <nick.scholten@...> wrote:
>
> David J. Peterson wrote:
> <<<<<
> What, for example, is the semantic difference between "The man
> pushed the panda" and "The panda was pushed by the man"?
> >>>>>
>
> Between those exact sentences there is no difference in meaning, only in
> point of view. I think the biggest difference between the two is that in 1
> you can leave out the accusative and in the second the oblique (or well,
> agent). And, I think you're right in thinking this is because of conjoining
> VPs: The man pushes, kicks and slaps. vs. The panda is pushed, kicked and
> slapped. (I feel here that English is not so happy with the first sentence...)
>
> <<<<
> Note, though, that the antipassive without
> the oblique argument is the equivalent of (c), which is why (c)
> is bizarre (kind of doesn't have a place in the language).
> >>>>
>
> I understand now why (c) doesn't make much sense, but is an antipassive
> without the oblique argument the equivalent? I think there is a difference
> because in that antipassive the agent is in the absolutive and the verb also
> has a specific mark for the situation.
>
> <<<<
> One thing to note: Your example sentences above suggest that
> your language is sensitive to the subject position, even though
> it's an ergative-aligned language.  Even if there were some
> rule requiring a pre-verbal nominal argument in all instances,
> you might expect the order of NP's in (e) to be flipped (unless
> the ergative argument is a true adjunct, and can appear in either
> place, like a prepositional phrase or an adverb).
> >>>>
>
> Actually, my language is basically SOV, I used that strict SVO word order
> for those examples. Does what you're saying also apply to SOV word order,
or
> doesn't this very basic way of describing some syntax apply at all to
> ergative languages?
>
>
> Nick Scholten
>
Difference between active and passive -- and, some say, in general,
differences in voices -- is not purely semantic (not merely to do with
meaning), but partially or totally pragmatic.  In my view most of the
pragmatics difference is one of "Information Packaging".  Both utterance
contain the same information; but they pacage it differently.

In multi-clause speech, or in a connected narrative, or even in a connected
discourse, one of the main purposes of passivizing is to share the passive's
subject with another clause.  The subject is the syntactically most-privileged
participant; and for many purposes in many languages, if you want to do
something syntactic to a shared participant of two clauses, it must be the
subject of one or the other or both of them.

A passive clause uttered in isolation, with no other clauses before or after it to
share participants with, just topicalizes or focuses the patient as opposed to
the agent.

In syntactically ergative languages, there is a good case to be made that the
absolutive argument is the subject; the default or unmarked subject is the
patient rather than the agent; so the usual way of saying a transitive clause
is _already_ "passive", in the sense that the patient is the subject; and
therefore "passivization" is not necessary.  However if one wants to make the
_agent_ be the subject, one must "anti-passivize" to promote the agent to the
subject position.  In connected speech the main reason to do this is to
accomplish some syntactic process involving another clause in which the
agent also participates (possibly not as an agent).  However if one utters the
antipassive clause all by itself, the effect is to emphasize the agent and de-
emphasize the patient (perhaps even to make the patient implicit, or not
specify it at all).

It's possible even in syntactically ergative languages to have a de-
agentivizing, detransitivizing voice-like operation that just drops the ergative
agent of a transitive clause.  This would not have much syntactic use; its
main use would just be to indicate that the speaker either knows nothing
about the agent, or considers the agent so unimportant that nothing need be
said about it to the addressee.  Thus, as far as speech-act goes, it would
have the same effect as passivizing an isolated clause has in English and other
accusative languages.  But there is no syntactic use for passivizing in
syntactically ergative languages; while in accusative languages, passivizing
has the result of preparing the clause to take part in some multi-clause
construction involving some other clause, which has as one of its participants
the patient of the passivized clause.