Re: Translation pattern of `to have'?

Henrik wrote:

> One problem (seems to be frequent in active languages) is:
> no-one acts.  `A has B' seems to have a similar structure
> as `A loves B', but to speak of an experiencer in the first
> sentence in order to justify agentive slot (or case) for A
> (as in the second sentence) still seems overinterpreting.
> Maybe not.  I don't know.  So I don't know how to find a good
> verb here.
If you go for a verb "have", the use of AGT or PAT case would
depend on whether the active alignment is based on control or
eventiveness (or something else) and what kind of verb "have"
is derived from. If it comes from "hold" -- a state -- and
you have the distinction "event" = AGT and "state" = PAT, the
possessor should probably be marked as PAT. If it comes from
a verb "grasp" -- an event -- it should be marked as an AGT.

OTOH, if the distinction is control vs. non-control, then the
possessor should always be marked as AGT, since the possessor
is in control (although it might be possible to think of an
example where it isn't in control). Thus:

* event/state *

I.PAT have/hold a dog.
I.AGT have/grasp a dog.

* control *

I.AGT have/hold a dog.
I.AGT have/grasp a dog.

If you go with the "on me is a book" alternative, you can ignore
the above discussion since that isn't a problem anymore.

daniel

--
<> "Lea eica waenaidh mae bwochath waenë,   <>
<>  ja jordhëchaidh mae gothëje jordhëchë." <>
<>       www.geocities.com/conlangus        <>
<>       daniel.andreasson@telia.com        <>