double accusative (was Re: passive of causatives)

[Georg von Simson]

Jonathan Silk wrote:
>I really very much hesitate to enter this discussion, which I but poorly
>understand, but my friend Georg said something which prompted a thought:
>
>> But is it not
>>more sound to take case endings seriously
>
>Is not the really crucial issue the one alluded to by Boris, namely
>"semantic roles"? In other words, at least from the Indian viewpoint should
>not the discussion be framed in terms of kaaraka theory, rather than case
>endings?  (But probably I have thoroughtly misunderstood the argument, in
>which case please ignore my babbles.)
>
Of course, we are talking about semantic roles or kArakas. But case endings
are *normally* correlated with a semantic role, and my point was that we
should not without necessity disregard such a normal correlation. The "low
transitivity" Boris Oguibenine was talking about, is not the same as the
role of "indirect object", and his analysis

> It is however useful to recall that "tam pustakaM darzayati/dadAti" can be
> analyzed as possessing three semantic roles.The agent "he" acts upon "a
> book" which is an affected object or patient and upon "him" which is the

does not quite convince me because you cannot say "taM pustakaM dadAti". It
is true that with transitive verbs there is a primary object and a
secondary object, but, as examples with intransitive verbs show ("dUtaM
grAmaM gamayati"), the person affected can and should still be taken as a
direct object (see also the conversion into passive, discussed earlier). It
is therefore I suggest in the case of darzayati a semantic shift from "he
makes someone (acc.) see something (acc.)" to "he makes something (acc.)
seen=visible to somebody (dat., later also gen.)." That seems to me a
better strategy than to jump to the conclusion that an accusative ending
can express the role "indirect object".

   Georg v. Simson