> Panini also
>assumes that the agent of the absolutive is identical with that of the
>finite verb. I cannot recall any deviation from this usage in canonical
>literature. The prose of the Jataka, however, is an entirely different
>story.

Examples of this? The only exceptions I've seen are in complex sentences where a participle can take an agent. Then there can subordinated gerunds that don't have the same agent as the sentence's finite verb, but it doesn't actually violate the paninian rule as I see it.

It looks to me like Burmese scribes sometimes didn't understand this and made incomprehensible changes to texts as a result.

/Rett