> But, speaking of translation, I'm wondering now about adjectives
> and nouns in relation to case agreement. If one wants to translate
> "on this foot", would we say */kesmi pedi/ (with case agreement)
> or would we say */kos pedi/ (without)?
My totally ignorant instinct is to say it depends how the language looks at
it. If a language thinks (on- this) + (on foot), then we get the IE
approach kesmi pedi; but if a language takes (kos ped) as all one phrase,
then kospedi is possible. The second treats the phrase as a single unit
(as happens in Polynesian languages, and Semitic) and (at least in Semitic)
this is reflected in the accent.
We don't see the unifying of words into phrases in IE, so it would be less
likely, inmyarrogantanduninformedopinion, in PIE. IE languages do show
redundant marking elsewhere (eg 3rd person verb endings with declared
subject), so why not here?
Quite apart from all that speculative drivel of mine, is there any IE
language which does not seem to come from a proto with noun-adjective
agreement?
Peter