On Tue, Nov 24, 2009 at 5:06 AM, Christophe Grandsire-Koevoets <[log in to unmask]> wrote: > 2009/11/23 Peter Bleackley <[log in to unmask]> > >> I have previously heard on this very list that Latin has to do this in this >> situation, as prepositional phrases can only be used as predicates, not as >> attributes. The Greek, I believe, simply says "Our Father in heaven", and >> the traditional English translation is based on the Vulgate, rather than the >> Greek. >> >> > Indeed. Actually, it's something I've often wondered. Modern European > languages seem to be quite free in allowing prepositional phrases as > attributes, but I've always wondered whether that freedom was common across > the world. > > I know I've played a lot with the idea to restrict this freedom. My Moten, > for instance, only allows genitive phrases to complete other nouns. Its > solution to get other types of phrases to complete nouns isn't relative > clauses, but *surdéclinaison*, i.e. the fully inflected form is inflected > again in the genitive case to allow it to complete a noun (I copied this > from Basque :) ). In Maggel, attributive prepositional phrases are formed > differently from predicative prepositional phrases, and a noun with an > attributive prepositional phrase must be in the construct state. > > What do you all do in your conlangs? Ilion puts all nouns used as noun adjuncts into the genitive case (which means some prepositions, such as those for location, can govern either the oblique or genitive case, depending on whether they modify a verb or another noun). Ekmartenkar just uses a bunch of cases, and suffixaufnahme on those that modify nouns. Kulaqil doesn't really have anything equivalent to PPs: coverbs with meanings like "exist/be-at" would probably be used. Samadurian grammar is still in a pretty nebulous stage right now, so I don't know yet.