daniel andreasson <[log in to unmask]> writes: > Jörg Rhiemeier wrote: > > > 1. Some fictional languages display combinations of features not > > attested in natlangs. For example, I have heard multiple times that no > > active language has yet been found that marks nouns for case that way - > > those that are known are all head-marking. But several case-marking > > active languages exist in the realm of fiction (e.g. Nur-ellen and I > > think also Chevraqis). > > And Rinya! :) And Georgian. Of course, Georgian could be argued > to be a head-marking language because of all the stuff you put on > the verb root, but it does inflect NPs for case. And Georgian displays > active alignment in the aorist tense by marking the NPs with different > case suffixes. So it should count as a case-marking active natlang. Of course, how could I forget?! Where did I leave my mind? Georgian is pretty familiar to me, I used it once as a counterexample against Marcus Smith's claim that active case-marking systems don't exist! It also plays a role in my musings about what historical linguists think what Nur-ellen is related to in my con-universe. In that conworld, some linguists try to link Quendian to Kartvelian exactly on the ground of that. The case marking patterns are really strikingly similar, down to the use of the dative for subjects of verbs of perception, even though I didn't know about the Georgian system when I designed Nur-ellen. There are differences, though: Georgian allows inanimate nouns in agentive, and the system of degrees of volition I have designed for Nur-ellen is AFAIK not paralleled by Georgian. > Svan and Laz also work as Georgian. Ts'ova-Tush (or Bats) has optional > free-standing pronouns which are marked for case with active alignment > (although there is obligatory agreement marking on the verb) so you > _could_ count Tso'va-Tush as well. > > That makes for at least four natlangs and at least three conlangs. And that ain't all of them. There is Winfired P. Lehmann's hypothesis that Proto-Indo-European might have been active (with case marking) at some earlier stage, which he claims manifests in the peculiarity that PIE neuter nominatives fall together with the accusatives: these cases arose from agentive and objective, respectively, and the neuters are a remainder of an old inanimate class which did not have an agentive case, such that the objective was pressed into service as both. I haven't read Lehmann's argumentation, but I am quite fond about this hypothesis; this model would come even closer to my Quendian model than the Kartvelian one, especially considering that the use of dative subjects with verbs of perception is attested in quite a number of IE languages and can thus be safely reconstructed for PIE as well. In my con-universe, again, this hypothesis has gained much acceptance, and gives even more credibility to the widespread assumption that IE and Quendian are related to each other. Greetings, Jörg. -- "Bus, bi, bo, bum, bo; bi, borum, bis, bos, bis!"