by Owehn » Sun Feb 08, 2009 9:49 pm
I did forget to reattach the re- to doyhah in the second option, but only because it's left separate in the first for a good reason: since the head noun is doyhah, not kormahn, it doesn't strictly make sense to attach re and kormahn. Rather, re is the article for the whole phrase, so I borrowed the rule that okh seems to use in these cases and left it separate. (Furthermore, there is at least one case where re is left unattached even under normal circumstances, so it doesn't necessarily latch on to the nearest word.)
As for a prepositional phrase involving tso, I didn't like suggesting it because A tso B seems to only be used when B is the recipient of an action that A specifically entails. You could argue that in context, there's only one possible way the machine could be acting on the descriptive books, but I tend to think it's more of a syntactic rule than a semantic one. I'd be happier using okh in place of tso, but that may be mostly because okh seems to be a catch-all for the "of"s I don't know what to do with, so I didn't suggest a prepositional phrase at all.
In the end, what we know about d'ni is so small limited that there often remain many choices consistent with what we know about it, and ranking those choices by how well they match the existing examples can be a rather subjective process.