On Sat, Jan 16, 2010 at 3:20 PM, David Peterson <[log in to unmask]> wrote: > > Right. This is where the misunderstanding is. You're talking about the > class > markers in Oneida being used as standalone nouns. This isn't what's > happening. > Standalone nouns can be used on the fly as classifiers in Oneida. You were > looking at > it backwards, and that's why I believe you were thinking your classifiers > "should" also be able to exist as standalone nouns. That interpretation > isn't > warranted here--though whether they can or not is still an interesting > question. > They certainly could, if you wanted them to. > > Ah, I see! That makes sense. I'll certainly have stand-alone versions of the class markers for use as pronouns/determiners/adjectives, but they won't be the forms used within the verb. > Okay, so I have to ask the same question: What pragmatic function does it > serve to have these locative phrases inside the verb as opposed to outside > the verb? Or is there no way to have a locative phrase outside of the verb? > (In which case, how would you say, "The strangers are living on a plain > that > lies just on the other side of those mountains over there where we used to > go to gather lumber and nettles"?) > > As in Navajo, incorporation will be optional, I think--I still have the locative case for use in freestanding nouns. Pragmatically, locatives would be incorporated to show that the action or state of the verb was closely tied to the location where it occurred. Due to the general mindset of the speakers, this would probably happen most of the time. You never just "go"; you always go to somewhere. You never just sit; you always sit somewhere. On the other hand, the act of crying or eating is only tied to its location in very particular circumstances. The difference between locatives tied to the verb and not tied to the verb would also show up in cases where two or more locatives were present. E.g., Kaukaheleeretae felieuri. walk<1S-PAST-ACT.IND-friend<LOC>> forest<LOC> "I walked with a friend through the forest." Kaukafeleiritae eleeurae. walk<1S-PAST-ACT.IND-forest<LOC>> friend<LOC> "I walked through the forest with a friend." In the first version, the act of "walking-with-a-friend" occurs in the vicinity of a forest. In the second version, the act of "walking-in-a-forest" occurs in the company of a friend. So, the unincorporated noun essentially becomes the focus of the sentence. Regarding more complex syntactical situations, I expect that this added synthesis will come at the cost of some syntactic flexibility, but I'm okay with that. The sentence you gave would probably gloss something like, plain<NOM> across<FAMILIAR-mountains<LAT>> gather<1P-PAST-familiar<LOC>> nettles<ACC>, territory<UNFAMILIAR-familiar<LOC>> By classing "plain" as familiar in the verb "across," the speaker marks it as the antecedent of the references incorporated into "gather" and "territory." > This does give me a thought, though. An interesting design goal might be > to have a language where every utterance is maximally one word long. It > might get difficult to define an "utterance", but it'd still be > interesting. > Indeed! The gymnastics involved would be interesting, but figuring out the philosophy and grammar would be buckets of fun.