Obviously, it also can be done with some other quantifiers - a few pieces each, a gallon milk each, etc. But let's go elsewhere with this.
Let us distinguish on plural datives whether they're distributive or shared. This, per se, does not create all that fascinating stuff, but let's go with it:
he gave them-distr.dat a hundred dollars
he gave them-share.dat a hundred dollarsThus far it is a bit boringly 'specific' as far as a case goes, but we could of course imagine some more general uses: distributive is intensive, shared is regular if no meaningfully divisible object is present.
he showed them-distr.dat = he really showed them (who's the boss or other implicit thing)
he showed them-share.dat = he showed them (something)We could of course imagine a similar distinction in other cases as well, particularly the accusative and the nominative:
he sold the cups.distr.acc = he sold the cups, as separate units
he sold the cups.share.acc = he sold the cups, as a single bunchNotable here might be that some plural nouns more naturally are treated as collectives, and some are more naturally treated as separate entities. Might be that the case marking is less marked for the more natural meaning for that particular noun, or for the majority of the nouns of that class (or prototypical nouns of that class).
We could do some fun things: conflate cases in some way. Dative and all other oblique cases are conflated for the atypical marking, for some nouns even the atypical accusative is conflated with the dative. Maybe subjects and accusatives are conflated for some noun classes for the atypical marking.
We could also mix in some split-alignment stuff here! Maybe the distributive object case is more likely to behave in an absolutive-fashion, whereas the shared one is more likely to behave in a nominative fashion - and this of course gives a neat reason to have some voice shenanigans, for the instances where the absolutive behavior prevents us from having a transitive subject that is distributed.
Of course, we have still not specified what it means to have a distributive object - in this case, I'd go for 'each is affected in a similar way, but not acted on as a group'. I also find it likely that with big numerals, the distinction would fall by the wayside.
Not a particularly clear post, but some quirks to work with.
Isn't this sort of how the Finnish partitive-as-object works?
ReplyDeleteAlthough there's probably some basic similarities - i.e. both would sort of be 'differential object markings', there's some differences:
Delete- this would differentiate objects (or subjects (and even singular ones!)) by the marking on the recipient - not the object -, primarily - c.f. 'give them an apple each' -> give them.dat.distr apple.acc'; however, maybe this could be a nice reason to bring in plural forms of 'one'!). The same case distinction applied to the object cases would do some stuff more similar to the Finnish case system, but ...
- this would differentiate non-objects too (i.e. as already mentioned, the distribute dative, and the like - Finnish very seldom has any dative-like behavior for the partitive. I also mentioned 'conflate lots of cases with this case', we'll look at that a bit further down)
- the partitive isn't in that sense central to the 'n pieces each' construction in Finnish (which instead is dealt with by a separate pronoun, 'kukin' in the appropriate case)
- although the partitive (plural) does sort of 'undifferentiate' a plural group in some sense, its main role in object marking is rather to indicate the absence of any out of bunch of rather specific conditions - telicity and affirmative truth value being the two main conditions for the use of the accusative, but a few others also being in there (greater likelihood for accusative if it is a kinetic action, greater likelihood for accusative if it's a definite noun, if it's a concrete noun, etc etc.)
The Finnish partitive does manage to pull in some almost split alignment stuff (existential intransitives vs. everything else)
- the Finnish partitive is a fairly 'distinct' case by itself - the cases suggested here would have some level of case syncretism with regards to either the distribute or non-distributive case, depending on which type of noun is involved. So e.g. 'dative-distributive' might be conflated with 'locative-distributive' and 'instrumental-distributive', whereas 'dative-shared' is differentiated from both of those for some nouns, whereas the opposite distribution holds for other nouns.