It occurs to me that cardinals and quantifiers are used in two distinct ways in some languages: English, for instance, distinguishes between "the three men" and "three of the men" in a fairly obvious way - the latter is partitive, the former is not. (Is there a word describing the former usage? I'm tempted to say "holonymic", but that already means something.)
In Japanese, certain quantifiers (e.g. cardinals, subete "all", takusan "many") can be used in at least two ways: directly as a nominal modifier, e.g. sannin-no gakusei "(the) three students", or appositively, e.g. gakusei-tachi sannin "the three students", or quasi-adverbially, e.g. gakusei-wa sannin iru. "There are three students." (I say adverbially because it can be separated from the NP: gakusei-wa koko-ni sannin iru. "There are three students here.") The latter usage is sometimes partitive, and the former two are never partitive, but the distinction is perhaps better understood as indefinite vs definite. The UDHR begins subete-no ningen-wa "all humans (are)", but ningen-wa subete "humans (are) all" is also possible, albeit with a slightly different meaning. Other quantifiers cannot be used as a modifier - minna-no uta "everybody's song" vs uta-wa minna "every song".
Does anyone else have any observations to add, or know of any deeper analyses?
Use of cardinals and quantifiers
Use of cardinals and quantifiers
書不盡言、言不盡意