Re: properties of masses

[jorge@phyast.pitt.edu]

pc:
> I agree with xorxes that my description of the use of mass terms fits ill
> with the default quantifiers for mass expressions. But I have some doubts
> about the default quantifiers for all the essentially singular descriptors
> (I have doubts about all the default quantifiers, if the truth were known,
> but I can make them stick in these cases). Consider the claim that the
> man touched the door. If you take quantifiers seriously (possibly a
> mistake), you have to say that that mean that some part of the man touched
> the door -- clearly all of him did not. But if you say that the man
> entered the room, you presumably mean that all of him did, not just a part
> (though, of course, each of the parts did too). So, what is the default
> quantifier here, given that we say "the man" both times?

{ro}, each of the men I have in mind, in both cases. No masses here,
the parts of the man are not even mentioned. It is the individual man
that is in relationship with the individual door, or with the
individual room. The parts of the man, the door and the room play
no part at all.

But the same example works for the mass case. If I say {lei ci nanmu
cu pencu le vorgai} then I suppose it is enough that one of the men
touches the door to say that they did. (Perhaps the door was on the
ceiling and they had to stand one on top of the other to reach it.)
But it would be strange to say that they all touched it as a group
unless they were functioning as one in some sense.

In the case of entering the room, I would expect {lei ci nanmu cu
nerkla le kumfa} to mean that all the three parts did, but again
in some context it may be enough that some of them do.

This still doesn't allow us to say in general what properties a mass
has from the properties of its components.

> Notice, by the
> way, that if we try to get hyperaccurate and say that the man's hand (or
> fingertips) touched the door, we end up saying something very different
> from when we say that the man did it.

I agree. ({pencu} has a place for the instrument of touching anyway.)

> I suspect that the correct answer
> about default quantifiers (assuming we want to mess with them at all) is
> that they are contextually defined, another kind of conventional
> implication, often largely conditioned by the main predicate (as in the
> cases above).

I don't think those cases present much problem with the quantifiers.
If I say that part of the mass of men touched the door, then I'm saying
that only some part was involved in the touching, even if only one of
that part actually made contact. The same if I say that part of the
mass entered the room. If I say that the whole mass of men touched the
door, then I'm saying that the touching was a group action.

> In short, quantifiers or no, the description I gave of the
> behavior of mass sumti vis a vis the underlying individual and submass
> sumti is correct in broad strokes.  The picky details require almost case
> by case work, again largely depending upon the predicates involved.

Well, I prefer to say that the mass should be viewed as a new individual,
and that the properties of the components may or may not be relevant to
the properties of the mass, depending on the predicate. I guess it comes
to the same thing, though.

Jorge