Components of a mass (was: Quantifiers)

["Dylan P. Thurston" <dpt@abel.MATH.HARVARD.EDU>]

la xorxes. cusku di'e
 > > I beg to disagree. {re lu'a le nanmu ku joi le ninmu ku joi le verba}
 > > can't be "the man's ear and the woman's nose". The mass is composed of
 > > three elements: the man, the woman and the child, and you are selecting
 > > two of them. Neither the man's nose, nor the man's going to the market,
 > > nor the man's grandparents are members of the mass. Otherwise, where do
 > > you stop? Please don't invoke inalienable possession or anything like that. > > Those are possessions of the mass, not its components.

la djan. cusku di'e
 > Masses don't have discrete "elements", they have components.  (Sets have
 > discrete elements.)  The whole purpose of this distinction is that you can
 > synthesize a mass in one way and analyze it in another.  The mass
 > of my cats can be dissected into the Max-component and the Freddie-component; > but an equally meaningful dissection is into the heads-component, the
 > legs-components, the trunks-component, and the tails-component.  (Each of
 > these components may themselves be viewed as masses, of course.)

Um.  I don't see how this answers Jorge's point.  Let's switch to masses
of something non-physical, say, {lei valsi}, to understand things
better.  Could you decompose {lei valsi} (suppose this refers to three
English words, "gay", "Robert", and "fiscal") into parts, say, the
meaning, the syntax, and the phonology?  What if you realize you've left
out some (e.g., I didn't list anything about pragmatics of usage)?
There are, as Jorge pointed out, probably an infinite number of
properties associated to each one.  How are you going to decide which to
include?

For physical objects, there's a possible way out that you mentioned
(considering dividing the matter up arbitrarily); but I don't think even
that works terribly well.  After all, a cat is much more than the
collection of atoms contained in it.

If it's true, as you say in the refgrammar, that "every property of a
piece is also a property of the whole", and it's also true that
arbitrary slicing is allowed, you end up with some very odd things
indeed--for instance, it would have to be true that {lei mlatu cu
rebla}, which seems like nonsense to me.  What you say in the refgrammar
("Using the mass descriptor "lei" signals that ordinary logical
reasoning is not applicable") seems like an utter cop-out to me.  We can
do better than that in defining the meaning of masses, can't we?

Intuitively, it seems to me that it's _not_ true in general that the
properties of the individuals extend to the mass, but rather that a
property of the mass has to involve all the individuals in some way.
Are there problems with this?

--Dylan Thurston