Date:  Wed, 10 Dec 1997 21:35:28 -0500 (EST)
Message-Id: <199712110235.VAA01994@locke.ccil.org>
Reply-To: JORGE JOAQUIN LLAMBIAS <jorge@INTERMEDIA.COM.AR>
Sender: Lojban list <LOJBAN@CUVMB.BITNET>
From: JORGE JOAQUIN LLAMBIAS <jorge@INTERMEDIA.COM.AR>
Subject:      Re: semisummary: countability
X-To:         lojban <lojban@cuvmb.cc.columbia.edu>
To: John Cowan <cowan@LOCKE.CCIL.ORG>
X-Mozilla-Status: 0011
Content-Length: 2082
X-From-Space-Date: Wed Dec 10 21:35:32 1997
X-From-Space-Address: LOJBAN@CUVMB.CC.COLUMBIA.EDU

And:
>{pi su`o lu mi klama li'u} and {pi su`o lo'u mi pi ku klama ka cu
>le'u} have sense and selmaho.

Yes.

>{pi su`o zo sp} doesn't.

Actually, if it's grammatical then it does, and it's a name. I think
that by definition, anything zo-quotable is a valsi.
The question we're trying to answer is whether anything
that is valsi be bau la lojban must be zo-quotable.

>I also think that {pi su`o zoi x. she has sp x.} is wordage, but
>not {pi ro zoi x. she has sp x.}. But it is notoriously hard to
>reason about these things.

Especially because sp could become an English word anytime
without you noticing it. It could be something like cd.

 >> I would want {ro da poi valsi cu cmima lo'i valsi} to be true.
>> "Every x that is a word is a member of a set of words."
>
>If you applied that to {djacu}, it would make {lo`i djacu}
>infinite, even though there is only a finite amount of water.

Well, mathematically speaking, if there is a finite number of
water molecules then there is a finite number of collections
of them, even if you count different arrangements as different
collections. But yes, the cardinality of {lo'i djacu} would
indeed be quite large.

It is not clear that there is only a finite amount of water, though.
Is yesterday's water the same as today's?

In any case, what I think that you're pointing out is that some water
would belong to many of the members of {lo'i djacu}, whereas
the members of {lo'i valsi} would be more independent.

>Is that desirable? (Mind you, I think it's inevitable.)

I don't see a problem with it.

>I think I might prefer: {ro da poi ke`a me/du pa valsi cu
>cmima lo`i valsi}.

Hmmm... but {lei ci valsi cu cmima lo'i valsi} would also be true,
no? Wouldn't they be memberage?

 >Anyway, I'm coming to think that the "single-word option", where
>{pi ro lei ci valsi cu valsi} is false (even though {pi ro lei
>ci djacu cu djacu} is true) is probably a bit more straightforward.

I think it is, though the other doesn't seem to be internally
inconsistent. At least I can't find an unavoidable inconsistency.

co'o mi'e xorxes