On Thu, Feb 27, 2003 at 12:09:13PM +1100, Nick Nicholas wrote:
> cu'u la djordan.
>
> >The changes people want to loi aren't backward compatible. They
> >range from complete gadri overhauls, to redefining the meaning of
> >"lo".
>
> Since I spent two months trying to make my solution backwards
> compatible (most of it after you gave up on the discussion, I
> believe), I'll thank you not to speak for me.
I was speaking about xod's solution and And's solution (at least
the ones that I looked at).
> Craig, your veto is overstated. *Changes* need to be strongly
> justified, this I agree with, and it is in the charter. (Contra
> Jordan, I think there is a problem with loi, though, because I have a
> different underlying philosophy of lojban, which we ay eventually
> need to thrash out --- namely, that while Grice is well and good for
> every day use, there must whensoever possible be explicit mechanisms
> of disambiguation in place if people need them.)
[...]
> loi can express four things, and there are no compelling
> disambiguations in the grammar (though you can approach it with
> paraphrase): collective (a bunch of sharks), substance (some shark
> [meat]), the Kind of shark ("Mr Shark"), the Any Shark (I'm
> drawing/seeking/needing a shark, any shark). Jordan contends that
> Grice should always tell you the difference between collective and
> substance; but if I am to have only Grice at my disposal, Lojban is
> much less clear than English (which allows you to have mass and count
> nouns); and no, I do not want to settle for that. The jboskeist core
> want to have different gadri for collective and substance; but that
> means loi is not backward compatible, which I cannot accept either.
We should be realistic about this though: Simply based on vocabulary
alone, it is unlikely that Lojban will ever be as clear as English.
Lojban is something to have fun with, and constant redesign and/or
unrealistic goals will prevent it from ever getting off the ground.
If you want a fully expressive language, pick a natlang, if you
want a formal system, use one. Lojban is a deformed baby. We can
only expect so much from it.
That said, I actually think Nick's solution below may have some
promise as a sane way to satisfy all sides (provided it does preserve
(correct) existing usage, and also provides And and whoever else
with what they want). And I do agree that Kind and Any are real
problems.
[...]
> My solution (to be refined and what-not):
>
> Collective: loi [so'a/su'eci'ino] finprcarka
> Substance: loi [ci'ipa] finprcarka
> Kind: [tu'o lo finprcarka] => lo'ei carka (new LAhE, but
> paraphrasable as normal individual sumti with quantification turned
> off)
> Any: either Propositionalism (what Lojban does now --- prenex
> of embedded clause), or Kind, depending on the selbri; [fi'u ro loi
> finprcarka] (in the right contexts).
Does this preserve the "loi nanmu pu klama le lunra" usage? I can
understand this in terms of lojbanmass, but not in terms of either
substance or collective, and I think it is a very useful usage to
have. Do you consider this to be Kind (which is what I thought it
was)? But you use "lo'ei" for Kind, and my understand of xorxes'
"lo'ei" doesn't fit with that.
> The Collective/Substance distinction is fully optional (so both still
> get to be lojbanmasses), and stated on the inner quantifier; but the
> distinction can be made if people choose to. (Right now, that just
> plain isn't possible.) The Kind ("Mr Shark") is disambiguated from
> the lojbanmass by giving it a new LAhE, though it can also be stated
> (prolixly) in terms of existing sumti structures and turned off
> quantification. (Anything true of the Kind is true of the lojbanmass,
> but I'm not convinced the converse is true.) The Any problem (how to
> say Any shark as distinct from A shark in the completely general
> case) admits of several solutions, none perfect, although we're now
> putting more thought into it; when we go into non-existing entities,
> we add something like {tu'o lo se ka co'e} or something (to be
> thrashed out), as distinct from {lo co'e} (which commits to existence
> of the referent; And, this was the coup John and I pulled on you in
> NYC.)
We need a new gadri for Any. tu'o doesn't work with lo (viewed it
as a mo'ezi'o), because of the existing semantics of lo.
> OK. You'll see a proper proposal in a few months. The BPFK will still
> start slow, and will start in a week or so. Back to your regularly
> scheduled flamewar.
vi'o
--
Jordan DeLong - fracture@allusion.net
lu zo'o loi censa bakni cu terzba le zaltapla poi xagrai li'u
sei la mark. tuen. cusku
Attachment:
pgpmUXRQvl8xS.pgp
Description: PGP signature