Re: Gadri for you

xod:
> 7On Fri, 1 Aug 2003, And Rosta wrote:
> > xod:
> > > On Fri, 1 Aug 2003, And Rosta wrote:
> > >
> > > > Xod (offlist):
> > > > > I am having trouble remembering what the competing meanings for
lo'e
> > were.
> > > >
> > > > By the time the debates ended, & according to my understanding at
that
> > > > point:
> > > >
> > > > 1. Something very much like CLL's typicality gadri. This seemed to
me
> > > > neither essential nor parrticularly useful, but did (after a number
of
> > > > further clarifications from Nick about how he thought it should
work)
> > > > seem logically coherent.
> > > >
> > > > 2. The Kind gadri. The only reason for assigning this to {lo'e} was
> > > > prior usage (xorxes's), and simply assigning Kind to {lo'e} doesn't
> > > > of itself allow us to express everything we wanted it to. (More
details
> > > > when we reopen the subject properly.)
> > >
> > > Does Kind = Mister? Where does Unique fit into this, is that a 3rd or
4th
> > > category?
> >
> > Kind = "Mister", yes, but I can't vouch for that being everybody's
> > understanding, and "Mister" has in the past been used for Substance, so
> > we need to tread cautiously if trying to use 'Mister' as a technicalish
> > term. As for Unique, I would say that this is equivalent to, or
superseded
> > by, Kind.
>
> The examples you offered for Unique are better suited to Kind? (you wrote:
> "the one and only broda"; all broda are treated as the same one broda
> :*Blue* is my favourite colour, *Monday* is the first day of the week, *B*
> is the second letter of the alphabet, I like *sherry*, *Sherry* is sweet,
> *The platypus* lays eggs, *The Afghan* is a dangerous foe, Is *the pope*
> catholic?")

Yes, these are all exx of Kind. Every Kind is unique (though it has
infinitely
many subtypes -- a Kind is like a node in a taxonomy). The reasons I prefer
the notion of Kind are that it is a notion not unknown to linguistics and
that it better captures the important element of intensionality/ideality/
conceptuality (a Kind exists solely by virtue of being imaginable).

> Typical is the non-mathematical equivalent of the statistical mode, isn't
> it? The most common type?

I don't know. I only know that "lo'e cinfo cu xabju le friko" means "lions
typically live in Africa". Personally I'm okay with this notion of
typicality
being vaguely defined.

> What does Prototype mean? Isn't that Stereotype?

Prototype is, roughly, Kind-in-Prototype-Theory` whereby Kinds/Prototypes
are defined by psychologically default (usu typical) properties. I suppose
it is close to the notion of Stereotype, but Prototypes are strictly
psychological, whereas Stereotypes can be cultural.

> > Regarding the things we need gadri to express (in 'gadri rows'), I think
> > we need just Collective and Kind (alongside Quantified).
> >
> > We might also need Substance if we decide that brivla don't encode the
> > countability of their own sumti places. (E.g. if x1 of valsi is neutral
> > between "is a single word" and "is a single amount of wordage", then
we'd
> > need to distinguish between these by means of gadri. But I am opposed to
> > using gadri for this.)
>
> Technically we should be able to use le tu'o or lo tu'o for substances.
> But, historically, lei/loi was used, so a backwards-compatible solution
> should branch substance off lei/loi.

lV tu'o still works, doesn't it, for expressing Substance. So there's not
really a need for a new gadri for it.

> > In other words, it's not that we need a really complicated gadri system;
> > it's that the simple gadri system that would suffice bears little
> > resemblance
> > to the current system. The complexity comes in bridging the gap between
the
> > two systems.
> >
> > > What signals the opening of proper debate? I was only waiting for the
> > > member's meeting and jboselsla to end.
> >
> > We can reopen the debate now, but unless we have some idea of what the
> > majority of Lojbanists will and won't accept, I don't see the point. If
> > we had license to simply throw away the current system and create a
> > new one, then there'd be no problem. But we don't have such a license
> > and I have no idea what is and isn't acceptable to the community.
>
> We can only find out by writing what we think is a decent proposal and
> testing the reaction.

Okay. For my part, I think "decent" will have to be weakened to "unbroken",
because I can't help but feel that the length of a gadri should be
proportional
to its frequency/importance/basicness. I suppose I'd be more comfortable
with
discussing what the gadri system should look like in the abstract, without
proposing actual forms (except patently silly ones like li'i'i'i).

--And.