Re: TECH: fuzzy logic

[ucleaar <ucleaar@UCL.AC.UK>]

> >I too wish to declare that, like everyone else, I am not making
> >grammar change proposals either. I think an immediate baseline should
> >be declared, so your tension levels drop a bit. Instead, I am
> >engaged in discussions about how the grammar should change if it
> >ever were to change.
> Does this mean I do not have to read and respond to your postings?
> because I am not interested in anything beyond the baseline until
> after the baseline. %^)

It's not for me to say whether you have to read and respond to my
postings, though my opinion is that you don't have to. I'm amazed
at your persistence. *You* call *me* stubborn! I don't think I ever
say much that requires the scrutiny of the LLG president.

> >> BUT na'e is not solely used for scales on brivla, so we have to be
> >> careful about adding to the set of scalar variables.
> >Could you explain more?
> Look at references to NAhE in the machine grammar.

To my shame and frustration I have never managed to think my way into
the formalism. Maybe when a thicko's guide to it comes out I'll be
able to pretend to be an expert on it.

> Allow your construct in any of them. Do they make sense in all.
> Likewise NA (which among other things MAY mean that your fuzzy XVV
> construct can end up in a logical connective in place of, say "na.a").

I see. Well for my xoi/xio proposals I did think of this matter.
If the semantics of {na a} and {jaa a} is compositional, then
{xoi a} ought also to have predictable meaning (if the basic logic
is understood). But it may be that the meaning of {na a} is idiomatic,
in which case there need be no expectation that {jaa a} or {xoi a}
will mean anything. [My impression is that {na a} is idiomatic, but
that it's fairly easy to extrapolate to {jaa a} and {xoi a}.] The
same goes for NAhE, but even more so. If every member of NAhE has
a meaning in every context in which NAhEs are permitted, then so
will {xio}.

> >> Otherwise we will get something or another that will use And's
> >> proposal and have to be interepreted isiomatically just as you fear
> >> as the case for NAhE xi quantifier.
> >& explain this too...?
> Well, see the above reference to na.a, and imagine replacing na by
> your construct.  Come up with a conventional interpetation.  Promise me
> that no other conventional interpreetation will EVER conflict with it.
> Repeat for all occurances of NA and NAhE to cover all of your proposal.

I answer this above.

> This is similar to the question of putting lambda in KOhA.  Does it
> have meaning in ALL places where KOhA can appear.  If not, you need
> conventions, and those conventions need to be immutable as possible
> (e.g. the convention on ke'a, which is anothe KOhA that has no meaning
> in many contexts)

The selmao of the lambda variable is an irrelevance. Since {ka} does
not have a selmao to itself, there is no way to constrain its occurrence
to within ka...kei. Selmao CUhE is in this respect no better than KOhA.
As you say, there is a convention for {kea}, which is why it made so
much sense to use {kea} for the lambda variable. I proposed a perfectly
straightforward convention that would cover all uses of {kea}. OK - I
failed to persuade you there was a semantic basis for making {kea} do
all these jobs, not least because my arguments were very unpersuasive
(though even when they are very persuasive you never get persuaded),
but the fact remains that the {kea} solution would have worked okay.

coo, mie and