[YG Conlang Archives] > [jboske group] > messages [Date Index] [Thread Index] >
Jorge:
> la and cusku di'e
>
> > > > I don't disagree with anything you've said (except that it needs
> > > > to be clarified, IMO, that .9 entails "not (wholly) true").
> > >
> > > It all depends on how you define NOT for continuous truth values.
> >
> >I don't see why it depends on this. If pi ro = (wholly) true, then
> >.9 entails "not pi ro".
>
> If you're working with binary logic, yes. It either is the case
> that .9, or it is not the case. If you allow continuous truth values
> then the claim "it is .9 true" itself will have a fractional truth
> value in turn, and it will probably not entail "not piro" with truth
> value pi ro.
My own view is that 3-valued truth functions applies to a simple
proposition is very useful, but applied to other truth functions
recursively is mind-bogglingly complicated and not useful in
practice.
Of the 27 possible unary operators for 3-value logic, I'm in not
in favour of those that output a 0 (e.g. {+,0,0}), except for
{+,0,-} and {-,0,+} (varieties of NA and JA'A).
> > > A value of .9 is not a value of 1 just as it is not a value of .8.
> >
> >Yes. I think "not .9" should be true if the value is .8 or ro or 0.
>
> Ok, so you are using binary logic applied to claims with these
> modifiers.
Yes. That stops my mind boggling.
> > > But we probably want a softer "NOT" for continuous truth values.
> > > For example, a function that maps value x to value 1-x.
> > > Then {.9 <bridi>} does not entail {not 1 <bridi>} = {0 <bridi>}.
> >
> >to'e?
>
> No, I don't think this has to do with {to'e}, that changes the
> predicate word. As pc pointed out, I was thinking of truth
> values in (0,1). I don't really like talking of truth values
> greater than 1, I prefer to use a mapping into the open (0,1)
> interval. But a system that goes from 0 to 1, then keeps going,
> and calls "true" all the values above 1 is too weird for me.
> In a (0,1) range there is no "true" and no "false". Instead of
> that there are infinitely many values, which can be called more
> or less true.
Okay, but 'your system' and 'my system' both seem to me useful
but distinct.
Your system is appropriate where there is no threshold between
true and false. My system is appropriate where there is
a threshold.
> > > Yes, I agree. I would say almost any proposition is susceptible
> > > of both treatments. {ko'a clani} can also be seen as a yes/no
> > > proposition in some contexts. I'm saying that whichever kind
> > > of modifier we use determines how we're treating the proposition
> > > for the purposes of truth evaluation.
> >
> >Fair enough. I wonder if there will be ambiguous cases, when
> >pi PA values are ambiguous between (a) how much p is happening,
> >and (b) the extent to which p satisfies the threshold criteria
> >for being true at all. For example, {ko'a ja'a xi pi bi melbi}
> >might mean that ko'a's beauty measures .8 in millihelens, or
> >it might mean that ko'a is not quite beautiful but is close to
> >the threshold of beauty. I'd prefer to stick with the latter
> >reading only.
>
> You mean something like "almost beautiful"? I think I'd prefer
> {ko'a na xi pi re melbi} for the "not beautiful" side.
No, I mean something like "beautifulish, on the threshold of
being wholly beautiful, but not quite wholly beautiful, but
closer to being wholly beautiful and to being wholly not
beautiful, e.g. satisfying .8 of the criteria for being
beautiful, where something definitely beautiful satisfies .ro,
and something definitely not beautiful satisfies .no".
"Almost beautiful" would be ambiguous between "slightly less
than pi ro" and "slightly less than pi no". It's like the
difference between being outside a room but close to the
entrance and being most of the way across the threshold.
--And.