[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lojban] More about quantifiers



In a message dated 3/12/2002 2:10:36 PM Central Standard Time, jjllambias@hotmail.com writes:


{me'iro} is hard to think about. I wonder whether there's no
simple word for it because it is hard, or whether it is hard
because there's no simple word for it... Probably the first.


I suspect that there is no simple word for it because it is so rarely useful as opposed to {su'o...naku...} and, when it is, "not every" works fine.

<Anyway, here's what I do to understand what it says:
First introduce a {naku naku} between the two terms:

     no broda naku naku me'iro da poi brode cu brodi

Now, {no broda naku} reduces to {ro broda} in both our
systems (not in Aristotle's though). And {naku me'iro da
poi brode} is any of {ro brode} or {ro da poi brode} in
my system, and I think {ro brode} in yours. That means
it reduces in both systems to:

     ro broda ro brode cu brodi

This is right, "no broda is a brodi to not all brode"
is the same as "all broda are brodi to all brode."

So, in my system it has import for neither broda nor
brode, while in your system it has import for both. Did
having the form {no broda me'iro da poi brode cu brodi}
really help with that?>

Interesting.  This shows that either system is going to have to spend some time on multiple quantifier cases, since this kind of recalculation just won't work in real time. There seems to be problems all over the place with second -- and probably later -- quantifiers.  As patterns emerge, it should no doubt be possible to get some pretty tight rules on this.
Why does Aristotle's system not have {no broda naku} = {ro broda}? It seems that if it works in my system it should in his, since I was taking his as just being one foursome out of the possibilities that the notation offered, in which case {no broda} as opposed to {no da poi broda}  (never mind for now that this distinction apparently has to go and you all have to find a new form for non-importing quantifiers) is just {ro ... naku ...}

<I am very comfortable with making transformations in my
system, but making any transformation in yours almost always
gives me a headache, so I will keep using mine. You can use
yours and call it official (though it departs from the
Book about as much as mine does), and we'll just have to
take the risk that if we ever communicate in Lojban we
might in some marginal case misunderstand each other. (Not
that this would be anything new.)>

In light of &'s comments, I think we need to consider a further possibility, namely the Traditional Logic pattern.  The pattern is invalid as it stands, but operates under the implicit assumption that all subject terms are non-null.  It then gives the very tidy system we all know and love -- and would like to come as close to as possible.  then, on the rare occasions when we want a free quantifier -- because we aren't sure whether the subject class is empty or not and it makes a difference -- we can use the cautious {ganai de S gi} prefix, ugly as it is, or something tidier that comes along