RE: except the cat

[pycyn@aol.com]

pc said
<<{ro da onoi le mlatu} say.>>

oNAI (bad typing habits and ghosts of loglans past combine)
As xorxes would soon point out no doubt, this does not work on the ususal X 
sumti1 ek sumti2 Y => x sumti1 Y ijek X sumt2 Y transformation (though one 
could pull some Gricean logic to help it along)  but presupposes an earlier 
rule, something of the sort X Qx ek L Y => X Qx Y ijek X x=L Y.  I suspect 
that there ae restrictions on at least Q and ek (probably corestrictions) and 
maybe on L in all this.  But it is an interesting rule to look at and 
develop.  Indeed, I suspect that a number of problems that are basically 
solvable by logical usage can be dealt with in these rules.  But "even" isn't 
one of them even though there is a theorem of logic that every property has a 
least likely (and a most likely) participant.  The problem is that "even" 
does not actually require the least likely thing, only a sufficiently 
unlikely one.