RE: More about questions and the like (was:What I have for dinner...")

["And Rosta" <a.rosta@lycos.co.uk>]

> From: Pycyn@aol.com
> 
> The "for all x, if x went to the party then John knows that 
> x went to the party (perhaps with a relative completeness 
> guarantee or something like "and if x did not go to the 
> party then John does not believe that x went to the party") 
> has to be used with care.  One should not infer from it 
> that, if Paul went to the party, John knows that Paul went 
> to the party, at least in the sense that if John is asked 
> "Did Paul go to the party?" he will say yes, even if he is 
> being as cooperative as possible.  the problem is the 
> intensionality of "know," for John may not know Paul under 
> the name "Paul" and may even know him under the mistaken 
> guise "Bill's father" (when he is actually Joan's 
> father).  To be sure, the quantifier outside the context 
> guarantees that the generality is indifferent to the 
> disguises, but the individual cases have no such obvious 
> external connection.  We need either explicitly write "under 
> some concepty" somewhere here or flag the cases with the opposite 
> of subject raising (which I think Lojban has).  The 
> set-of-answers version of questions does not work so well in 
> this case, unless the questions are subdivided into identity  
> classes and all that is required is that John know at least one 
> member of each of the appropriate classes.

I don't see that we're falling into this trap, at least on
the "for all x, John knows that x is..." treatment as opposed to 
the set of answers one. This is precisely because the criterion
by which x is to be identified from whatever is not x is not
included within the proposition that John knows.

> A couple pages on from this point in the Handbook of Philosophic (i.e., 
> freaky) Logic is the reminder that every natural language sentence is a 
> dependent of a (usually unexpressed) performative, usually "I 
> tell you that" or some such.  However, some of these performatives 
> may also be intensional, in which case every term in the surface 
> sentence is in that cloud-cuckoo-land where Leibnitz's law fails along 
> with existential generalization and universal instantiation.  In 
> particular, sentences mentioning non-existent objects which are 
> nonetheless held to be true are under performative like "I 
> now recite to you a bit of myth that..."  This being the case 
> (and it sure solves a lot of problems), Lojban needs to dig into its 
> small stock of unused cmavo for a flag of this sort for when context is 
> not enough.  Remembering that the term-length flag of this sort is 
> something like tu'a, I suggest the corresponding x form, xu'a.  
> this refers to a different enduring problem form ages past.

I don't understand the problem or the proposed solution. Or at least,
I am aware that performatives are necessary, but assuming they are
available (linguistically), then I see no outstanding problems. And 
I don't see how you envisage {xu'a} working. What we need is a way 
to identify (a) something as a performative, and (b) its scope. And 
since the baseline has been set, all we can actually do is ask "do 
(a) and (b) exist?".

--And.