[FOM] Re: "Leibniz's Law"

Richard Heck heck at fas.harvard.edu
Mon Jun 9 11:30:51 EDT 2003


> > > I meant
> > >     (x) (y) [R(a, x) & x=y --> R(a,y) ]
> Richard (Heck)
> > That depends upon whether you mean it schematically, or whether you mean
> > "R" to be a higher-order variable.
> What does "mean it schematically" actually mean?  I suppose it means,
> "R(a,-)" stands in for a gappy sentence.  But there must still be a relation
> between the gappy sentence, and the object named by the symbol which leaves
> the gap. 

No, that simply does not follow. Simply take the predicate to be: - was
so called because of his size. There is no relation that obtains between
this predicate and the denotation of the name substituted into the blank
iff the resulting sentence is true. Otherwise The Fridge would both
stand in this relation to the predicate and fail to stand in this same
relation to the same predicate, which is impossible. The schematic form
of Leibiniz's Law of the Indiscernibility of Identicals (as opposed, as
Bill Tait points out, to the Indentity of Indiscernibles, which is more
controversial) is therefore completely implausible. No-one, so far as I
know, has ever endorsed it.

> How else do you explain the difference between "someone is bald"
> and "Socrates is bald"?  You want some story like: the proper name picks out
> an object (Socrates) such that when the gap is filled up by the name, then
> the sentence is true iff Socrates is bald.  But then that's a relation.  The
> name refers to the object, the name fills the sentence in the "right" way.
> Unless you want to give any different account of reference?

But of course I do not deny, in this case, that there is a relation like
the one you mention (and which is usually called "satisfaction"). I only
deny that we have reason to suppose that there is such a relation in
every case which can be used to give the semantics of the relevant sort
of construction. In which cases satisfaction can be used to give the
semantics and in which cases in cannot is the difficult question that
was discussed earlier.

> Richard:
> > Most people who work on this problem take themselves to be focused upon
> > that-clauses in general, not on belief-reports in particular. Anyone who
> > is familiar with the literature knows as much.
> Show me an article that deals with that-clauses in reports about evidence
> and the like.

I don't see why any sensible semanticist would focus upon the use of
that-clauses in evidentiary claims. Indeed, my point was that much of
the best work on complement-clauses in semantics is focused on the
interpretation of complement clauses in general, not upon how they
function in the one context of belief-reports, which is of no special
interest, from a semantical point of view. My own view, for what it's
worth, is that so-called ILF views are on the right track. These do not
deliver the counter-intuitive consequences that Soames's view does. But
the matter is, of course, extremely controversial.

On the other hand, and on a more conciliatory note, perhaps your idea is
to give the following sort of argument against Soames et al: You can't
make any sense of evidential relations if you treat propositions as
Russellian. Or: You can't make the kind of sense you need to make of
evidential relations if you are going to do epistemology. If you could
develop such an argument in detail, that would be a real accomplishment.
As it happens, I have for some time been working on an argument along
these lines, though I tend to frame it in terms of an agent's reasons
for her beliefs, or even in terms of mental causation. But I have not
absolutely convinced myself that the argument will work, so it sits in
my pile of notes....

But we are, and long have been, far beyond anything of relevance to FOM,
so I shall say no more.

Richard Heck




More information about the FOM mailing list