[hist-analytic] Davidson's Hume

Roger Bishop Jones rbj at rbjones.com
Wed Jun 3 03:25:22 EDT 2009

On Friday 29 May 2009 16:02:06 steve bayne wrote:

[RBJ wrote]
>> "'but a lot depends on whether 'the number of planets' is, purely,
>> referential.'Indeed.  But why should we suppose that it is?"

> My point is, merely, that there is ambiguity.
> If we take the description one way, we get
> one view of its status and another view if read another way. As for the
> "right" reading that's much like the de dicto/de re situation. Some verbs I
> can read only one way, transparent; others I can read either way. Here is a
> description we can read either way and what follows depends on the
> reading.

Perhaps you can read a description "either way" but I don't
myself think that the two readings are equally plausible,
and I don't think there is substantive ambiguity in this case.

I will assume that you mean by calling a description
"purely referential" that it is in Kripke's terms
a "rigid designator" (correct me if I am wrong).

It seems to me plausible that names (real names, i.e. proper nouns)
are rigid designators, but not plausible that descriptions are
(generally!). Here's why:

The fact (if it is one) that a name is a rigid designator
means that to understand the name you have to discover what
it refers to, and you do not need to know anything specific
about that thing, any way of identifying it will do.
Kripke has these stories about christenings I believe.

If descriptions were rigid designators then we would have
to understand them in like manner.  We would have to know
when the definitive use of that description had occurred
and what in that context it had referred to.
I don't believe that is the way we understand descriptions
and if it were it would be disasterous for mathematics,
and we would have to invent a new kind of description
which was not rigid.

The language works nicely if descriptions are not
rigid and names are.  Then if you want a rigid
designator for something you give it a name.

Thus, if we say:

Let n be the number of planets.

Then, necessarily n>9.

Unless the number of planets <10, in which
case necessarily n is not >9.

But if the descriptions were themselves rigid then
we would be in trouble.

Nothwithstanding "the Colossus of Rhodes",
which probably is rigid.

Roger Jones

More information about the hist-analytic mailing list