Re: Roles and dependence (Pat Hayes)
Message-id: <>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Date: Wed, 20 Sep 1995 17:46:09 -0600
To: (John F. Sowa),,
From: (Pat Hayes)
Subject: Re: Roles and dependence
Precedence: bulk
At  6:01 PM 9/19/95 +0500, John F. Sowa wrote:
>Pat Hayes:
>>Ah, careful. It does, I suggest, imply that some THING must be an instance
>>of another prediacte and have a relation to X. For example a Person must
>>have their time of birth. Every concept we have gets its meaning from the
>>network of relations that bind it to other concepts. There arent any
>>'primitive' ones. 
>We must, however, draw a distinction between what is physically necessary
>and what is logically necessary.  There is no logical reason why a
>human being might not be an uncreated immortal living in a vacuum
>without any need for sustenance.  That is, in fact, some people's
>concept of God or of the blessed who are resurrected at the end of
>the world.  That idea might be physically hard to swallow, but it is
>not self-contradictory.

Very, very little is *logically* necessary. But the point of my
purple-orange example is that if we change too many of the 'extrinsic'
properties and relations of something, it gets to be simpler just to say
that it isnt one of those things any more. I'd be inclined to say that a
thing living eternally in a vacuum, etc., wasnt a human being, for example.
The idea may not be logically impossible, but it is very hard to reconcile
with other things I know about human beings.

Side remark: discussions of counterfactual conditionals often, it seems to
me, fail to appreciate the need for consistency with the entire mass of our
knowledge. Eg its one thing to say, if Hannibal hadnt crossed the alps,
then...; but I find it impossible to imagine Putnam's 'twinearth' world in
which H2O is replacedf by XYZ, since I know that no other molecule would
have the properties of water. Replacing H2O with xyz is a logical
possibility, but doing that and believing in chemistry isn't. 

>>I report here the complete definition of (strong) conceptual dependence
>>taken from Simons' book (see DD2 p. 297, DD5 p. 303)
>>A predicate A is (strongly) conceptually dependent on B iff:
>   necessary ( (A x) implies (exists y) ((B y) and not (y Part-of x)) ) and
>   posssible ( (exists x) (A x) ) and not necessary ( (exists x) (B x) )
>Note the modal operator "necessary".

If that is logical necessity, this definition doesnt seem to work. For
example, it is never the case that logical-necessary ((exists x) (B x) )
unless B is tautological. I assume its meant to be more like some kind of
practical necessity. 

>Although everything in the universe may be related to everything else,
>there is a big difference between the various kinds of relations.
>Let me use the terms intrinsic/extrinsic instead of the terms
>By extrinsic relation, I would therefore include all influences of
>anything outside my own body and any historical event that may have
>occurred at any time prior to the present. 

Heres an intellectual exercise. Consider ancient Rome around the time of
Caesar's assassination by Brutus. Now, try to imagine a 'twinearth' version
in which Caesar and Brutus are reversed, but all their extrinsic relations
and properties are kept the same. In this alternative world, theres a guy
who kills another guy, and the killer is in all external respects like
'our' Brutus (he has the same date of birth, mother and father, career,
memories, he looks the same, he smells the same, everyone calls him
'Brutus' and he recognises that as his name, etc.;) and the one killed is
in all external respects like 'our' Caesar. That world, in fact, is in all
observable respects completely indistinguishable from our world. Its just
that, unknown to each of them, Brutus is 'really' Caesar, and vice versa;
this world differs only in that the 'intrinsic identities' of Brutus and
Caesar have been somehow switched. Now, does this make any sense whatever?
I suggest not. There is no such 'intrinsic' property which is somehow
attached to a thing independently of all its relations to other things.

>.... To get back to Peirce's triads, ....



Beckman Institute                                      (217)244 1616 office
405 North Mathews Avenue              (415)855 9043 or (217)328 3947 home
Urbana, Il.  61801                                     (217)244 8371 fax