Re: Responding to all those messages
phayes@cs.uiuc.edu
Message-id: <199305101726.AA22591@dante.cs.uiuc.edu>
Date: Mon, 10 May 1993 12:28:38 +0000
To: sowa <sowa@turing.pacss.binghamton.edu>
From: phayes@cs.uiuc.edu
X-Sender: phayes@dante.cs.uiuc.edu
Subject: Re: Responding to all those messages
Cc: cg@cs.umn.edu, interlingua@ISI.EDU
John, I know I said I wouldnt say anything else, but...
Your new talk of 'depictions' is just an example of the kind of muddle
that is going to emerge if we let you get away imposing your constructivist
religion on our formalisms. These things are not properly defined
anywhere, are supposed to be made of symbols but to be something like
images; like databases but able to play the role of possible worlds,
to be built of datastructures but in 1:1 correspondence with reality,
and to act as a kind of unifying information blackboard in a robot.
To insist that these strange things MUST serve as (or even be involved
in) our semantic theories is ridiculous. I simply reject these
things: to hell with them, I don't need them. The only reason they have
been hypothesised is because John Sowa has religious objections to
the established and widely used semantic account of Krep formalisms.
Let me briefly mention some of the problems with making sense of
these ideas. First, a TMT model (a possible world) is not a
representation, it is an account of how a representation might be
understood to mean something. If a database is taken to assert all
those facts - treated as a collection of ground assertions - then
its operational role is very different from that appropriate for a
model (eg as a counterexample). This distinction, between assertion
and counterexample, has been understood and used in reasoners for
25 years. Second, you merge the ideas of image-like representation
and database; but these are very different, at least in the way that
'image-like' is usually understood, in which such a representation
represents by being in some ways structurally similar to the world
it describes. Databases don't usually have 3-d geometry. Third, much
of what we want to represent in our computers (and what we think about
every day) has no particular connection with perception, so
the architecture of robots should not be given a central place in
semantic theory.
We can argue about philosphical matters for ever, but the central issue
is that you refuse to let me talk about how representations represent,
and I insist on keeping this clear. You believe I am muddying
the purity of Tarski's thinking with unattractive details of pattern
recognition, and exhibit a zealot's fierceness in defending him.
But this is a mistake, arising from a position your formalist
tastes have brought you to.
This entire discussion started because you asserted that you would
not permit people to say that representations could modelled the
world. They can.
Pat Hayes
PS. One last remark . I don't want to start another >>/>>> battle, but
in your last message you adopted a pose of 'reasonable man' which suggested
that your side in this argument is the one which properly interfaces
with all of computer science. I reject this implicit claim also. We must
check with Len, but I think you misunderstand Schubert. Your mappings to
databases do not respect their functional meanings (see Reiter on this),
and you cite robotics, virtual reality, etc etc without any authority.
Existing robot architectures use many representations and translate between
them in complex ways which deserve study and analysis, but do not routinely
take some to denote others, as you insist must be done (on metaphysical
grounds).
OK, thats really all for now from me.
----------------------------------------------------------------------------
Beckman Institute (217)244 1616 office
405 North Mathews Avenue (217)328 3947 or (415)855 9043 home
Urbana, IL. 61801 (217)244 8371 fax
hayes@cs.stanford.edu or Phayes@cs.uiuc.edu