Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!seismo!mcnc!duke!mps
From: mps@duke.cs.duke.edu (Michael P. Smith)
Newsgroups: sci.philosophy.tech
Subject: Re: Logic and Coercion
Message-ID: <9975@duke.cs.duke.edu>
Date: Wed, 29-Jul-87 00:13:04 EDT
Article-I.D.: duke.9975
Posted: Wed Jul 29 00:13:04 1987
Date-Received: Thu, 30-Jul-87 05:00:45 EDT
References: <68@thirdi.UUCP> <9962@duke.cs.duke.edu> <573@rlgvax.UUCP>
Reply-To: mps@duke.UUCP (Michael P. Smith)
Organization: Duke University, Durham NC
Lines: 22
Keywords: explanation, transcendental argument, Kant, Nozick
Summary: Nozick claims his explanations not transcendental arguments

In article <573@rlgvax.UUCP> cliff@rlgvax.UUCP (Cliff Joslyn) writes:
[regarding Nozick's favoring of *explanation* over *proof* in philosophy]:
>
>I believe that this is very similar to something expressed by Kant in the
>Prolegmonena (although it has been many years since I read it), that
>explaning how something might be possible is really the task of metaphyics.

Indeed it is, but Nozick tries to distinguish his method from Kant's
in a footnote on page 15.  Here's his summation:
	There is a difference between explaining *p* via *q*, and
	proving *q* is the correct explanation of *p*.  A transcendental
	argument [Kant's technique] attempts to prove *q* by proving
	it is part of any correct explanation of *p*, by proving it a
	precondition of *p*'s possibility.

----------------------------------------------------------------------------
"As regards those who adopt a *scientific* method, they have the choice
of proceeding either *dogmatically* or *skeptically*; but in any case
they are under obligation to proceed *systematically*.   I. Kant

Michael P. Smith	ARPA: mps@duke.cs.duke.edu
----------------------------------------------------------------------------