Newsgroups: comp.ai.philosophy,sci.cognitive
Path: cantaloupe.srv.cs.cmu.edu!nntp.club.cc.cmu.edu!miner.usbm.gov!rsg1.er.usgs.gov!stc06.ctd.ornl.gov!fnnews.fnal.gov!uwm.edu!math.ohio-state.edu!howland.reston.ans.net!news.sprintlink.net!cs.utexas.edu!utnut!utgpu!pindor
From: pindor@gpu.utcc.utoronto.ca (Andrzej Pindor)
Subject: Re: FIRST order? was: why Ginsberg grouses
Message-ID: <DCnCK1.4C3@gpu.utcc.utoronto.ca>
Organization: UTCC Public Access
References: <3v63ad$p14@nntp5.u.washington.edu> <DCDusL.7Lr@gpu.utcc.utoronto.ca> <3veo62$cvb@percy.cs.bham.ac.uk>
Date: Tue, 1 Aug 1995 19:24:00 GMT
Lines: 107
Xref: glinda.oz.cs.cmu.edu comp.ai.philosophy:31186 sci.cognitive:8765

In article <3veo62$cvb@percy.cs.bham.ac.uk>,
Aaron Sloman <A.Sloman@cs.bham.ac.uk> wrote:
>pindor@gpu.utcc.utoronto.ca (Andrzej Pindor) writes:
>
>> Date: Thu, 27 Jul 1995 16:21:57 GMT
>> Organization: UTCC Public Access
>
>>...
>
>> My interpretation of Popper is as following: a theory is considered
>> a scientific theory if it makes predictions amenable to experimental
>> verification. If predictions are not confirmed by observations, the theory
>> is falsified.
>> Our minds can create all sorts of things. The falsification, as described
>> above, is an only method of making sure that our concepts correspond to
>> reality and not to an internally created fantazy land.
>> ...................
>
>I think that if you read Popper carefully you will find a number of
>additional points.
>
>(a) He claims merely to be making a demarcation between what is and
>is not empirically testable. The fact that something is not
>empirically testable does not make it "an internally created fantasy
>land". He does call some such things metaphysics, but for Popper
>metaphysics is not always bad.
>
Whatever you call these untestable things is of secondary importance. My main
point is that experimental tests is what imposes restrictions on our models
of nature, which cannot be gotten around. Parts which are not restricted can
be built almost any way we want, this is why I called it "fantasy". And even
though I used the words "a bad literature", there is also "a good literature".
I did not mean to imply that methaphysics is always bad, rather wanted to 
stress that if there are no empirical ways to decide (falsify) one view or
another, claiming that one is right and the other one is wrong is not 
a scientific claim. There is no reason to get all worked up about such
controversies, there are a metter of personal taste, like religious views.

>(b) Although he says that some statements that are not empirically
>testable should be called metaphysical, he acknowledges that they
>can play an important role in the early stages of a science, for
>instance the earliest (ancient Greek) forms of the atomic theory of
>matter.
>
Absolutely, but they play this role only if they do not rule out a possibility
of an empirical verification. 

>Lots of people have found problems with his demarcation between
>science and metaphysics. For example existentially quantified
>statements such as "Electrons exist" may or may not be
>confirmable but they definitely are not falsifiable. That's because
>no amount of observations of facts of the form ~P(x) cannot refute
>    (Ex)P(x).
>
Yes, it is diffcult to prove that something does not exist, but a statement 
that something exists only makes sense if it can be falsified, i.e. if it leads
to experimental predictions which can be checked out.

>As far as I recall, all attempts to find a precisely definable and
>satisfactory logical demarcation between science and metaphysics
>have foundered, and that's one of the reasons why Lakatos, Popper's
>pupil, favoured long term evaluation of research programmes rather
>than cut and dried classification of statements or theories.
>
I have no quarel with the above, very few things in real life are cut and dry.

>My own view of science (expounded at length in Chapter 2 of The
>Computer Revolution in Philosophy, Harvester Press and Humanities
>Press, 1978) is that the deepest advances in science are those that
>extend our ideas of what is possible. E.g. Newtonian mechanics
>suggested that it was possible for objects with no connecting medium
>to exert forces on each other; the special theory of relativity
>suggested that it was possible for an object's length to be
>different for different observers moving relative to each other;
>quantum physics suggested that it was possible for the same thing to
>behave partly like a wave partly like a particle; and AI shows that
>various kinds of processes governing behaviour are possible that
>were never thought of previously.
>
Perhaps I would not formulate things this way, but basically I agree.

>Claims about what is possible, like claims about what exists, cannot
>be refuted by observation (all you can argue is that attempts to
>create or find an instance "so far" have failed) and they often
>cannot be conclusively demonstrated empirically because often
>difficult new concepts are involved and the observations are open to
>different interpretations, though if even one example is found the
>claim is conclusively established. (Which is not the case for claims
>about laws of nature, which most people mistakenly take to be the
>central stuff of science, not noticing that the role of laws is
>to delimit previously conceived of possibilities.)
>
I agree in principle, although with some reservations, see above comments about
criteria for something to exist.

>> Andrzej
>> --
>Aaron Sloman, ( http://www.cs.bham.ac.uk/~axs )

Andrzej
PS On Thursday I am going away for 10 days, so if you would like me to know
your reply, please email it to me.
-- 
Andrzej Pindor                        The foolish reject what they see and 
University of Toronto                 not what they think; the wise reject
Instructional and Research Computing  what they think and not what they see.
pindor@gpu.utcc.utoronto.ca                           Huang Po
