Patrick Durusau wrote:
> John F. Sowa wrote:
>> There is no difference between those two claims:
>>> It is one thing to claim that a theory is compelling, etc.
>>> but quite another to claim that it correctly describes
>>> "the true nature of reality."
>> If you have a theory that makes correct predictions in every
>> case that anyone has ever been able to devise for many centuries,
>> then it "correctly describes the true nature of reality." That
>> does not mean it is a *total* description or that there are no
>> other equally correct theories. But it *does* mean that you have
>> discovered an important part of the truth.
> Why isn't it sufficient to say that a theory makes correct predictions
> in every case we have encountered? Isn't that enough?
> I am not sure what more you could ask of any theory, that is that for
> every case encountered it made the correct prediction.
> It certainly supports the use of such a theory should be relied upon as
> a basis for other theories.
> Moreover, I can use a theory that has always given the correct
> prediction whether or not I think it describes some "truth" or not. In
> other words, use of the theory does not depend upon its elevation to the
> category of being "truth." (01)
Yes. I think John has conflated a theory's being true and a theory's
being empirically adequate. a theory may perfectly fit the data and
allow for usable predictions, even if it is wrong about the nature of
the phenomena addressed. (02)
I thus support the statement that there *is* a difference between a
theory's being correct and its being accurate, and consequentially, that
there is a difference (beyond the obvious syntactic one) between saying
that a theory is true and that a theory is accurate. (03)
If you could prove that a theory necessarily makes correct predictions
in every possible case, you could claim that the theory is true (but I
am not still convinced this would be correct). You can't (not in
empirical sciences, perhaps in mathematics); theories are induced or
abduced from partial data (data about only some part of the reality).
They are not tested on every conceivable input, since what is the set of
all conceivable inputs is only another theory. (04)
> By the same token, if someone offers a theory that does not give correct
> predictions or that relies upon what I suspect are fanciful premises,
> i.e., airplanes fly because the motors attract fairies that carry the
> plane aloft, I can ignore that theory because it is less useful than
> some other theory. Such as calculating the thrust needed by a jet engine
> to propel the next generation of jet aircraft. I don't ever have to
> reach the issue of "truth" but can rely upon theories that I find useful
> for some particular task.
>> Newtonian mechanics, in fact, is such a theory. During the past
>> century, physicists have discovered phenomena of relativity and
>> quantum mechanics, for which Newtonian mechanics makes incorrect
>> predictions. However, for macroscopic phenomena at nonrelativistic
>> speeds (i.e., for cars and airplanes) Newtonian mechanics is a true
>> description of reality. I trust my life to Newtonian mechanics
>> whenever I drive a car or fly in a plane.
>>> A friend of mine sent me the following example:
>>> There is an elementary model in electrical engineering, called
>>> the "4-terminal network". The thing is a closed ebony container,
>>> with an input, an input return, an output, and an output return.
>>> The student is given a set of inputs and outputs, and asked to make
>>> the simplest thing that he can which could be substituted for the
>>> actual contents of the container. The problem of what is _actually_
>>> inside the container is dismissed as impossible.
>> That is an excellent example.
>> If anyone defines a theory (i.e., a set of axioms) that correctly
>> predict the output for every conceivable input, then I would say
>> that the theory is a true description of the behavior of that box.
> First, "every conceivable input" isn't really possible. Testing is
> always with a finite set of inputs.
> Second, I would say that a theory (set of axioms) that correctly
> predicts all the inputs we have tried is simply accurate. That is it
> agrees with the inputs and predicted outputs.
> Calling a theory "true" or "part of truth" doesn't make it any more or
> less accurate. The aether theory was at one time thought to be "true"
> but that did not save it from being superceded.
> I suppose that is what I am missing. What claim is it, beyond accuracy
> (agreement of theory with observations), that you want to make by saying
> something is "part of the truth?" (If any claim at all. I may simply be
> over-reading your statement to imply more than it is actually saying.)
> Hope you are having a great day!
Wacek Kusnierczyk (07)
Department of Information and Computer Science (IDI)
Norwegian University of Science and Technology (NTNU)
Sem Saelandsv. 7-9
tel. 0047 73591875
fax 0047 73594466
Message Archives: http://ontolog.cim3.net/forum/ontolog-forum/
Shared Files: http://ontolog.cim3.net/file/
Community Wiki: http://ontolog.cim3.net/wiki/
To Post: mailto:ontolog-forum@xxxxxxxxxxxxxxxx (010)