[extropy-chat] what is probability?

Benjamin Goertzel ben at goertzel.org
Tue Jan 9 01:48:01 UTC 2007


> > if you did some background reading in the philosophy of science it would
> > make more sense to you
> Probably, but I doubt it would be worth my time. If I was able to translate
> your remarks into English I have a very very strong hunch it would be one of
> 2 things:
> 1) True but trivially obvious.
> 2) Wrong.

Well, it is simply BS that all of "philosophy of science" ever written
is either trivially obvious, or wrong.

Let me ask you this: How would you tell the story of the "disproving"
of classical gravitation theory and its replacement by Einsteinian
gravitation?  I will be curious to compare your story with the
analyses of Lakatos and other historians/philosophers of science,
whose work you denigrate as "obvious or wrong" **without even being
familiar with it**.

> But I'm not just picking on you, I think that is largely the case for all
> philosophers of science. The thing is, if they were really onto something,
> if they really did have a better understanding of the scientific method then
> mere scientists you'd think they would have made major contributions to our
> understanding of how the universe works just like those silly scientists
> have. But I can't think of a single philosopher of science that has done
> that. Isn't that strange? They remind me a little of movie critics who
> go on and on why a movie is terrible but are incapable themselves of
> making even a crappy movie.

I don't know why you view philosophy of science as being about arguing
that scientists are "silly"?  That may be the perspective of some
philosophers of science, but only a small minority.

Your argument does not hold up at all, anyway.  To riff on your "movie
critic" analogy ... For instance, there may be a physicist who
understands more about the physics of golf than Tiger Woods, yet
cannot play golf as well as Tiger Woods.  Declarative knowledge about
science [or golf[ does not necessarily translate into procedural
ability to do science [or golf] (to use some lingo from cog sci).

I am primarily a scientist, not a philosopher of science, personally.
I became interested in philosophy of science out of a genuine interest
in understanding the enterprise of science better.

Note that up until the last century or so, many eminent scientists
**were** also philosophers of science.  The rigid division between
science and philosophy is a relatively recent invention, and one with
plusses and minuses.

> > Philosophy of science has its jargon like any other discipline, and as
> > usual, replacing the jargon with everyday words buys apparent
> > comprehensibility at the cost of precision.
> That is true for real science, but in social science and philosophy jargon
> has a quite different purpose, to conceal the fact that what you are trying
> to say is so obvious it's a downright cliché or it's just plain stupid.

Actually, the main reason for specialized vocabulary in philosophy is
the desire for precise expression.

I suppose that if you take a precise statement in philosophy and
translate it crudely into imprecise English, sometimes you may come up
with a cliche'.  So what.  Even if so, that doesn't prove that the
precise philosophical formulation was a cliche'.

> > The point is that assessing the truth or falsehood of a theoretical
> > scientific statement is not a simple thing.
> Excellent, you did not use one word of jargon and yet your sentence was
> clear precise and true. Unfortunately it is also obvious.

Yes, that statement was obvious.  What some philosophers of science
have done is to probe into the particular nature of the assessment of
truth or falsehood of scientific statements.  Because their
conclusions have NOT been so simple and obvious, they are not so easy
to concisely describe in conversational English.

> > Criteria for validation vary from one scientific approach to another, e.g.
> > cultural norms
> So in some cultures this bridge will collapse if I march over it, and in
> other cultures it won't.

Consider the mass of the top quark, for example.  This is calculated
by averaging the mass estimates obtained thru various observations of
the top quark.  However, there is some artfulness and judgment
involved in defining which empirical observations are to be considered
actual observations of the top quark, versus which are to be
considered noise generated by the experimental equipment.

It turns out that some physicists who think the top quark has a
different mass than the mainstream of physics -- also would like to
count some additional observations as top quark observations rather
than noise, relative to what the mainstream would like.  So, the
"empirical data points" utilized by physicists, in this case (as in
many others) turn out to be somewhat theory-dependent.

Next, consider General Relativity.  It is accepted by physicists
primarily because of its mathematical intuitive elegance.  There are
(obscure) classical theories of physics that do not involve curved
spacetime yet that also explain the same observations it does.  But,
they are uglier.

In principle, yeah, we could do science by
-- collectively gathering a huge table of data
-- agreeing which data points are valid and which are not
-- formalizing in great detail exactly how the data points were
gathered, i.e. the correct interpretation of each data point relative
to our shared experiential understanding of shared empirical everyday
-- then assessing various patterns in the data points.

Arguments between rivals would then come down to arguments regarding
which patterns were more prominent in the dataset, which would not
always be resolvable since there is no objective definition of the
"prominence" of a pattern, there are many different definitions.

But in reality this is not how science works.  Many of the data points
scientists look at are in reality theory-dependent, as in my top quark
example.  And in reality, many of the criteria people use to judge
theories have to do with their sense of beauty and elegance, as in my
General Relativity example -- not just with the prominence of the
theory as a pattern in a huge dataset.

So in reality, what happens is that we have a collection of scientific
research programmes, each one consisting of a body of theories and a
body of data accepted as "real", and gathered and interpreted
consistently with the theories.  One of these research programmes is
refuted only when it gets boring, or dramatically fails to provide an
intuitive explanation for some major phenomenon.

This is how science really works.  For better or for worse.  You can
maintain your idealization of science if you prefer, though.

-- Ben G

More information about the extropy-chat mailing list