Re: Explaining "explain".

[Leigh Palmer <palmer@sfu.ca>]

> In order to minimize semantic entanglements, I propose that there is a sense
> in which Physics both "explains" and "describes" reality.  What does it mean
> to "explain"?
>
> I propose that to explain means to describe in terms of ideas already
> explained.  We avoid a "regressio ad infinitum" only by ending with
> something which needs no explanation,  either by definition (God) or because
> we have grown used to it and have ceased to question it.
>
> In all of this the criterion is not "truth", but usefulness for
> describing/explaining reality in human terms (ie.; ultimately God and/or
> other accepted notions/processes).  This includes both empirical and
> conceptual usefulness.  Note that the usefulness criterion is not
> fundamentally disturbed by the assertion that the axioms are "unexplained".

I certainly accept Bob Sciamanda's view, expressed above. It is what I
mean when I say I am explaining something. What I don't wish to forget
at any time is that at the foundation of all our explanations lies a
small number of principles that are tentatively accepted on faith. In
my lifetime I have seen such articles of faith overturned in the
sciences, yes, even in physics. We must always recognize the potential
ephemerality of the tenets of our Faith, or it will truly become a
religion. (Yes, I do think my belief system is better than religion,
and I do not apologize for that opinion.)

My point is that we explain Nature within a belief system. Our
criterion for judging the validity of an explanation is that it
describes a natural phenomenon successfully. For example, Newton's
law of universal gravitation, together with his laws of motion, are
sufficient to describe the motions of all of the planets with periods
greater than one year to the limits of our capacity to measure those
motions. We know that both universal gravitation and the laws of
motion are flawed, but no one would propose explaining the orbit of
Jupiter (the first one for which Kepler's earlier explanation fails)
using general relativistic quantum mechanics, and only partly because
general relativistic quantum mechanics doesn't yet exist. (The Kepler
problem hasn't even been solved in ordinary GR.) The older methods
suffice to describe the motions sufficiently well that spacecraft can
be targeted and maneuvered with great accuracy using them alone. The
old ways provide an excellent description of the motion of Jupiter
even though we know that they are certainly not correct, and thus
cannot even explain that well observed motion in terms of currently
accepted axioms.

Fermat's principle (or the closely related principle of least action)
is itself as close to being one of those fundamental axioms as any
other I know. Any neoplatonist would rush to embrace it and to defend
it as an article of faith. The response I made was to the rejection
of an explanation made by invoking that principle. If that was not an
explanation then nothing else is, either.

Leigh