Re: [asa] The Parsimony Principle (aka Occam's Razor) and atheism

On 7/11/07, David Campbell <pleuronaia@gmail.com> wrote:
>
> > The argument goes something like this. Because the laws of physics and
> the
> > theory of evolution can explain, in principle, everything there is,
> > including all the complexity of life, etc, there is no need to invoke a
> > Creator to explain it all. A universe without a Creator is a simpler,
> and
> > therefore more parsimonious, explanation than a universe with a Creator.
> > Hence the god-less universe is to be preferred.
>
> This makes the assumption that the physical universe is all that there
> is, or at least all that's interesting. Thus, it's circular
> reasoning.

I don't think the argument as given here makes such an assumption - however,
the assumption is hidden in there for the reasons I gave concerning
probability theory. The parsimony argument applies to distinguishing
between alternate models as the best "explanation" of the data & the most
parsimonious model is also the most probable. So to be able to apply the
principle, one must indeed make God part of the model, IOW part of the
physical universe. This is the problem I've had all along with Dawkins's
arguments in The God Delusion - they are indeed, as you say, circular. Of
course there's no God if the material universe is all there is - by
definition in fact.

If one wants a solid basis for morality (not merely an
> explanation of its evolutionary origins but a reason why I ought to do
> something), or accepts evidence of things besides the physical
> universe, then physics, etc. do not explain everything.
>
> > However, I think this argument is flawed, for the following
> reason. Occam's
> > Razor, or the principle of parsimony is certainly to be used and
> encouraged
> > in trying to decide between different scientific models. For example, I
> > have a bunch of data and I have to fit a polynomial curve through it.
>
> Why bother with the model or the curve? Why not simply amass a pile
> of data points? All the real world information is there. Isn't a
> model-less universe simpler than one with a model? No, because
> science also seeks to produce models, and fewer general models is seen
> as more parsimonious and more desirable than numerous special case
> models or a mere pile of data.

This is because the combination of model + differences from the model
prediction admit a more compact description of the data than the raw pile of
data itself. The raw pile can't tell you anything about the nature of the
data, what laws are obeyed etc. But the fact that it approximately fits,
say, an exponential curve tells you quite a lot.

Continuing such reasoning, God
> provides an ultimate explanation of everything (though of course we
> can't directly reason from Him to, e.g., E=mcc) and thus including Him
> makes the model more parsimonious than excluding Him.

I don't find this valid in the strict probabilistic sense of parsimonious -
the kind of model I'm talking about can make a prediction of the value of
the data at any given values for the input parameters. But God as an
explanation can't as you say predict E= mc^2, or the gravitational
attraction between two bodies etc.

Another example of misuse of parsimony comes from some advocates of a
> particular approach (called maximum parsimony) to the generation of
> evolutionary trees. Some lose sight of the fact that the goal is to
> represent the data and advocate more parsimonious against less
> parsimonious (at least by the particular metric, which is not entirely
> appropriate as a comparison) but reflecting the uncertainty in the
> data.

Interesting - I was aware of this principle, but didn't know it was
considered a misuse. On the face of it, it does fit my expectation of a
reasonable use of it - trying to decide between different postulated
evolutionary paths. The probabilistic interpretation of Occam's razor
implies that the simplest explanation is the most probable (though that, of
course doesn't rule out the brute possibility that a less probable path may
be the one that actually occurred).

In his review of Conway Morris in Books and Culture, Dembski accepts
> the atheistic Occam argument and therefore claims that ID is necessary
> for religion.

Oh dear. But pretty much as I expected from my original post. God is not
part of the model - God defines the model. Hence Occam doesn't apply. I
like the following quote from H. Allen Orr's devastating review of The God
Delusion:

Orr writes:

Second, the fact that we as scientists find a hypothesis question-begging—as
when Dawkins asks "who designed the designer?"— cannot, in itself, settle
its truth value. It could, after all, be a brute fact of the universe that
it derives from some transcendent mind, however question-begging this may
seem.

--- indicating, it seems that he is unlikely to accept the atheistic Occam
argument either.

He goes on:

What explanations we find satisfying might say more about us than about the
explanations. Why, for example, is Dawkins so untroubled by his own (large)
assumption that both matter and the laws of nature can be viewed as given?
Why isn't *that* question-begging?

--- Maybe this does reflect your point that God as the ultimate explanation
of everything could be seen as more parsimonious than a pile of laws and
constants that just happen to produce matter and life as we know it.
(Though I still see it as a bit of a category mistake).

Iain

To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.
Received on Wed Jul 11 16:00:56 2007