Re: [asa] YEC and ID arguments

On 10/22/06, Dawsonzhu@aol.com <Dawsonzhu@aol.com> wrote:
>
> dopderbeck@gmail.com wrote:
>
>
> That's why Dembski talks about a "design inference," just like your
> inference concerning the latent variable. In fact the method of inference
> you describe sounds just like the method folks like Dembski and Behe
> describe.
>
>

In addition to what Wayne has written below, which seems very sensible to
me, I would add that there is a subtle but very important distinction
between the kind of inference I'm talking about and the kind that (I think)
Dembski is talking about. Dembski specifies a "universal probability bound"
stating that events with a probability less than 10^(-150) are impossible,
and therefore must be due to external intervention (there is a whole lot
more subtlety than that - embodied in the notion of "specified complexity",
but I won't go into that now - I have in the past on the list).

But there is a long-running "religious war" in the field of statistics
between so-called "Fisherian" or "frequentist" statistics, and the more
modern formulation, namely "Bayesian" statistics. Dembski is, I believe
opposed to Bayesian statistics.

Latent variable theory makes extensive use of Bayesian techniques in order
to infer the value of the hidden variable.

But I think Bayesian theory exposes the weakness of Design arguments. Let's
go right back to Paley's watchmaker. In Bayesian statistics, you start with
a "prior probability" which represents your degree of belief (or doubt)
about the value of a variable (and the variable may be binary-valued as in
true/false). After observations are made, Bayes's theorem gives you a
framework from which you compute a "posterior probability", reflecting the
way your belief has changed given the observation.

Now, with Paley's watch that he finds in the field, ones prior belief in the
existence of the watchmaker has probability 1. We know humans who make
watches exist.

For an atheist, the probability of God existing might well be zero (I
actually met an atheist who asserted this). So however incredible seeming
the event is (existence of a molecular motor), the atheist's posterior
probability is still going to be zero.

So I think the Design argument only works if you already believe in God.
And if you already believe in God, why do you need to make the argument in
the first place?

The only possible utility for an ID argument might be to strengthen one's
degree of faith - the posterior probability is greater than the prior
probability. However, I think this also is fraught with danger. Suppose
someone comes up with a naturalistic explanation of the flagellum, which
means it's a whole lot less improbable than you thought before. Does that
then weaken your faith again?

By contrast, consider the following. On my "windows wallpaper", I have a
stunningly beautiful picture of the planet Saturn, showing the rings at an
unusual angle, and one of the satellites above the rings. It was one of the
first images from the Cassini probe. The NASA website dubs the picture
"nature's canvas". I prefer to call it God's canvas. On top of the
picture, I have superimposed the words from Psalm 143 "I meditate on all
your works and consider what your hands have done". I see no problem in
attributing it to "God's hands" even though I know that God didn't
physically form the planet with his huge hands - that most likely it
condensed out of gas from the remnants of a supernova. I don't need an
argument that says how incredibly unlikely it is to attribute it to the
hands of God.

In a quick executive summary:
> Dembski's aim in "The Design Inference" was to find a base line
> where you could say, "this cannot occur by chance". To discuss
> his idea as he worked toward this goal, he used well defined
> and indisputable examples such as tossing a fair coin, rolling
> fair die, or shuffling a fair card decks. They were all
> probabilities that can be decided irrefutably and with absolute
> certainty.
>
> Now, technically, there are some uncertainties if you enter
> in the physics of shuffling, rolling a die, what surfaces
> are used, past history, individual factors, etc. This would
> be much closer to what Iain is talking about, I think.
>
> Behe does not provide any quantitative estimation for his arguments
> (at least in Darwin's Black Box). However, I suppose you could
> take irreducible complexity to mean "the likelihood of structure X
> appearing
> by chance crosses Dembski's baseline", and therefore, you "infer" design.
> However, I recall the rhetoric was certainly a bit stronger than that.
>
> At any rate, we can understand that this is an application of
> fully estimable quantities. It is not dealing with the
> underlying uncertainties of real systems and unexplained
> factors that seem to influence the outcome. Gambling casinos
> basically use this. They can "infer" pretty well if someone is
> cheating because they know the odds and their "baseline" is
> well above the noise level where the physical uncertainties of
> cards and dice may become significant.
>
> A major problem is that, in application to biology, these
> quantities are not so easy to estimate (like card games),
> and I suspect are biased by all parties to their advantage.
>
> So, OK, I admit that Dembski does not say "this PROVES" directly,
> in what I have read, and possibly, we can take the word "infer" to
> mean a weaker position. However, somewhere between these statements
> and rhetoric that comes out elsewhere, this is what they are insinuating
> they have in their hand. And so, to anyone watching such a confrontational
>
> style as often comes out, this sure looks like they are saying "this
> PROVES".
>
> ... and what is it that all we religious folk really long to hear? So
> there you go....
>
> by Grace we proceed,
> Wayne

-- 
-----------
After the game, the King and the pawn go back in the same box.
- Italian Proverb
-----------
To unsubscribe, send a message to majordomo@calvin.edu with
"unsubscribe asa" (no quotes) as the body of the message.