Snoke's reply

From: Randy Isaac <randyisaac@adelphia.net>
Date: Tue Aug 16 2005 - 17:43:54 EDT

Continuing the dialogue I posted previously, this afternoon I received the following reply from Dave.
I may take a while before responding since I'd like to go back and do some more reading on some of these aspects. And hear some of your thoughts and suggestions.
Randy

You raised so many points in your last two letters that I hope you
don't mind that it has taken me some time to get back to you. I
really appreciate the deep thought you are putting into this, and I
hope that the ASA and ID communities will have substantial dialogue
and overlap.

As you have said in numerous ways, a lot revolves around the question
of whether "that looks designed" is a meaninful statement. We agree
that in the case of human origin, there is no question. You then
argue, however, that if we do not have an "original" to compare to,
then there is nothing we can say about design. That would seem to
reject the program of SETI, that even if the designer is alien, we
could define some basic characteristics of intelligent beings which
we could look for. Are you making that strong claim-- that SETI is
worthless, not just because the odds are low of finding alien
intelligent life, but because even if there is alien intelligent
life, it is intrinsically undetectable, because we have no originals
to compare to?

The argument that Paley and others have made is that "design", like
"heat", is something we can perceive and recognize intrinsically.
The scientific question is how to define that perception
quantitatively. At one time, heat was a purely qualitative
perception-- you "knew" when something was hot but couldn't say a lot
more. Only in the 1800's was there a quantitative theory of heat. But
long before that, I think people were being scientific in talking
about heat and its effects.

Dembski addresses a lot of the questions you raise. His "filter"
requires all three of the following for a design inference:
contingency, complexity, and specificity. I will not go into that
because he has done a good job himself. Where he is weakest, is in
the area of specification-- can we justly claim a specification
_after_ the fact. To do that he invokes a kind of "platonic"
specification, the idea that some things are just intrinsically
specified. For example, if we found a binary encoding of the number
Pi carved into ancient rock, we would take that as specified even if
no one predicted that we should find the number Pi.

There are other possible ways to try to quantify design. I wrote a
paper (unpublished) a while back, "Toward a Quantitative Theory of
Design"(available at http://www.cityreformed.org/snoke/design.pdf). I
encourage you to look up that article, since it addresses several of
the questions you raise, including how to define undesign in a
universe which is designed.

You may not like my approach any more than Dembski's. The larger
point is that I do not see an _inherent_ reason why "design" is
intrinsically unquantifiable any more than "heat." Just because we
have not got a universally satisfying standard does not mean it is
intrinsically non-objective. Both Demski's and my (much less well
thought out) approaches are stabs at that task.

Fundamentally, what attracts me to ID is that everybody I know, who
is in any way objective, when looking at the inner workings of the
cell, sort of gasps and says "that looks like machinery." This is my
experience over and over at biophysics talks-- everyone in the
audience is blown away by the apparent design. They are not using a
quantitative theory of design, but they "see" it anyway-- just like
they could feel hot, without having a thermometer or a kinetic theory
of heat.

I do not quite understand one issue you raise. You say that the
information in DNA is not seperable from its physical embodiment. But
I thought that was what the whole Human Genome project was about-- to
copy the information of the DNA into another form, namely computer
code. And I have been told about "designer genes" in which you can
type in any DNA sequence you want and they can grow it by injecting
it into bacteria DNA. So clearly the information is separable from
the DNA, though it is not usable in that form, just as a program for
running a robot is not usable until it is actually in the robot.
I also don't see how information being attached to a physical
embodiment reduces the design aspect. An artist might make an artwork
out of rocks, and we would say that a picture does not convey all the
information, you need the rocks themselves. But how does that make it
less designed?

Your analogy of the butterfly wings making the alphabet would clearly
fail Dembski's filter, and so would Bible Code. In each case has
searched through a large set of random things until one finds
something that fits. So specification is lost. Of course, if one went
to a zoo and found a cage of 26 butterflies who had every single
letter, one would conclude design. Or if one found a single butterfly
with all 26 letters.

I still think you may be missing the point of my analogy with the
luminscence rings. In the analogy, I found some aspects which didn't
agree with the standard theory, and in addition, some aspects which
_looked like_ something else. So two things are going on: 1) tearing
down an old theory, which as I argued, is scientific even in the
complete absence of a rival theory, and 2) the initial grabbing at a
new theory by way of picture models. That, I argue, is also
scientific. It is silly to say that my argument is "we don't
understand it, therefore it must be God." In my experiments, the data
"looked like" a puddle, and the excitation data "looked like"
something was hopping over barriers. In ID, stuff "looks like" it is
designed. By Occam's razor, one proposes a theory that "fits", that
invokes a picture-model consistent with what one is seeing. There is
no reason to invoke God's direct intervention in the case of
luminescent rings (or any other type of round puddle, for that
matter) because they don't "look" designed. But some things do.

Finally, in trying to propose a standard for predictions, I said we
could argue "God is like this, therefore I expect such and such." You
are right that we have to be careful with that, because nature
reveals who God is, so I can't force the data into a preconceived
mold of how God would do things on the basis of just wishful
thinking-- that is why some people have problems with carnivorous
animals. But I think it is valid, and scientific, to make predictions
the following way: I sample nature, find that in my sample, certain
things always seem to be true, make a generality, and predict that in
future samples these things will still be true. That is the essence
of the scientific method. So, for instance, I might find that in all
cases where I study, organs in the body have a purpose, and therefore
I expect that in cases of "vestigial" organs and "junk DNA" I will
find some purpose for them if I look hard enough. Maybe I will have
to modify that, to allow for some vestigial organs. But it would
still be worlds away from what would seem an evolutionary prediction
of lots of vestigial organs and random junk. Another prediction,
which I develop in my book "A Biblical Case for an Old Earth", is
"balance". I expect everywhere we look to find "cleaner-uppers," or
"destroyers," not just builders. We don't like that idea, but it is
all through nature and Scripture.

These are wild stabs at predictive theories, which are far from
complete, but I include them just to say that I don't see any
_intrinsic_ reason why ID cannot form the basis of a predictive,
quantitative theory.

I can't finish without taking a little poke at evolution. Jay Parker writes:

"I have a problem applying the concept of predictive science to any
kind of macro-species biology. That is, the examples I'm most
familiar with in other sciences, like physics, are richly predictive
because at root there is a theory of continuous equations (for
example, partial differential equations based on things like
inverse-square force laws). These equations allow for interpolations
between observations and extrapolations to combinations of conditions
not yet probed (with caution, but also considerable success).

"I'm not saying I deny biology can make any predictions; but I don't
know how to think about their range of applicability and determining
if tests are valid. Where's the math? Or, how do you make a strong
argument without math?"

I (DS) and many other ID advocates have often pointed out that
evolution itself is not a predictive theory. We are constantly
reading of major surprises that don't fit the model, but which are
worked into the model anyway. What we have is a documented set of
homologies-- things which are like other things-- and this is useful
for science. But I can't think of any cases in which a history of the
origins, i.e. a theory of a specific path of changes, was used to
generate useful predictions. But maybe Jay and I are just being naive
hard-sciences types and there are many such examples.
Received on Tue Aug 16 17:46:57 2005

This archive was generated by hypermail 2.1.8 : Tue Aug 16 2005 - 17:46:58 EDT