At 06:59 PM 08/13/2000, you wrote:
>I followed the link given by Wesley, to Jonathon Wells's web pages. Besides
>the usual discredited ID arguments, I found one that I haven't seen before:
<snip of the argument and Richard's excellent reply to it>
Chris
The argument is based on the idea that, if the DNA does carry the
information, it must be in a kind of "blueprint" of the resulting organism.
This in turn seems to derive from early crude metaphors used to explain
genetics to neophytes. Richard correctly points out that all the
information that is needed to produce a cake is the recipe; we don't need a
picture of it, or a schematic of it.
Technically, a better metaphor might be that of a computer program that
directs the development of the organism.
But, either way, "blueprints" and "floor plans" are not necessary.
Interestingly, we don't even *have* to have blueprints in cases where we do
in fact use them. The entire procedure for specifying the building of a
house, for example, could be encoded in a sequence of instructions that
would never include a picture.
I bring this up because I have in fact been recently working on what I am
calling, for lack of a better term at the moment, a Universal Rigorous
Specification Method, or URSM. This would be a generalization of a design
technique used for efficiently designing provably correct software. By
suitably constraining the primitive operations and control structures, a
design of a computer program can be mechanically (and rapidly) verified to
be internally correct and complete (or not, of course). For the interested,
James Martin's book, "System Design from Provably Correct Constructs"
describes this method in detail.
What I am attempting to do is generalize the technique to the rigorous
specification of *anything*, including physical structures, systems of
ideas, and, of course philosophical and scientific *theories*. Obviously,
both evolutionary theory and ID theory (such as it is) are candidates for
such formalization and verification of internal correctness and completeness.
But, one of the issues I've been considering is that of specifying things
according to how they may be constructed vs. according to identifying the
components and relationships among the components. Even a "blueprint" can
be specified as a computer program that would cause the computer to print
out the resulting image (and yet, if you examined the program, you would
find *nothing* analogous to the resulting printed image).
Besides the obvious technical value of a rigorous method of specifying
things, there is the added major benefit that a theory specified in such
terms is utterly unambiguous. There would be little or no room for the
likes of Jones and Johnson to pretend that they made "honest mistakes" when
they create their farcical parodies of evolution to refute.
This is one of the reasons why I hope to write some computer programs to
prove the viability of various aspects of evolutionary theory (such as the
fact that random variations *do* generate new information (in fact, a
random variation is *usually* new information, relative to the "parent"
information that is being randomly varied)).
But, a URSM specification would do as well, partly because it would be even
more rigorous than a computer program as a specification of what evolution
is and how it works, and partly because it could be converted to a program
mechanically, so that, in effect, it would be both a rigorous specification
of evolution (or core aspects of it) and a computer program to re-create it
in software.
And, no "floor plans" or "blueprints" would be allowed.
Chris Cogan
This archive was generated by hypermail 2b29 : Mon Aug 14 2000 - 10:17:45 EDT