The last section of Dembski's book has an interesting section on the
comparison of computer programming to evolution with a critique of Orr on
this issue. Orr claims that there is a resemblance between evolution and ad
hoc character of much programming.
This is by no means a new criticism by Dembski, and in fact anyone who has
programmed a fairly large size program (preferably from scratch, so you don't
forget you designed it) knows the point Orr is making, but that this is
misleading in the extreme.
This is then computer hacking of the worst kind, but it's true people program
this way, whatever the books on program design say.
Beyond a certain complexity, survival of the workable program line, by
increments, against rejection of the debugger is the only way to get
OK. But there are exceptions to this, plus the fact that you must know the
overall nature of the program, just to stay out of trouble. Although that is
still al. Further, the initial loop_main must be designed properly from the
start, and that is very unlikely to be done in increments. Far from it, at
that stage, the overall picture is still clear and must be created as one
And these increments of code are increments of meaningful code, but they are
The list of false comparisons goes on and on, and the irony is that the
analogy works perfectly the other way around.
Evolution could be a lot like programming with a balance of the two factors,
design and incremental expansion.
Website on the eonic effect
[log in to unmask]