Unraveling the Ideas of William A. Dembski

William A. Dembski (dembski@discovery.org) has written a long tract Intelligent Design as a Theory of Information (copy made 2001 May 29) in which he claims to use information theory to show that so called "intelligent design" is the only explanation of information in biological systems. The bottom line is that he claims that natural selection cannot generate information. This is completely disproven by my paper Evolution of Biological Information. Given that an organism already exists, new information is easily gained by random mutation and selection. The computer simulation, which is freely available, completely destroys Dembski's argument.

[2001 May 29]. Evidently Dr. Dembski disagrees with about the above claim. Dr. Dembski talked at a creationist meeting, "Design, Self-Organization, and the Integrity of Creation" Conference (24-26 May 2001, Calvin College, Grand Rapids, MI). Someone who was there wrote this note: "Tom Schneider claims he can generated information from scratch, but actually information is being inserted." These notes were taken quickly and so may not be accurate. My correspondant said:

If you would like to see exactly what Dembski said, you can order a videotape ($25) or audiotape ($7) of his talk from

Esther Waid
c/o Calvin College Bookstore
Calvin College
3201 Burton St. SE
Grand Rapids, MI 49546

Add $3.00 for shipping and handling. Specify "Seminars in Christian Scholarship", Spring Conference, "Design, Self-Organization, and the Integrity of Creation", May 24-26, 2001, William Dembski talk.
On-line book order form.

Presuming that the note is accurate, I challenge Dr. Dembski to:

1. state the claim in written form so that anyone can evaluate it. Publication in a peer reviewed scientific journal would be appropriate. If a paper on the topic cannot be published, an email to me will do (see note below).
2. state whether or not he had carefully read the paper on the topic before making this claim.
3. state whether or not he had examined the documentation for the ev program before making this claim.
4. state whether or not he had examined the ev program code before making this claim.
5. state whether or not he had run the ev program before making this claim.
6. explain what he means by "information".
7. explain how he would measure the information in binding sites of proteins on DNA.
8. Identify the location in the ev code where the information is "inserted".
9. Explain the mechanism of "insertion".
10. Explain specifically why the ev program does not simulate what occurs in vivo.

2001 June 6. On 2001 June 6 William A. Dembski posted some statements about the ev program. Basically he is objecting to the selection function. (He calls it a fitness function, but that is not correct. It merely defines which organism is to survive relative to the others.) He neglects to point out that this is exactly the same selection function that occurs in nature. He needs to read more molecular biology. For further details see my rebuttal.

His other objection is to the SPECIAL FUNCTION described in the code. I inserted this to avoid bias in the selection from the sorting routine. It should (theoretically) not affect the final results. This was tested. Dembski's claim was found to be incorrect.

2001 June 7. Wesley R. Elsberry has a very extensive page on William A. Dembski.

Tom Schneider's Home Page