November 19, 2001
On Tuesday, November 13th, at the University of New Mexico in Albuquerque about 500 to 600 people attended a remarkable event. William Dembski and Stuart Kauffman had a public encounter in which Kauffman, the preeminent self-organizational theorist of the Santa Fe Institute, publicly admitted that intelligent design was a legitimate intellectual and scientific project and that research projects like SETI (the Search for Extraterrestrial Intelligence) couldn't even get off the ground without it. To be sure, Kauffman was quick to add that he saw the application of design-theoretic methods to biology as unsuccessful. But ID's success or lack of it can be contested, and Dembski did an admirable job defending ID and questioning the adequacy of self-organizational methods to account for biological complexity.
Kauffman is precisely the sort of insightful critic the ID movement needs to develop a flourishing program of research. Kauffman stands in marked contrast to Massimo Pigliucci, whom Dembski had debated two weeks earlier at the New York Academy of Sciences (see previous update). Whereas for Pigliucci there is no legitimate scientific or academic debate about biological design, Kauffman made it clear that he had no patience for the sort of censorship of ideas that Pigliucci represents. Having himself withstood dogmatic Darwinists, Kauffman knows only too well the dogmatism within scientific ranks. He values dissent within science, for it is the only way to keep science honest. By this reasoning intelligent design is the best thing that's happened to science in a long time.
In the actual encounter, Dembski began by overviewing his methods for design detection as developed in The Design Inference. He then sketched how these methods apply to biology and why he thought they provided in-principle objections to undirected natural causes, including Kauffman's self-organizational processes, generating the key item that his methods detect, namely, specified complexity or complex specified information. Dembski developed his argument as follows: the only way for self-organizational, Darwinian, and other naturalistic methods to generate specified complexity is by in essence dissolving the actual complexity (or improbability). What seems highly improbable is therefore no longer improbable once one knows the right naturalistic process. Dembski indicated that there were good reasons to think that no such reduction of complexity or improbability would be possible in the case of systems like the bacterial flagellum. He also adverted to some of the techniques developed in his forthcoming book titled No Free Lunch for computing the probabilities of discrete combinatorial objects like the bacterial flagellum.
Kauffman went second. He began his presentation by openly stating that Dembski's project of design detection is perfectly legitimate. He then added that he saw it as not properly applying to biology. To argue for the sufficiency of natural processes to produce biological complexity he focused on self-organizational processes that can generate complexity without design. He started with examples outside biology, like Benard cell convection and random graphs, as an intuition pump. He then focused on some of his own research on autocatalytic sets, arguing that these could account for the complexity we see in biological systems. One noteworthy moment in Kauffman's presentation, though it might have seemed inconsequential to some observers, was where Kauffman wanted to emphasize the efficacy of natural processes to generate biological complexity and thus added that it be done "without shenanigans." Having said it, he stopped, and out of respect for the issues on the table, corrected himself and said "rather, without design." This shows that design is slowly going from an "unthinkable option" to a "implausible but not completely crazy option." That's progress.
The rebuttal time was friendly, with both Kauffman and Dembski "rigging" the first question to each other. Dembski asked Kauffman to elaborate on the No Free Lunch theorem and how he saw it as relevant to the discussion (he had mentioned it during his presentation but ran out of time to elaborate on it then). Kauffman knows this field well and gave a nice overview of NFL. Kauffman in turn asked Dembski how he made sense of biological specification at the level of the whole organism. Dembski indicated that this was a difficult problem at that level of analysis, but that more tractable levels of analysis existed -- like molecular machines and individual enzymes. Dembski also indicated that independently given biological specifications were not a problem for some systems at that level -- for instance, outboard propellors driven by rotary motors had been invented by humans before they were discovered in the bacterial flagellum.
Perhaps the most striking admission from Kauffman during the rebuttal period was that for Kauffman's project to succeed, the probabilities of getting certain functional polymers had to be between one in a million and one in a trillion or so (essentially, the sorts of numbers that ribozyme engineering experiments can handle). This represents a huge advance for intelligent design, for in essence it is saying that the sort of probabilistic considerations that Dembski has been putting forward are essential to assessing the validity of Kauffman's own self-organizational approach. However the numbers come out, it puts ID in a privileged position of laying the ground rules for whatever program for a general biology will be successful. Dembski predicted that in the next five to seven years the complex specified information of certain biological systems would be definitely nailed down and that the improbability problem would come to be seen as irremovable.
One extended footnote is worth adding in conclusion. Dembski also spoke at the University of New Mexico the Monday night before his encounter with Kauffman. Dembski talked that night at the Continuing Education building of UNM where he laid out some of the new ideas he is presenting in his forthcoming book No Free Lunch and also gave a lay description of the design filter and irreducible complexity. In attendance at both talks were individuals from the local skeptics group, New Mexicans for Science and Reason (NMSR). During the Monday night presentation, members of this group asked standard questions about imperfect design, vestigial organs, and possible gradual routes to irreducible complexity.
Their questions were perfectly legitimate, but the demeanor of those who asked them Monday night was hostile. They laughed and giggled during the talk and gave airs that said that they did not consider Dembski's work to be science. There were other questions such as "who designed the designer" by a person who was clearly agitated during the entire lecture. He kept grumbling under his breath and slapping his notebook as Dembski spoke. Many in the audience were not happy about Dembski's Monday lecture and one person even turned red. Dembski told her that her question was a non sequitur, which she did not like that at all, and those sitting with her were ready for a fight.
The difference in the attitudes of the audiences between Dembski's Monday night seminar and the Tuesday Kauffman-Dembski encounter was therefore quite striking. On Tuesday, the NMSR folks were still there in numbers, but they were completely silent during the Q&A. This time the audience was not grumbling but was politely listening to both sides. The questions were very even for the most part and dealt mostly with the theories themselves. What was the crucial difference? I submit the difference was this: Stuart Kauffman went out of his way to state that what Dembski is doing is legitimate science (although Kauffman doesn't agree with it's application to biology). This set the tone for the audience and forced them to be respectful during the Tuesday encounter. Soon those participating in the hostility against ID will begin to look foolish, even to their own members.
Copyright 2001 Phillip E. Johnson. All rights reserved. International
copyright secured.
File Date: 11.19.01