Wednesday, September 23, 2009

William Dembski and Jello

In the gufuffle about the new Dembski/Marks paper that, according to them and the Discovery Institute, constitutes evidence against evolution, several people across the blogosphere have commented on the paper by David Wolpert from 1997 that took Dembski's application of the NFL theorems to the woodshed. It is every bit as devastating as the testimony by Jeffrey Shallit for the Dover-Kitzmiller trial. This is even more true because Wolpert is the co-creator of the No Free Lunch theorems, along with William Macready1. Wolpert's article is titled William Dembski's treatment of the No Free Lunch theorems is written in jello, a title that does not leave much to the imagination. So what are the No Free Lunch theorems? Wolpert writes:
These theorems, loosely speaking, say that the performance-weighted measure of domains in which some search algorithm A beats some contender algorithm B exactly equals the measure of domains for which the reverse is true. So, for example, in attempting to find a high point on a surface, a hill-ascending algorithm will perform no better than random search, and in fact no better than a hill-descending algorithm, over the space of all surfaces one might search. In short, according to these theorems there is no free lunch; without tailoring one's algorithm to the domain at hand, one has no assurances that that algorithm will perform well on that domain.
Dembski has taken the theorems and argued that if search algorithms are analogous to genetic changes in the environment and that genetic change is an "unguided random process," then no amount of evolutionary change will amount to anything. But Wolpert has concerns about how Dembski has applied the theorems and how well it applies to the biological world:
Indeed, throughout there is a marked elision of the formal details of the biological processes under consideration. Perhaps the most glaring example of this is that neo-Darwinian evolution of ecosystems does not involve a set of genomes all searching the same, fixed fitness function, the situation considered by the NFL theorems. Rather it is a co-evolutionary process. Roughly speaking, as each genome changes from one generation to the next, it modifies the surfaces that the other genomes are searching. And recent results indicate that NFL results do not hold in co-evolution.
The reason there is a marked elision of the biological details is that Dembski has demonstrated numerous times that he does not understand the biological details, having never worked in the field or studied how organisms change over time. As Wolpert and other people have pointed out, Dembski assumes the same fitness function for all environments. This in no way describes biological reality or how selection works. Why is it that Dembski refuses to acknowledge this?

1D. H. Wolpert and W. G. Macready, “No free lunch theorems for search,” Santa Fe Institute, Sante Fe, NM, Tech. Rep. SFI-TR-05-010, 1995.

Now playing: Afro Celt Sound System - When You're Falling (feat. Peter Gabriel)
via FoxyTunes


  1. Anonymous11:48 AM

    Cornelius Hunter seems to have launched something of an attack on Steve Matheson, .
    Just the day before he did the same to Darrel Falk, .

    He appears to be a rather disturbed individual.

  2. He did write me back after I asked him about how much education he had in palaeontology and geology and his response was insulting and condescending. I will have to have a look.