Dembski's continuing contradictions

Dembski has ‘responded’ to Wesley Elsberry’s and Mark Perakh’s criticicsm at ARN

Panda’s Thumb.

Does the discussion at the Panda’s Thumb advance the discussion we had on this board about that paper? As I mentioned in another post, that paper will be the basis for my technical lecture at the Trotter Prize Lecture Series at Texas A&M coming up the beginning of April. I’d enjoy meeting any critics on this board there (as well as supporters, of course).

Other than the usual self inflation, Dembski has little to say about the critiques themselves.

When pressed for details as to how Dembski ‘abuses’ critics, Dembski responded:

I’m happy to acknowledge my critics where I think they are being insightful. There tends to be a disconnect, however, between the criticisms I regard as insightful and those that my critics regard as insightful. I’m afraid that Wesley Elsberry and Mark Perakh do not rank high among those I regard as insightful critics. Since I’m quite busy and have plenty of critics, they tend to fall low in the queue. Consider, for instance, that Tom English on this board at least engaged the mathematics in my article. I’ve seen no indication that Elsberry or Perakh could even state the gist of it in plain English.

So why does Dembski start a thread about Wesley and Mark when he does not consider them to be insightful critics?

In fact, both critics have contributed significantly to the demise of Dembski’s arguments, which may help explain why Dembski is reluctant to address their claims.

As RB observes in the same thread

Of course neither can Bill, but then that gives him the fall back position of “you just don’t understand”. Nor can Bill discuss, in plain english, the application of his math to the realities of evolution and molecular research and theory.

Dembski obviously excels at self-promotion, and although his academic accomplishments seem to be rather limited, his popular contributions seem to have been embraced somewhat uncritically by many ID proponents (caveat emptor!!).

Dembski has consistently ignored his most formidable critics: For instance it was Wesley Elsberry who showed how Dembski failed to explain how to distinguish between apparent and actual CSI. In Dembski’s latest piece, he has made an almost incredible concession to his critics, one overlooked by most of the ID proponents, namely that both evolutionary mechanisms and intelligence can create CSI.

Dembski’s latest paper, undermines even more his explanatory filter approach since it has shown how false positives are a real possibility and provides no way to detect false positives. Which makes the explanatory filter, in Dembski’s own words, useless.

Both Elsberry and Perakh were quick to point out the problems with using NFL theorems as well as the displacement problem.

Dembski, sadly enough does not seem to give his critics much credit for this. Let’s hope that when/if his latest paper is published, the reviewers require a more thorough list of references.


Douglas Theobald has continued the thread to expose some of the deeply rooted ignorance amongst ID proponents. Surprisingly to many perhaps, the ignorance seems to be more about Dembski’s arguments this time than about evolutionary theory.

Douglas Theobald wrote:

I really don’t mean to be catty, but seriously, have any of the ID-ists here read Dembski’s latest paper, the topic of this thread?

But Theobald does far more than showing the level of unfamiliarity amongst ID proponents with Dembski’s arguments

Douglas Theobald: Once again, I will try to get this thread back on target.

Douglas Theobald: Dembski’s “FTID” is premised on a targeted search. According to Dembski, a “target” of an algorithmic search is a subregion of the search space that is pre-specified before the search begins. Dembski’s math concerns the probability of finding that pre-specified target, and that target alone.

Douglas Theobald: However, evolution and RMNS are not targeted searches by Dembski’s own usage, so Dembski’s math does not apply to evolution. Unless it can be shown that certain biological targets were pre-specified, Dembski’s math is irrelevant to evolution.

Douglas Theobald: Would anyone here (who has read Dembski’s paper, others need not apply) like to try to refute that?

Douglas Theobald: Specifically:

  • Can you provide examples of what the targets are in biology?
  • Can you explain how you know they are targets?
  • Who or what pre-specified these biological entities as targets?
  • Can you explain why the biological targets you propose are not simply labeled as such after the fact?

Douglas Theobald: This is the challenge I have presented, and nobody has answered it yet. Even Dembski acknowledges this problem in his paper, and he also gives no answer for it.

It follows that assisted search, even with so modest a problem as finding a specific protein 100 amino acids in length, requires a considerable amount of information if it is to surpass blind search and successfully locate a target. How are we to explain this net increase in information? One way is to explain it away by suggesting that no targets are in fact being searched. Rather, a space of possibilities is merely being explored, and we, as pattern-seeking animals, are merely imposing patterns, and therefore targets, after the fact (see, for instance, Shermer 2003).

(Dembski, pp. 15-16 of “ Searching Large Spaces “)

Douglas Theobald: Dembski goes on to explain how targets are valid in engineering, by specifically giving an example about finding a useful polymer. Yet he never provides even a hint of an argument for why pre-specified targets are valid in evolution or biology in general.

Douglas Theobald: In Dembski’s human engineering example, it is easy to justify the target as valid: I can easily answer each of the four challenges above in the engineering case.

Douglas Theobald: The target is a polymer with a minimum resilience and strength. We know such a polymer is a target because we have independent knowledge that human engineers pre-specified the target, indicating a desire for finding such a polymer. These desirable polymers are not after-the-fact targets, because they were specified by human engineers before they went and searched for them. Success of the algorithm is measured as whether it finds these polymer targets or not. Dembski’s math applies here. If Dembski’s math is valid, we can conclude that these engineers knew something about their search space in order to choose a successful targeted search algorithm.

Douglas Theobald: Now how about evolution and biology in general?

Douglas Theobald: The answer should ideally be as clear and concise as the above engineering example.


For those fans of Salvador, you may enjoy his attempts to distract from the conversation with his usual Turing machines, or other irrelevant arguments. For someone who once said he was willing to “take the grenade” for Dembski “so that Dembski can continue to not respond to critics”, he seems to ignore the implosion his own arguments have caused for ID.

Seems that Bill as withdrawn himself from the discussion, I am curious how he is going to fix his paper before he presents it at the Trotter Prize lecture series?