to review of
The Design Inference: Eliminating Chance Through Small Probabilities.
William A. Dembski. Cambridge University Press, Cambridge, UK, 1998. 272 pp.
Massimo Pigliucci's review of William Dembski's The Design Inference puzzles me. What puzzles me most is lack of substantive comment about the book itself--Pigliucci's "review" reads for the most part like an extended diatribe against the intelligent design (i.e. creationist) movement. Do not misunderstand me--I am unsympathetic towards creationists of all stripes. But the style in which Pigliucci attacks Dembski's book is analogous to badmouthing the Principia Mathematica because Newton spent substantial time working on alchemy, or to trashing Dirac's Principles of Quantum Mechanics because many crackpots appeal to quantum mechanics to advance their personal pseudoscientific agendas.
Although Dembski's book never attacks evolution (not even in the only seven pages in the book in which he talks about evolution at all) Pigliucci clearly represents the book as an attack on evolution. The Design Inference is supposedly "part of a large, well-planned, and not at all secret conspiracy whose objective is nothing less than the destruction of modern science and its substitution with a religious system of belief." And indeed, it is no secret that Dembski promotes creationism, and that he is a fellow at the Center for the Renewal of Science and Culture, which wishes to push precisely the agenda Pigliucci claims. But none of this says anything about the book. One cannot simply dismiss the book by virtue of guilt by association--one must assess what it actually says. Pigliucci fails to make that assessment time and time again:
1. Dembski's book has been hailed "as a revolutionary contribution to design theory." Indeed it has. But so what? Every New Age guru under the sun has hailed quantum mechanics as a revolutionary contribution to his or her nonsensical doctrines. Does that make quantum mechanics suspect? Does even the fact that some of the founders of quantum mechanics dabbled in mysticism make quantum mechanics suspect? Evaluation by association is uncalled for. Besides, if we are going to evaluate books by association, then the fact that The Design Inference has been published in the prestigous "Cambridge Studies in Probability, Induction, and Decision Theory" series--the editors of which are no slouches when it comes to mathematics and logic--should be sufficient to overrule any taint creationist connections may place on it.
2. Dembski's book is "soon to be followed by a more explicit attack on evolution entitled Uncommon Descent." This may well be true (Dembski has edited anti-evolution books in the past). And should Dembski write such a book, and should its attack on evolution be unsuccessful, the scientific community will point that out, as it has done for other flagship pieces of the intelligent design movement, such as Darwin on Trial and Darwin's Black Box. But attacking a current book because it is to be followed by a different book is just nonsensical.
3. Dembski's book attempts to define and defend a "universal small probability" of 0.5x10150 "that can be used as a universal yardstick for inferring design." According to Pigliucci, (i) this is done to undergird an attack on the probability of DNA forming naturalistically -- "if [Dembski] can demonstrate that the probability of a molecule of DNA forming in the primordial soup approaches...this 'universal small probability', then life does not evolve by chance" but (ii) Dembski "missed the solution to this riddle," namely, that biological evolution (and presumably DNA formation) is a product of "necessity, which can be caused by a non-random, deterministic force such as natural selection." There are two problems here: (1) The design inference (an argument form, mind you, not an argument itself) which Dembski seeks to establish in his book explicitly requires ruling out the possibility of the phenomenon under consideration being produced by a regularity (see Premise 4 on page 222). So anyone who wishes to make a Dembski-style design inference to the conclusion that DNA was designed, is automatically required to consider Pigliucci's natural "necessity"--the format of the design inference does not contain the oversight Pigliucci attributes to Dembski. (2) More importantly, Dembski never argues against naturalistic DNA formation in The Design Inference to begin with. Will he do so in Uncommon Descent? Fine, then attack that argument in a review of that book. The merit of the "universal small probability" itself is not addressed by criticism of bad creationist arguments which employ it, any more than Kolmogorov's axioms are unseated by virtue of bad creationist arguments which depend in part upon them.
Pigliucci, in fact, makes only one criticism which is directed at the actual content of The Design Inference. It is that Dembski "cloak[s] his logic with semi-obscure (and totally useless in practice) pseudo-mathematical jargon and symbolism." Presumably the bulk of the book (for instance, the demonstration of Caputo's rigging) is trivial and obfuscates with jargon that which should just be left to common sense. I have heard many of my students make similar criticisms of formal logic. They simply do not appreciate the importance of attempting to make precise and set on a rigorous foundation patterns of reasoning which appear common-sensical. Pigliucci fails to appreciate Dembski's attempt to do the same for the patterns of reasoning we employ when we attempt to discern whether an intelligent agent is responsible for certain phenomena we observe. Dembski's attempt may or may not fail (I am agnostic here) but his project is neither trivial, nor a gratuitous (or "pseudo") use of mathematical symbolism.
The aspect of Pigliucci's review which irritates me the most personally is the insinuation that the dust jacket endorsers--Wiliam Wimsatt and Jon Jarrett--are part of the intelligent design "conspiracy" by virtue of their praise of the book and their affiliation with two of the universities from which Dembski matriculated. Dembski may have received degrees from the universities at which they work, but we're talking about UIC and the University of Chicago here, not Bob Jones University and the Institute for Creation Research Graduate School. In personal correspondence, Wimsatt wrote "I am an evolutionist to the core, and am not at all bothered by Dembski's book." Furthermore, he wrote, "I think [the book] is a double-edged sword, and--properly understood--should give no comfort to Creationists! i.e., I don't think, by proper interpretation of the evidence that we have anything like evidence for intelligent design, but instead evolution emerges by far as the best explanation of the data." Jarrett, far from being a co-conspirator, was completely unaware of Dembski's religious views even after reading The Design Inference, until I showed him a CRSC pamphlet with Dembski's picture on it (this is an excellent demonstration of precisely how devoid of creationism the book itself is).
That Pigliucci's accusations of conspiracy unwittingly extend to people on his own side constitutes, in my opinion, further evidence that his zeal against creationism (as understandable as it is) has generated in him an illegitimate negative bias against anything that is in any way associated with creationism. Once again, I believe Pigliucci's strong dislike for creationism is well-founded, but it appears to have caused him to miss the point of The Design Inference, whether the point of the book turns out to be correct or incorrect.
1. Not that I think Dembski's book is of the same caliber as Newton's and Dirac's. In fact, I'm sure it couldn't possibly be of the same caliber as the Principia Mathematica. Beyond that, though, Dembski's book is too far over my head for me to make any assessment of it.
2. Bill Wimsatt points out that Pigliucci (and Dembski to an extent) is sloppy in his use of the word "necessity." Natural selection will not bring about the products we see around us today with necessity. Nevertheless, it acts as a "probability amplifier [as any multi-stage selective filter will do], in which the information in the product of a process can grow linearly (and its improbability decline exponentially) with time." Presumably, Dembski means to capture processes like this under the rubric of "necessity." It is unclear whether Pigliucci would agree with the characterization of natural selection as an amplifier, or stick to his guns and declare it a "deterministic force."
Copyright 2000 Mark I. Vuletic; Internet Infidels. All rights
reserved. International copyright secured.
File Date: 7.05.00