dna green
Digital illustration of  DNA

Sean Carroll Fails to Scale The Edge of Evolution: A Rebuttal to Sean Carroll’s Anti-ID Book Review in Science

A Response to Sean B. Carroll's "God as Genetic Engineer" in Science Originally published at Evolution News

[Editor’s Note: This article originally appeared as a four-part series on Evolution News and Views, as Part 1, Part 2, Part 3, Part 4.]


A few months ago I posted a review of Sean B. Carroll’s book The Making of the Fittest: DNA and the Ultimate Forensic Record of Evolution, the book in which biologist Sean B. Carroll intimates that the salvation of our species hangs upon accepting Darwin. Carroll has now invoked his own religious metaphors in his review of Michael Behe’s book The Edge of Evolution: The Search for the Limits of Darwinism in Science. Carroll postures himself as Thomas Henry Huxley debating Bishop Samuel Wilberforce in a famous 19th century debate over evolution. Carroll even opens the review by invoking Huxley, saying, “The Lord hath delivered him into mine hands.” Science is a leading scientific journal published by the American Association for the Advancement of Science, an organization with a long history of knee-jerk political opposition to intelligent design (ID), so it isn’t surprising that they invited a negative review with such theatrics against Behe’s book. But in his eagerness to attack Behe with the approval of “[t]he Lord,” Carroll completely fails to engage Behe’s actual arguments.

Michael Behe himself has responded to Carroll here, but there are many other errors in Carroll’s review of Behe’s The Edge of Evolution. Carroll’s mistakes include: misrepresenting Behe’s arguments, failing to engage Behe’s arguments by citing small-scale examples of evolution that Behe readily acknowledges can occur, citing papers that fail to support Carroll’s contentions, and in some cases, citing papers that actually support Behe’s view rather than opposing it. These mistakes will be discussed further below.

Part 1: Carroll’s Fundamental Misrepresentations of Behe’s Arguments

Carroll ignores that Behe quite contently acknowledges that at times multiple amino acids can change in a protein when there is a selective advantage for each mutation. Specifically, Carroll’s mistake begins when he claims Behe says that “multiple-amino acid replacements therefore can’t happen”:

Behe states correctly that in most species two adaptive mutations occurring instantaneously at two specific sites in one gene are very unlikely and that functional changes in proteins often involve two or more sites. But it is a non sequitur to leap to the conclusion, as Behe does, that such multiple-amino acid replacements therefore can’t happen. Multiple replacements can accumulate when each single amino acid replacement affects performance, however slightly, because selection can act on each replacement individually and the changes can be made sequential.

Sean B. Carroll, “God as Genetic Engineer,” Science, Vol. 316:1427 – 1428 (June 8, 2007).

In Carroll’s eagerness to attack Behe, he somehow fails to acknowledge that Behe makes precisely the same point throughout The Edge of Evolution. Behe repeatedly explains that when there is an advantage along each small step, evolution takes place. Early in his book Behe explains that “variation, selection, and inheritance will only work if there is also a smooth evolutionary pathway leading from biological point A to biological point B.” (p. 5) Behe later states:

The Darwinian magic works well only when intermediate steps are each better (“more fit”) than preceding steps, so that the mutant gene increases in number I the population as natural selection favors the offspring. Yet its usefulness quickly declines when intermediate steps are worse than earlier steps and is pretty much worthless if several required intervening steps aren’t improvements).

Michael Behe, The Edge of Evolution: The Search for the Limits of Darwinism, p. 112, (Free Press, 2007).

Behe makes this point impossible for any serious reviewer to miss:

This point is crucial: If there is not a smooth, gradually rising, easily found evolutionary pathway leading to a biological system within a reasonable time, Darwinian processes won’t work.

Behe, 2007, p. 7.

Behe again concedes that evolution can sometimes occur when there are stepwise advantages along each mutational step of evolution:

Although it hasn’t yet occurred in nature, we shouldn’t be at all surprised to see resistance of mosquitoes to the new insecticides arise and spread by Darwinian processes. The necessary preconditions are all there: tiny, incremental steps—amino acid by amino acid—leading from one biological level to another.

Behe, 2007, p. 76.

In each of these quotes, Behe acknowledges that evolution can happen when there is an advantage along each small step of an evolutionary pathway. Carroll thus completely misrepresents Behe’s position to claim that Behe says that mutations “can’t happen,” even when “each single amino acid replacement affects performance.”

But what happens when there is not an advantage gained at each step? This will be discussed further below.

Part 2: Carroll’s Citations Actually Confirm Michael Behe’s Arguments

One of the papers Carroll cites in an attempt to refute Behe actually explicitly confirms Behe’s position that there are limits to the creative power of Darwinian processes. Carroll argues that Behe claims that “multiple-amino acid replacements … can’t happen.” As noted above, in contrast to Carroll’s misrepresentation, Behe’s actual position contends evolution can proceed forward where there is a stepwise advantage gained with each mutation. But Behe also contends that evolution gets stuck when intermediate states becomes harmful or do not increase fitness: “If two mutations have to occur before there is a net beneficial effect — if an intermediate state is harmful, or less fit than the starting state — then there is already a big evolutionary problem.” (Behe, The Edge of Evolution, p. 106) To oppose Behe’s arguments, Carroll claims to provide examples of “cumulative selection changing multiple sites in evolving proteins.” But Carroll’s citation of the paper Weinreich et al. (2007) actually confirms Behe’s precise position.

Weinreich et al. (2007) demonstrates that 5 amino acid sites can change during the evolution of bacterial resistance to the antibiotic drug penicillin. Yet this paper actually confirms Behe’s view, as it reports that 102 of the 120 possible mutational combinations don’t occur naturally for precisely the reason Behe says they won’t work: they don’t give stepwise mutational advantages. Consider how similar this finding is to Behe’s argument above:

However, we demonstrate that 102 trajectories are inaccessible to Darwinian selection and that many of the remaining trajectories have negligible probabilities of realization, because four of these five mutations fail to increase drug resistance in some combination.

Daniel M. Weinreich, Nigel F. Delaney, Mark A. DePristo, Daniel L. Hartl, “Darwinian Evolution Can Follow Only Very Few Mutational Paths to Fitter Proteins,” Science, Vol. 312:111-114 (April 7, 2006), emphasis added.

In other words, the evolution stopped when the evolutionary pathway encountered a point where any further mutations cause bacterial resistance to the anti-biotic drug to drop, or did not increase it. This implies that when random mutation and natural selection is asked to either do a random walk or traverse a drop in fitness, it gets stuck. Carroll’s example of bacteria evolving resistance does not address Behe’s arguments. In fact, it confirms precisely why Behe argues there is an edge to evolution.

Part 3: Carroll’s Examples are Well-Within the Edge of Evolution’s Capabilities

Many of Carroll’s cited papers report types of evolution that Behe readily concedes can occur, and are unimpressive examples within the “edge” of evolution. As noted above, Carroll cites the evolution of antibiotic resistance in an attempt to refute Behe. Yet Behe finds that bacterial mechanisms of antibiotic resistance present a challenge to natural selection that pales in comparison to the challenge posed by true complexity of the cell:

Where is it reasonable to draw the edge of evolution? … One the one side … several mutations can sequentially add to each other to improve an organism’s chance of survival. An example is the breaking of the regulatory controls of fetal hemoglobin to help alleviate sickle cell disease. … On the other side are the examples of what random mutation and natural selection clearly cannot do. … The structural elegance of systems such as the cilium, the functional sophistication of the pathways that construct them, and then the total lack of serious Darwinian explanations all point insistently to the same conclusion: They are far past the edge of evolution. Such coherent, complex, cellular systems did not arise by random mutation and natural selection, any more than the Hoover Dam was built by the random accumulation of twigs, leaves, and mud.

Michael Behe, The Edge of Evolution: The Search for the Limits of Darwinism, pp. 111-112, (Free Press, 2007).

It’s Easier to Tear Down Walls than to Build Them

In another misplaced attempt to refute Behe, Carroll also cites the evolution of toxins. But Behe provides a ready discussion regarding the evolvability of destructive proteins because they entail merely evolving the ability to break things in the cell — a relatively simple task. Behe writes:

Foreign proteins injected into a cell by an invading virus or bacterium make up a different category. The foreign proteins of pathogens almost always are intended to cripple a cell in any way possible. Since there are so many ways to break a machine than to improve it, this is the kind of task at which Darwinism excels. Like throwing a wad of chewing gum into a finely tuned machine, it’s relatively easy to clog a system—much easier than making the system in the first place. Destructive protein binding is much easier to achieve by chance.

Michael J. Behe, The Edge of Evolution: The Search for the Limits of Darwinism, p. 149 (Free Press, 2007).

Behe has a ready rejoinder to Carroll’s mention of the evolution of toxins.

In another incredibly misplaced citation, Carroll cites the evolution of malarial resistance to drugs as a rebuttal to Behe. Yet Behe spends multiple chapters discussing the evolution of resistance to malaria and how it generally entails unimpressive genetic changes that, in actually, demonstrate that there is a limit to evolutionary change. Carroll simply isn’t engaging Behe’s arguments. He cites papers that discuss the evolution of biological functions that Behe already acknowledges are within the “edge” of what Darwinian evolution can produce.

Part 4: Carroll’s Mistakes Protein Sequence Similarity as Evidence for Random Mutation and Natural Selection

Carroll has thus far failed to engage Behe’s actual arguments. But Carroll does make a coupler attempts to tackle the origin of a complex biological features. Yet these attempts fail because they confuse the evidence for common descent from sequence similarity with evidence for random mutation and natural selection. Again, Behe anticipates this mistake and provides ample rebuttals to Carroll’s citations. As discussed below, Carroll badly mis-cites one paper as showing that “new protein interactions … can evolve fairly rapidly.”

Carroll Spins the Flagellum

When discussing the flagellum, Carroll cites Pallen and Matzke’s review paper on the evolution of the flagellum. Yet this paper itself admits that “the flagellar research community has scarcely begun to consider how these systems have evolved.” Those that read the Pallen/Matzke paper will find vague generalities and nothing remotely approaching a specific step-by-step model for the evolution of the flagellum. In fact, readers of the Pallen/Matzke paper will find generally find misplaced attempts to give evidence of protein homology as evidence for random mutation and natural selection. But Behe readily anticipates this fallacious strategy. In his book, Behe cites to Pallen’s own work and notes that “modern Darwinists point to evidence of common descent and erroneously assume it to be evidence of the power of random mutation.” (Behe, Edge of Evolution, p. 95) None of the work cited by Carroll remotely attempts to explain the evolution of the complexity of flagellar assembly discussed in Appendix 3 of The Edge of Evolution.

Another Questionable Citation

Behe’s recognition that evidence for common descent is not evidence for natural selection also provides a poignant rebuttal to one of Carroll’s prized citations. Carroll relies on his claim that “new protein interactions … can evolve fairly rapidly,” giving as one citation Budovskaya et al., “An evolutionary proteomics approach identifies substrates of the cAMP-dependent protein kinase,” Proceedings of the National Academy of Sciences, Vol. 102:13933-13938 (Sept. 27, 2005). But this paper epitomizes Behe’s recognition that “modern Darwinists point to evidence of common descent and erroneously assume it to be evidence of the power of random mutation.” (Behe, Edge of Evolution, p. 95)

The study’s methods were essentially as follows: (1) take a protein that is acted upon by a particular enzyme (called a cAMP-dependent kinase), (2) search protein databases to find proteins with similar sequences, and then (3) study those proteins with similar sequences to determine if they are acted upon by the same enzyme, the cAMP-dependent kinase. In fact, the paper didn’t even look for a function as such of these proteins; it merely demonstrated that the proteins identified by sequence analysis are likely targets of the same kinase. Budovskaya et al. (2005) even admits that it merely “uses sequence information to identify the biologically relevant occurrences of a protein motif of interest.” In other words, the raw data entails the mere comparison of proteins through sequence similarity, and there’s no direct testing of random mutation and natural selection whatsoever.

Indeed, the paper recognizes that its results support a strong correlation of “structure-function,” due to the fact that one can completely remove common descent and Darwinian evolution from this picture and get the same results: All the study truly found was the mundane result that proteins with similar sequences tend to be acted upon by similar enzymes. Big deal. The paper then adds a lot of evolutionary gloss, but that’s all it is: inference based upon assumptions, not hard evidence. It in no way demonstrates the power of random mutation and natural selection, and at best provides an example of how “Darwinists point to evidence of common descent and erroneously assume it to be evidence of the power of random mutation” (Behe, Edge of Evolution, p. 95) (and even then, this assumes that sequence similarity is necessarily evidence for common ancestry).

Even if this paper had demonstrated natural selection, it still would fall well within Behe’s edge of evolution. By citing this paper, Carroll is referring to the interaction between enzymes that modify or cleave other proteins, and the protein binding sites recognized by those enzymes may be quite short. Carroll equivocates over the term “interaction motifs”: Behe is talking about the kinds of protein interactions that build enzyme complexes or structures like the ribosome or the cilium or the replication machinery of the cell. These kinds of temporal interactions are stabilized by many amino acids, not the short motifs studied in this paper. These kinds of short motifs may be necessary for building molecular machines, but they are far from sufficient to build molecular machines where two or more proteins must dock together and stably interact.

Finally, even if Budovskaya et al. (2005) had demonstrated random mutation and natural selection (which it doesn’t), Carroll cites this paper to claim that protein-protein interactions “can evolve fairly rapidly,” yet the paper studied proteins in “a group of budding yeast species that are separated by up to 800 million years of evolutionary distance.” 800 million years is not “fairly rapidly” — in fact, it represents nearly 1/4 of the entire history of life on earth. There appears to be no legitimate grounds whatsoever for Carroll citing Budovskaya et al. (2005) to claim that “new protein interactions … can evolve fairly rapidly.”


Sean B. Carroll’s review in Science against Michael Behe’s book The Edge of Evolution talks big, but carries a small stick. In the end, what is starkly missing from Carroll’s review is anything that actually demonstrates the evolution of something that Behe argues is beyond the edge of evolution. It seems that Carroll is afraid of heights — and his errors include the following:

  1. Misrepresenting Behe’s arguments and position
  2. Citing Articles that Confirm Behe’s Arguments Rather than Refuting Them
  3. Citing Studies that Demonstrate Small-Scale Evolution of the Type Behe Readily Acknowledges can be within the “Edge of Evolution”
  4. Citing Papers that Demonstrate Mere Protein Sequence Similarity and Mistake that as Evidence for Random Mutation and Natural Selection

The crucial piece missing from Carroll’s attack on Behe’s book is anything remotely close to meeting Behe’s challenge to explain the step-by-step Darwinian origin of a highly complex biological feature.

Casey Luskin

Associate Director and Senior Fellow, Center for Science and Culture
Casey Luskin is a geologist and an attorney with graduate degrees in science and law, giving him expertise in both the scientific and legal dimensions of the debate over evolution. He earned his PhD in Geology from the University of Johannesburg, and BS and MS degrees in Earth Sciences from the University of California, San Diego, where he studied evolution extensively at both the graduate and undergraduate levels. His law degree is from the University of San Diego, where he focused his studies on First Amendment law, education law, and environmental law.