History repeats itself. If reactions to Michael Behe’s previous books foretell anything about reviews of his latest, the forthcoming Darwin Devolves, then a lot will be said, misrepresentations will abound, and things will get ugly and personal.
This is exactly what has happened with a review in the journal Science. Behe already pointed out that this prepublication article offers “no response to the very central argument of the book.” That argument is what Behe calls the First Rule of Adaptive Evolution, namely that Darwinian evolution tends to “Break or blunt any gene whose loss would increase the number of offspring.”
Science is of course a major venue and we have only just begun in responding. As John West notes, the review by biologists Nathan Lents, Joshua Swamidass, and Richard Lenski wrongly charges that Behe “ignores evidence,” “misrepresents theory,” and “avoids evidence that challenges him.” The not-so-subtle subtext of the review is that Behe is an untrustworthy scholar who ignores criticism, so don’t pay any attention to his book. West showed how their review fails to acknowledge where Behe has responded to critics — directly on the topics they cover. Here’s a telling example of this problem: Their paragraph on the evolutionary origin of chloroquine resistance.
Use Your Words Carefully
In case you’re counting, Lents et al.’s paragraph on chloroquine resistance is in fact the longest of their review’s ten paragraphs. Yet the subject is not a topic of Behe’s present book but a major topic of Behe’s previous book, The Edge of Evolution (2007). How strange. Not enough room to squarely refute the main argument of the book under review — “We left out important points…in the interest of space,” co-author Swamidass explains — but enough room for this?
All right, fine. Here’s what they write:
Behe doubles down on his claim that the evolution of chloroquine resistance in malaria by random mutations is exceedingly unlikely because at least two mutations are required, neither of which is beneficial without the other. His calculations have already been refuted (5), and it has long been known that neutral and even deleterious mutations can provide stepping stones to future adaptations. Indeed, a 2014 study, unmentioned by Behe, reported discovery of two genetic paths through which malaria has evolved chloroquine resistance through multiple steps (6).
Lents et al. make it sound like Behe resurrects this topic in his new book, when that’s not at all the case.
In fact, in Darwin Devolves, Behe spends just two sentences—less than one half of one paragraph—on it. He spent less space (about 30 fewer words) talking about chloroquine resistance than Lents et al. spend on the topic in their one-page review!
Here’s what Behe says in his new book:
As I noted in The Edge of Evolution, in an astronomical number of malaria cells exposed to the antibiotic chloroquine, no fancy alternative evolutionary mechanisms helped the parasite develop resistance. Only a couple of classical random point mutations in the gene for a single protein plus run-of-the-mill Darwinian natural selection were effective. (Darwin Devolves, p. 252)
In other words, Darwinian evolution didn’t create something extravagantly new in malaria-causing parasites to make them resistant to chloroquine. Behe is just making the reasonable observation that “none of the mechanisms of EES [Extended Evolutionary Synthesis] proponents were anywhere to be seen” in helping this feature to evolve, because “a couple of classical random point mutations in the gene for a single protein plus run-of-the-mill Darwinian natural selection” were sufficient.
Lents et al. overstate what Behe is saying:
Behe doubles down on his claim that the evolution of chloroquine resistance in malaria by random mutations is exceedingly unlikely because at least two mutations are required, neither of which is beneficial without the other.
Now, Behe might agree with everything in that sentence. But do these words accurately represent what he wrote in his latest book? No. In his new book Behe doesn’t even enter the debate over whether sequential or simultaneous mutations are needed to build chloroquine resistance — i.e., whether “neither” mutation “is beneficial without the other.” His main point there is that EES mechanisms didn’t do anything helpful, and that what Darwinian mechanisms did build wasn’t anything extravagant. Lents et al.’s description of Behe’s argument captures none of this and attributes arguments to Behe that, while he may agree with them, he does not make in this book.
There’s a long tradition among ID critics of grossly, uncharitably attacking and misrepresenting what Behe says about chloroquine resistance. Here’s some of the abuse that Behe received over The Edge of Evolution:
- Ken Miller wrote in Nature that Behe had perpetrated a “breathtaking abuse of statistical genetics,” which revealed “the intellectual desperation of the intelligent-design movement as it struggles to survive in the absence of even a shred of scientific data in its favour.”
- In The New Criterion, Paul Gross repeated Miller’s allegation, claiming “it would need a book longer than The Edge” to correct its “errors and omissions”
- Sean Carroll wrote in Science that Behe’s arguments were “so poorly conceived and readily dispatched that he has unwittingly done his critics a great favor in stating them.”
- Jerry Coyne in The New Republic charged that Behe displayed a “wilful ignorance of the evolutionary process” that is “disingenuous,” amounting to a “pathetic” attempt to “resurrect his campaign for ID.”
- Richard Dawkins called Behe in the New York Times “the disowned biochemist from Lehigh University” who should “re-establish his scientific credentials and start over.”
What did Behe argue in The Edge that pushed critics over it?
Behe’s Argument in The Edge
To summarize, Behe began by observing that chloroquine resistance is quite rare: public health data suggests that it arises once in ever 1020 cells of the most virulent malarial parasite Plasmodium falciparum. This rarity is not a calculation based upon statistical odds of the trait arising. It’s an observed data point from public health studies.
Because of the trait’s observed rarity, Behe inferred that this resistance is probably a complex trait that requires multiple mutations to arise. He didn’t specifically state whether those multiple mutations must be “stepwise” (where each successive mutation in the pathway yields a successively greater resistance advantage), or “simultaneous” (where multiple mutations would be needed before any resistance arose). He did point out that the empirically observed rarity of the trait suggests that it doesn’t evolve easily and that multiple mutations are required. Behe called this a “chloroquine complexity cluster” or CCC. He noted that based upon empirical observations, such a trait would require about 1020 organisms to arise.
He then asked a hypothetical question:
What if a problem arose during the course of life on earth that required a cluster of mutations that was twice as complex as a CCC? (Let’s call it a double CCC.) For example, what if instead of the several amino acid changes needed for chloroquine resistance in malaria, twice that number were needed? (The Edge of Evolution, pp. 62-63)
The answer, Behe argued, poses a serious problem for Darwinian evolution:
In that case the odds would be that for a CCC times itself. Instead of 1020 cells to solve the evolutionary problem, we would need 1040 cells.
Workers at the University of Georgia have estimated that about a billion billion trillion (1030) bacterial cells are formed on the earth each and every year. (Bacteria are by far the most numerous type of organisms on earth.) If that number has been the same over the entire several-billion-year history of the world, then throughout the course of history there would have been slightly fewer than 1040 cells, a bit less than we’d expect to need to get a double CCC. The conclusion, then, is that the odds are slightly against even one double CCC showing up by Darwinian processes in the entire course of life on earth. (The Edge of Evolution, p. 63)
It was this argument that incurred the wrath of Behe’s critics. But many of their criticisms were based upon a misreading.
An Observation, Not a Calculation
The first major mistake the critics made was attacking Behe’s claim that malarial resistance arises once in every 1020 malarial parasites. Paul Gross ridiculed it as “a mere guess.” Yet the statistic actually came from a review article, “Antimalarial drug resistance,” by a top malaria expert Nicholas White in a well-respected medical journal, The Journal of Clinical Investigation. White deduced the following:
Resistance to chloroquine in P. falciparum has arisen spontaneously less than ten times in the past fifty years. This suggests that the per-parasite probability of developing resistance de novo is on the order of 1 in 1020 parasite multiplications.
This seems like a completely legitimate number for Behe to cite. Behe didn’t conclude that the likelihood of evolving a CCC was 1 in 1020 organisms due to probability calculations; it is an empirically observed data point based upon public health data. He writes in The Edge of Evolution:
We estimated the odds of a CCC — one in a hundred billion billion (1020) — by looking at the number of malarial parasites needed to develop the double mutation of a particular protein of a particular gene. (The Edge of Evolution, p. 61)
The key words there are “looking at the number of malarial parasites needed to develop the double mutation” — the number was obtained by “looking” at the real world.
Writing on his Amazon Blog, Behe noted that his critics wrongly thought the 1 in 1020 organisms is a probability calculation based upon some assumptions about the evolutionary pathway required to obtain resistance:
The number of one in 1020 is not a probability calculation. Rather, it is statistical data. It is perhaps not too surprising that both Miller and Coyne make that mistake, because in general Darwinists are not used to constraining their speculations with quantitative data.
Lents et al. claim that Behe’s “calculations” about “the evolution of chloroquine resistance in malaria” have “already been refuted.” They refer to an exchange in the journal Genetics in 2008 and 2009 between Behe and two critics, mathematicians Rick Durrett and Deena Schmidt. But it turns out that Behe was not refuted, nor did he ignore their critique.
Behe “Refuted” or Behe Vindicated?
Lents et al. want the reader to think that in Darwin Devolves Behe is reusing (i.e., “doubles down on”) old inaccurate “calculations [that] have already been refuted.” But Behe offers exactly zero calculations in Darwin Devolves about the evolution of chloroquine resistance.
Moreover, one of his main claims on the topic — that in chloroquine resistance “at least two mutations are required, neither of which is beneficial without the other” — is uncontested by Durrett and Schmidt. Their initial article in Genetics replying to Behe developed a population genetics model for calculating the waiting time for two mutations to arise when both are needed before some trait (like a binding site) provides a benefit! They don’t deny Behe’s suggestion that multiple mutations might sometimes be required before an advantage is gained. They simply take issue with his calculations about how long it would take for such a double-mutation trait to evolve.
Durrett and Schmidt spend about two paragraphs applying their model to Behe’s arguments about the evolution of chloroquine resistance. They note that Behe derived his probability of a CCC arising from real-world public health observations:
Arguing that (i) there are 1 trillion parasitic cells in an infected person, (ii) there are 1 billion infected persons on the planet, and (ii) chloroquine resistance has arisen only 10 times in the past 50 years, he concludes that the odds of one parasite developing resistance to chloroquine, an event he calls a chloroquine complexity cluster (CCC), are 1 in 1020.
It’s good that they acknowledge Behe’s 1 in 1020 statistic is empirically derived data. However they go on to ignore this and proceed to make the very mistake that Behe was trying to avoid, and that Behe’s critics (wrongly) accused him of committing: they calculate the probability of CCC mutations using a model. They disregard real-world observations and try to re-model something that has already been modelled by the real world.
Lents et al. want to paint Behe as a shady scholar who ignores criticism. Yet they leave unmentioned that Behe actually replied in Genetics to Durrett and Schmidt, directly engaging with this critics. John West has already noted this. There, Behe observes that Durrett and Schmidt’s theoretical model is not as solid as his empirically observed data point:
Their criticism compares apples to oranges. My figure of 1020 is an empirical statistic from the literature; it is not, as their calculation is, a theoretical estimate from a population genetics model. Generally, when the results of a simple model disagree with observational data, it is an indication that the model is inadequate. … The difﬁculty with models such as Durrett and Schmidt’s is that their biological relevance is often uncertain, and unknown factors that are quite important to cellular evolution may be unintentionally left out of the model. That is why experimental or observational data on the evolution of microbes such as P. falciparum are invaluable — because they can constrain our models. Whatever we speculate about what may be the usefulness of a new transcription factor-binding site, gene duplication, meiotic recombination, protein domain swap, or anything else, none of them were of much use in helping the malarial parasite fend off an evolutionary challenge. The data show that in 1020 chances only several point mutations in PfCRT were useful to it in effectively combating chloroquine.
Durrett and Schmidt then replied to Behe, and there is a back and forth over the “waiting time” required to evolve a double-CCC. The population genetics math here gets complicated. But in brief, Durrett and Schmidt claimed that Behe massively overestimated the number of trials required to generate a double CCC. To this, Behe answered that their own math included an error that overestimated the overestimate by a factor of about 30.
They conceded this mistake. Durrett and Schmidt also conceded that the length of the waiting time is highly dependent upon the degree to which the first mutation is deleterious, an empirical question they say is for “biologists to debate.” Most importantly, Behe points out in his reply that even if Durrett and Schmidt are right in their criticisms of his calculations, their result still shows that it would take an unreasonably long time for two specific mutations to arise in mammalian organisms like humans:
Durrett and Schmidt (2008, p. 1507) retort that my number ‘‘is 5 million times larger than the calculation we have just given’’ using their model (which nonetheless gives a prohibitively long waiting time of 216 million years).
To be specific, after citing their 216-million-year statistic, Durrett and Schmidt write in their original paper that “as our new results show, a coordinated pair of mutations that first inactivates a binding site and then creates a new one is very unlikely to occur on a reasonable timescale” (emphasis added).
If showing Behe was “refuted” also entails making calculations that fatally challenge Darwinian mechanisms, then this is hardly the sort of refutation that ID critics should want to cite. But that’s exactly what Lents et al. have done.
Supposedly it is Behe who “avoids evidence that challenges him.” The challenge for his critics is: Why didn’t their review acknowledge that Durrett and Schmidt, in allegedly showing Behe was “refuted,” uncovered serious mathematical challenges to Darwinian evolution?
After citing public health data showing the rarity of chloroquine resistance, Behe inferred that this trait must require multiple mutations that are also collectively rare. This too is legitimate reasoning. In fact, as reported here at Evolution News, Richard Dawkins used the same type of logic in his book The Greatest Show on Earth. He took the rarity of a trait as evidence that a complex mutational pathway is required to generate it.
Summers et al. (2014): Don’t Throw Behe in that Briar Patch!
In 2014, Behe’s inference that multiple mutations are needed to produce chloroquine resistance was massively vindicated by a paper in PNAS by Summers et al. They found that at least two mutations were in fact required before chloroquine resistance could arise. Behe wrote about this paper when it first came out:
A recent paper in PNAS confirms a key inference I made in 2007 in The Edge of Evolution. Summers et al. conclude that “the minimum requirement for (low) [chloroquine] transport activity … is two mutations.”
A minimum of two mutations sufficed for (low) CQ transport activity, and as few as four conferred full activity. … The findings presented here reveal that the minimum requirement for (low) CQ transport activity in both the ET and TD lineages of CQR PfCRT is two mutations.
Despite the protests from critics, it turned out that Behe reasonably inferred that chloroquine resistance not only requires multiple mutations, but multiple mutations must be present before resistance can evolve. His critics mistakenly thought this was a crucial plank in his argument, when it wasn’t. But Summers et al. (2014) showed that Behe’s ID-inspired suspicions were right all along.
In this regard, the review by Lents et al. is extremely misleading. The review claims that “a 2014 study, unmentioned by Behe, reported discovery of two genetic paths through which malaria has evolved chloroquine resistance through multiple steps (6).” Their statement is incomplete and inaccurate for multiple reasons.
Citation (6), as you might have guessed, is to Summers et al., which reported that “the minimum requirement” for the trait that confers chloroquine resistance in malaria parasites “is two mutations.” So yes, as Lents et al. state, this paper did find that “multiple steps” are needed to evolve chloroquine resistance. However, it did not find that chloroquine resistance can evolve in a “stepwise” manner, where each successive mutation confers a greater advantage. When the paper states that “the minimum requirement” for chloroquine resistance is “two mutations,” that means that multiple mutations must exist together before the trait can arise. Ironically, the citation provided by Lents et al. in their review of Darwin Devolves supports Behe’s position, not theirs. Why didn’t the reviewers disclose that Summers et al. vindicated Behe’s prediction that multiple mutations must be present before chloroquine resistance will appear?
But that’s not the worst of it.
The Unmentionable Mention
Lents et al. state that Summers et al. is “unmentioned by Behe,” insinuating that this is somehow a paper that doesn’t support Behe and he’s trying to ignore it. Wrong on both counts. Quite to the contrary, this paper does support Behe and Behe has written about it extensively! Immediately after it was published in 2014 Behe wrote a post here on EN, “A Key Inference of The Edge of Evolution Has Now Been Experimentally Confirmed.” He explained there how Summers et al. vindicated his prediction in The Edge of Evolution:
However, at the time the book’s chief, concrete example — the need for multiple, specific changes in a particular malarial protein (called PfCRT) for the development of resistance to chloroquine — was an inference, not yet an experimentally confirmed fact. It was really an excellent, obvious inference, because resistance to chloroquine arises much, much less frequently than to other drugs. For example, resistance to the antimalarial drug atovaquone develops spontaneously in every third patient, but to chloroquine only in approximately every billionth one. About PfCRT I wrote, “Since two particular amino acid changes [out of four to eight total changes] occur in almost all of these cases [of chloroquine resistance in the wild], they may both be required for the primary activity by which the protein confers resistance.” The result would be that “the likelihood of a particular [malarial] cell having the several necessary changes would be much, much less than the case [for atovaquone] where it needed to change only one amino acid. That factor seems to be the secret of why chloroquine was an effective drug for decades.” Still, the deduction hadn’t yet been nailed down in the lab.
Now it has, thanks to Summers et al. 2014. It took them years to get their results because they had to painstakingly develop a suitable test system where the malarial protein could be both effectively deployed and closely monitored for its relevant activity — the ability to pump chloroquine across a cell membrane, which rids the parasite of the drug. Using clever experimental techniques they artificially mutated the protein in all the ways that nature has, plus in ways that produced previously unseen intermediates. One of their conclusions is that a minimum of two specific mutations are indeed required for the protein to be able to transport chloroquine.
(Interestingly, one of the two mutations I discussed in The Edge of Evolution as possibly required, at position 76 of the protein chain, is in fact one of the two that Summers et al. proved to be needed. But the other one I discussed, at position 220, isn’t. Although that change can help, Summers et al. found that the second required mutation is at either position 75 or position 326. They also showed that, although proteins with just the two required mutations could pump chloroquine past a cell membrane in their test system, the rate was significantly less than for some proteins with additional mutations. What’s more, the two required ones weren’t necessarily enough to allow malarial parasites to survive better in the presence of chloroquine in the lab. What that means for malaria in the wild is still unclear.)
The need for multiple mutations neatly accounts for why the development of spontaneous resistance to chloroquine is an event of extremely low probability — approximately one in a hundred billion billion (1 in 1020) malarial cell replications — as the distinguished Oxford University malariologist Nicholas White deduced years ago. The bottom line is that the need for an organism to acquire multiple mutations in some situations before a relevant selectable function appears is now an established experimental fact.
Behe also discussed this paper in a number of other posts at EN linked earlier. Far from the study going “unmentioned by Behe,” he immediately discussed Summers et al. in 2014 and has discussed it extensively since.
Doubly Wrong, Twice
Why do Lents et al. insinuate that Behe ignored this paper? As we’ve seen, it’s because they want to attack Behe as essentially deceptive and refusing to acknowledge contrary evidence. To repeat their review’s conclusion:
Ultimately Darwin Devolves fails to challenge modern evolutionary science because, once again, Behe does not fully engage with it. He misrepresents theory and avoids evidence that challenges him.
But the evidence cited in their paragraph about chloroquine resistance doesn’t challenge Behe — it contradicts Behe’s critics. And Behe has far from ignored that evidence.
One thing Lents et al. are right about is that Behe doesn’t mention Summers et al. in his new book. Is there a benign explanation for this? Of course, and it takes us back to the beginning of this post: Behe doesn’t mention the Summers et al. paper in Darwin Devolves because he devotes less than half a paragraph in his 300+ page book to discussing chloroquine resistance. Chloroquine resistance is simply not a topic of his latest book.
The Science authors thought they had an opportunity to impugn Behe’s scientific and scholarly integrity, so they jumped on it. But in their eagerness, they jumped prematurely: In this one paragraph Lents et al. twice cited sources that they thought contradicted Behe, and twice suggested he ignored or dismissed the contrary evidence. They are wrong on all counts.
Photo: Michael Behe, in Revolutionary: Michael Behe and the Mystery of Molecular Machines.