H. L. Mencken once wrote, “There is always a well-known solution to every human problem—neat, plausible, and wrong.” (William Dembski mangles the quotation in The Design Revolution [2004], as Jeffrey Shallit observed.) When it comes to the problem of creationism, the solution that meets Mencken’s criterion might be falsifiability. Certainly there’s no end of people who have misused falsifiability in attacking or defending creationism or attacking or defending evolution. Unlike the character in a classic xkcd.com cartoon, I need my sleep, so even when someone is wrong on the internet (or even in print) about creationism and falsifiability, I’m not especially inclined to try to argue about it. But Michael Behe, in a paper entitled “The modern intelligent design hypothesis” (published first in Philosophia Christi in 2001, and reprinted in Neil A. Manson’s anthology God and Design in 2003), is so wrong that it’s worth examining why.
For sheer chutzpah, you have to hand it to Behe. “ID is quite susceptible to falsification,” he argues, while “Darwinism” (as he calls it) “seems quite impervious to falsification.”
The reason for this can be seen when we examine the basic claims of the two ideas with regard to a particular biochemical system like, say, the bacterial flagellum. The claim of ID is that “No unintelligent process could produce this system.” The claim of Darwinism is that “Some unintelligent process could produce this system.” To falsify the first claim, one need only show that at least one unintelligent process could produce the system. To falsify the second claim, one would have to show the system could not have been formed by any of a potentially infinite number of possible unintelligent processes, which is effectively impossible to do. (emphasis in original)
Thus he hopes not only to dispel the complaint that “intelligent design” is unfalsifiable but even to turn the tables on it. How neat, how plausible, how wrong!
The first problem is that there is absolutely no reason for anybody to be interested in “Darwinism,” as Behe formulates it. Scientists who actually are engaged in investigating the evolutionary history of the bacterial flagellum aren’t interested in what processes could or could not produce the system; they are interested in what processes in fact produced the system. And their hypotheses about those processes are falsifiable. For example, in a 2006 paper in Nature Reviews Microbiology, Mark J. Pallen and Nicholas J. Matzke suggested that the ur-flagellum arose “from mergers between several modular subsystems: a secretion system built from proteins accreted around an ancient ATPase, a filament built from variants of two initial proteins, a motor built from an ion channel and a chemotaxis apparatus built from pre-existing regulatory domains,” and was honed by selection for motility. That’s just a sketch; they provide further details.
But the details of Pallen and Matzke’s hypothesis aren’t relevant here. What matters is that it is falsifiable. (For example, if any of the subsystems they mention were discovered to postdate the flagellum, then it would be falsified.) Given that the hypothesis appeals only to unintelligent processes, it’s true that the hypothesis logically entails the claim that Behe ascribes to “Darwinism.” But nothing about the falsifiability of the hypothesis itself follows. Compare: if it’s true that Janet Parker died from smallpox in Catherine-de-Barnes, England, in 1978, then it’s true that someone died from smallpox somewhere sometime. The latter hypothesis, by Behe’s lights, is unfalsifiable, just as the claim he ascribes to “Darwinism” is, for there are a potentially infinite number of deaths (including deaths in the future and elsewhere in the universe) to investigate. But the hypothesis about Janet Parker’s death is clearly falsifiable nevertheless.
The second problem is that “intelligent design,” as Behe defines it, isn’t what he wants. In the first place, given that no unintelligent process could produce the bacterial flagellum, it still doesn’t follow that some intelligent process could produce the bacterial flagellum. Perhaps no process whatsoever could produce the bacterial flagellum. It would presumably be necessary to decide then whether the bacterial flagellum doesn’t exist at all—hoax, myth, illusion?—or whether it existed, uncreated, from all eternity. Either way, it would not have been produced by a process, intelligent or unintelligent. Certainly neither of these seems like a tremendously plausible scenario, but since Behe was, in Darwin’s Black Box (1996), willing to entertain the possibility that the bacterial flagellum was produced by extraterrestrial aliens or by time-traveling cell biologists from the far future, he’s not really in a position to complain about implausible scenarios.
Therefore, “intelligent design” with respect to the bacterial flagellum would be better defined as “Some intelligent process”—rather than “No unintelligent process”—“could produce the bacterial flagellum.” But that’s not quite what Behe wants either. For the structure of bacterial flagella (at least for certain bacteria) is understood in exquisite detail, and even if it is not currently feasible to assemble a replica atom by atom, the barriers are, as far as I know, purely technical. It’s not like it would take a miracle to do it. So “intelligent design” as redefined isn’t in conflict with “Darwinism” as defined by Behe. Really, “intelligent design” with respect to the bacterial flagellum would be best defined as “Some intelligent process did in fact”—rather than “could”—“produce the bacterial flagellum.” That final redefinition is enough to restore the head-to-head conflict between “intelligent design” and “Darwinism”—or is it?
It isn’t, because the meaning of “intelligent process” isn’t given. It’s not inconsistent with the latest redefinition of “intelligent design” to construe the processes to which Pallen and Matzke appeal as a manifestation of divine intelligence, for example. (Whether such a construal would be philosophically satisfactory or theologically acceptable is a separate question, not addressed here.) Importantly, by the same token, it would be a mistake to claim that “intelligent design,” on its latest redefinition, is shown to be falsifiable by the fact that Pallen and Matzke’s hypothesis is falsifiable. Rather, “intelligent design” on its latest redefinition is, simply, unfalsifiable. But that’s not necessarily a problem: after all, “Darwinism” as defined by Behe is unfalsifiable, but that’s not important because particular instances of it, like Pallen and Matzke’s hypothesis, are falsifiable. Are there particular instances of “intelligent design” that are falsifiable?
Sure, although none seems particularly plausible. (The hypothesis that I produced the bacterial flagellum, for example, is eminently falsifiable—in fact, it is easily falsified.) Of course, proponents of “intelligent design” have failed, and indeed barely seem to have tried, to present a plausible falsifiable version of “intelligent design” with regard to the bacterial flagellum. Instead, their efforts have focused on either arguing from irreducible complexity to unevolvability and then to design—which, for familiar reasons, limps at every stage of the argument—or on taking scientific hypotheses not involving design to task for not meeting their arbitrary demands for detail. But, returning to the topic at hand, namely Behe’s argument that “intelligent design” is and “Darwinism” is not falsifiable, it’s clear that it is a miserable failure. Presenting it, he suggested that it might seem counterintuitive. Those were intuitions that should have been heeded.
