Philosopher James Stump resigned recently from his faculty position at Indiana’s Bethel College to join BioLogos, citing differences of views over theological questions relating to human origins. BioLogos, of course, is the well-known center of theistic evolutionary advocacy. The move caught the attention of education reporters, and it adds some interest to a generally favorable review by Stump of a recent book by Benjamin C. Jantzen, An Introduction to Design Arguments (Cambridge University Press).
The book is marred by major errors on irreducible complexity, the bacterial flagellum, and more. You can find Stump’s review over at Christian Century.
I had actually picked up the book by Jantzen, a philosopher at Virginia Tech, soon after it came out. I found it to be a helpful resource, despite a sometimes condescending tone towards intelligent design. Obviously it’s very rare for a book to cover a topic with 100 percent completeness, but I immediately noticed that something important was missing: any discussion of Stephen Meyer’s method of inferring design via an inference to the best explanation. Scanning the index, and I found that Meyer — the foremost philosopher of science in the ID movement — goes entirely unmentioned. Stump notes the same problem:
In the first half of the book, Jantzen outlines many other interesting treatments of design arguments throughout history. But for many readers the most important and helpful part will be the large section on the contemporary intelligent design movement. It is a curious omission that there is no treatment at all of intelligent design’s currently most prominent figure, Stephen Meyer. His books Signature in the Cell (2009) and Darwin’s Doubt (2013) have become the leading edge of the movement. I suspect that Meyer’s work came along too late in the development of Jantzen’s book, which is clearly the fruit of many years of engagement with the material. It is regrettable, though, that what aims to be a comprehensive treatment of design arguments does not include the most important contemporary exemplar.
Stump is right. Instead, Jantzen focuses on irreducible complexity and specified complexity — mainstay arguments of ID to be sure — but even there his analysis is highly inaccurate. A preview of what’s wrong with Jantzen’s treatment is found in Stump’s review:
In his argument for irreducible complexity Behe appeals to natural structures that have many different parts and perform a certain task, like causing the blood to clot or propelling bacteria. He argues that it wouldn’t be possible to complete such a task without all the parts functioning together; therefore evolution, which works through gradual accretions, can’t account for such structures. So, Behe claims, there must be a supernatural designer who brought about these irreducibly complex structures.
That’s not quite right. For one thing, Behe doesn’t claim that we can go from irreducible complexity all the way to a “supernatural designer.” As Behe writes in Darwin’s Black Box:
The conclusion that something was designed can be made quite independently of knowledge of the designer. As a matter of procedure, the design must first be apprehended before there can be any further question about the designer. The inference to design can be held with all the firmness that is possible in this world, without knowing anything about the designer.
(Michael Behe, Darwin’s Black Box, p. 197 (Free Press, 1996).)
Elsewhere Behe makes clear that his argument for design doesn’t specify whether the designer is natural or supernatural:
[ID] is not an argument for the existence of a benevolent God, as Paley’s was. I hasten to add that I myself do believe in a benevolent God, and I recognize that philosophy and theology may be able to extend the argument. But a scientific argument for design in biology does not reach that far. Thus while I argue for design, the question of the identity of the designer is left open. Possible candidates for the role of designer include: the God of Christianity; an angel — fallen or not; Plato’s demiurge; some mystical new age force; space aliens from Alpha Centauri; time travelers; or some utterly unknown intelligent being. Of course, some of these possibilities may seem more plausible than others based on information from fields other than science. Nonetheless, as regards the identity of the designer, modern ID theory happily echoes Isaac Newton’s phrase hypothesis non fingo [“I frame no hypothesis”].
(Michael Behe, “The Modern Intelligent Design Hypothesis,” Philosophia Christi, Series 2, Vol. 3, No. 1 (2001), pg. 165 (emphasis added).)
Likewise at the Dover trial, Behe testified:
Q: So is it accurate for people to claim or to represent that intelligent design holds that the designer was God?
Behe: No, that is completely inaccurate.
Q: Well, people have asked you your opinion as to who you believe the designer is, is that correct?
Behe: That is right.
Q: Has science answered that question?
Behe: No, science has not done so.
Q: And I believe you have answered on occasion that you believe the designer is God, is that correct?
Behe: Yes, that’s correct.
Q: Are you making a scientific claim with that answer?
Behe: No, I conclude that based on theological and philosophical and historical factors.
(Michael Behe, October 17 Testimony, AM Session.)
Jantzen makes another mistake when he tries to frame irreducible complexity as a strictly negative argument against Darwinian evolution:
The idea is to eliminate a small set of plausible explanations (e.g. natural selection, chance, etc.) and from this negative result to argue for intelligent agency. … [W]e can’t be confident that we have eliminated all other possibilities. (p. 195)
Here we see how unfortunate it is that Jantzen didn’t engage with the arguments of Stephen Meyer. Stump lets Jantzen off the hook by pointing out that Meyer’s work may be too recent to include. Is it possible that Jantzen’s book had been sitting on a shelf awaiting publication since before 2009 when Meyer’s landmark book Signature in the Cell was published? That seems unlikely, since according to his CV, Jantzen only got his PhD in 2010. So he was surely studying up on the subject of intelligent design as Meyer released his work. Anyway, as far back as 2004 — ten years before Jantzen published his book — Meyer explained how irreducible complexity can provide a positive argument for design:
In all irreducibly complex systems in which the cause of the system is known by experience or observation, intelligent design or engineering played a role [in] the origin of the system. . . . Although some may argue this is a merely an argument from ignorance, we regard it as an inference to the best explanation . . . , given what we know about the powers of intelligence as opposed to strictly natural or material causes.
(Scott A. Minnich & Stephen C. Meyer, “Genetic Analysis of Coordinate Flagellar and Type III Regulatory Circuits in Pathogenic Bacteria,” in Proceedings of the Second International Conference on Design & Nature, Rhodes Greece.)
Meyer’s argument as an “inference to the best explanation” shows that we don’t have to eliminate every single possible known or unknown material cause before we infer design. As I have noted before, Meyer only claims that “known” material causes cannot explain certain structures. He then shows that this means we can infer design using standard methods of historical sciences via the principal of uniformitarianism. We don’t have to (and in fact never could) eliminate every possible unknown material cause, since those causes aren’t known.
Jantzen seems unaware of this forceful rebuttal, for he then writes:
Whether or not this objection is decisive, there are other grounds on which to reject Behe’s argument, irrespective of whether all possibilities have been accounted for. Specifically, one could argue that in fact natural selection can produce a system that is irreducibly complex. (p. 196)
Jantzen then produces the old example of the Type III secretory system as an “intermediate” to the bacterial flagellum:
Behe claims that systems which are irreducibly complex with respect to some capacity cannot be favored by selection when a part is removed. If he is right, then the flagellum — which loses the capacity to propel the bacterium when one of its component proteins is removed — cannot have intermediate structures that are favored by natural selection. However, [Ken] Miller points out that a subset of the proteins in a typical flagellum make up a very important structure found in many disease-causing bacteria. That is, a flagellum missing some of its pieces is found to be highly favored by selection for an altogether different capacity than motility. The structure in question is the Type III injectisome which bacteria use to pierce the cell membranes of other organisms and inject a variety of proteins. In injectisome is really just a flagellum missing the long filament of the propeller and the proteins that bind the propeller to the drive-shaft. So an intermediate system that contains most of the parts of a flagellum could have been favored by natural selection for its capacity to function as a syringe. To get a system favored for motility, only one or two additional proteins related to the filament would have to be introduced. So even though the flagellum is irreducibly complex with respect to the capacity for swimming, taking away one part does not give a ‘nonfunctional’ system — it give a system with other capacities favored by natural selection. (p. 199, emphasis added)
I dealt with this argument just last week in a lengthy piece looking at structural, phylogenetic, and ecological reasons why the Type III secretory system (T3SS) and the injectosome it helps build cannot be an “intermediate” to the bacterial flagellum. Beyond that, there are many logical and factual errors in Jantzen’s statement.
First, it’s simply not the case that the flagellum is nothing more than a Type III injectisome plus “only one or two additional proteins related to the filament.” This is an incredibly incorrect claim! True, there are diverse types of flagella, but they all have an irreducibly complex core of components — one paper notes that even many of the proteins in the basal body (i.e., the part of the flagellum that doesn’t include the filament) are different from the T3SS:
Of the roughly 20 proteins needed to form the flagellar basal body, about half have clear counterparts in the injectisome (Table 1).
It’s news to nobody familiar with such things that about 10 proteins in the flagellum are homologous to those of the injectisome’s T3SS. And what about the other 10 or so proteins in the flagellar basal body that aren’t found in the T3SS? What do they do and how are they different from the T3SS? Many of them, like MotA and MotB help form the motor — i.e., the spinning — function for the flagellum, and these crucial proteins are entirely absent from the T3SS. If you added merely a filament on top of a T3SS, you wouldn’t go anywhere because without the motor components, the filament wouldn’t spin!
Then there’s FlgI which makes up the P-ring, and FlgH which makes up the L-ring, both of which lack known homologues in the T3SS or anywhere else.
Moroever, there is more to a flagellum than a basal body (which includes the T3SS and the motor) and the filament. There’s also the flexible hook that couples the rotary motor to the filament:
The structure of the flagellum can be divided into three parts: I) the basal body as the rotary motor and the Type III export apparatus; II) the flexible hook that couples the rotary motor to III) the rigid filament. (Erhardt, 2010; internal citations removed)
The hook and its connector component are composed of multiple proteins like FlgD, and FlgL, but there’s also the rod that transmits force from the motor to the hook composed of proteins like FliE and FlgB, FlgC, FlgF, and FlgG. And don’t forget the cap proteins that are crucial for the assembly of rod (FlgJ), hook (FlgD) and the filament in the flagellum (FliD), all of which are absent in the T3SS.
I could go on. The point here is that there’s a lot more to a flagellum than a T3SS plus “only one or two additional proteins related to the filament.” In fact, what the flagellum requires is a whole suite of proteins specifically related to the motor apparatus of the flagellum and the various components that transmit its spinning motion to the filament, and the filament itself. All of these lack counterparts in the injectisome’s T3SS.
Those are some of the scientific problems with Jantzen’s argument. There are also severe logical problems. The first is that Jantzen misstates how we test for irreducible complexity. He contradicts himself when he says that something “irreducibly complex with respect to some capacity cannot be favored by selection when a part is removed” and yet also says that “a flagellum missing some of its pieces is found to be highly favored by selection for an altogether different capacity than motility.” This commits the same mistake about how we test irreducible complexity as Matt Baker made a few weeks ago.
In assessing whether a system is irreducibly complex, we focus on the function of the system itself and the impossibility of its being built in a stepwise fashion, not on a possible function that one particular sub-part could have elsewhere. Michael Behe asserts that a system is irreducibly complex when it stops functioning upon the removal of one part. This is the appropriate test of Darwin’s theory because it asks the question, “Is there a minimal level of complexity which is required for functionality of this system?” Behe never suggests that some subsystems cannot play some other role in the cell — in fact he suggests the opposite. Behe simply argues that evolution requires the total system be built up in a slight, step-by-step fashion, where each step is functional.
For example, I could probably use my laptop’s power cord to power my toaster, and its screen as an alternative to my television. But that doesn’t mean that there aren’t key systems in my laptop related to the CPU and its processing functions that are not irreducibly complex. It certainly doesn’t mean that you can evolve my laptop in a gradual step-by-step manner wherein every part remains functional along every minute step of the evolutionary pathway.
Thus, Jantzen has applied a common but fallacious straw man test for irreducible complexity. The test does not ask “Can one subcomponent of the macrosystem be used to do something else?,” but rather “Can the system as a whole be built in a step-by-step fashion which does not require any ‘non-slight’ modifications to gain the final target function?” Any non-slight modifications of complexity required to go from functional sub-part(s), operating outside-of-the-final system, to the entire final functional system, represent the irreducible complexity of a system.
Jantzen certainly does not provide anything like a step-wise evolutionary explanation for the origin of the flagellum, nor does he attempt to explain how the T3SS and flagellum could interconvert or give rise to each other. Given that he’s a philosopher who is just looking to see what is logically possible or not, his quick, uncritical dismissal of the problem of irreducible complexity is most unfortunate.
Thus, I find it intriguing that Jantzen writes that “If he [Behe] is right, then the flagellum — which loses the capacity to propel the bacterium when one of its component proteins is removed — cannot have intermediate structures that are favored by natural selection.” But that’s exactly what Scott Minnich’s genetic knockout experiments showed. According to the results of Minnich’s research, the flagellum is irreducibly complex.
Despite all these problems, Stump seems to endorse this very argument from Jantzen. Stump writes:
The idea of irreducible complexity has had remarkable intuitive staying power among ID followers, but when the intuition is converted into an argument, it has considerably less persuasive force. First, almost all biologists think Behe is wrong about the specific examples of structures that he says are unexplainable by evolution. But most people’s intuition is guided by a caricature of how evolution works. They think that each structure or trait develops in isolation. In reality, natural selection operates on combinations of traits, not merely on isolated structures. Half-developed wings won’t help an insect fly, but they might help it do other things that contribute to its survival, like skim across the surface of water. Contrary to the ID claim about irreducible complexity, you don’t have to get the whole thing at once.
In reality, Behe has written quite a bit about indirect evolutionary routes, how they might not be absolutely impossible but are highly implausible, and are never demonstrated to be true. But I think the real reason irreducible complexity has demonstrated its staying power is that attempts to refute it have been so weak and have misstated how we test irreducible complexity. Arguments that haven’t been refuted tend not to go away.
The Danger of Capitulating to So-Called “Settled Science”
In light of Jantzen’s serious factual and logical errors about irreducible complexity and the flagellum, it’s ironic that Stump closes his book review by stating:
The ID camp does a disservice to the predominantly conservative Christian community to which it appeals by conditioning that community to mistrust science. Its arguments depend on accepted, settled science getting things wrong.
I don’t agree that ID encourages anyone to “mistrust science.” Instead, we encourage people to think for themselves. If the consensus turns out to be right, then fine! But if we’re never willing to question what some label “settled science,” then we end up with precisely what we see here: Cambridge University Press publishing wildly inaccurate arguments against intelligent design, and terribly false claims about the flagellum. If we’re going to seek and find the truth, then Jantzen’s errors need to be corrected. But it’s the ID camp — not Dr. Stump’s theistic evolutionist camp — that is equipped to correct such errors.
Unfortunately, Dr. Stump’s philosophy — which I have commonly observed among my theistic evolutionist friends at BioLogos — seems to be to avoid questioning what they call “settled science.” This default approach can be a major barrier to the search for truth. I prefer ID’s perspective, which holds that the consensus deserves respect and very serious consideration, but that skepticism can also be called for.
Stump’s capitulation to the “settled science” of Jantzen’s book goes so far that he even says in his review, “by the end I almost felt sorry for design advocates as the soft underbelly of their arguments was exposed.” Cornelius Hunter has an excellent post analyzing Stump’s triumphalistic claim, and it seems clear that Stump’s philosophy — to eschew questioning the “settled” consensus view — has led to erroneous conclusions about the strengths of ID arguments. This approach is bad for science. Hunter writes:
Unfortunately what the philosopher [Dr. Stump] demonstrates here is not a helpful and insightful commentary on design arguments but rather the usual sequence of evolutionary misrepresentations.
It begins with Stump’s appeal to authority. This is a common evolutionary argument, but the fact that a majority of scientists accept an idea means very little. Certainly expert opinion is an important factor and needs to be considered, but the reasons for that consensus also need to be understood. The history of science is full of examples of new ideas that accurately described and explained natural phenomena, yet were summarily rejected by experts. Scientists are people with a range of nonscientific, as well as scientific influences. Social, career, and funding influences are easy to underestimate. There can be tremendous pressures on a scientist that have little to do with the evidence at hand. This certainly is true in evolutionary circles, where the pressure to conform is intense.
In contrast, ID’s motto is “Follow the evidence where it leads.” The theistic evolutionary motto seems to be more like “Follow the evidence so long as it leads to the consensus.” None of this means that we mistrust science or tell anyone else to do so. It means we put the pursuit of truth ahead of the pursuit of universal agreement. That’s the best way to foster the progress of science.