This article is from
Journal of Creation 35(3):29–33, December 2021

Browse our latest digital issue Subscribe

Irreducible complexity remains a fatal problem for Darwinian evolution

A review of: A Mousetrap for Darwin: Michael J. Behe answers his critics by Michael J. Behe
Discovery Institute Press, Seattle, WA, 2020



Author Michael J. Behe is a leading activist in the Intelligent Design (ID) movement. He is Professor of Biological Science at LeHigh University. Behe is a biochemist, and much of this work is rather technical and biochemically oriented in his back-and-forth polemics with leading evolutionists. Since biochemistry is not my field, I do not elaborate on this subject in my review. Rather, I focus on broader principles.

The author addresses many evolutionary objections to ID. The reader who has followed this issue for many decades, as I have, may be impressed by the fact that the same bogus arguments used against creationists are now being recycled and used against ID.

Behe has a sense of humor. When an evolutionist objected that ID members extensively quoted evolutionists [BTW, the same objection once used on creationists], Behe retorted, “What is a book concerning evolution supposed to contain if not quotes from evolutionists? Quotes from accountants?” (p. 56).

The author’s personal journey

It was disillusionment with Darwinian evolution, and not religious teachings, that led Behe to embrace ID. He writes:

“Perhaps I can help. After all, I used to believe that a Darwinian process did indeed build the wonders of life; I had no particular animus against it. Yet I believed it on the say-so of my instructors and the authority of science, not on hard evidence. When I read a book criticizing Darwin’s theory from an agnostic viewpoint it startled me, and I then began a literature search for real evidence that random mutation and natural selection could really do what was claimed for them. I came up completely empty. In the over thirty years since then, I’ve only become more convinced of the inadequacy of Darwinism, and more persuaded of the need for intelligent design at ever-deeper levels of biology, as detailed in my books” (p. 450).

Nothing has changed. Behe still concludes that “Random mutation/natural selection works great in folks’ imaginations, but it’s a bust in the real world.” (p. 109).

The mousetrap as an example of irreducible complexity: frivolous objections

Author Behe points out that he chose the mousetrap as an example of intelligent design because the layperson can more easily grasp its irreducible complexity than that of a folded protein. It is plain to see that the mousetrap must have all its parts in place before it will work at all: There is no such thing as a half-mousetrap.

Some evolutionists have argued that there are many ways to kill a mouse other than the mousetrap (for example, glue-covered cardboard). This is true, but is completely irrelevant to the existence of the mousetrap and its putative origins without an intelligent designer. In addition, many different kinds of metal-and-wood mousetraps can be built (figure 1), but this does not vitiate irreducible complexity. In fact, each type of trap has its own irreducible complexity.

Figure 1. Although many different kinds of mousetraps can be built, this does not vitiate irreducible complexity.

Other evolutionists, invoking exaptations, have noted that the mousetrap components, when acting alone, have various functions. For instance, the spring is used in many different devices. They posit that the different components got together and acquired new functions once they combined in the mousetrap. But this does not eliminate the need for an intelligent designer; it only changes its role a bit. An intelligent designer is still needed to assemble the parts and harmonize their hitherto-disparate functions; to get each component to work near-perfectly, and to do so simultaneously, with every other component in the mousetrap. Otherwise, there is no functioning mousetrap at all.

Attempts to assemble the mousetrap step by step

Some evolutionists have argued that a simpler mousetrap is possible. One suggestion involved the use of a spring-alone ‘mousetrap’, wherein the spring ends are abutted against each other, in perpendicular fashion, and with just the right tension, to become uncocked when the mouse takes the cheese.

Behe takes a closer look at this. To begin with, the spring-alone ‘mousetrap’ is itself an example of irreducible complexity. Unless the spring-ends are abutted against each other in a very precise way, and with just the right amount of tension between them, there is zero possibility of catching a mouse. Intermediate stages are useless and have no selective value. So even a spring-alone ‘mousetrap’ could not arise by small, successive, naturally selected changes as required by Darwinism.

In addition, the spring-alone ‘mousetrap’ is incapable of being straightforwardly imported into the regular-mousetrap design. The ends of the spring-alone ‘mousetrap’ are deployed in a manner completely different from the way they are deployed relative to the trigger mechanism and the lever in the regular mousetrap. Again, no series of small, successive Darwinian changes, each naturally selected, will transform a spring-alone ‘mousetrap’ into a workable component of the regular mousetrap. This owes to the fact that the intermediate states will not function as mousetraps, and so natural selection will not favour their survival.

So we are right back to an intelligent designer. This time an intelligent designer is needed to bend the spring ends so that they jibe perfectly with the trigger mechanism and the lever in the regular mousetrap. Clearly, then, the postulated spring-alone ‘mousetrap’ is in no sense a simplified version of the regular mousetrap. Nor is it in any way a potential stage in the hoped-for gradual Darwinian appearance of the regular mousetrap.

Intelligent Design (ID) is not a religious teaching

Opponents of ID never tire of trying to dismiss it as a religion and as something requiring a prior belief in a deity. It is not and does not. Behe writes: “The conclusion of intelligent design flows naturally from the data itself, not from sacred books or sectarian beliefs” (p. 31).

In my review, I use the word ‘Designer’ with a capital ‘D’ in deference to the Christian belief that God is the Designer. However, a designer or Designer can be inferred solely from reason and observation, just as a designer can be inferred from situations that have nothing to do with religion. For instance, Behe cites the example of SETI (Search for Extraterrestrial Intelligence). When ‘listening’ to radio waves coming from space, attempts are made to identify an intelligently designed radio signal from the background, unintelligent ‘noise’.

Now some evolutionists have argued that we can accept a designer for the mousetrap because we see designers making them, but we cannot accept a Designer for living things because we never see this process at work. The SETI example, above, dispenses with this silly argument. We can discern a designed radio signal even though we cannot comprehend, let alone see, the extraterrestrial entity that made it.

Design or non-design are inferred in everyday situations. Behe writes:

“Forensic scientists routinely make decisions of whether a death was designed (murder) or an unfortunate accident. Archeologists decide whether a stone is a designed artifact or just a chance shape” (p. 59).

Again, this has nothing to do with any kind of prior commitment to religion.

Ironically, perhaps, it is evolutionists that introduce religion into the picture, as when they regularly make ‘poor design’ arguments. Behe will have none of that. He comments:

“On a different note, I’m glad Professor Doolittle likes Rube Goldberg too, but unfortunately it supplies what I think is his rock-bottom reason for deciding that natural selection produced the [blood] clotting cascade: ‘no Creator would have designed such a circuitous and contrived system’. Well, Doolittle is a good scientist, but he’s no theologian, and he doesn’t serve science well when he lets his theological presuppositions influence his scientific judgment” (p. 59).


ID is not creationism

Quite a few evolutionists hope that the ‘bad smell’ of creationism will attach itself to ID. So they repeat the hoary argument that ID is just ‘religion in disguise’—a ‘dressed up’ version of creationism. It may be good propaganda, but it is completely contrary to the facts.

Behe spurns the label ‘creationist’ (p. 62). In fact, most of his views are indistinguishable from those of the standard evolutionist. He believes in an old Earth and in the common ancestry of all living things (p. 218). He indicates that he would be just as surprised as anybody else at the discovery of a human skeleton in the Precambrian. So he accepts the entire evolutionary-uniformitarian package, and believes in the reality of phylogenies. For example, he concurs with the premise that tetrapods evolved from fish (p. 417).

In order to understand Behe’s position, one must not confuse the fact of design with the process of design. He stresses that “design focuses on the endpoint, not the process. Arguments about process require further evidence beyond that for the bare conclusion of design” (p. 474). For this reason, there is, at this stage of evidence, no room for any kind of supernatural, fiat creation in Behe’s thinking. He rejects the claim, attributed to him, that the irreducibly complex system must have originated ex nihilo by Divine fiat (pp. 63, 466). Furthermore, he provisionally rejects any kind of miraculous origin for irreducibly complex systems (p. 473). Rather, Behe thinks in terms of some form of guided evolution to account for irreducibly complex systems.

There is the argument (also long used against creationism) that ID relies on negative evidence: The perceived inability of Darwinism to account for a living feature thus becomes evidence of ID. But, not mentioned by Behe, the same can be said of all other situations. Thus, in SETI, the failure of random ‘noise’ to account for certain extraterrestrial radio signals becomes evidence for a designed source of these radio signals. Likewise, the failure of accidental death to account for an unexpectedly dead person becomes evidence for a purposeful death (murder). Again, this is everyday reasoning, and has nothing to do with religion.

Behe realizes that conclusions of intelligent design, like all conclusions in science, are tentative. It is possible that such things as the self-organization of matter will eventually prove capable of explaining design without a designer (pp. 139–140). However, science must be based on what is presently known, and not on what perhaps may eventually develop. Right now, the best explanation for irreducible complexity is intelligent design, and it should be embraced.

ID is not just ‘natural laws in action’

Having seen that ID is not creationism, we must avoid the other extreme—of wrongly supposing that ID considers the presumed creative powers of nature sufficient to account for all living things. It does not.

In evolutionary thinking, Nature is sometimes endowed with creative powers. Behe points out that ID adds a layer of design to that provided by nature. He writes:

“By ‘intelligent design’ I mean to imply design beyond the laws of nature. That is, taking the laws of nature as given, are there other reasons for concluding that life and its component systems have been intentionally arranged?” (p. 130).

Defining away ID as science

Some evolutionists have argued that science is whatever the scientific community says it is—since prominent scientific organizations have decreed that ID is not science, therefore it is not science. This sociological view of science, and reasoning against ID, was adopted by the judges in the US court cases that involved ID.

Behe points out that saying in advance what is and is not science only risks conflating the presumptions and prejudices of the current group of practitioners with the way that physical reality must be understood. Science should be an unrestricted search for the truth about nature based on reasoning from physical evidence (p. 482). It should not start with preconceived notions.

Others have argued that scientists should stick to methodological naturalism and thereby exclude ID as science. This, too, means that science is being ruled by preconceived notions.

Finally, there are those who say that invoking a designer is variously an argument from ignorance, an act of ‘giving up’, and essentially a copout. This, too, is a preconception—that a naturalistic, non-intelligent cause is the correct one (p. 146).

ID is falsifiable

For a theory to be scientific, it has to allow for it to be potentially proved wrong. ID qualifies. Behe comments:

“To falsify design theory a scientist need only experimentally demonstrate that a bacterial flagellum [figure 2], or any other comparable complex system, could arise by natural selection. If that happened I would conclude that neither flagella nor any system of similar or lesser complexity had to be designed. In short, biochemical design would be neatly disproved” (p. 50).

Figure 2. The bacterial flagellum, an example of irreducible complexity.

Some evolutionists, such as Jerry Coyne, have argued that ID proponents ‘move the goalposts’. According to their argument, whenever such-and-such a complex system is shown not to need a designer, ID proponents just latch on to some other complex system and proclaim that it is the one that needs a designer. This ‘moving the goalposts’ argument is fallacious. From Behe’s quote in the preceding paragraph, it is evident that proof of a non-designer for a complex system would apply to all systems of comparable complexity, and not just that particular one.

The party line, as enunciated, for example, by the (US) National Academy of Sciences, is that ID is not falsifiable, and therefore not a part of science (p. 130). However, evolutionists such as Russell Doolittle and Kenneth Miller have claimed that their arguments have falsified ID. So, evolutionists try to have it both ways; simultaneously arguing that ID is not falsifiable, and then arguing that ID has been falsified (pp. 130–131). They cannot have it both ways! [Note that exactly the same evolutionistic gambit was tried against scientific creationists decades ago.]

Arguably, non-design is not falsifiable

Accusations of non-falsifiability can be turned around, and validly attached to evolution, just as they were decades ago in creation–evolution debates. Behe updates this consideration. He comments:

“If a scientist went into the laboratory and grew a flagellum-less bacterial species under selective pressure for many generations and nothing much happened, would Darwinists be convinced that natural selection is incapable of producing a flagellum? I doubt it. It could always be claimed that the selective pressure wasn’t the right one, or that we started with the wrong bacterial species, and so on. Even if the experiments were repeated many times under different conditions and always gave a negative result, I suspect many Darwinists would not conclude that Darwinian evolution was falsified” (p. 132).

Do not confuse survival of the fittest with the arrival of the fittest

In his response to evolutionists, Behe points out that they resort to speculation and just-so stories. He remarks:

“From my point of view, in all of the cited papers the evolutionary explanation takes the form ‘system X developed because it would help the cell do Y,’ without noticing the difficulties of making X by a blind process. It’s like saying, ‘Air conditioners developed to enable more people to work indoors in the summertime’” (p. 61).

Evolutionists often mix processes with inferred outcomes. Behe comments:

“… the argument for intelligent design in biology has little to do with protein-sequence similarity or common ancestry, for the same reason that knowing all the parts are made of metal does not explain the mousetrap … doesn’t even begin to explain how a mousetrap could be built step by step by a random process” (pp. 177–178).

Let us take this further. Behe quotes evolutionist Jerry Coyne, “First, as both Dawkins and I point out, if random mutations cannot build complexity, how can they possibly have been so effective in artificial selection of plants and animals?” (p. 227). Clearly, Dawkins and Coyne are assuming what they need to demonstrate, as unmasked by Behe:

“Because, of course, the genomes of many plants and animals already contain much developmental plasticity. Turning some existing genes or regulatory elements on or off, or tuning them up or down, or changing them slightly by simple, single mutations, can certainly affect the shapes and other properties of organisms somewhat. Artificial selection for such variants can easily explain dog breeds and such, as I noted in Chapter 9. But of course that begs the question of where the complex systems controlling the organisms’ development came from” (p. 227).

The blood-clotting cascade: gene duplication alone does not explain it

Behe perceptively comments:

“But does gene duplication lead straightforwardly to the blood clotting cascade? No. The important thing to keep in mind is that a duplicated gene is simply a copy of the old one, with the same properties as the old one—it does not acquire sophisticated new properties simply by being duplicated” (p. 93).

Aerobic citrate metabolism: an observed evolutionary novelty?

Ever since 1988, Richard Lenski, of the University of Michigan, has been conducting the LTEE (long-term evolution experiment). He has, so far, cultured over 73,500 generations of E. coli bacteria, observing the changes that have occurred in this period of time.

Normally, E. coli only metabolizes citrate in an anaerobic environment. One day, Lenski discovered a new laboratory strain that did this in an aerobic environment. The evolutionistic triumphalists hyped this as an evolutionary novelty. Not quite. Here is how Behe assesses this situation:

“The gene duplication which brought an oxygen-tolerant promoter near to the citT gene did not make any new functional element. Rather, it simply duplicated existing features. The two FCT’s comparing the oxygen-tolerant citrate transporter locus—the promoter and the gene—were functional before the duplication and functional after” (pp. 279–280).

Darwinian evolution stumbles with changes requiring even two mutations

Evolutionists glibly speak of neutral mutations becoming fixed in the population, and then a sequence of such mutations coming together with a successive mutation in order to make a substantial evolutionary change. It may be that way in theory, but the facts are very different.

Consider drug-resistant malaria. A certain mutation, K76T, considered to be neutral but now shown to be deleterious (p. 361), by itself confers not even a tiny bit of resistance to chloroquine, so we are not talking about an incremental, step-by-step increase in chloroquine resistance. Instead, a second mutation arises, and acts with K76T in order to make that malarial parasite resistant to chloroquine.

Behe analyzes this situation:

“The need for two specific mutations neatly explains the approximately billionfold increase in difficulty for the [malarial] parasite to evolve resistance to chloroquine versus other antimalarial drugs, such as atovaquone, which require only one … . It also illustrates the feebleness of the Darwinian mechanism when confronted with the need for even the tiniest amount of coordination—just two simple point mutations. The difficulties go up exponentially with the number of mutations required” (p. 405).

Some evolutionists have objected that there are many ways for a malarial parasite to acquire chloroquine resistance. Actually, there are potentially only a few. Behe remarks:

“Instead of a chloroquine pump appearing, maybe a chloroquine-degrading enzyme could have arisen, or maybe malaria’s membrane could somehow be altered to stop chloroquine from entering the cell, or maybe the cell could become dormant until the chloroquine passed. But none of these scenarios happened. Why not? Because any evolutionary pathway leading to these outcomes was even less probable than the pathway that occurred … 1 in 1020” (p. 354).

Behe’s 1 in 1020 probability is based on the number of cells per malaria patient (a trillion) times the number of ill people over the years (a billion), divided by the number of independent events (10). (p. 332)

ID will win out in the end

Behe discusses the politics that has gone on in order to suppress and discredit ID. He remains unfazed by all this, and has this forecast:

“Nonetheless, I’m serenely confident that the Darwinian paradigm will collapse, and the idea of the intelligent design of life will eventually carry the day, even in the snootiest intellectual circles, even in state-controlled schools. This will happen not because of any particular thing that I or any other ID proponent has done or will do. Rather, it’s simply because that’s the way the biological data are headed. Virtually every week stunningly sophisticated new features of life are being uncovered by scientists, features that in any other context we would immediately recognize as designed. This evidence cannot forever be denied” (p. 461).


Darwinian evolution continues to be an inadequate explanation for living things that exhibit irreducible complexity. To the open-minded, intelligent design remains a viable explanation for this.

Posted on homepage: 23 December 2022