Darwin’s Black Box by Michael Behe

Michael Behe. Darwin’s Black Box: The Biochemical Challenge to Evolution. The Free Press: New York, 1996.

Summary: An argument premised on the author’s lack of imagination.

Biochemist Michael Behe’s 1996 book Darwin’s Black Box is, to a significant extent, the fulcrum and rallying point of the modern “intelligent design” movement. Partially this is because his book is one of the few ID works that contain any genuinely original material, and therefore others tend to rely on him (the work of other prominent supporters of ID, such as William Dembski, is entirely dependent on Behe’s argument, and Behe’s claims regarding the bacterial flagellum have become iconic to ID), but also because Behe is one of the very few advocates of intelligent design who possesses any relevant scientific credentials whatsoever. Although the movement is awash with lawyers, mathematicians, philosophers and theologians, ID supporters who actually have academic degrees in evolutionary biology or some related field are conspicuously lacking. By contrast, the National Center for Science Education was able, in a very short time and with no advance notice, to put together a list of hundreds of practicing, qualified scientists who unequivocally support evolution – all of whom are named Steve. Of course, scientific disputes are ultimately decided by the evidence, not by appeals to authority, but it is nevertheless telling that advocates of ID have attracted so little support from precisely that group that is best positioned to judge the quality of their arguments. Although Behe is at best a drop in the bucket against this trend, he has been lionized by the creationist community. It is therefore worthwhile to show that even his work, despite his legitimate degree, contributes nothing to the cause of overturning evolution. As I said, scientific disputes are ultimately decided by the evidence, and the evidence shows that Behe’s argument does not hold up.

There are several themes that recur throughout the book, and it will hopefully prove more illuminating to treat each of them comprehensively rather than in piecemeal fashion. Therefore, this review, unlike others on this site, will be organized by topic rather than following the order of Behe’s book.

Behe’s Views on Common Descent

Before delving into Behe’s argument, however, there is an important fact about his beliefs worth pointing out. Although intelligent design advocates in general, including Behe, are aptly described by the term “creationist” (despite their strenuous disavowals of the word), not all creationists are of the extreme young-earth variety that believes in a 6000-year-old universe, a literal Garden of Eden, and a worldwide flood survived by a breeding pair of each species carried on an ark. The prominent ID advocate William Dembski does not believe this, for example, and neither does Behe. Rather than put words in Behe’s mouth, I will allow him to say it himself:

“For the record, I have no reason to doubt that the universe is the billions of years old that physicists say it is. Further, I find the idea of common descent (that all organisms share a common ancestor) fairly convincing, and have no particular reason to doubt it. I greatly respect the work of my colleagues who study the development and behavior of organisms within an evolutionary framework, and I think that evolutionary biologists have contributed enormously to our understanding of the world” (p.5).

“Like the sequence analysts, I believe the evidence strongly supports common descent” (p.176).

Many creationists have also expressed the (false) view that evolution necessarily implies atheism. Later in the book, Behe rebuts them:

“Ken Miller… is like myself a Roman Catholic, and he makes the point in public talks that belief in evolution is quite compatible with his religious views. I agree with him that they are compatible” (p.239).

Although Behe still doubts certain aspects of evolutionary biology, it is reassuring to hear that he is not as anti-science as some of his ideological allies. Those who expect him to support their more extreme views may well be dismayed at this.

Irreducible Complexity

The key argument Behe presents in his book is centered around a concept called irreducible complexity. He writes that this consists of:

“…a single system composed of several well-matched, interacting parts that contribute to the basic function, wherein the removal of any one of the parts causes the system to effectively cease functioning. An irreducibly complex system cannot be produced directly (that is, by continuously improving the initial function, which continues to work by the same mechanism) by slight, successive modifications of a precursor system, because any precursor to an irreducibly complex system that is missing a part is by definition nonfunctional. An irreducibly complex biological system, if there is such a thing, would be a powerful challenge to Darwinian evolution” (p.39).

Since evolution can only select for what works at the current moment, not what may be useful in the future, Behe claims that irreducibly complex systems could not have evolved because they would provide no selective advantage at all until they were fully assembled, and thus must have come into being all at once.

Such armchair hypothesizing, while it may sound attractive to opponents of evolution, fails the test of the evidence. Irreducible complexity has been shown to evolve both in living organisms and in computer simulations that use evolution as a problem-solving technique. For examples of this, see this post detailing the evolution of a novel, irreducibly complex multi-step enzyme pathway to break down a toxic man-made chemical, or this article concerning the creation, by a genetic algorithm, of an irreducibly complex circuit composed of 37 logic gates. Both these systems consist of multiple interacting parts all of which are required for them to carry out their primary function, thus meeting Behe’s definition; and yet, both these systems are indisputable products of evolution.

How can an irreducibly complex system evolve? There are three major pathways. The first can be summed up as scaffolding: extra parts which support a partially functional system until it is completely assembled, at which point the extra parts become unnecessary and are pruned away by selection. The second is the case of improvement becomes necessity, where an adaptation is at first merely beneficial, but as later changes build on it, it becomes indispensable. The third, possibly the most important, is change of function, also called cooption, a key mechanism of evolutionary change whose occurrence has been common knowledge among biologists ever since Charles Darwin. A system which originally evolved to perform one function may take on a new function, starting out with multiple functioning parts rather than having to acquire them one piece at a time. Even Behe acknowledges that this can happen:

“Even if a system is irreducibly complex (and thus cannot have been produced directly), however, one can not definitively rule out the possibility of an indirect, circuitous route” (p.40).

– although he fails to deal with its implications, other than a handwaving claim that such a process is too unlikely to be relevant. In reality, cooption is a hugely important and very common mechanism of evolution, and there is no excuse for him to pay as little attention to it as he does.

The Argument from Ignorance

Having presented the basic idea of irreducible complexity, Behe attempts to apply it to five biological systems: the waving cilia and rotating flagella that some single-celled organisms use to swim, the enzyme cascade that causes vertebrates’ blood to clot and seal wounds, the transport systems that cells use to move proteins from where they are manufactured to where they are needed, the antibody-creating mechanism that forms the “adaptive” arm of the vertebrate immune system, and the biochemical processes responsible for the synthesis of molecules such as AMP, a precursor to adenine (the “A” in the four-letter genetic code of DNA). All of these systems are highly complex, as Behe correctly points out.

However, that is by no means equivalent to a claim that such systems cannot evolve. There is no reason why evolution cannot produce systems of great complexity, and indeed plausible, detailed evolutionary explanations have been advanced for several of these systems as well as others. Behe claims that biologists are utterly baffled by these systems and have no idea how they could have come into existence, but this is false. For example, see these pages on vertebrate blood clotting, the bacterial flagellum and the antibody system.

Behe makes no attempt to engage the evolutionary models for the origin of any system in detail or show what, if anything, is wrong with them. The closest he comes to dealing with them is the following complaint relating to one such explanation:

“Doolittle appears to have in mind a step-by-step Darwinian scenario involving the undirected, random duplication and recombination of gene pieces. But consider the enormous amount of luck needed to get the right gene pieces in the right places. Eukaryotic organisms have quite a few gene pieces, and apparently the process that switches them is random. So making a new blood-coagulation protein by shuffling is like picking a dozen sentences randomly from an encyclopedia in the hope of making a coherent paragraph” (p.93).

This is a classic creationist fallacy, and someone with Behe’s education should know better. Such a probability calculation would only be relevant if evolution had been aiming specifically at this function and this particular method of achieving it, which is not how evolution works. There are undoubtedly a vast number of different ways to evolve a functional blood-clotting system, and even if there were none, then such a system would simply never have evolved and living things would just have had to make do without it. Attempting to calculate the probability of a unique event after the fact is a pointless exercise, akin to shuffling a deck of cards and then expressing amazement and concluding design must be at work because the particular permutation you obtained was so improbable.

This, then, is the fundamental paradox of intelligent design: Behe claims that evolutionary explanations have not been given for the systems he describes, but when such explanations are given, he simply says they are not detailed enough to meet his standards. It is a continual raising of the bar, a standard deliberately set to be impossible to meet. Meanwhile, all the time, he holds intelligent design apparently exempt from having to meet any evidentiary standards at all. As far as Behe is concerned, evolution’s failure to explain exactly how a biological system came into being, down to the specific genes that mutated and the exact probability of those mutations (a level of detail that would be impossible to obtain with anything less than a time machine), is a crippling blow against it; however, ID’s complete failure to explain anything about how an irreducibly complex system was created, when, where, or even by whom is not considered a problem. Even the least detailed evolutionary explanation possible at least has a mechanism – natural selection acting on random genetic mutation. ID does not even have that.

In essence, irreducible complexity is an example of the basic logical fallacy of argument from ignorance, in which a proposition is claimed to be false because it has not been proven to be true. In this case, Behe claims that because evolutionary theory has not produced detailed, step-by-step explanations of these systems, evolution must not have produced these systems.

If it were true that evolution could never produce an IC system, then Behe would have a point. But that is not true. There are well-understood principles by which evolution can produce such a system; evolutionary processes have been observed to produce such systems; and as the above section shows, Behe himself admits that this is possible. His argument, therefore, is essentially the same for each of these biochemical systems: describe how complex it is, express bewilderment at the idea that such a thing could have arisen through evolution, and conclude that God must have created it. This is an argument from ignorance in its pure form.

Publish or Perish

“Publish or perish” is a basic rule of science – for an idea to become widely accepted, its authors must do the research to support it and present their results in scientific journals for review. If supporting evidence is not forthcoming, the idea must be rejected. Behe asserts that this is precisely the case with evolution:

“There has been virtually no attempt to account for the origin of specific, complex biomolecular systems, much less any progress” (p.x).

“There has never been a meeting, or a book, or a paper on details of the evolution of complex biochemical systems” (p.179).

“In effect, the theory of Darwinian molecular evolution has not published, and so it should perish” (p.186).

This claim is flatly false, and was false at the time Behe’s book was published. There is a huge amount of published scientific literature on the evolution of biochemical systems, with more being added all the time. A search of scientific databases such as PubMed or the Proceedings of the National Academy of Sciences on simple phrases such as “biochemical evolution” returns hundreds of results. There is no excuse for any competent researcher to overlook the existence of material this extensive, and worse, to then base an argument on the presumed lack of such material.

If there is any proposal that is ailing for lack of published evidence, it is intelligent design. After 15 years, ID can boast a grand total of one supportive article in a scientific journal, and that article contains numerous factual errors and was published without the knowledge or consent of the journal’s governing council or editorial review board. Behe himself has never published any articles on the subjects of irreducible complexity or intelligent design, instead going straight to the public with a book that did not undergo scientific peer review, and even this book makes no attempt to explain the origin of any biochemical system beyond “Some intelligent designer did it.”

Later in the chapter, Behe continues to claim that no detailed evolutionary models of the origin of a given adaptation have been produced:

“After describing the whale the textbook remarks, ‘Thus we see in the sperm whale a remarkable anatomical and biochemical adaptation, perfected by evolution.’ But that single line is all that’s said! The whale is stamped ‘perfected by evolution,’ and everybody goes home. The authors make no attempt to explain how the sperm whale came to have the structure it has” (p.181).

“Of course, the possibility of design should cause researchers in biology to hesitate before claiming that a particular biological feature has been produced substantially by another mechanism, such as natural selection or transposition. Instead, detailed models should be produced to justify the assertion that a given mechanism produced a given biological feature” (p.230).

This is special pleading with a vengeance. Why does intelligent design not have to produce equally detailed models? Behe does not write as if ID is under any such pressure, and he certainly makes no attempt whatsoever to account for the origin of any complex biochemical system, nor does any other ID advocate. Can ID offer an explanation for any biological system any more detailed than “perfected by design”? If so, what is it and where has it been published? If not, why should ID be taken seriously as a possible alternative to evolution? After all, as Behe himself says so well:

“If a theory claims to be able to explain some phenomenon but does not generate even an attempt at an explanation, then it should be banished” (p.186).

Why Intelligent Design Is Not Science

The point made at the end of the last section ties into the most fundamental problem with intelligent design: no matter how strenuously its advocates protest otherwise, it is not science. There are two primary reasons for this, both of which are clearly exemplified in Darwin’s Black Box.

Consider biological adaptations such as the transport system which are discussed in Behe’s book. Behe does not ask if there are any species that don’t have this adaptation but manage without it, if there are any species that have more primitive versions of it, if there are any species that have versions of it that function with fewer parts, or if there are any species that have systems that perform different functions but are similar to it. He does not ask if its proteins are homologous to proteins that do different tasks in this species or other species, or if the genes that code for those proteins have any sequence similarity to any other genes that have been sequenced. Evolutionary scientists seeking to understand the origins of a system do ask these questions, as well as many others, but Behe does not. He seems content simply to declare it complex and then abandon all further efforts to investigate its origins. The difference is that where a scientist looks at a particular adaptation and sees a challenge, an interesting puzzle that can be solved if we ask the right questions, Behe looks at the same adaptation and immediately declares all such efforts hopeless, encouraging others to give up as well.

Of course, this defeatism is tactical, not genuine. The reason for it is because ID, as shown in the earlier section “The Argument from Ignorance”, is fundamentally a God of the Gaps explanation – it sees God’s activity only in things we do not understand. Behe admits as much in this quote:

“The most relevant laws are those of biological reproduction, mutation, and natural selection. If a biological structure can be explained in terms of those natural laws, then we cannot conclude that it was designed” (p.203).

Therefore, the more natural phenomena Behe can declare off-limits to scientific investigation, regardless of whether science can actually explain them or not, the more “breathing room” he can create for his god. Conversely, as far as he and other ID advocates are concerned, to be able to explain some phenomenon in natural terms is equivalent to saying that God was not involved in creating it.

This tactic illuminates the fundamentally anti-scientific nature of ID. Once an adaptation is declared irreducibly complex, what then? Are scientists supposed to just give up on it and study something else instead? But evolutionary explanations have been published for several such systems both before and since the publication of Behe’s book. Had we done as he suggested, these advances in our knowledge would never have come about.

Good science is intellectually fertile. A real scientific explanation genuinely adds to our knowledge, often explains other apparently unrelated phenomena, and always suggests new directions for our further research. ID does none of these things. It is intellectually sterile, scientifically lifeless – an empty labcoat. It “explains” natural phenomena with vague platitudes rather than insight and understanding, and rather than opening up new avenues of research, tries to close them all down.

The second major reason why ID is not science has to do with its persistent invocation of the supernatural – of miracles – as its own solution to every problem. Behe himself says as much, although in the coy way typical of ID advocates who are not speaking to Christian church groups:

“The dilemma [for the scientific community] is that while one side of the elephant is labeled intelligent design, the other side might be labeled God” (p.233).

Although he is well aware that science as traditionally defined cannot appeal to the supernatural (for reasons discussed in the article Naturalism In Science), Behe argues that this rule is not a fundamental component of science, criticizing comments by biochemist Richard Dickerson, who writes that the principle of not invoking supernatural explanations is the “one overriding and defining rule” of the “game” that is science:

“Certainly the taxpayers who fund science to the tune of several tens of billions of dollars a year would be surprised. They probably think they’re spending their money to find cures and treatments for cancer, AIDS, and heart disease. Citizens concerned about diseases they have or may acquire in old age want science to be able to cure the disease, not to play a game that has no bearing on reality.” (p.240)

I thank Behe for his insightful choice of words – I could not have said it better myself. Despite his equivocation over the meaning of Dickerson’s statement – Dickerson was obviously referring to science as a “game” in the sense of a competition with fixed rules, not in the sense of a frivolous activity that has no impact on the larger world – Behe has hit the nail right on the head. The fundamental reason for believing in science is not because of prior philosophical considerations or biases, but because science works. Explaining natural phenomena only in terms of natural laws, without allowing miracles to preserve a particular hypothesis from criticism or disproof, is the engine that has driven the most rapid and astounding progress humanity has ever known, all within the space of the last few centuries. And if we want to continue making this progress and enjoying its fruits, then we should continue to do what works – studying, research, investigation – rather than playing a pointless game that consists of labeling natural objects “designed” and then giving up. A person certainly could explain some phenomenon by appealing to the miraculous intervention of some supernatural being, but then the activity they were engaging in would no longer be science.

“…scientists should follow the physical evidence wherever it leads, with no artificial restrictions” (p.243).

It is true that scientists should follow the evidence wherever it leads, without preconception. But the exclusion of the supernatural as a scientific explanation is not an “artificial restriction” – it is a fundamental and necessary principle if one is to make sense of the data. The business of science consists of examining observed data to extract patterns and derive general rules – rules that can then be used to make predictions the next time something similar occurs. But the supernatural, by definition, follows no rules (if it did, it would no longer be supernatural, but natural). A phenomenon is supernatural precisely because it violates the rules of nature. Whether and when God will next intervene cannot be predicted, nor can we predict the form such intervention will take – even according to theists.

This is why science is restricted to natural explanations. This does not mean that science itself denies the existence of the supernatural; it is merely that, if the supernatural does exist, we cannot learn about it through science. There clearly are things that, were we to observe them, we would have to conclude that the supernatural was at work (see “The Theist’s Guide to Converting Atheists“); but Behe has not identified any such thing, nor has any other creationist.

If the supernatural were allowed into science, science as such would cease to exist, because it would lose all its explanatory power. All existing phenomena that were currently not understood could be explained in supernatural terms, and the reason to investigate further would disappear. This would have been the case if supernatural explanations had been accepted in Galileo’s day, when certain individuals argued that God framed the cosmos and did not intend for us to understand it, therefore it would be impious for us to try to find out whether the Sun orbited the Earth or vice versa (see the review of Paradise Lost). This would have been the case if supernatural explanations had been accepted in medieval times, when epidemic diseases were thought to be God’s punishment of sinners rather than the result of poor hygiene. And it would be the case if supernatural explanations were allowed in science today.

Behe does not agree with this perspective:

“The fossil record shows that about 60 million years ago, the dinosaurs all died out within a geologically brief time period. One theory offered to explain this is that a large meteor crashed into the earth, sending clouds of dust high into the atmosphere and perhaps causing many plants to die, disrupting the food chain…. The hypothesis has been accepted by many scientists. Nonetheless, there has not been a rush to postulate meteors as the cause of all sorts of things. No one has said that meteors caused the Grand Canyon, or the extinction of horses in North America” (p.242).

The reason for this is precisely that natural causes follow laws which we can use to deduce that they were at work in one case but not another. The supernatural follows no such laws, and therefore once it is allowed in the door at all, there is no reason to restrict it to only one area. Indeed, there are theists who do believe that God was directly and miraculously responsible for every single thing that exists, and would prefer to see this taught in schools – they are called young-earth creationists. Behe should know this.

Behe’s final point in this section is a claim that only preconceived bias has prevented scientists from seriously considering the possibility of design:

“The…. most powerful reason for science’s reluctance to embrace a theory of intelligent design is also based on philosophical considerations. Many people, including many important and well-respected scientists, just don’t want there to be anything beyond nature” (p.243).

This is a somewhat strange argument for Behe to make, considering his claim several pages earlier that “scientists who believe in God… are much more common than popular media stories lead one to believe” (p.243) – for example, Kenneth Miller, the staunch opponent of ID mentioned earlier, is an evolutionary scientist and also a devout Christian. Even assuming that some scientists are atheists and would reject any hypothesis of intelligent design on metaphysical grounds, why would the theistic scientists not see its merit and point out the flaws in their arguments in peer-reviewed publications?

And what about the advocates of intelligent design? Are we to believe that they are pure of theological motives? That seems unlikely, considering the originators and defenders of that hypothesis are, without exception, evangelical Christians. By contrast, as Behe readily admits, scientists of all religious backgrounds and leanings can and do support evolution. Those who seek to remove the motes from others’ eyes would do well to first consider the beam in their own.

What seems clear is that advocates of ID want the respectability of science without having to do the hard work necessary to earn it. The theory of plate tectonics, for example, was not widely accepted when it was first proposed, but its authors did not immediately drop their research and create a public relations firm to convince the public that their ideas should be taught in schools. Instead, they continued to collect evidence and publish articles in peer-reviewed journals. When the evidence became strong enough, the scientific consensus tipped in their favor and their theory gained broad acceptance among the geological community. Then, and only then, did it make its way down into textbooks and schools. That is how the process of science works. The behavior of advocates of ID, by comparison, is revealing; they clearly wish to skip the entire tedious process of research, testing and criticism – since they have made no effort to participate in it. Instead, they promote their ideas by appealing directly to the religious public, and claim that institutional prejudice and atheism, rather than their own total lack of participation in the process and failure to support their claims with evidence, is what prevents their ideas from enjoying wide acceptance among scientists.

Behe, apparently without even noticing it, supports this argument and undercuts his own when he discusses a theory from another field of science:

“To many the notion of the Big Bang was loaded with overtones of a supernatural event…. Nonetheless, despite its religious implications, the Big Bang was a scientific theory that flowed naturally from observational data, not from holy writings or transcendental visions. Most physicists adopted the Big Bang theory and set their research programs accordingly” (p.244).

If this is the case, then why are biologists resistant to a different idea that also has “supernatural overtones”? The only plausible explanation is that qualified scientists reject intelligent design not because of the religious motivations of its sponsors, but because, unlike the Big Bang, it does not flow naturally from observational data. Evolution does.


To conclude this review, I will point out several apparent contradictions in Behe’s position. The first concerns how an observer can identify design.

“As the number or quality of the parts of an interacting system increase, our judgment of design increases also and can reach certitude” (p.199).

“Our ability to be confident of the design of the cilium or intracellular transport rests on the same principles as our ability to be confident of the design of anything: the ordering of separate components to achieve an identifiable function that depends sharply on the components” (p.204).

And yet:

“The most relevant laws are those of biological reproduction, mutation, and natural selection. If a biological structure can be explained in terms of those natural laws, then we cannot conclude that it was designed” (p.203).

These two sets of statements cannot both be true. Either a system that has an ordered arrangement of parts that work together to achieve a specific function, but whose origin can be explained as the operation of mutation and natural selection, should be judged the product of design or not. Which is it?

Another inconsistency comes in Behe’s view toward common descent. As detailed in the first section of this review, he states explicitly that he accepts its occurrence (“Like the sequence analysts, I believe the evidence strongly supports common descent” (p.176)). However, he also says this:

“But it is at the level of macroevolution – of large jumps – that the theory [of evolution] evokes skepticism. Many people have followed Darwin in proposing that huge changes can be broken down into plausible, small steps over great periods of time. Persuasive evidence to support that position, however, has not been forthcoming” (p.15).

Again, which is it? Does the evidence strongly support common descent or does it not?

Finally, there is one more comment that cannot be allowed to pass. Behe says this:

“Instead, a curious, embarrassed silence surrounds the stark complexity of the cell. When the subject comes up in public, feet start to shuffle, and breathing gets a bit labored. In private people are a bit more relaxed; many explicitly admit the obvious but then stare at the ground, shake their heads, and let it go at that” (p.233).

This is not mere dramatic license. This passage sounds very much as though Behe is saying he interviewed evolutionary scientists who themselves believe and admit that evolution is insufficient to explain the complexity of the cell. If that is so, who did he interview to obtain these reactions, and what exactly did they say? If, on the other hand, no such interviews took place, then Behe owes his readers an apology for misleading them with such a deceiving passage.

In summary, Behe’s argument, like that of many creationists, is fundamentally based on a lack of imagination. He sees an adaptation, marvels at its complexity, and then immediately concludes it is hopeless to try to understand how such a thing came to be. Of course, there is nothing wrong with appreciating the elegant majesty of nature – people of all beliefs and backgrounds can agree on this. Whether it is the dazzling color of a cardinal’s wing, the living green solar collectors of a tree, or the byzantine architecture of the molecules within a single cell, nature continues to astound us with its beauty and subtlety. Where he crosses the line is in asserting that these exquisite adaptations are so complicated as to defy natural explanations. That is not the case; nature is not bound by the limits of what any one person can readily imagine. The true wonder is not just that such things exist, but that their origins can all be traced back to the same process, an intricate interplay of variation and selection ultimately responsible for all the exuberant web of life’s diversity. That we have identified this process is one of the true triumphs of human ingenuity and reason, and when compared to the vast array of evidence supporting its occurrence, the creationists’ arguments are seen for the feeble and shallow sophistries they are. They give nature far less credit than it deserves.

“It seems to be characteristic of the human mind,” Behe writes, “that when it sees a black box in action, it imagines that the contents of the box are simple” (p.23). This characterization applies with a vengeance to the supporters of intelligent design. They see a black box – the origin of complex adaptations – and imagine that its contents are simple – God did it, end of story. But reality is rarely so simple, and the true answer, while not as easily grasped as ID defeatism, is far more compelling when truly understood. Darwin’s black box has indeed been opened, but the contents do not support the imaginings of the latter-day creationists. The sooner they accept this and cease their impassioned but ultimately futile crusade against science, the better we will all be for it.