Michael Behe and the Edge of Evolution
I read Michael Behe’s Darwin’s Black Box some years ago, and while I have tried to keep up with the popular literature on Evolution, I was not aware until sometime last year (2010) that he had written a second book in 2007, The Edge of Evolution. After reading it, I thought it might be interesting to look for critical reviews on the web. This note is my reaction to Behe’s book and to reviews by a couple of well-known authors of popular books on evolution that surprisingly, to me at least, offered poor or misdirected rebuttals of Behe’s major points while missing the weakest part of his argument.
Before considering the reviews, we ought to look at Behe’s central arguments. He accepts a basic tenet of Darwinian evolution, the concept of descent with modification or common descent—the succession of species from earlier species. Common descent presupposes long spans of time during which organisms were modified and Behe obviously accepts the geological evidence that the earth is 4.5 billion years old and that life has been present for at least 3.5 and perhaps 3.7 billion years. What we have learned of genetics—the similarities and differences in the DNA of organisms—in the past 60 years, Behe argues, makes common descent undeniable. For example, Behe would agree that land animals descended from fish and humans descended from earlier apes. He also accepts the notion that some of the simpler variations in phenotypes, observable traits of organisms, arise from chemical changes or mutations in DNA. He describes in great detail certain mutations in blood cells such as sickle cell anemia and thalassemia that enhance survival of individuals infected with malarial parasites. What he denies is that complex organic structures could arise through mutations. This was the primary point of his earlier book. He asserts that most biological complexity is “irreducible,” that is, it cannot have arisen piecemeal, as would be required by natural selection through random mutations, but must have arisen as a complex working structure.
The Irreducible Complexity of the Bacterial Flagellum. In his first book, Behe’s example of irreducible complexity was the bacterial flagellum, a little tail that whips around driving the organism through its fluid environment. This example soon became iconic and was reproduced in book after book arguing for and against Darwinian evolution. The complex molecular motor that drives the flagellum is composed many separate parts. The motor is obviously adaptive, increasing the survivability of the bacterium. Bacteria with swimming mechanisms can move more effectively to food or escape from predators. However, Behe argued that the individual parts that compose the motor are not themselves adaptive. A quarter or a half of a molecular motor is useless as a propulsive device and thus would be no more likely to be passed on to the bacterium’s descendants than any other random and nonfunctional mutation. If random mutation is the primary mechanism underlying natural selection, each part would have appeared as a separate mutation. To compose a molecular motor the mutations that produced each of its components would have to have occurred simultaneously and arranged as a working motor. The probability of this is infinitesimally small. Here is an abstract example. Suppose we observe three events whose individual probabilities of occurring are two chances out of a hundred (.02), three out of a hundred (.03) and one of a hundred (.01). The probability of all occurring simultaneously is the product of the three: .02 x .03 x .01 = .000006, which is 6 chances out of one million.
Behe’s Argument is Logical. Note that this argument is not empirical. We are not in a position to observe earlier forms of bacteria, like for example the skeletal structure of the precursors of whales or horses. There are no bacterial fossils. Rather his argument is logical. The logic is correct so long as we accept Behe’s assumption of simultaneity. The weakness of the argument is not the assertion of low simultaneous probability but the assumption that all the parts of the molecular motor had to appear at once. The question is whether simpler precursors of the flagellum could have evolved that could serve as parts that would eventually come to compose the developed motor. Because the argument is logical, all that is needed is a logical and plausible counter argument. Ian Musgrave in his “Evolution of the Bacterial Flagellum” in the book Why Intelligent Design Fails does just that. He shows how simpler secretory systems in single celled organisms could have evolved into swimming systems. He argues that the flagellum could have evolved in several stages of growing complexity from simpler secretory beginnings.
Does this mean that Behe’s arguments are without merit? Should we just dismiss them as religiously driven conclusions? After all, Behe is associated with Discovery Institute where promoters of intelligent design are at work. Yet, Behe is not the only thinker that has suggested the need for other mechanisms than random mutations to explain the emergence of new species. He considers, and rejects, some of these in the 8th chapter of his book. For instance, Stuart Kauffman has promoted self-organization as a mechanism that might underlie the evolution of complex structures. The interesting thing about this is that a preeminent and widely admired thinker, who is not formally associated with a religious organization, is effectively saying that more is needed to explain the evolution of biological complexity. Behe briefly considers the newer work called evo-devo in which, for example, the same or similar genes are shown to be at work controlling analogous but dissimilar structures in widely dissimilar animals. The same genes have been found at work in insects and mammals, for example. With a little tweaking the same genes can produce widely varying structures in the developing organism. This means that a much smaller number of mutations than had previously been supposed may be behind the incredible variety of organisms the have lived and disappeared in the past and are alive today. Behe rejects this as a full explanation arguing that it simply adds another layer of complexity to the causes of common descent. The important thing, however, is that here are other biologists arguing that earlier Darwinian and Neo-Darwinian approaches to evolution are inadequate. Behe cites others who support this contention. A book published after Behe’s also falls into this category, What Darwin Got Wrong, a book by a philosopher and a scientist who are at pains to make clear in preface that they are both atheists. The difference between Behe and others who question the mechanisms underlying natural selection is that they assume that answer will be found through scientific investigation, whereas Behe appears to have concluded on logical and probabilistic grounds that no such mechanisms are possible. We will see that in fact this is not Behe’s position.
The Response of Richard Dawkins. Now let us briefly consider the two reviews. The first is Richard Dawkins’ 2007 review in the New York Times. Passing over the invective and personal insults that permeate the review, Dawkins’ asserts that random mutations do not limit evolutionary change. He illustrates this with the appearance of a wide variety forms among domesticated dogs, all of which evolved from wolves in a few thousand years. Given the time-frames involved in natural evolution, this is like the blink of an eye. One weakness of this example is that dogs are not a new species. Dog breeds commonly interbreed with one another and sometimes with wolves, something that is impossible for distinct species. Surely, Dawkins could have come up with a better example. More to the point, Behe’s primary arguments are about mutations and changes at the molecular level. A convincing counterexample at this level is called for. Dawkins’ other primary argument is that mathematical geneticists—Fisher, Wright, Haldane, etc.–have shown that evolutionary rates are not limited by mutations, although he is content to cite authorities rather than explaining their analyses, something he could have done if he had not wasted the first third of the review on personal attacks.
The Response of Jerry Coyne. The second review is by Jerry Coyne, an evolutionary scientist and now a popular writer who has published an engaging book for the lay reader on evolutionary theory, Why Evolution is True. The review is long and I do not propose to summarize it. The heart of his argument, in my opinion, is found three-quarters of the way through the review where Coyne tries to counter the complexity argument. He gives the example of a throw of twenty dice in which a six would represent success.
Suppose a complex adaptation involves twenty parts, represented by twenty dice, each one showing a six. The adaptation is fueled by random mutation, represented by throwing the dice. Behe’s way of getting this adaptation requires you to roll all twenty dice simultaneously, waiting until they all come up six (that is, all successful mutations must happen together). The probability of getting this outcome is very low… But now let us build the adaptation step by step, as evolutionary theory dictates. You start by rolling the first die, and keep rolling it until a six comes up. When it does, you keep that die (a successful first step in the adaptation) and move on to the next one. You toss the second die until it comes up six (the second step), and so on until all twenty dice show a six… This sequential way of getting twenty sixes is infinitely faster than Behe’s method. And this is the way natural selection and mutation really work, not by the ludicrous scenario presented by Behe.
The obvious question in this example is why any one of the twenty mutations would survive to the next generation. The first would have to combine with the second to produce some adaptation that was more likely to survive. Then the third would combine with these two to produce a new adaptation with higher survival value, and so on. In Coyne’s example, the combinations would have to increase the likelihood of that the organism survival at each step of the process. Each step would occur in a separate organism that would reproduce and pass the genetic change on to its progeny. Coyne’s example is possible but given the requirement of neo-Darwinian evolutionary theory—that changes be adaptive to survive—it is necessary to explain why each the twenty successive combinations might be retained and passed on.
There is one way that the adaptation requirement might be set aside for some of the twenty stages in Coyne’s example. Stephen J Gould (in an essay that I cannot now recall) developed the idea of exaptation, which are changes in structure that are not necessarily adaptive yet are retained and reproduced in each generation of organisms. The changed structure may be utilized in a later generation for some purpose that makes survival of the organism more likely. An example I recall (but cannot remember the source… Gould?) is cranial sutures in newborn human babies, which consist of flexible plates that only join later in infancy. The gaps in the skull bone permit brain growth and expansion of the skull in infancy but it was suggested that they also permit the infant’s head to pass more easily through the narrow birth canal. It might be supposed that cranial sutures evolved as an adaption in humans to permit the birth of big-headed infants as well as later brain growth. However, it was known that cranial sutures existed in our earlier mammalian ancestors that were not big-headed. They may represent an evolved structure that was not adaptive but was passed on. Alternatively, they may have been retained for some unknown purpose only to be used later for a different purpose. Genetic changes producing non-adaptive structures that somehow get passed along to subsequent generations of organism might be an explanation for the survival of some of the intermediate structures in Coyne’s hypothetical example.
I alluded above to Musgrave’s article in which such an explanation of provided for the evolution of the bacterial flagellum. Coyne refers to the evolution of the whale which involves successive adaptions from a land creature millions of years ago through successive stages that are documented in the fossil record to the modern creature that lives its life entirely in the ocean. The empirical evidence for this progression is strong. Yet, the adaptations in the gross structure of the whale each must have involved complex series of intermediate changes of the twenty-dice-throw type that are not apparent in the fossils. The migration of the nose in the earlier precursor mammals up along the head to become a blowhole in whale, for instance, seems to me entirely possible and to fall into the category of gradual successive changes. On the other hand, I always wondered how the side-to-side motion of the tail of the earliest fossils in the series changed to an up-and-down movement of the tail characteristic of whales. To me it is not apparent how this could take place through a series of gradual changes because of the massive changes in bone, muscle and nervous system that would be required. Perhaps, someone has provided reasonable set of transitions.
Coyne could have used a real life example. Johnjoe McFadden presents just such a case in his book Quantum Evolution, A New Science of Life (page 267). AMP (adenosine monophosphate) is a precursor to ATP (adenosine triphosphate), which is a basic energy storage molecule made by living cells. McFadden writes:
…The synthesis of AMP from its precursor, ribose-5-phosphate, involves thirteen independent steps involving thirteen different enzymes (which we will represent as: A->B->C->D->E->F->G->H->I->J->K->L->M, where A is ribose-5-phosphate and M is AMP). Each enzyme involved in the pathway is absolutely essential for the biosynthesis of AMP. Darwinian evolution requires this complex system to have evolved from something simpler but, as far as we know, nothing simpler works.
We can see the similarity to Coyne’s dice throwing example. Maybe the reason he did not use this or one of many other examples available is that he also would have been forced to say something like the final sentence in the McFadden quote. McFadden also immediately goes on to disavow agreement with Behe and in the process presents yet another possible explanation of the evolution of complexity. He proposes to explain the evolution of complex structures through Quantum Theory, which is very interesting but not a topic that can be taken up here.
Deistic and Tinkering Intelligent Design. Both Coyne and Dawkins missed what I think is the weakest point in Behe’s argument. As we have seen, Behe admits common descent or descent with modifications. He accepts the notion that adaptation could occur through mutations at levels that require only one or two mutations, such as a single change in the chemical structure of blood cells in response to the malarial parasite. But he does not believe that more complex adaptions could have occurred through random mutations for the reasons just described. He rejects some of the alternate explanations to random mutations that other life scientists have proposed. He instead proposes an argument from design (see pages 231-2).
Those who opt for intelligent design must make one of two further choices. The first of these we might call tinkering intelligent design in which the designer constantly intervenes and directs each genetic change underlying the occurrence of each and every new complex biological structure in the history of the world. This position is hard to hold for anyone who represents himself to be a scientist. It is the biological equivalent of divine control of weather, and by that I do not mean belief in the occasional miraculous drought-ending rainstorm. Rather, I mean something akin to ancient beliefs in wind and rain gods. God, in this scenario, is at work making countless trillions of modifications in organisms over billions of years. Some lead to higher and more complex creatures and finally intelligent beings. Others end in extinction. Why would a designer go to all this trouble? Why not simply create humans immediately with no intervening steps. Maybe this is one of the reasons why the fundamentalists and literalists—Christian, Muslim and Jewish—advocate instant creation.
Behe asserts that in his opinion constant intervention by a designer is not required. Instead he advocates a deistic form of intelligent design. He says that “the universe is purposely set up to produce intelligent life.” It is set in motion from the beginning in a single creative act, and then unfolds “exclusively by the intended playing out of natural laws.” There is no need for random variation and natural selection for universal common descent to occur. Instead, Behe seems to be proposing a predetermined sequence of changes that culminate in intelligent life, a kind of deistic determinism.
This explanation is not anti-scientific, although some may want to quibble about how and by whom the “universe was purposely set up,” arguing that these are not scientific questions. Instead, Behe’s explanation has some clear scientific implications. Here are some of them. It would seem to imply that many (most?) of the processes in the nature that appear to be random are only apparently so and are in fact determined. For example, it might imply the survival of our organic precursors during various extinction events was not random. The survival of our organic precursors during the extinction that occurred after the explosion of life forms during the Cambrian period was not simply random but somehow directed. And by that I do not mean immediately directed by an intelligent designer but that “natural laws,” as Behe suggests, insured the survival of certain types of living creatures. In this light, we would have to reject the claim of Stephen J Gould that if history could be rewound the Cambrian extinction would not necessarily result in the same winners and losers. It may be that Gould’s claim can be rejected on other grounds—see Simon Conway Morris—but Behe’s approach begs the questions what exactly were the natural laws that led to the survival of our ancestors during various extinction events.
In addition, and more to the point, Behe appears to be claiming that the appearance of complex biochemical sequences represents the playing out of natural laws. Again, if such laws are present they ought to be discoverable and demonstrable? The scientist will seek to determine what they are and how they operate. He or she would not be satisfied to leave them vague and undefined. Rather, they would specify the laws and demonstrate their operation. Surely it would be possible to show how the playing out of natural laws occurs for one or two complex changes, like the evolution of the synthesizing sequence of ribose-5-phosphate into AMP. McFadden, rightly or wrongly, makes such an attempt. Behe does not.
Furthermore, does not this approach imply that the origin of life on earth from inorganic matter 3.7 billion (or so) years ago represents the playing out of natural laws? And if such laws existed then, is it not likely that they continue to exist today. And if they are present today, should not new life be springing up from nonliving organic material? Why have all the experiments in producing living organisms resulted in organic sludge rather than life? Did the natural laws that led to life somehow cease to operate 3.7 billion years ago?
I could go on, but I think the point is clear. Behe rejects random mutations and natural selection, perhaps with some justification. But by rejecting the notion of a constantly intervening designer in the common descent of life and positing natural laws that were intended by a designer to produce life and then intelligent life, Behe has put himself in the position of demonstrating that such laws exist. In the end, Behe is back in the camp of the many thinkers and writers that have proposed alternatives or additions to the theory of natural selection.