From Skeptic vol. 8, no. 4, 2001, pp. 54ff.

The following article is copyright ©2001 by the Skeptics Society, P.O. Box 338, Altadena, CA 91001, (626) 794-3119. Permission has been granted for noncommercial electronic circulation of this article in its entirety, including this notice.

Impossible Evolution?
Another Physicist Challenges Darwin

Review of Mathematics and Evolution, by Fred Hoyle, Memphis, TN: Acorn, 1999

Massimo Pigliucci

Is evolution by natural selection possible? Most biologists would consider this a rather anachronistic question, satisfactorily settled during the first few decades after the publication of Charles Darwin's Origins of Species in 1859. Not so, of course, for the hordes of creationists that infest the American pseudo-intellectual landscape and stubbornly try to legislate scientific ignorance in our public schools. Sir Fred Hoyle, a British cosmologist, also does not believe that evolution can occur. However, he is no friend of creationists, despite being often quoted by them, especially by Duane T. Gish of the Institute for Creation Research.

Hoyle has had a distinguished and rather peculiar scientific career. He is perhaps most famous in academic circles for proposing the theory of steady-state universe, thus triggering a large research effort that culminated with the rejection of his theory and the building up of a convincing amount of theoretical and empirical evidence in favor of its rival conception, the Big Bang. Later on, Hoyle went on record with his inseparable colleague, Chandra Wickramasinghe, to propose an even more esoteric hypotheses for an extraterrestrial origin of life on Earth, regarded by professional biochemists and biologists as Pindaric flights of fancy rather than serious scientific theories.

However, Hoyle is not a naive creationist with no scientific background, so if he claims that he has a mathematical demonstration of the fact that evolution cannot occur, an honest skeptic has to give him a hear, read his book, and figure out if and where Hoyle is wrong. That is exactly what I did, and my conclusion is that¨like other recent attempts at mathematically "proving" the errors of Darwin, the claim has been made a bit too hastily and without the necessary biological foundations. The major difference between Hoyle and, for example, William Dembski (author of The Design Inference), however, is that while the latter has a clear religious agenda which guides (and blinds) his efforts, Hoyle is in the business of doing real science¨albeit reaching once again the wrong conclusions.

In Mathematics and Evolution Hoyle seeks to put evolutionary theory on solid mathematical foundations or to disprove it in the attempt. This approach is laudable, given that mathematics has been the official language of the "hard" sciences since Galileo and Newton. However, there are two points that Hoyle neglects right at the start of his book: first, some objects of study, such as biological organisms or geophysical cycles, are much too complex for the mathematics that we have devised thus far. That is why biology, geology, and to an even larger extent, psychology, have always been less amenable than physics and chemistry to rigorous mathematical treatment. Instead, the so-called "soft" sciences are the realm of statistical analyses and probabilistic statements. While this state of affairs may be a regrettable one from the point of view of a mathematical physicist such as Hoyle, it is a simple and undeniable reality. Second, and perhaps most surprisingly, Hoyle seems to be only vaguely aware that a limited, yet vast and sophisticated¨mathematical theory of evolution already exists. Its foundations were laid down by people such as Sir Ronald Fisher, Sewall Wright, J.B.S. Haldane, Mooto Kimura and countless others who have published on the topic since the 1920s. While Hoyle does mention a few of these scientists here and there (the book comes with no references at all!), he dismisses them as a bunch of incompetents and delusional dilettantes, without ever giving the reader a detailed justification of such a harsh evaluation of these highly regarded scientists.

The core of Hoyle's argument is that natural selection is sufficient to explain what is usually termed ˘microevolution,÷ that is, variation within species and small, gradual changes of species over time. However, according to the British astronomer, Darwinians made the mistake of rashly extrapolating their findings to macroevolution¨the origin of new body plans and of major transitions in the history of life, something that is alleged in the book to be pure nonsense. In this, Hoyle comes close to the position that most creationists have entrenched themselves in, although again both the motivations and the arguments of the latter are dramatically different and infinitely less sophisticated. Instead, Hoyle suggests that the real explanation for major evolutionary changes is to be found in what he calls "genetic storms," hypothetical periodic events during which the Earth is suddenly flooded with genetic material from outer space carried by meteors.

Let's examine the two components of Hoyle's argument¨the inadequacy of natural selection and the necessity for imported genetic material, to see where he makes a good point and where he completely misses the mark.

On page 20 of Mathematics of Evolution, Hoyle claims that natural selection can at best keep a species from sliding down into the abyss of extinction, but cannot play any significant role in improving that species' adaptation to the environment, let alone create novel features for an organism: "The best that can be done is to hold the position, which is basically what bacteria have done for almost 4000 million years." This is, of course, the old conceptual misunderstanding of selection as a force that can weed out but cannot build, which countless authors ever since Darwin have attempted to dispel. Biologists, contrary to Hoyle, think that bacteria have remained bacteria simply because they are very well adapted to their environment, and their particular niche in that environment has not changed for billions of years, thereby not necessitating any further improvement. As my professor of biophysics, the late Mario Ageno, once put it, "the ideal of a bacterium is not to become a man, but to become two bacteria."

Of course, Hoyle's statement does not just come out of thin air, but is based on some good mathematical reasoning (at least, as far as I could tell). However, like all mathematics when applied to real life problems, it is as good as the assumptions on which it is based¨and these assumptions reveal little knowledge of biology. The underlying premise of Hoyle"s "demonstration" of the ineffectiveness of natural selection is that there are more mutations with negative or even lethal effects than there are mutations with positive effects. This, coupled with another of Hoyle's fundamental assumptions, that living organisms are so complex and finely tuned that any change in their machinery is overwhelmingly more likely to cause damage than benefit, leads to his conclusion that evolution by natural selection cannot possibly work. There are two problems with this reasoning. First, biologists have known now for decades that most mutations are neither positive nor negative, but neutral or quasi-neutral. This leaves much more room for natural selection to maneuver than in the tight scenario adopted by Hoyle. The reason for such a surprising amount of neutrality of mutational effects is related to the second problem embedded in Hoyle's argument: organisms are not designed according to stringent engineering principles where every part has to work exactly in a particular manner and interact with precision with all other parts. Rather, living beings are put together in a rather loose way, with a lot of redundancy and suboptimal design, exactly as one would expect if they were the result of a natural process instead of an intelligent designer. Modern molecular developmental biology, entirely ignored in Hoyle's book, has clearly demonstrated that genetic redundancy and sub-optimality are a universal characteristics of life on Earth.

Curiously, the bulk of Hoyle's book is actually about the advantages of sexual reproduction over the asexual alternative in terms of genetic variation, which partly explains why sexual organisms have become so much more complex than bacteria. While Hoyle makes a big deal of this to show that bacterial evolution is much more limited than its counterpart in more complex organisms where sex is the norm, his arguments would not raise any biologist's objection, and are hardly a challenge to the theory of evolution. It is certainly true that the ability to recombine existing genetic material has opened a whole new venue to evolution by natural selection compared to what is possible when all the genetic variation in a population must come from mutations alone. But that is not enough, according to Hoyle. He thinks that to explain the major transitions in evolution one needs something more, an additional source of genetic variation not thus far considered by biologists. And he thinks he has the answer. This "external incidence" is described on page 108: "External incidence appears to come in storms of rather short duration, the most recent very large storm being the one that occurred 65 million years ago. Species seem to vary considerably in their sensitivity to genetic storms."

In other words, salvation comes from outer space. Somehow, genetic material compatible with life on Earth is imported aboard meteorites that periodically destroy life forms such as the dinosaurs and simultaneously infuse new life into the terrestrial biosphere; if such a "theory" sounds rather like science fiction, that is because Hoyle is also a sci-fi writer, though in this case he is very serious.

Two thoughts come to mind in response to Hoyle's idea of genetic storms: first, the idea is almost certainly wrong; second, its roots are to be found in a genuine problem with which evolutionists have been struggling for most of the 20th century. Let me start by acknowledging the problem and discussing how it is in the process of being solved by means completely different from Hoyle's fanciful proposal.

One of the classical assumptions of the so-called neo-Darwinian theory of evolution that emerged during the 1930s and 1940s from a synthesis of Darwin's work and the new field of genetics is that evolution occurs very slowly over long periods of time. Furthermore, evolutionary change has commonly been thought of as the result of alterations in the frequencies of many different genes, each with a very small effect on the phenotype, i.e. on the way organisms look and behave. It was already clear at the time, however, that there must be more than that to evolution by natural selection. For example, noted geneticist Richard Goldschmidt published a severe critique of the neo-Darwinian model, concluding that something else was necessary to account for macroevolutionary changes. He proposed that a "genetic revolution" would have to occur from time to time to produce entirely new types of organisms, which he called "hopeful monsters." Does that sound familiar? Except for the fact that Goldschmidt's genetic revolutions were to be caused by major internal reshuffling of an organism's chromosomes, the idea is identical to Hoyle's genetic storms.

The evolutionary community rejected the hopeful monsters as unlikely on two grounds: (1) there was no known mechanism by which the postulated revolutions were actually possible. This was eventually confirmed by the explosion of the field of molecular biology and the elucidation of how genetic material actually changes over time. (2) Even if a genetic revolution could occur, the result would be disastrous because a major alteration of the molecular machinery of an organism would be fatal, again something that has been borne out through countless instances of experimental evidence. (Notice that Hoyle wants it both ways: he claims that even minor changes in the genes are bound to be fatal or highly detrimental, but then turns around and invokes "storms" of extreme genetic alterations.)

Yet the neo-Darwinians did not enjoy a lasting victory, being challenged again first in the early 1970s by paleontologists Niles Eldredge and Stephen Jay Gould, and then by a slew of developmental geneticists and evolutionary ecologists. The currently emerging consensus is that reality lies somewhere in the middle between the strict neo-Darwinian paradigm and Goldschmidt's monsters: while most every day evolution does occur because of incremental changes in the frequencies of genes with small effects, some major changes do occur from time to time and cause what we have referred to as macroevolution. This second category of change, however, is not triggered by any new genetic mechanism, nor does it cause the dramatic effects postulated by Goldschmidt (or, 60 years later, by Hoyle). What happens instead is that standard mutations occur in regions of the genome (the ensemble of all genes) that regulate the functioning of a cascade of other genes. In this way, one small change has multiple, far-reaching, yet coordinated effects because of the way genes interact with each other throughout the development of an organism, a sort of domino effect with consequences that may or may not be negative. This mixed model of evolution is still very much in the process of being shaped and probed experimentally. It will probably take decades to fine tune, but it is based on a very solid foundation of both theoretical and empirical evidence.

Another genetic mechanism recently shown to be more common than previously thought is lateral (or "horizontal") gene transfer, that is the shuttling of some genes from one species to another. This phenomenon has been known in bacteria for quite some time, and it has now been demonstrated to occur even between multicellular organisms. Furthermore, there is good, albeit preliminary, evidence that some such gene transfers have in fact played a role similar to Hoyle's genetic storms, on a much smaller scale. Ironically, Hoyle has put the finger on the right problem, but his complete ignorance of the biological literature has led him to propose outlandish theories, good for a mediocre episode of Star Trek. The point is that both the evolution of gene regulation and horizontal gene transfer still represent the mechanistic basis of good old evolution by natural selection, which Hoyle allegedly "demonstrated" to be impossible.

Hoyle's genetic storms, on the other hand, are truly impossible¨or at least unlikely beyond reasonable doubt. This is for a variety of reasons: first, contrary to what Hoyle suggests (without empirical evidence) it is very difficult to imagine how the genetic material embedded in meteorites would reach the Earth in a sufficiently unaltered state as to be useful to terrestrial organisms. Cosmic radiation, and possibly the temperatures generated by the entry of meteorites in our atmosphere, would severely damage the delicate structure of DNA that we now know so well from biochemistry. It is true that meteorites have been found to harbor organic material such as amino acids, but this is in no condition to be used by a metabolically active organism (and, incidentally, has been demonstrated to be of an entirely different chemical origin). Second, it is astronomically unlikely that genetic material wandering randomly through the universe would repeatedly happen to be compatible with its earthly equivalent which evolved in almost complete isolation for tens or hundreds of millions of years at a time. As Douglas Adams put it, "the universe is large, mind-bogglingly large." Third, of course Hoyle never explains, nor could he, where these genetic storms are coming from. Where and how did DNA originate in the cosmos? There is not a shred of evidence that it ever did, nor do we know of any mechanism that would make it possible. While the absence of evidence is certainly not evidence of absence, the burden of proof rests on Hoyle, given that he has set out to accomplish nothing less than the overturning of one of the major contributions to science and human thought on record, Darwin's theory of descent with modification.

Challenges to established science are not only possible, but a crucial component of the scientific endeavor. They are usually received with skepticism, some times even with acrimony, not only because people feel threatened whenever the accepted paradigm is challenged, but more importantly because scientists have to be conservative in order to minimize the likelihood that a crackpot theory might take over their field simply because it's new and sounds reasonable: a certain degree of resistance to new ideas is actually healthy. Nevertheless, I did take Hoyle's book seriously enough to read it to the end, which is probably more than a lot of other scientists will be inclined to do. His attempt is not a serious challenge to Darwinism, it is rather the simplistic outcome of an ego large enough to allow him to see himself far ahead of tens of thousands of others despite the fact that he has no training in the specific field and has obviously not bothered reading much about it. To overthrow a paradigm is much harder work than Fred Hoyle has been willing to put into his casual approach to biology, and his only lasting accomplishment will be to provide dubious justification to hordes of creationists whom, ironically enough, he despises as much as any evolutionist. This is not much of a legacy for one of the most brilliant minds of 20th century cosmology.