H. ALLEN ORR:

Having thoroughly enjoyed David Berlinski's recent book, A Tour of the Calculus, I am not eager to squabble publicly with him. But I am afraid I must. In his article, "The Deniable Darwin" [June], Mr. Berlinski discusses an essay of mine in which I criticize an extremist style of evolutionary thinking called "adaptationism." He concludes: "Like Orr, many biologists see an acknowledgment of their doubts as a cagey, a calculated, concession; but cagey or not, it is a concession devastating to the larger project of Darwinian biology."

I admit I briefly enjoyed the suggestion that my essay is so clever that it manages to undermine Darwinism. But, alas, Mr. Berlinski is exaggerating just a tad. The claim that my criticism of adaptationism, or, . . . for that matter, any other biologist's criticism of any other "ism," has pulled the rug out from under Darwin is just wrong.

But Mr. Berlinski has his own, fairly novel, criticisms of Darwinism. A couple of these are very clever and, to a nonbiologist, surely seductive. I am writing, therefore, to explain why I think these criticisms are wrong.

But first a few facts. Near the beginning of his essay, Mr. Berlinski states: "Evolution is a process, one stretching over four billion years. It has not been observed." While it is true that no biologist attended the last four billion years of evolution, the claim that evolution has not been observed is simply wrong. Examples are a dime a dozen. When antibiotics were first introduced, most bacteria were susceptible. Antibiotics were handed out like candy and anyone who had read a page of Darwin could have predicted the result: now, many bacteria are resistant. And Mr. Berlinski surely knows what happened when we threw DDT at insects: they evolved insecticide resistance. On a grander scale, botanists have documented the recent evolution of new species and some species have even been recreated from their ancestors in the lab. Though hardly packing the drama of reptiles metamorphosing into mammals (but what do you expect in 100 years of observation?), these are all iron-clad -- and witnessed -- examples of evolution.

But on to Mr. Berlinski's more novel criticisms of Darwinism. He has two worries. First, can random changes in an "alphabetic" system (like DNA) fuel evolution? And second, if evolution is fueled by such changes, how does it know "where to go"?

Mr. Berlinski's first question is more sophisticated than it might seem. He is not just rehashing the tired argument that Darwinism depends on random mutations and that such changes cannot build something as fancy as an eye. He knows that Darwinism is not just "random mutations" but "random mutations plus natural selection," which is a different beast altogether. His worry is more subtle. It is this: DNA is "alphabetic," a discrete language of A's, T's, G's, and C's that somehow encodes all the designs we find in organisms. But how can random perturbations in such a language yield usable material for evolution? In every other language we know of, Mr. Berlinski writes, "randomness . . . is the enemy of order." Random changes in English yield gibberish. Random changes in computer programs are even worse: we cannot even ask what a "modified program would compute; it just jams." And so, he argues, look what Darwinism really asks of us: it demands we believe that selection uses random changes in DNA, when -- by analogy with any other formal language -- such changes should yield mere gibberish, hopelessly "jamming" organisms.

Mr. Berlinski's objection is one of those beautiful theories that gets killed by an ugly fact. The fact is: whether or not random DNA changes should invariably jam organisms, they do not. While lethal mutations -- changes which so derail an organism that it dies -- are common, so are mutations of such benign and subtle effect that heroic chemical measures are required just to find them. The existence of subtle, functional, usable mutations in DNA is a simple fact that no amount of analogizing with computer programs can make go away. That random changes in computer programs -- but not DNA -- invariably jam things does not show that there is something wrong with Darwinism but that there is something wrong with the analogy.

What about Mr. Berlinski's second criticism? Even if random mutation could provide usable material for evolution, how does evolution know where to go? Mr. Berlinski retells the story of the proverbial monkeys who -- banging away at their typewriters -- try to create a phrase from Shakespeare. In Mr. Berlinski's version, a "Head Monkey" preserves any stray letter that matches the target phrase. For the evolutionist who usually tells the story, the point is that it does not take long for this cumulative sifting process to yield the desired phrase: mutation (typing) plus selection (save the matches) works. But Mr. Berlinski is not sold: "A target phrase? Iterations that most resemble the target? A Head Monkey that measures the distance between success and failure? If things are sightless, how is the target represented?" The Designer, allegedly tossed out of a job by Darwinism, has apparently reappeared as the guy who erects the target and measures the distance between here and there.

Mr. Berlinski certainly shows that the monkey analogy is imperfect. But that is true of any analogy . . . and the monkey analogy captures an important part of Darwinism. It shows that, by saving favorable random changes, evolution can gradually build fancy structures. One need not wait for all the "parts" to appear miraculously at once.

But the analogy completely flubs another part of Darwinism: evolution does not, of course, work toward any "target." So how, then, does evolution know where to go? The answer is the most radical and beautiful part of Darwinism: it does not. The only thing that "guides" evolution is sheer, cold demographics. If a worm with a patch of light-sensitive tissue leaves a few more kids than a worm that cannot tell if the lights are on, that is where evolution will go. And, later, if a worm with light-sensitive tissue and a rough lens escapes a few more predators, that is where evolution will go. Despite all the loose talk (much of it, admittedly, from evolutionary biologists), evolution knows nothing of "design" and "targets." . . .

In the end, I am afraid that Mr. Berlinski's criticisms do not fare any better than those of other anti-evolutionists. I will be the first to admit, though, that Mr. Berlinski is not a traditional anti-evolutionist. I know him well enough to know that he, unlike them, is neither anti-scientific nor doctrinaire. His criticisms -- like those from any good scientist -- are, I think, both sincere and tentative.

Department of Biology
University of Rochester
Rochester, New York

RICHARD DAWKINS:

David Berlinski's article reminds me of the tactics employed by a certain creationist with whom I once shared a platform in Oxford. The great evolutionist John Maynard Smith was also on the bill, and he spoke after this creationist. Maynard Smith was, of course, easily able to destroy the creationist's case, and in his good-natured way he soon had the audience roaring with appreciative laughter at its expense. The creationist had his own peculiar way of dealing with this. He sprang to his feet, palms facing the audience in a gesture eloquent of magnanimous reproof. "No, no!" he cried reproachfully, "Don't laugh. Let Maynard Smith have his say. It's only fair!" This desperate pretense that the audience was laughing at Maynard Smith, when in fact it was laughing with Maynard Smith at the creationist himself, reminds me of Mr. Berlinski's pretending to misunderstand Jacques Monod and me to the extent of thinking we disagree with each other over the issue of chance.

As for the identity of the creationist who tried to pull this little stunt on Maynard Smith, it was none other than David Berlinski. The audience, by the way, saw through his tactic instantly and treated it with hoots of derision.

University Museum
Oxford University
Oxford, England

DANIEL C. DENNETT:

I love it: another hilarious demonstration that you can publish bull---t at will -- just so long as you say what an editorial board wants to hear in a style it favors. First there was Alan Sokal's delicious unmasking of the editors of Social Text, who fell for his fashionably anti-scientific "proof" that according to quantum physics, the world is a social construction. Now David Berlinski has done the same to the editors of Commentary, who fell just as hard for his parody of "scientific" creationism. They must really be oppressed by evolutionary theory to publish such inspired silliness without running it by a biologist or two for soundness. Two such similar pranks in a single month make one wonder if this is just the tip of the Zeitgeist. What next? A hoax extolling the educational virtues of machine guns for tots in the American Rifleman?

I love the rich comic patina of smug miseducation Mr. Berlinski exudes: Latin names for species mixed with elementary falsehoods in about equal measure, the subtle misuse of "Doppelganger," the "unwitting" creation of a new term, "combinatorial inflation," and the deft touch of "betraying" his cluelessness by referring to Kim Sterelny as "she."

The hints are subtle but conclusive. No serious opponent of evolutionary theory would trot out the ill-considered remarks of the mathematician M.P. Schutzenberger -- a line of discredited criticism quietly abandoned by others years ago -- without so much as a hint about their standing. How could the heroic misunderstanding of Jacques Monod that enables our author to pit Monod against Richard Dawkins be anything but disingenuous? Could any actual professor of mathematics and philosophy "in American and French universities" misrepresent the import of the second law of thermodynamics with such poetic fervor, such blithe overconfidence?

Whoever this David Berlinski is, he is clever enough to fool Commentary, and I wouldn't even be surprised if some evolutionists take him seriously enough to rebut him in detail. Even better, some earnest creationists may clasp him to their bosom. That is, one presumes, his larger joke. The only reason I am exposing it now (killjoy that I am) is to make it clear that so far as I know, we evolutionists did not put him up to it. We feel no need to burden our critics with such agents provocateurs, but they are welcome to him if they want him.

Center for Cognitive Studies
Tufts University
Medford, Massachusetts

ARTHUR M. SHAPIRO:

. . . When I debate anti-evolutionists, I always warn the audience that a critic can raise more issues in a given amount of time than a defender can answer. The defender is thus vulnerable to the claim that he has failed to address this or that criticism. In replying to David Berlinski's eloquent but deeply flawed article, let me stress that I am not concerned with a point-by-point refutation. I am much more interested in his uses of rhetoric to create unjustified impressions in the mind of the reader. . . .

Most biologists I know are driven by curiosity, not ideology. We recognize that it is pointless to deny that a real world exists, and dangerous to place perfect faith in our own objectivity in interpreting it. That is why science has institutionalized skepticism in the process of criticism and competition. . . . We are indeed human in becoming enamored of or trapped in our own ideas, but we are surrounded by hungry rivals who work to keep us honest.

That is just as true in evolutionary biology as in other sciences. So how can Mr. Berlinski get away with accusing us of hushing up our disagreements in order to present a united front against our cultural foes? There is no more contentious, pugnacious, querulous bunch of professionals on earth than evolutionary biologists, . . . and intergenerational and ideological conflicts within academia keep the pot boiling in a very public way. In the process, anti-evolutionists are provided with a mountain of very quotable quotes, usually presented to the public out of context. If we were so concerned with papering over our differences, why would we debate them before the public? . . .

I was at the Wistar Institute in 1966 when Murray Eden and M.P. Sch¸tzenberger presented their arguments, cited approvingly by Mr. Berlinski. If he had read the proceedings of that meeting, he would know that it was organized at the instance of the biologists, who wanted the opportunity to hear and discuss the criticisms of their mathematical colleagues. He would also know that some of the objections he cites now were actually dealt with effectively from the floor. Instead, he disingenuously implies that those arguments still bear as much force now as they did then -- that they somehow have been swept under the rug or ignored by biologists.

Mr. Berlinski ignores the fact that mathematics has been the indispensable handmaiden of evolutionary biology for over 70 years. Does he really think nothing interesting has happened in evolutionary biology for 30 years, or that anything interesting that has happened has merely weakened the Darwinian paradigm? Perhaps. After all, with the avalanche of technical, semipopular, and popular books on paleobiology published in recent years, Mr. Berlinski sends his readers off to consult Alfred S. Romer's classic treatise on vertebrate paleontology, also published in 1966, and showing its age badly.

In another display of disingenuousness, Mr. Berlinski wants evolutionary biologists to explain from first principles why one species of widow spider commits sexual suicide while another does not. He knows perfectly well that the chain of causes is too long, with too many opportunities for historical contingency to operate, for anyone with less than the omniscience of the deity to do that. But just as physics rules out mountains shaped like inverted cones, so, too, does Darwinism rule out certain things. For example, it rules out any organism displaying a structure or behavior that enhances the fitness of some other species while reducing its own: the cartoonist Al Capp's altruistic "Shmoos" were a Darwinian impossibility. That is precisely why "altruism" was such a hot topic among evolutionists in the 1960's and 70's, when numerous biologists tried and failed to demonstrate its existence in nonhuman species.

One final point, which, while seemingly a trivial matter of semantics, illuminates the deepest problem of Mr. Berlinski's essay: our ability to reconstruct the history of life (phylogeny) depends upon the "tree of life" not only not being "densely reticulated," but not reticulated at all. "Reticulated" means net-like; branches diverge, then anastomose. But in phylogeny, branches, once separate, stay separate. It is this fact that enables us to reconstruct phylogeny through the method of nested, shared resemblances, now usually called "cladistics." That method fails if branches really do anastomose -- either through hybridization, which is rare in many groups but frequent in others, or through a more arcane process such as the vectoring of genetic information across taxonomic lines by viruses.

If the tree of life were very reticulate, formal phylogenetic reconstruction would be impossible, and we truly would have a problem -- not with the Darwinian paradigm per se, but with our ability to know whether it was true or false.

Mr. Berlinski could write a good essay about a living, breathing issue like that -- were he not so preoccupied with rehashing the debates of 30 years ago. . . .

Division of Biological Sciences
Section of Evolution and Ecology
University of California
Davis, California

PAUL R. GROSS:

Berlinski No. 1 ["The Soul of Man Under Physics," January] was amusing: especially his silly similes, such as Jupiter's moons in the guise of testicles. Most of the science was OK, although marred by glosses that were themselves errors. . . .

Berlinski No. 2 is much more presumptuous and more seriously wrong. How could Commentary not have let some biologist or other read it? I do not have the space to deal with all his unsupported or dead-wrong assertions, so I will just make a few counterassertions which perhaps will reduce the encouragement Mr. Berlinski has given creationists and other consumers of anti-science who might be among Commentary's readers.

The claim that intermediate fossil forms are absent is simply false. It is the oldest canard in the creationist handbook. Picking and choosing from among a few of the more florid statements of Stephen Jay Gould and Niles Eldredge does not reveal to the innocent reader . . . why these two are and have always been convinced of the essential truth of organic evolution as Darwin first described it and of natural selection as at least one mechanism of it.

The "fitness" argument is dealt with, seriously, in every introductory course in biology, including the ones I have taught. Mr. Berlinski's fixation on it was out of date a very long time ago.

"What good is 5 percent of an eye?," is a stupid question if asked in the context in which Mr. Berlinski places it. Moreover, there is new and astonishing evidence of the genetic, hence selective, mechanism underlying the embryology of all eyes -- insects, mice, humans.

Mr. Berlinski grossly misrepresents Jacques Monod's argument, as he does the way in which "chance" is used and understood in modern biology. It is possible he really does not know, but I am not sure; maybe he is a postmodernist at heart.

"Bauplan," like a few other fancy words, is comically misused.

I could go on and on, . . . but for a small fee I would be glad to support all these assertions with citations of solid science from the current literature of molecular and evolutionary biology. A bit of that literature Mr. Berlinski may have mined; but if so, he has done it the same way that "Intelligent Design Theorists" mine it: to extract a few sentences here and there for the comfort of creationists. This second Berlinski opus amounts to a charge that biologists are in a conspiracy to hide from outsiders the bankruptcy of the central principle of biology. . . .

As I did once before, I deplore Commentary's giving such comfort to Luddites.

Center for Cell Signaling
University of Virginia
Charlottesville, Virginia

RANDY M. WADKINS:

. . . It would take an article longer than David Berlinski's original one just to correct and/or clarify his many misstatements, and I will address only a few of them here. I would strongly suggest that anyone who thinks the points made by Mr. Berlinski seriously detract from evolutionary theory should read the talk.origins FAQ's (http://earth. ics.uci.edu:8080/origins/faqs. html) for a brief synopsis of why virtually everything he says on the matter of evolution is suspect. . . .

Mr. Berlinski states that "before the Cambrian era . . . very little is inscribed in the fossil record." Yet the oldest known fossils are of bacteria and stromatolites (containing blue-green algae), both of which are with us today, and are still among the simplest forms of life known. The Cambrian explosion occurred much later, after simple life (and even simple multicellular life) had already appeared. This is one of the many points in the fossil record that support evolution.

Mr. Berlinski fails to mention the specific cases where transitional fossil forms are found in abundance, and demonstrate quite readily how animals we know today have, without doubt, evolved from earlier forms. The exhibit on horse evolution at the Smithsonian Natural History Museum is quite convincing, and the evolution of horses is perhaps the most well-documented of any animal. . . .

Although he points out that gradualism is not considered to be an accurate description of evolution, Mr. Berlinski uses it anyway to argue against evolution. He presents several examples of related species which exhibit traits that, according to him, could not have come about if one came from the other (or from a precursor). But he neglects to point out that a single amino-acid mutation in a single protein is enough to cause amazing changes in a developing animal, such as eyes that grow on an insects' legs, or chickens that develop webbed duck feet and hair. . . .

Mr. Berlinski further asserts that evolution is random. It is not quite that simple. Evolution is accomplished by random mutations of DNA. However, the process of natural selection is not random at all. Any physician treating a relapsed cancer patient knows full well that the cancer . . . will be resistant to the original drugs used as therapy (and this resistance may be due to the overexpression of a single protein). This is not random in any sense. Similarly, any physician treating a patient infected with the new strains of antibiotic-resistant bacteria can tell you that there is nothing random about the selection of these bacteria: they grow where their predecessors could not.

More generally, when a red, orange, yellow, green, blue, and purple ball are all placed in a bag, there is a one-in-six chance that I will draw out the green one. If, however, I look into the bag, there is a 100-percent chance I will draw the green one. There is nothing random about this, and it is this nonrandom process (natural selection) that drives evolution. Mr. Berlinski does not see this at all. . . .

Just to show how ineffective Mr. Berlinski's "Advent of the Head Monkey" section is, . . . while Richard Dawkins was unable to write a computer program that simulates evolution, scientists more adept at programming have indeed been able to do just that. The natural-selection algorithm in Keen and Spain's Computer Simulations in Biology generates the phrase basic biological modeling is fun from a string of random letters, based on Dawkins's suggestions. Furthermore, the entire burgeoning field of genetic algorithms is based on similar, but more complex, models of evolution. . . .

The theory of evolution deserves constant criticism, as do all scientific theories. But . . . it deserves some challenges that are less trivial than the ones Mr. Berlinski proposes. . . .

Silver Spring, Maryland

KARL F. WESSEL:

The heart of David Berlinski's argument has been experimentally refuted, a result that is of more recent vintage than the obsolete mathematical metaphor he employs. In a 1986 experiment performed by Marshall Horwitz and Lawrence Loeb of the University of Washington, 19- base long messenger RNA promoter sequences were deleted from the genomes of E. coli bacteria and replaced by randomly synthesized sequences. Of the approximately 10¹¹ possible sequences of the type, it turned out that many promoted the function of the deleted natural sequence (which confers resistance to tetracycline) as well as or better than the original. This was true even for a small subset of sequences randomly generated from two bases, i.e., from far fewer than 10⁶ of the 10¹¹. Nor did some of the most efficient sequences at all resemble the original.

Given this outcome and others like it, it is clear that something is radically wrong with Mr. Berlinski's analogy of biological genomes to computer programs. Either ge-nomes are nothing like programs, or else at least some programs are far more robust in the face of effects resembling natural selection than he imagines.

In point of fact, for several years people like John Koza of Stanford University have been using analogs of natural selection to evolve computer programs. Many of these evolved programs perform their optimizing tasks better than the best intentionally designed ones, providing the reductio ad absurdum of Mr. Berlinski's criticism. . . .

Mr. Berlinski seems to have been misled in his opinions by the unbiological prejudices of mathematicians accustomed to working with formal systems. Koza has addressed several of these biases in print. One of the most characteristic is the belief that the genome must be parsimonious, like the shortest proof of a mathematical theorem or the code of a highly compressed computer program. Maximally compressed programs, in the sense of algorithmic information theory, are necessarily both rare and randomly distributed within the space of all possible programs; in fact, such programs cannot be obtained through evolutionary processes, since genetic distances do not correlate nonrandomly with differences in fitness in this kind of regime. . . .

The fact is that biological genomes are not highly compressed in the information-theoretic sense but are, rather, extravagantly redundant, something that was not appreciated in the era (the 1960's) from which Mr. Berlinski draws his argument. Redundancies in the code help enhance evolvability by reducing the impact of deleterious mutations. At the same time, sexual recombination works to filter out deleterious mutations in the homozygote while helping fixate beneficial ones. . . .

When Mr. Berlinski's anti-Darwinian rhetoric is deflated, essentially nothing of his argument is left. Natural selection is neither more nor less "forcelike" than the Pauli Exclusion Principle; in both cases, probability distributions mimic the effects of forces. Gould and Dawkins are no more argumentative than Mach and Boltzmann, Einstein and Bohr, Glashow and the superstringers. States of punctuated equilibria fall out gracefully from highly plausible computer models of natural selection. William Paley's (and David Berlinski's) intuitions about the Design of the World are neither more nor less reliable than those of flat-earthers, goat-entrail readers, or believers in the Oedipus complex. . . . Incidentally, DNA was not discovered by James Watson and Francis Crick but by Friedrich Miescher (1869).

Rancho Palos Verdes, California

PHILIP H. SMITH, JR.:

In his elegant, lyrical critique of Darwinism, David Berlinski shrewdly derides the smugness of some mainstream evolutionists who consider the Darwinian theory of evolution "fact," . . . but the theory is far more robust than his attacks suggest.

Darwinism is a theory not unlike the Indo-European theory in linguistics, which holds that most of the languages of Europe and the Near East descended from a common parent language by a process of evolution. Each theory notices a number of facts -- some contemporary, some fossil -- and attempts to link them logically. Neither pretends to predict the future, either among biological organisms or among European languages.

In comparing Darwinism unfavorably with the science of geology, Mr. Berlinski indicates that he expects predictive natural laws from evolutionary theory, similar to the laws of physics. But Darwinism is much more a theory of history . . . than a predictive theory. In this it resembles the Indo-European theory. Although that theory talks in terms of sound-change "laws," these laws are really only observed historical regularities, and claim no predictive power.

Mr. Berlinski says that Darwinism "suggests a view in which living creatures are spread out smoothly over the great manifold of biological possibilities, like colors merging imperceptibly in a color chart," but he finds life in fact to be quite different. Evolution must, of course, proceed by myriad small changes, but not all the resulting life forms will be visible at one time. Indeed, as these numerous small changes occur, the less successful forms will, of necessity, no longer prevail, leaving the field to the latest-evolved forms. It would be unrealistic to expect a continuous rainbow of all life forms. In exactly analogous fashion, the linguistic landscape of Europe shows sharp discontinuities between the Romance and the Germanic areas, between Germanic and Slavic, and so on. . . .

Like many another critic of Darwinism, Mr. Berlinski applies the second law of thermodynamics to evolution. He suggests that life "appears to offer at least a temporary rebuke to the second law of thermodynamics." Not so. He has forgotten the phrase "in a closed system," which belongs in the statement of the second law. Life is not a closed system; it draws energy from the increasingly entropic universe to enhance its own order, no less than we do when we forge iron and other materials to make a highly organized automobile (or a watch). . . .

Both Darwinism and the Indo-European theory have been around for well over 100 years, and each has constructed an interlocked edifice of observed fact (plant, bone, rock; language, document inscription) and consistent explanations of their relationships, and in both cases the theories hang together pretty well. . . . Each waits to be torn down by a more satisfactory theory. This I fear Mr. Berlinski has not managed to do in his article, despite the eloquence of his prose.

Chapel Hill, North Carolina

SHELDON F. GOTTLIEB:

. . . David Berlinski fails to understand the role of determinism in evolution, not in the sense of an Intelligent Designer but in the inherent properties of molecules themselves. Sidney Fox and his colleagues have demonstrated that amino acids, when heated (to speed up the rate of the reaction) at temperatures found on Earth give rise to "thermal proteins" (TP's). When placed in water, TP's organize themselves into protocells. Protocells demonstrate the major properties associated with living organisms, including metabolism, electrical excitation, reproduction, and so forth. . . .

Fox's work shows that very early stages of evolution can occur rapidly, within minutes, and are highly reproducible. Sites where such reactions can and probably did occur include the hydrothermal vents found on the surface of the Earth and in the oceans. Around the ocean vents, highly complex ecosystems evolved. . . .

Like other creationists, Mr. Berlinski seems to refuse to accept the fact that the introduction of the supernatural removes the discussion from the realm of science. . . . In the world of the supernatural, anything goes, and the only limitation is the extent of one's imagination. No evidence is required to substantiate any claims. . . .

In science it is rarely appropriate to ask why. Scientists ask questions such as: what is the nature of _______?, or what is the mechanism whereby ______? Such questions take science out of the realm of the supernatural into the material world and provide the basis for constructing testable hypotheses and for obtaining factual evidence as to the physical, chemical, and biological makeup of the world. . . . But Mr. Berlinski asks: "Why did the firefly discover bioluminescence?" To the best of my knowledge there is no evidence to suggest that fireflies went into laboratories and discovered bioluminescence. Mr. Berlinski also asks: "Why is the Pitcher plant carnivorous but not the thorn bush?" . . . To which biologists respond: what are the evolutionary forces at work that resulted in the Pitcher plant adopting carnivory? In examining this question, biologists have found that carnivory is a means whereby Pitcher plants can grow in nutrient-poor soils: carnivory provides a means of obtaining phosphates and nitrates. The story of thorn bushes is different: for these species, found in dry, arid environments, the more pressing evolutionary forces apparently involved protection from desiccation and excessive herbivory. Thus, the more immediate survival mechanisms for each species evolved differently and conferred different survival advantages. . . .

Finally, Mr. Berlinski asks: ". . . who on the basis of experience would be inclined to disagree" with his conclusion of deliberate design? I would!

Department of Biological Sciences
University of South Alabama
Mobile, Alabama

ROBERT SHAPIRO:

In "The Deniable Darwin" David Berlinski has captured, in elegant prose, the complexity of living things. This complexity is displayed in both their structure and behavior, and is already overwhelming at the level of one-celled organisms. Mr. Berlinski quite justifiably scolds those biologists who exchange skepticism for dogma and proclaim that our current level of understanding is enough to provide a full account of the origin and development of life.

As his article develops, however, we recognize that the author . . . has a hidden agenda, the promotion of an idea that falls outside the bounds of scientific investigation: Intelligent Design. In advocating his cause, Mr. Berlinski juxtaposes Darwin and Design as if they were the only possible explanations for the intricacy of life, and then defines the Darwinian side to his own satisfaction. If we choose this answer, we must presume that natural selection is the only mechanism that drives evolution, and that only "sheer dumb luck" lies at its heart.

On the other side, no details are offered concerning the identity of the Designer or the means and mechanism by which He created. I suspect that the author does not wish to specify a power that operates within this universe and according to its laws, such as the intelligent silicon chip once proposed by Fred Hoyle. Mr. Berlinski prefers a supernatural explanation. He is not content to sustain his religious convictions by faith alone but, in the style of the creationists, harasses those who seek a natural explanation for origins and evolution. The creationists offer an even more polarized two-way choice, of course. They would have us select between Darwin and chance on the one hand, and a literal interpretation of the Bible with a 6,000-year-old universe on the other. . . .

Fortunately, there is a third alternative. . . . In this view, an organizing principle exists in nature and governs the rise of complexity in life, and in many other natural phenomena. The roots of this idea lie in several philosophies, but it has taken on a new respectability with the development of mathematical theories of complexity.

An excellent popular account of this field and its application to evolution is given by the physicist Stuart Kauffman in his book, At Home in the Universe. Kauffman argues that both self-organization and natural selection have driven evolution; the development of increasing complexity in life is an expected result of these mechanisms.

Unlike Intelligent Design, self-organization can be subjected to the test of experiment. Few efforts have been made to explore this process, perhaps because complexity theory is new and scientists have been too devoted to the more traditional form of Darwinism. Yet I believe the concepts have considerable potential, particularly in the origin-of-life field, and I hope they will gain greater attention in the future. Questions of origins and evolution can then move out of the debating hall and into the laboratory, where they belong.

Department of Chemistry
New York University
New York City

PAUL H. RUBIN:

One reason for reading Commentary is that it supplies a rational, intellectual analysis of important issues. Therefore, it was surprising and dismaying to read David Berlinski's anti-intellectual attack on Darwinism, one of the great intellectual accomplishments of the human mind. . . .

"We never attribute the existence of a complex artifact to chance," says Mr. Berlinski. But evolution does not imply only chance. Mr. Berlinski makes it sound as if the alternatives were either chance or conscious direction; many systems, however, are complex and yet not planned. In my field of economics, for example, the entire structure of a market economy is neither designed nor planned. . . . Language itself, which Mr. Berlinski spends much time discussing, is not planned and yet is a highly complex artifact, and one which Stephen Pinker and others have shown to be a product of Darwinian processes.

Mr. Berlinski asks: "Could a system we do not completely understand be constructed by means of a process we cannot completely specify?" He answers, "we do not know." But it is not only evolutionary biology which presents this conundrum. All science is an attempt to explain systems we do not fully understand with processes we cannot completely specify. Unless we proceed as if the answer were yes, we cannot do science. . . .

Mr. Berlinski calls the human mind "conscious, flexible, penetrating, inscrutable, and profound." He is of course correct. But today we are at the threshold of truly understanding the source of this mind, as the new science of evolutionary biology is beginning to explain it. Indeed, it is particularly perverse to publish a critique of Darwinism just as its practitioners are beginning to use it to explain human behavior in a remarkable and truly profound way. . . .

Emory University
Atlanta, Georgia

JOHN M. LEVY:

David Berlinski does not like the philosophical and theological implications of Darwinism and so tells us that it plays the same role in the "economy of belief" as does the biblical account of creation. In that it supports a world view, perhaps that is so. But what is missing from his article, and what may not be apparent to the reader without some education in biology, is how massive is the evidence for an evolutionary scenario. The student in comparative anatomy cannot but be struck by how often similar forms appear in the structure of one species after another and how the same structure gradually changes form and sometimes function as one proceeds from lower to higher forms.

The student of embryology sees traces of evolution throughout the subject. At one stage human embryos have fishlike features. The early human embryo has an arterial pattern like the arteries in the gill arches of fish. As the embryo matures, that pattern changes to produce the mammalian pattern of the arteries in the thorax and neck. A comparable transition can be seen if one looks across species from fish to amphibian to reptile to mammal. Biology students are taught "onto-geny recapitulates phylogeny" (the history of the individual repeats the history of the phylum) because the evidence of it is so overwhelming. . . .

Mr. Berlinski makes much of gaps in the fossil record, but consider the following: for many years biologists were convinced that aquatic mammals had evolved from terrestrial mammals but there were many gaps in the record. They could not provide the path. Then pieces of fossil evidence started turning up in Egypt, Pakistan, and elsewhere. Commenting on these, Stephen Jay Gould . . . states:

I am absolutely delighted to report that our usually recalcitrant fossil record has come through in exemplary fashion. . . . The embarrassment of past absence has been replaced by a bounty of new evidence -- and by the sweetest series of transitional fossils an evolutionist could ever hope to find.

Terrestrial mammals returning to the sea to feed on fish and then evolving the aquatic features of whales and porpoises sounds like a "just-so" story. But the fossil evidence suggests that it really is "just so." To base one's rejection of Darwinism on gaps in the fossil record is to place it on a very erodable base. . . .

Living creatures in their incredible complexity are mind-boggling. How a sequence of bases in a double helix can specify that you will have a dimple like your grandmother's is mind-boggling. But where the evidence for the hypothesis is so strong, and where various different types of evidence reinforce one another, I do not think that rejection-because-of-bogglement is the wisest course.

Blacksburg, Virginia

MARTIN GARDNER:

David Berlinski's spirited attack on evolution, like Phillip E. Johnson's earlier book, Darwin on Trial, contains one huge, glaring omission. Nowhere does he tell us what brand of creationism he supports. It is as if someone wrote an article blasting evidence for the earth's roundness but refused to say what shape he thought it was.

Like Johnson, Mr. Berlinski seems to think that the punctuated evolution of Stephen Jay Gould and his friends somehow damages the Darwinian view that all life evolved by gradual small changes. The jumps in Gould's theory are, of course, jumps only relative to the extremely long periods during which certain species remained stable -- trilobites, for example. Gould's jumps are tens of thousands of years, arising from tiny mutations that for reasons still unclear occur more rapidly than usual. Darwin's bulldog, Thomas Huxley, was well aware of such jumps, and they have provided fuel for creationists from Darwin's day until now. . . .

I assume that . . . Mr. Berlinski is not a young-earther who thinks all the fossils are records of life that perished in Noah's Flood. But what exactly is his creation scenario? There are three possibilities.

1. God created each individual species by fiat -- that is, they had no ancestors. Because there are millions of insect species alone, this requires God to perform many millions of miracles. I cannot believe that . . . Mr. Berlinski favors such a view.

2. Creations occurred over millions of years but only for broad classifications of life, such as plants, fishes, reptiles, birds, mammals, and, above all, humans. Gradual evolution could then take place within those broad categories to give us different species of reptiles, birds, and so on.

3. Each mutation (Darwin, of course, did not know about mutations) is a creative act of God. Instead of relying on a combination of chance and natural laws, as almost all liberal Christians and Reform Jews believe, God's method of creation was to direct each mutation in a way it could not have occurred naturally. This is difficult to defend because, in the light of the beliefs of theists like myself that all natural laws were created and are upheld by a transcendent deity, there is no more need for God's special interventions than there is a need for Him to give periodic pushes to stars and planets, as Newton argued, to keep their orbits from deteriorating. . . .

One final question. Does Mr. Berlinski think the first humans had parents who were beasts, or no parents at all? . . .

New York City

HERBERT GINTIS:

. . . What is David Berlinski trying to say? He never denies evolution. He never denies that natural selection occurs, or that it is important. He never provides an alternative theory of genetic variation. Actually, his purpose is stated almost as an afterthought, in the final paragraph of his piece. Here he quotes Richard Dawkins's pronouncement, "Darwin made it possible to be an intellectually fulfilled atheist," and Mr. Berlinski himself remarks that "the theory of evolution" plays a "singular role in the life of our secular culture. . . ."

Fortunately, it is not necessary to deny Darwin in order to affirm the validity and importance of a spiritual (i.e., nonsecular) culture. Spiritual life has a validity of its own, neither supportive of nor in conflict with natural science. Faith, love, beauty, and other cherished dimensions of human life are experienced realities whose truth does not depend on this or that fact of nature.

Moreover, if one's belief in God depends on the inability of science to explain the world (I do not believe it should), science gives one ample support for one's faith. Even with Darwin, we are far from able to explain the emergence of consciousness from the history of the modification of the structure of DNA molecules. . . .

I think Darwin made it easier to be a nonbeliever, undercutting one of the most powerful arguments for the existence of God -- the so-called "argument from design." I remember personally, in first grade, being made to repeat every morning in our school prayer, "for it is God who has made us, and not we ourselves," and thinking how strong an argument that was. But by the time I became an adult, and came to appreciate the theory of natural selection for the brilliant monument to the human intellect that it is, I had already gone far beyond the argument from design in grounding my own spiritual life. Denying Darwin is not only intellectually impossible, it is spiritually unnecessary.

University of Massachusetts
Amherst, Massachusetts

DAVID P. BABCOCK:

David Berlinski is refreshing in attacking evolution without assuming the correctness of its most frequently offered alternative, the literal truth of Genesis 1. Indeed, he is somewhat dismissive of creationism. So far, so good. But where Mr. Berlinski then gets into trouble is in failing to discuss what possibilities other than evolution remain after creationism is set aside. He seems to accept the general validity of science; fine. But given that, he might be expected to offer some explanation for the observation that species have arisen at different times. He does not.

Perhaps we are to infer, from his acceptance of the Argument from Design, that his designer from time to time flings new species onto the earth out of whole cloth. This is not an inherently ridiculous idea. It is, however, rather more embarrassed by lack of evidence than is the evolutionary hypothesis. Moreover, it is just the understanding of such evidence as does exist that caused evolutionary biologists to suspect that Piltdown Man was a fraud when their intellectual opponents had no scientific reason for doing so. And the evolutionary biologists were right. If their opponents have ever had such success with their inferences from the fossil record, history does not record it.

Boulder Creek, California

EUGENIE C. SCOTT:

. . . The content of David Berlinski's article does not differ from more traditional creation-science material, though his tone is more genteel and his writing a lot more literate. . . . But true to the creation-science genre, his approach consists of constructing strawmen, then knocking them down with misinterpreted, faulty, or nonexistent data as well as carefully se-lected quotations from evolutionary scientists. . . .

For example, the "intermediate/ transitional-fossil" strawman is a staple of anti-evolutionist rhetoric. . . . Because their theology says God created all living things as separate "kinds," creationists assert that there can be no transitional fossils, and no descent with modification. As a result, they persist in their inability to recognize mosaic specimens. Show them a fossil like the part bird/part dinosaur Archaeopteryx, and they will claim that because it has feathers, it is "100-percent bird," regardless of the fact that specimens which lack clear feather imprints are indistinguishable from small, toothed, clawed, bipedal dinosaurs. Show them a hominid fossil with a small brain, teeth similar in some respects to those of an ape and in others to those of a human, and a pelvis requiring bipedal locomotion, and they will see "just an ape." . . .

Readers can write me at NCSE for a bibliography of articles and books that refute the Cambrian-explosion argument, refute the impossibility of amino acids coming together to form a protein, explain the difference between chance and evolution, discuss why William Paley's argument from design has been supplanted in science by arguments based on natural causes. . . .

One can be a believer and accept evolution. Articles like Mr. Berlinski's confuse rather than elucidate issues. Mostly, however, they perpetrate bad science in a society that needs more science literacy, not ignorance.

National Center for Science Education
El Cerrito, California

EDWARD T. OAKS, S.J.:

David Berlinski's superbly written expose of the loaded logic embedded in Darwin's "explanation" for the origin of species raises once more the issue of its replacement: if Darwinism is as inadequate as Mr. Berlinski claims, what is the engine of evolution?

I was raised in a Jesuit school system that pretty much took evolution for granted, and indeed seemed rather proud of the work of the Jesuit paleontologist Teilhard de Chardin in reconciling Christian doctrine with evolution. . . .

I awoke from my own Darwinian dogmatic slumbers only late in life, when I first read Gertrude Himmelfarb's tour de force of a biography, Darwin and the Darwinian Revolution, now sadly out of print. Miss Himmelfarb makes many of the same points that Mr. Berlinski does and documents how these same objections were very much alive among Darwin's fellow scientists in the Royal Society. . . . Yet still, Darwin's theory won out, and that requires explanation, for which I can offer no other hypothesis than the even greater weakness of its perceived alternative: creationism.

According to the Himmelfarb thesis, nothing was more crucial to the eager hearing the Origin received than the appearance the year before of Omphalos by Philip Gosse, a professional geologist and (speaking avant la lettre) fundamentalist Christian minister. This peculiar book argued -- for the first time and in explicit answer to Darwin, with whose theories Gosse was already professionally familiar -- that there had been no incremental change in the earth's surface or gradual development of organic life. The semblance of change and development had been imprinted on the earth at the very moment of the act of creation. The reaction of the scientific and educated public to this bizarre, even grotesque, thesis was predictable. In Miss Himmelfarb words: "The howl of execration that [might have] awaited Darwin was let loose on Gosse [instead]. . . ."

One of the most fair-minded historians of this controversy is Peter J. Bowler, who notes in his book Evolution: The History of an Idea, that Darwin's proposal needed creationism as a foil:

Why, then, did paleontologists accept evolution? The answer seems to be that enough fossil evidence turned up to convince them of the futility of creationism, even if it did not prove evolution. The evolutionists could predict that if, for instance, birds have evolved from reptiles, then one might hope to find a fossil with characteristics intermediate between the two classes. Creationists, by contrast, had a vested interest in maintaining the absolutely distinct characteristics of existing groups and were unwilling to postulate intermediates in the past. . . .

Of course, as Mr. Berlinski points out so effectively, Darwinism lacks predictive power too, but I cannot help feeling that we are dealing with two different orders of unpredictability here. After all, meteorology and history lack true, rigorous predictive power, but no one accuses either discipline of alchemical chicanery just because the weatherman cannot predict a hurricane two weeks from now or a historian cannot predict where the next tin-horn dictator will overthrow a constitutional government. After all, evolutionary biology is, like history, a retrospective science, one whose search for causes must by definition terminate in the present pre-given status quo. . . .

I can only say in closing that if I am to join the law firm of Himmelfarb, Berlinski & Bowler, I would first have to say: "Give me a replacement and I'm yours."

Regis University
Denver, Colorado

KENT GORDIS:

The strangely florid style and lack of coherence of David Berlinski's article obscures a rather simple situation.

When Darwin's theory of evolution was formulated in the 1850's, it was a good example of a Popperian scientific hypothesis -- in Karl Popper's words, "an imaginative or even mythological conjecture about the world." Today, this hypothesis has been convincingly refuted on both empirical and theoretical grounds. The empirical grounds are the total absence of fossil evidence of the transitional forms predicted by the theory, the existence of which is practically the definition of "evolution."

The theoretical grounds are the vanishingly low probability that random selection could explain the existence of life forms, the simplest of which is of extreme complexity. This point is vividly conveyed by Fred Hoyle's metaphor, quoted by Phillip E. Johnson: "that a living organism emerged by chance from a pre-biotic soup is about as likely as that a tornado sweeping through a junkyard might assemble a Boeing 747 from the materials therein."

Mr. Berlinski quotes Richard Dawkins, one of the leading contemporary Darwinists, as follows: "Darwin made it possible to be an intellectually fulfilled atheist." In fact, the refutation of Darwin's theory would threaten the faith of all those whose world view is based on "naturalism," the view that only natural or material phenomena are real, a class which includes not merely atheists like Dawkins but most agnostics, as well as many people professing "modern" religious beliefs. This, of course, is the reason Darwin's theory of evolution, although refuted as a scientific hypothesis, has become a naturalistic dogma whose transmission from generation to generation is seen as a sacred duty of the scientific clergy. . . .

New York City

J. R. DUNN:

One point not mentioned in David Berlinski's article is the fact that natural selection is the sole scientific theory that cannot be superseded. Other theories, no matter how well-established, are commonly overthrown or subsumed in light of later discoveries, much as Newtonian mechanics became a special case within the framework of Einsteinian relativity. But natural selection is presented as the last word in its field, monumental and unassailable, with all ensuing work subordinated to the original thesis. This is a complete reversal of the accepted method of scientific inquiry, one that effectively turns Darwin into a prophet and his theory into a pseudo-religion -- in a word, an ideology.

This explains why vicious infighting goes on among evolutionary biologists, so similar to doctrinal battles among religious sects; why natural selection is presented as a universal doctrine, dominant far beyond the boundaries of biology. . . and why contradictory findings are jammed into the basic theory with no attempt at consistency. . . . Lastly, it goes a long way toward explaining the incessant hostility that evolutionary biologists display toward religious faith -- one does not see many astrophysicists or cosmologists crowing about their victories over the fundamentalists. . . .

The vast majority of scientific theories prevalent in the mid-19th century have been supplanted. How much longer will Darwin's immunity to fundamental scientific principles continue?

New Brunswick, New Jersey

ARTHUR B. CODY:

It was courageous of Commentary to publish David Berlinski's much-needed questioning of evolutionary theory. It is important to have serious inquiry rather than religious dogmatism allowed into a debate over what seems to be a doctrine scientists are reluctant to examine. One unfortunate thing creationism has brought about is the hunkering down of biological scientists and others in a position which will allow no place for expressions of uncertainty. As a result, writers in all sorts of fields -- biology, ethics, law -- do not hesitate to justify their observations by teleological reasoning under the guise of Darwin's theory of natural selection, confident that none will propose an embarrassing query. . . .

One hopes that David Berlinski's analysis of an important element in the neo-synthetic theory of evolution will be taken as needed ventilation, letting both light and air into this dark and darkening area of science. I fear instead what he will get are hysterical vituperation and angry shouts to close the window.

Santa Cruz, California

TOM SOUTHWICK:

It is surely beyond exasperation for science writers, being deeply invested in the conventional wisdom of blind evolution, to encounter doubts not only from a science-illiterate general public, but also challenges from their own ranks. They would prefer to identify all skeptics of evolution with young-earth biblical literalists. I am sure David Berlinski will be chastised for his brilliantly evocative tour through the evidential ruins of Darwinism, with its grand extrapolations from essentially trivial observable facts, and its simply unbelievable claim that chance -- unguided natural processes -- could organize the information requirements of even basic biological structures, let alone complete functioning organisms.

As Mr. Berlinski points out, chance ("pure dumb luck") is the "beating heart" of evolutionary theory, just as Jacques Monod clearly stated, and Richard Dawkins in turn has presumed to deny. The truth is that no hint of design is permitted; those who would propose a hybrid, as in "theistic evolution," are not really in the debate; and their accommodation to the regnant evolutionary view assumes that the "fact" (read, "proven fact") of evolution is inarguable.

But chance is not a satisfactory engine of the generation of life from nonlife, or the presumed grand procession of creatures from simpler antecedent forms. I appreciate the highly serviceable presentation Mr. Berlinski makes of the staggeringly great improbability involved: happening by blind chance upon the one useful configuration of chemicals, in a spot the size of a dime, on an earth-sized planet otherwise covered by uselessness.

I would add that this dime must be found for just one successful molecule to emerge; just one amino acid string to form one useful protein. To make a second protein, we must start over; and just happen to find that dime again, having no idea where we found it before, or even that we did find it; and having found it, our second spectacular success must also accidentally set aside our result such that it will combine usefully with our first success . . . and so on, passing quickly into the realm of plain impossibility.

Actually, it is even worse: even allowing for the eons of time required to give verisimilitude to the evolutionary model (typically four billion years), not one useful strand of DNA could have organized itself by accident. As the microbiologist Michael Denton puts it in Evolution: A Theory in Crisis, to suppose that chance processes could create life (a far greater challenge) "is simply an affront to reason." In 1981, when Francis Crick (Mr. DNA) seriously proposed the it-came-from-outer-space explanation for life on earth, it was at least in part to expand (greatly) the time available to evolution. . . .

It seems clear that intelligence designed life; and even the graybeards of the evolutionary cathedral must propose, as Mr. Berlinski puts it, a Head Monkey, overseeing vast numbers of ordinary monkeys banging uncomprehendingly at their keyboards, "striving" for that line from Shakespeare, and the Head Monkey discerning (accidental) "targets" being reached for, so as to preserve the results of successful aiming, thereby carrying out the selection process. Dumb luck must have also just blundered, an incalculably great number of times, into this highly sophisticated decision-making mechanism. Nothing recognizable as "evidence" is presented for this; evolution simply requires it, or something like it. This is philosophical musing, not science. But it is proposed, with no cognitive dissonance, as having the imprimatur of truth that science confers.

The debate over origins will persist, not because biblical creationism is convincing science, but because Darwinism (chance) is finally seen as an alluring impossibility. Even as it is passionately resisted, the God solution will stand as a reasonable interpretation of the emerging design paradigm. . . .

Milltown, New Jersey

JOHN WIESTER:

Thank you for David Berlinski's insightful article. His examples of Darwinian "just-so" stories and fuzzy logic are both instructive and amusing. As a science educator and parent I am not amused, however, by the extent to which both science and reason are being abused by the Darwinist establishment in the education (?) of our children. The following excerpt from an official 1995 Statement on Teaching Evolution illustrates my point:

Evolutionary theory, indeed all of science, is necessarily silent on religion and neither refutes nor supports the existence of a deity or deities. Accordingly, the National Association of Biology Teachers, an organization of science teachers, endorses the following tenets of science, evolution, and biology education: "The diversity of life [i.e., all life] on earth is the outcome of evolution: an unsupervised, impersonal, unpredictable, and natural process of temporal descent with genetic modification that is affected by natural selection, chance, historical contingencies, and changing environments."

Now that is just double-talk. Of course the claim that life is the product of an "unsupervised" process of evolution has important religious implications, and of course it goes way beyond the empirical evidence. The bogus claim of "silence" on religion serves only to protect these anti-religious statements from the critical analysis they deserve. . . .

Science Education Commission
Buellton, California

NANCY R. PEARCEY:

David Berlinski's piece on Darwinism was colorful and witty, but his postscript topped it all, with its mock claim that all modern novels have descended from Don Quixote by a series of copying errors. Mr. Berlinski weakens his case unnecessarily, however, when he denies that DNA is "completely comparable" to a language. The application of information theory to biology reveals that the genetic code is structurally identical to human codes. That is, DNA is not merely like a language, it is a chemical language. To explain its functioning, biologists have been obliged to borrow the terminology of editors and linguists. They speak of DNA as a code or symbol system; they speak of molecules that copy and translate the message; they speak of proofreading functions and error correction.

The upshot is that the origin of life and its myriad forms must be recast as the origin of biological information. The DNA of even the simplest bacterium contains as much information as the entire Encyclopedia Britannica. Mr. Berlinski's spoof puts the onus on Darwinists to explain how new information can emerge through the accumulation of errors in a text. And if that is a preposterous way to account for the origin of Madame Bovary or Anna Karenina, it is even more implausible when it comes to the origin of living things.

The Wilberforce Forum
Reston, Virginia

DAVID J. MANDEL:

David Berlinski's article, "The Deniable Darwin," is brilliant, clear, witty, wonderful. His postscript on the evolution of Ulysses from Don Quixote per Borges is marvelous. . . .

New York City

LARRY EUBANK:

David Berlinski has made many good points about evolutionary theory. I am always taken aback by the unquestioning certainty displayed by evolutionists in the face of a theory so fundamentally, intuitively preposterous; by the solemn assurances of TV broadcasts on nature that their fairy-tale rationalizations are "science"; and by the intellectual thuggery of college professors who insist that evolution is scientific fact and that we must believe it. . . .

Jeffersonville, Indiana

HENRY SHERMAN:

David Berlinski's brilliant "The Deniable Darwin" riveted my attention. I have waited 40 years to read such an article.

Admittedly no better schooled than most other aspiring intellectuals in the fine points of Darwin's theory, I have never been able to accept its explanation of how the fantastic profusion of living and long-dead species evolved from that first lonely cell, aimlessly drifting in the primordial soup.

Something always stuck in the craw -- it would just not go down. Why did the known fossil record fail to document a single example of the continuous evolutionary change proposed by the theory? If evolutionary mutation took place in response to changes in the external environment, how could one account for the bewildering abundance and variety of species surviving side by side for untold millennia in the same environment? Why, as Mr. Berlinski suggests in his closing sentence, does acceptance of Darwin's theory require an act of faith similar to that demanded of those who confidently place their trust in biblical (and other) accounts of creation?

For me, there was always too much that distinguished a trout from a tiger; an unfathomable distance between a mouse and a Mozart. Unhappily, there appeared to be nothing between the no-longer acceptable beliefs of the creationists and the Darwinism which seemed to have cleared the secular field of any other explanation of life on our planet. Bereft of the consoling beliefs of my fathers in divine intervention; dismissive of the pallid nostrums of the deists who see God in everything and thus, ultimately, in nothing; somehow sensing in Darwinism the illusionist's trick, which makes the impossible appear plausible (Mr. Berlinski's witty "ta-da!"), I found it rather lonely out there.

To be sure, Mr. Berlinski does not offer an alternative to the theory he so drolly declares deniable. Nor does he offer much comfort to those of us left out in the cold by our inability to accept any of the other answers to the perpetual questions: What are we doing here? How did we get here? Where are we going? However, by shaking the Darwinian tree with such vigor, he may have helped once again to make respectable to nonbelievers the proposition that man, after all, may have more soul than a snail. And, for that, we must thank him.

New York City

TOM BETHELL:

David Berlinski's article is but a harbinger of the great brewing controversy over evolution. The truth is that, although we have long been told that evolution is a "fact," there is almost no evidence to support it. In 1981, Colin Patterson, senior paleontologist at the British Museum (Natural History), said that he did not know of any such evidence; subsequently he has amended this only slightly. The claim that evolution is a fact depends on a definition that is so weak as to be valueless: evolution becomes "change over time." The frequent reference to those biston moths in England -- in a polluted environment, the dark or camouflaged variety became relatively more numerous than the light variety -- shows how feeble the evidence really is. If the evolutionists had something more persuasive, we would have heard about it on a daily basis.

Darwin's great claim to fame was his discovery of the alleged mechanism of evolution -- natural selection. But as Mr. Berlinski notes, this turns out to be a disguised tautology. The fittest survive, and survival is the criterion of fitness. Change was equated with improvement and the Darwinian discovery involved little more than the importation of progress into biology. Now we no longer believe in the idea of progress, and faith in biological evolution may be jeopardized as a result.

Hoover Institution
Stanford, California

RABBI DANIEL LAPIN:

Publishing David Berlinski's "The Deniable Darwin" was an act of great intellectual courage. You have fired a shot in what is becoming a great moral revolution, and it will be heard around the world. Discovering that Darwin is deniable might tell us a little of how primitive life began. It would tell us everything about how modern life should continue.

Today's greatest question is whether humans have been touched by the divine and thus possess moral judgment or whether we are just sophisticated animals. Upon the answer depend most important matters of public policy. Our view of criminals must depend upon whether we consider them to be acting naturally or in a depraved and evil manner. Seventh-grade sex-education classes would look very different if we knew of no reason to expect children to behave like animals. Come to think of it, all classes and indeed all schools would look different, too. The debate, thus, is not really over the mating habits of the male redback spider (Latrodectus hasselti), it is over the mating habits of American urban men (Homo sapiens).

On July 2, 1863, a thirty-four-year-old professor at Bowdoin College led his regiment to glory at Gettysburg where he ordered the brilliant charge that saved Little Round Top and avoided a Union catastrophe. After the war, Joshua L. Chamberlain was elected governor of Maine and later served as president of Bowdoin College. He wrote these words: "I do not fear these men of science, for after all they are following in God's ways. . . . I would say that [scientific] laws are God's ways seen by men." Remember that The Origin of Species had been published just four years earlier.

Today, Cornell University's William Provine, a leading historian of science, insists that the conflict between science and religion is inescapable. Specifically, he states: "Modern science directly implies that there are no inherent moral or ethical laws, no absolute guiding principles for human society . . . we must conclude that when we die, we die and that is the end of us."

Understandably, passions flare over this debate. It will ultimately resolve whether people must choose between science and religion or whether they can simultaneously rejoice in the Creator's morality and His methods.

Chamberlain and Provine cannot both be right. America will ultimately follow the vision of only one. Which one will be decided by the outcome of the intellectual and moral war you have courageously joined. Until now the war was being ignored. Commentary's June 1996 issue forever eliminates that option.

Toward Tradition
Mercer Island, Washington

MICHAEL J. BEHE:

David Berlinski and Commentary both deserve much credit for shining a light on the woefully underexamined theory that supports much of modern thought. I found his arguments to be quite compatible with my own thinking concerning the systems I encounter daily as a professional biochemist and about which I have written in Darwin's Black Box. Indeed, just as the pleasing shape of a jetliner belies the complexity of its internal organization, so the complexity of life mushrooms as one gets closer to its foundation. The shape of the eye, which Darwin tried to explain, pales in comparison with the interactions of rhodopsin, transducin, arrestin, rhodopsin kinase, and other proteins in the visual cascade. Explaining how the swimming behavior of a whale might be produced gradually (if anybody ever succeeded in doing this) would be a walk in the park compared to explaining the bacterial swimming system -- the flagellum, which requires more than 40 gene products to function.

Most people consider Darwinism, whatever its faults, to be science. Yet in an interesting way they are wrong -- at least in my area of biochemistry, the study of life's foundation. Scientific results get reported in science journals, like Nature, the Journal of Biological Chemistry, and so forth. But if you search the science journals, as I have done, for detailed explanations of how particular, complex biochemical systems (such as, say, the blood-clotting system or intracellular-transport) were produced, you come up completely empty-handed. Astonishingly, science's own journals contain no explanations. Since science is found in science journals and Darwinian explanations are absent, Darwinism is not science.

Department of Biological Sciences
Lehigh University
Bethlehem, Pennsylvania

PHILLIP E. JOHNSON:

The media stereotype of the evolution controversy is that it pits "science" against fundamentalist "religion." It would be more accurate to say that Darwinism is the most important of the materialist ideologies -- Marxism, Freudianism, and behaviorism are others -- which have had done so much damage to science and society in the 20th century. The theory persists in spite of the evidence because Darwinism cannot survive without it. Commentary deserves great credit for helping to break the stereotype by publishing David Berlinski's fine essay.

Public understanding of the defects of Darwinism is limited because the educators think it their duty to indoctrinate, and prestigious propagandists like Richard Dawkins protect the theory effectively with ridicule and bullying. (This method of dealing with criticism is itself a hallmark of bad science.) Readers who are not intimidated should be sure to read Michael J. Behe's Darwin's Black Box, which shows why the mutation/selection mechanism has been all but abandoned as an explanation for the irreducible complexity found in the biochemistry of organisms. There is also a new journal called Origins and Design.

The fall of Darwinism will be the big story of the early 21st century -- learn about it now, and be ahead of the curve!

School of Law
University of California
Berkeley, California

JEFFREY SATINOVER:

A theory becomes widely accepted because it explains so much. This is exactly why Darwinian evolutionary theory has triumphed, especially its core proposition: that new living forms emerge over time exclusively via competitive, "natural" selection. The theory is so powerful that it has been able to account for variations at the level of DNA and protein morphology, even though it was formulated long before anyone knew anything about either. Indeed, "genetic algorithms," mathematical techniques modeled on competitive selection, have recently been developed and used to solve previously intractable problems in a growing number of nonbiological domains.

So massive is the accumulation of natural data adequately accounted for by competitive selection, so impressive is its capacity to mimic goal-directed, problem-solving intelligence in spite of being a "dumb," mechanical method based upon random variation, that it should come as no surprise when critics of Darwin are perceived as willfully ignorant, or worse. This must certainly be the gist of what is being leveled at David Berlinski and at Commentary.

In fact, the challenges to the principle of competitive selection are of varying sophistication. The crucial challenge to Darwin is not the one which asks whether God created the world in 144 hours, and then cleverly buried under its surface a bunch of rocks made to look like animal bones so as to ensure tenure to a group of professors 5,700 years later. The essential challenge is, rather, this: is there any evidence of anything at all at work in the world which is, however, not of the world? And if there is evidence of any such foundational anomalies, is there further evidence that this "something" has a central (not incidental or unnecessary) role in the processes by which the forms of life have come into existence? For if there is such evidence, then however machine-like much, or even most, of the universe is, it is not wholly a machine, and neither is man.

There are such anomalies, of both sorts. They may be put together as follows: the degree of unitary coherence in the stage-by-stage emergence of higher and higher levels of nervous-system organization (that is, brain structure) does not appear wholly explainable by purely physical influences, however much can in fact occur in the way of self-assembly via nonlinear, stochastic (random) processes given (a) the amount of time since the emergence of life of earth and (b) the absence of any form of teleology or preprogramming (in which knowledge of a desired end influences the choice of and application of means).

So much does competitive selection explain that it has long been presumed to be merely a matter of time before this anomaly will be reduced as well, and it has therefore been shunted aside as unimportant. After all, most of the interesting and fruitful projects in modern science have borne their fruit by presuming as a working hypothesis that the machine-model of existence is true, and by following up on the consequences, and there is simply no gainsaying the power of the successes thus achieved. But the fact remains that the most interesting discoveries of 20th-century science have in fact provided evidence of exactly the opposite.

For 80 years now, ever since the great and utterly intractable anomalies of radiation and radioactive decay confounded the almost completed structure of determinism and forced the adoption of quantum mechanical principles (an oxymoron, since quantum, by definition, means not mechanical), the world's greatest scientific minds have been tying themselves in knots. They were quickly able to show just how outrageous the implications of quantum mechanics are; they waited patiently for a technology sophisticated enough to do the experiments they were certain would confirm the theory's falseness. . . . But in the last ten years, as the technology has reached the necessary level, many quantum hy-potheses have been tested, and been proved correct. . . .

Now, Darwinian theory and its variants are corollaries to the machine-model, and David Berlinski and the editors of Commentary deserve kudos for attacking head-on the notion that the fundamental principle of Darwinian theory is beyond rational dispute. The simple fact is that, however uncomfortable people may feel at having to take seriously points of view that remind them of religious fundamentalism, modern science itself has long recognized that the unquestioned hegemony of mechanical determinism is open to question, and this directly affects evolutionary theory. . . .

The question may be put as follows: can one treat all of life as a gigantic random machine, a huge genetic algorithm "aimed" at solving the "problem" of survival and self-assembly, whose outputs (morphologic categories; species) can be sufficiently accounted for by purely random events operating according to the principles of "nonlinear" (chaotic) dynamics? Or are the results more akin to the solutions produced by the emerging, quite serious field of quantum computation: "impossible," noncomputable results that require "something more" than the physical universe to arrive at? Many serious scientists believe that the latter proposition is not on the level of the Scopes "monkey trial," but of the utmost significance.

Commentary is not a scientific journal, but let me give a brief taste, just to let it be known that the debate is real. In 1992, Koichiro Matsuno of the department of bioengineering at the Naguaka University of Technology in Japan published an article in BioSystems (a high-level journal of theoretical biophysics) entitled, "The Uncertainty Principle as an Evolutionary Engine." In it he demonstrates that there is a peculiar uncertainty relationship between local quantum fluctuations (these give rise to gene-level mutations) and final global morphologies. In brief, the "needs" of the final forms "influence" (in the peculiar meaning that this word takes on when dealing with quantum phenomena) the fluctuations necessary to generate the solutions.

It is precisely this kind of "impossible," nonmechanistic influence which quantum mechanics has repeatedly demonstrated at the heart of all physical systems. Serious scientists now entertain the possibility that the universe itself might be rather like a freely-acting agent (a pantheistic point of view, as it were); some consider that this "agency" might more precisely be spoken of as residing elsewhere, or being other, than any physical system, however large (a more theistic point of view, and less popular, though not for reasons of evidence).

The proposition that evolution is in some way guided by a kind of "intentionality" analogous to our own (not wholly determined: just regular nudges to keep it on this track instead of that one) is far from absurd. . . . Matsuno and others may be correct or they may be wrong, or partially both. But it is long overdue for readers to understand that they are not fools who insist that there is more at work in the universe than meets the modernist eye. Just what kind of central organizing principle will in fact arise to replace the emptiness of today's still-regnant Cartesian model remains to be seen, but construction is well under way.

Kudos again for the first public notice to appear in so prestigious a venue as Commentary; and thanks to David Berlinski for risking it.

Westport, Connecticut

RUSSELL ROBERTS:

I want to thank David Berlinski for his elegant expose of the Darwinian emperor and his meager amount of clothing. Biologists and the rest of us often confuse what I would call micro-evolution (animals have coloring that makes them hard to spot in the wild) with macro-evolution (the tree of life that Mr. Berlinski mentions where man makes his orderly climb out of the primordial soup). There is a great deal of evidence for the former and precious little for the latter. Where are those pesky intermediate forms?

People often speak about the theory of evolution as if it were a "fact" or "proven." Alas, it is only a theory, a useful way of organizing our thinking about the real world. When theories stretch too far to accommodate the facts, a paradigm shift is usually forthcoming. As Mr. Berlinski notes, biologists have trouble imagining an alternative paradigm to evolution. This may explain the vehemence with which they greet criticism. Perhaps they too are uneasy with the emperor's wardrobe.

Olin School of Business
Washington University
St. Louis, Missouri

M. P. SCHUTZENBERGER:

Molecular biology has provided classical fields such as embryology with powerful new tools, but it has not brought any confirmation to the Darwinian theory of evolution. Therefore, many biologists have quietly turned away from this simplistic view of life.

Commentary is to be congratulated for inviting David Berlinski to present the most recent scientific argument against neo-Darwinism. In so doing it gives voice to a growing silent minority. This article will have a considerable impact.

Academy of Sciences
Paris, France

DAVID BERLINSKI:

Some readers seem to have been persuaded that in criticizing the Darwinian theory of evolution, I intended to uphold a doctrine of creationism. This is a mistake, supported by nothing that I have written.

Confronted with a complex human artifact like a watch, William Paley inquired into the source of its complexity. Insofar as such a complex object is unlikely, Paley reasoned, its existence can be explained only in terms of a human act, one in which material objects (gears, springs, levers) are deliberately arranged in a particular configuration. The same pattern of observation and inference, Paley went on to argue, indicates that complex biological structures are likewise the products of a deliberate act of design, the designer in such cases being the Christian deity.

Darwinian theorists accept the first of Paley's inferences, but reject the second. Biological artifacts are complex, they say, but not designed. Their existence may be explained in terms of random variation and natural selection. I dispute this claim, without endorsing Paley's theological inference. It is not necessary to choose between doctrines. The rational alternative to Darwin's theory is intelligent uncertainty.

A number of letters raise similar points; I have distributed my comments over a number of responses.

IN MAINTAINING that evolution is a process that has not been observed, H. Allen Orr writes, I appear to have overlooked examples of evolution like the speckled moth, which undergoes mimetic changes in wing coloration as the result of environmental pollution, or the development of antibiotic resistance in bacteria. Mr. Orr is correct that there are such examples; I scruple only at the conclusions he draws from them. Changes in wing color and the development of drug resistance are intraspecies events. The speckled moth, after all, does not develop antlers or acquire webbed feet, and bacteria remain bacteria, even when drug-resistant. The most ardent creationists now accept micro-evolution as genuinely Darwinian events. They had better: such are the facts. But the grand evolutionary progressions, such as the transformation of a fish into a man, are examples of macro-evolution. They remain out of reach, accessible only at the end of an inferential trail.

In calling attention to "species [that] have even been recreated from their ancestors in the lab," Mr. Orr is, no doubt, referring to the recent work of L.H. Rieseberg and his co-workers ("Role of Gene Interaction in Hybrid Speciation: Evidence from Ancient and Experimental Hybrids," Science 272, 1996). The example is pertinent to my critique. Rieseberg and his co-authors reproduced under artificial conditions the genetic changes that have historically led from H. annus and H. petiolaris to H. anomalus. The plants in question are sunflowers. What is remarkable is the extent to which this experiment contravenes Darwinian doctrine. Given the crucial role played by random events in evolutionary theory, many biologists have drawn the conclusion that the tape of life, if rewound, would produce "a different array of evolutionary end products" (Stephen Jay Gould, Wonderful Life). The number of crossing schemes notwithstanding, the tape in this experiment ran to precisely the same genetic end product each time it was played.

Mr. Orr contends that in my discussion of the role played by randomness in formal systems, I appear to be upholding an analogy at the expense of the facts. Random events, he writes, do occur in molecular biological systems; so much the worse, then, for my analogy. But the interpretation of molecular biological facts in formal terms is hardly a matter of analogy. It is molecular biologists themselves who have found unavoidable the language of codes and codons, information, algorithms, organization, complexity, entropy, and the like. There is little by way of analogy in all this. DNA is not like a code; it is a code. It follows, then, that circumstances known to degrade meaning or information in formal systems should be the source of alarm in the context of theoretical biology.

No one denies that random events take place within molecular biological systems. The relevant question is how. In a formal context, the matter is not a mystery. Codes may be designed to remain robust in the face of background noise; what is required is redundancy, and the genetic code is, in point of fact, highly redundant. In communication systems, redundancy appears as a matter of design. It does not arise spontaneously. In the case of the genetic code, according to one commentator, "strong selection pressures" created the requisite redundancy. But this is to dispel one mystery by promoting another, as the familiar Darwinian circle again makes its appearance, the tail of one concept lodged firmly in the mouth of another. I will return to this point in a number of other responses, but let me repeat what I stressed in my essay, that I am not advancing an argument on this issue, only expressing intellectual unease.

In inserting a Head Monkey into Richard Dawkins's thought experiment, my aim was to show how the mechanism of design, purged on one level of Darwinian analysis, makes a stealthy reappearance at another. Mr. Orr is unpersuaded. "The monkey analogy," he believes, "shows that by saving favorable random changes, evolution can gradually build fancy structures." Such indeed is the perennial hope of Darwinian theorists, but Mr. Orr has, I believe, underestimated the force of my criticism. Favorable changes are one thing; changes that will be favorable, another. If the mechanism of Darwinian evolution is restricted to changes that are favorable at the time they are selected, I see no reason to suppose that it could produce any fancy structures whatsoever. If the mechanism is permitted to incorporate changes that are neutral at the time of selection, but that will be favorable some time in the future, I see no reason to consider the process Darwinian.

This is hardly a matter of semantics. A system conserving certain features in view of their future usefulness has access to information denied a Darwinian system; it functions by means of alien concepts. But this is precisely how Dawkins's experiment proceeds. My estimable Head Monkey conserves certain alphabetic changes because he knows where the experiment is going. This is forbidden knowledge; the Darwinian mechanism is blind, a point often stressed by Darwinian theorists themselves (see George C. Williams, Natural Selection, 1992). I develop this in more detail in responding to Randy M. Wadkins.

In his final argument, Mr. Orr repeats what is certainly the current orthodoxy, namely, that evolution has no targets and so like the rest of us is not going anywhere at all: "The only thing that 'guides' evolution is sheer, cold demographics." Although this is not a point I discuss in my essay, I remain unconvinced. There are certainly long-term trajectories visible in the progression of life. The development of neurological complexity is the obvious example.

RICHARD DAWKINS has succumbed to the endearing weakness of revising the history of an unpleasant encounter in one's own favor. I have done as much myself. But a public charge calls for a public response. In 1992, Mr. Dawkins, John Maynard Smith, and I did share a podium at Oxford University. His hands trembling with indignation, Mr. Daw-kins proposed to attack organized religion; I proposed to attack Richard Dawkins; and John Maynard Smith, seeing that it was required, proposed to defend Mr. Dawkins from my attack. The intellectual drubbing that Mr. Dawkins imagines I received, I recall in distinctly different terms. But why argue over the past? I have a videotape of our encounter, which I would be happy to make publicly available. If he wishes to debate again, Mr. Dawkins need only set the time and the place.

In remarks that have by now become well known, Jacques Mo-nod observed with some sorrow that under Darwin's theory, it is chance that plays the crucial role in the emergence and evolution of life. Mr. Dawkins proposes to deny this. His views and those of Monod are in conflict, a point clear to anyone able to read the English language. Mr. Dawkins's continuing insistence that two contradictory propositions are mutually consistent is evidence of an alarming logical deficiency.

In fact, Mr. Dawkins has simply misunderstood the fundamental character of the theory to which he has committed his passionate defense. Darwin's theory is both random and deterministic. True enough. Mutations occur randomly, but once they have occurred, natural selection acts deterministically to cull the successes and discard the failures. By and large, true again. Nonetheless, Darwin's theory is essentially stochastic, a term which in statistics refers to a process in-volving a random sequence of observations.

Let me call a random mutation together with its deterministic consequences an evolutionary episode. The proto-tiger develops claws; he lives to mate successfully. Such is a single evolutionary episode. According to Darwin's theory, evolutionary episodes are independent. A snapshot of any given episode does not suffice to determine the character of future episodes. And for obvious reasons: future events are contingent on further random events. It follows that the episodes must themselves be represented by what probability theorists (and everyone else) call a random variable. And processes represented by a random variable are by definition stochastic. These facts are understood by anyone in possession of the requisite technical concepts. For all his flaws as a philosopher, Monod was quite clear about the character of Darwin's theory.

On one important matter, Mr. Dawkins's letter requires a word of reproof. At our debate, I was asked by a member of the audience whether I held to any creationist beliefs or doctrines. I replied unequivocally that I did not. My views of Darwinism, I said, were negative, but rational. On the videotape, as I utter these forthright words, Richard Dawkins may be seen sitting placidly on the podium, staring somberly into space.

DANIEL C. DENNETT is under the curious impression that the best rejoinder to criticism is a robust display of personal vulgarity. Nothing in his letter merits a response.

Still, one general point deserves attention. Both Daniel Dennett and Richard Dawkins have fashioned their reputations as defenders of a Darwinian orthodoxy. Their letters convey the impression of men who expect never to encounter criticism and are unprepared to deal with it. This strikes me as a deeply unhealthy state of affairs. Ordinary men and women are suspicious of Darwin's theory; Dennett and Dawkins hardly go far here in persuading them that their intellectual anxieties are in any way misplaced.

CONTRARY TO what Arthur M. Shapiro asserts I do not doubt that evolutionary biologists are contentious; I said as much in my essay. What I deplored was their tendency to conceal their differences from the public. This is another matter entirely.

Evolutionary biologists have a habit of ignoring the most pertinent criticisms of their theory until they can decently call them out-of-date. My references to papers by M.P. Schutzenberger and Murray Eden not surprisingly prompt Mr. Shapiro to the conclusion that my arguments are anachronistic. In fact, Sch¸tzenberger and Eden enter my essay unobtrusively in largely a stage-setting role -- Sch¸tzenberger to call attention to a conceptual problem at the very heart of evolutionary theory and Eden to offer, for the first time, a quantitative assessment of the space within which evolutionary searches must be undertaken. Their papers are historically important, the points they make no longer controversial.

Does Mr. Shapiro doubt that randomness introduces an alien and discordant note into the dynamics by which very complex objects change? Or that combinatorial inflation blows up the space of possible proteins? These themes have been pursued in countless papers, monographs, and books. Thus, Hubert Yockey, arguing that the discovery by chance of a single molecule of iso-1-cytochrome c requires a miracle, continues the line initiated by Schutzenberger and Eden (Hubert Yockey, Information Theory and Molecular Biology, 1992. But see also Gregory J. Chaitin, "Toward a Mathematical Definition of 'Life,'" in The Maximum Entropy Formalism, eds. R.D. Levine and M. Tribus, 1979; Francis Crick, Life Itself, Its Origin and Nature, 1981; Max Delbruck, Mind from Matter?, 1986; Robert Shapiro, Origins: A Skeptic's Guide to the Creation of Life on Earth, 1986). The language of choice has changed -- a fragile consensus is emerging that Kolmogorov complexity is a natural measure of biological complexity or specificity* -- but the problems remain the same. The space of possible objects is entirely too large to be successfully searched by random means, a theme pursued yet again in Michael Denton's Evolution: A Theory in Crisis (1986).

I agree that much has happened in biology over the past 30 years -- who could doubt it? Developments taking place within molecular genetics, molecular biology, and biochemistry do seem to me to be profoundly at odds with the Darwinian paradigm, and those within paleontology flamboyantly so. I mentioned some points of conflict in my essay; I refer to others here.

Consider, for example, the question of how an undifferentiated cell manages the task of specialization, becoming over the course of time a neuron or a muscle cell or any other particular and peculiar biological object. The requisite information is contained, of course, within the cell's genetic apparatus; the problem is one of specificity. What regulates the expression of some parts of that apparatus, while simultaneously suppressing the expression of other parts? One suggestion of long standing is that regulatory mechanisms are switched on and off by means of biochemical signals sent from neighboring cells.

It is this suggestion that experiments conducted by Chiou-Hwa Yuh and Eric Davidson seem to support (C.-H. Yuh and E.H. Davidson, "Modular cis-regulatory Organization of Endo 16, a Gut-specific Gene of the Sea Urchin Embryo," Development 122, 1996). A gene in the sea urchin Strongylocentrotus purpuratus contains over 30 binding sites for more than 13 regulatory factors. These regulatory factors turn parts of the genetic apparatus on and off. They are sensitive to signals from adjacent cells, and, what is more, they are clustered in discrete modules; different module combinations lead to different patterns of cellular development.

The system that results is one of extraordinary combinatorial intricacy and complexity. What of randomness in all this? We have no idea how the general mechanism for cell-specific transcription came into existence; to argue otherwise would be sheer dogmatism. But one might have thought that a system of such delicacy, once it came into existence, would be unusually sensitive to random perturbations. Not so. The regulatory apparatus seems designed to incorporate mutations. Certain sites within each module function as key switches, turning the module on or off; mutations have a specific, but highly discrete effect. Modules are largely independent, functioning as more or less complete sets of instructions, like blocks of code. As one commentator puts it, the transposition of cis-regulatory modules from one gene to the next "may be a convenient way for nature to develop novel patterns of development" (Wade Roush, Science 272, 1996).

As is so often the case in molecular genetics, the description of a specific biological system reveals a pattern at odds with the one demanded by Darwinian theory. Differentiation is a highly sophisticated, enormously complex, and stable process, one in which the system is protected from noise by its very design. Mutations play a role in developmental change, but not a driving role. Rather, the facts suggest a system in which there are a finite number of combinatorial possibilities, with mutations serving to initiate certain carefully stage-managed sequences. The possibilities for module combination are fixed from the first; random changes serve simply to throw the various switches.

Examples of this sort could be multiplied at length; as our knowledge increases, the crude Darwinian scheme seems progressively remote from the evidence. How, for example, to account for the astonishing fact that from the point of view of the informational macromolecules, human beings and chimpanzees are virtually identical, their sequences in alignment to 99 percent (M.C. King and A.C. Wilson, "Evolution at Two Levels in Humans and Chimpanzees," Science 188, 1975)? The remaining slight difference evidently has controlled the development of a bipedal gait, with the profound neurological changes that this requires; the fully formed hand; the formation of the organs of speech and articulation; and, of course, the elaboration of the human mind, an organ unlike anything else found in the animal kingdom. Nothing in this suggests even remotely a continuous accretion of small changes. The evolutionary promotion of our ancestors seems to have been under the control of powerful regulatory genes, instructional blocks capable of coordinating a wide variety of novel functions. We simply have no idea how any of this works.

Still, the real infirmities of Darwin's theory are conceptual and not empirical. By the standards of the serious sciences, Darwin's theory of evolution remains little more than a collection of anecdotal remarks. Criticism is often a matter of clarification. Thus, making a point unrelated to the evidence, I argued that Darwinian theory is logically troubled. When I maintain that both Messrs. Dennett and Dawkins introduce by means of the backdoor the element of design they have ostentatiously booted from the front, my criticisms are again intended to show that the theory is deficient if only because it fails to meet its own standards for success.

The redback spider gives Mr. Shapiro pause; my request that its dining habits be deduced from first principles -- from any principles at all -- strikes him as disingenuous. "The chain of causes," he remarks, is simply too long to be mastered by anyone other than an omniscient deity. I do not for a moment doubt that that is so, but precisely this circumstance prompts the request for scientific theory in the first place. The chain of causes is everywhere too long to be explored and then mastered; it is the purpose of a theory to abbreviate and compress the data. Imagine a biologist who, on the grounds that the chain of causes is too long, failed to explain the fact that while men develop arms, geese develop wings!

The fact that the redback spider commits sexual suicide is interesting; one wishes to know why it is so. In this, and in countless similar cases, evolutionary theory simply has no explanation. What good is it, then?

Finally, I am astonished that Mr. Shapiro should think to object to Romer's Vertebrate Paleontology as a reference. The subject at hand is not particle physics; the main lines of vertebrate development have been clear for more than a century. My aim in suggesting Romer was to point the reader toward his stratigraphic charts: these plot vertebrate groups against time, the very many dotted lines between groups indicating purely hypothetical phylogenetic relationships.

PAUL R. GROSS is anxious lest in criticizing Darwinian theory I give comfort to creationists. It is a common concern among biologists, but one, I must confess, to which I am indifferent. I do not believe biologists should be in the business of protecting the rest of us from intellectual danger.

I did not say in my essay that the fossil record contains no intermediate forms; that is a silly claim. What I did say was that there are gaps in the fossil graveyard, places where there should be intermediate forms but where there is nothing whatsoever instead. No paleontologist writing in English (R. Carroll, Vertebrate Paleontology and Evolution, 1988), French (J. Chaline, "Modalites, rythmes, mecanismes de l'evolution biologique: gradualisme phyletique ou equilibres ponctues?," reprinted in Editions du CNRS, 1983), or German (V. Fahlbusch, "Makroevolution, Punktualismus," in Palaontologie 57, 1983), denies that this is so. It is simply a fact. Darwin's theory and the fossil record are in conflict. There may be excellent reasons for the conflict; it may in time be exposed as an artifact. But nothing is to be gained by suggesting that what is a fact in plain sight is nothing of the sort.

I do not doubt that Stephen Jay Gould and Niles Eldredge are convinced of the "essential truth of organic evolution"; or that they believe natural selection to be "at least one mechanism of it." The difficulty with this observation is that it is compromised by its qualifications. At any given moment, if the phases of the moon happened to be right, I might align myself with Mr. Gross's essential truth of organic evolution; as for natural selection, nothing at all remains at issue if it is demoted from its central position in Darwin's theory. The idea that evolution proceeds by means of many different forces is both unanswerable and uninteresting. To his credit, this is something Richard Dawkins recognizes.

It may well be true that my concerns for the logical niceties of Darwinian theory are out of date, as Mr. Gross suggests. So much the worse for evolutionary biology. To those of us on the outside, Darwin's theory will continue to seem seriously infected by conceptual circularity. (In Concepts and Methods in Evolutionary Biology, the philosopher Robert Brandon begins by at least recognizing the problem.)

The pattern of self-deception that I mocked in my essay is on display in any number of publications. In the first chapter of The Causes of Molecular Evolution (1991), John H. Gillespie is concerned to determine why a certain kinetic parameter (the Michaelis constant, K_m) reaches intermediate values in a certain class of fish, the ectotherms. "If natural selection is responsible for the evolution of K_m," Gillespie writes, "we should be able to understand why it would be maladaptive to exhibit values that are much less or much greater than the intermediate value." In fact, the two biochemical explanations Gillespie considers are flatly in conflict. They cannot both be true, although both may be false. To Gillespie, though, it hardly matters. ". . . [A]t this point in our discussion," he writes, "it is important merely to accept that there are plausible reasons for K_m to be evolutionarily adjusted to intermediate values. This forms the theoretical basis of our acceptance of the conservation of K_m in ectotherms as evidence for the action of natural selection in response to different thermal environments" (emphasis added). If the discussion has proceeded beyond the observation that K_m reaches intermediate values in ectotherms, the fact is not discernible by me.

On the matter of the eye, Mr. Gross has misunderstood me. I did not propose to champion Stephen Jay Gould's question, "What good is 5 percent of an eye?" Instead, I argued that the question is hopelessly premature. Without knowing how the visual system works, we cannot determine whether it is accessible to a Darwinian mechanism. This seems to me an incontrovertible point, if also one that in my experience evolutionary biologists indignantly deny.

Let me offer an analogy. Starting from one and adding by two's, I cannot hope to reach the number eighteen. So far, so good. Starting from one and adding by two's, can I hope to reach the number n, where n is some number or other? Who knows? The question is underdetermined. Starting from some principle of addition or other, could I expect to reach some number or other? The question is now doubly underdetermined, functioning as a single equation in two unknowns. So, too, the question of whether a Darwinian mechanism with unspecified properties could reach a mammalian visual system whose properties are not yet completely understood. It is only credulous philosophers who imagine that a Darwinian mechanism is universally competent.

I am in agreement with Mr. Gross when he refers to "new and astonishing evidence" about the origin of the eye. Herewith the facts. Halder, Callaerts, and Gehring's research group in Switzerland discovered that the ey gene in Drosophila is virtually identical to the genes controlling the development of the eye in mice and men. The doctrine of convergent evolution, long a Darwinian staple, may now be observed receding into the darkness. The same group's more recent paper, "Induction of Ectopic Eyes by Targeted Expression of the Eyeless Gene in Drosophila" (Science 167, 1988) is among the most remarkable in the history of biology, demonstrating as it does that the ey gene is related closely to the equivalent eye gene in Sea squirts (Ascidians), Cephalopods, and Nemerteans. This strongly suggests (the inference is almost irresistible) that ey function is universal (universal!) among multicellular organisms, the basic design of the eye having been their common property for over a half-billion years. The ey gene clearly is a master control mechanism, one capable of giving general instructions to very different organisms.

No one in possession of these facts can imagine that they support the Darwinian theory. How could the mechanism of random variation and natural selection have produced an instrument capable of anticipating the course of morphological development and controlling its expression in widely different organisms?

I deny that I have in any way misrepresented Jacques Monod's arguments; his words are clear and unequivocal, in English and in the original French. If the word "chance" has an idiosyncratic use among evolutionary biologists, the secret has been closely kept. As far as I can determine, evolutionary theorists make use of the same technical concepts as the rest of the mathematical community, a point evident in any current text -- Mathematical Evolutionary Theory, ed. R. Feldman (1989), for example.

The notion of a bauplan (der Bauplan, die Bauplanne), or body plan, has had some currency in the English-speaking world; Stephen Jay Gould and Richard Lewontin use the term and so does J.W. Valentine. But it is in general dismissed as a concept by Darwinians -- George C. Williams, for example. The idea of a body plan gained currency in the work of the great turn-of-the-century embryologists (Dreisch, Child, Boveri, Spemann), and before that in the writings of pre-Darwinian French biologists (Saint-Hilaire, Cuvier, Serres). I cannot imagine why Mr. Gross thinks it a term I have misused. (For reasons that are obscure to me, both he and Daniel Dennett carelessly assume that they are in a position to instruct me on a point of usage in German, my first language.)

I do not for a moment suppose, nor have I ever written, that biologists are in conspiracy "to hide from outsiders the bankruptcy of the central principle of biology." For one thing, the theory of evolution is hardly the central principle of contemporary biology. That description must surely be reserved for the thesis that all of life is to be understood in terms of the "coordinative interaction of large and small molecules" (James Watson, The Molecular Biology of the Gene). Rather, the theory of evolution functions as biology's reigning ideology. And no conspiracy is re-quired to explain the attachment of biologists to a doctrine they find sustaining; all that is required is Freud's reminder that those in the grip of an illusion never recognize their affliction.

WHAT RANDY M. WADKINS affirms about the pre-Cambrian era is true enough (but see E.H. Davidson, K. Peterson, and R. Cameron, "Origin of Bilaterian Body Plans: Evolution of Developmental Regulatory Mechanism" Science 270, 1995) for a real sense of the inadequacy of our grasp of fundamental facts concerning the Cambrian explosion). I only wonder why he imagines that his ob-servations are in conflict with anything I have written.

In the same spirit, Mr. Wadkins calls me to task for failing to cite "the specific cases where transitional fossil forms are found in abundance." The fossil record does contain many intermediate forms; a recent publication on the Internet (Kathleen Hunt, Transitional Vertebrate Fossils, FAQ [Frequently Asked Questions], jespah@u.washington.edu) lists more than 250. But Mr. Wadkins has misunderstood the nature of the argument. My concern was to state the obvious: the fossil record contains gaps, places where the continuity assumptions of Darwinian theory break down. That there are places where the gaps are filled is interesting, but irrelevant. It is the gaps that are crucial.

Classical physics suggests that the spectral distribution of intensity in black-body radiation should be a continuous function of temperature. Experiments conducted at the end of the 19th century indicated otherwise. Continuity is an essential aspect of classical mechanics, impossible to discard without discarding the entire theory. Since the anomaly of black-body radiation could not be understood within classical mechanics, physicists sensitive to the evidence were persuaded to attach their allegiance to the new quantum theory.

The classical Darwinian theory of random variation and natural selection requires a continuous distribution of animal forms, one that must be reflected in the fossil record. The assumption of continuity is a crucial aspect of Darwinian theory; it cannot be carelessly discarded. This again is something that Richard Dawkins has rightly emphasized. The fossil record does not appear to support the assumption of evolutionary continuity, or anything much like it. Why is it that evolutionary biology is immune to evidence of the sort that falsified classical physics?

It is upon the horse that Mr. Wadkins pins his best hopes: its evolution, he suggests, comprises an unassailable sequence, one bright bursting beast after the other. To a certain extent, however, the neat evolutionary progression from Eohippus to Equus is an artifact of selection. The original groupings of species are far thicker (or bushier, to use the term of choice) than first thought, so that the sequence depends on a judicious selection of horse-like organisms at each stage of development. It is rather as if one were to explain the emergence of the word cat by selecting the "c" from cattle, the "a" from abattoir and the "t" from tattle. Although almost 50 years old, G. G. Simpson's discussion in The Major Features of Evolution still repays study; for the modern point of view there is B.J. MacFadden, "Horses, the Fossil Record, and Evolution: A Current Perspective" (Evolutionary Biology 22, 1988).

But setting these reservations aside, what follows if the Equus sequence is accepted as a Darwinian progression? Very little. There are no more than three or four evolutionary sequences that, under the best of circumstances, suggest a complete progression of forms. By contrast, there are thousands upon thousands of species whose significant morphological features are not explained by complete or even highly suggestive sequences. Are the existing evolutionary sequences representative, or anomalous? In view of the striking discontinuities in the fossil record, I urge that they are anomalous; it would be interesting to know why Mr. Wadkins demurs.

When I observed that Richard Dawkins was unable to write a computer program that simulated his linguistic thought experiment, I did not mean that the task at hand was difficult. It is impossible. Mr. Wadkins commends the discussion in Keen and Spain's Computer Simulation in Biology as a counterexample; it is no such thing. What Keen and Spain have done is transcribe Dawkins's blunder into the computer language Basic. Here are the steps they undertake. A target sentence is selected -- basic biological modeling is fun. The computer is given a randomly derived set of letters. The letters are scrambled. At each iteration, the computer (or the programmer) compares the randomly derived sequence with the target phrase. If the arrays -- sequences on the one hand, target phrase on the other -- do not match, the experiment continues; if they do, it stops.

There is nothing in this that is not also in Dawkins, the fog spreading from one book to the next. The experiment that Keen and Spain conduct is successful inasmuch as the computer reaches its target; but unsuccessful as a defense of Darwinian evolution. In looking to its target and comparing distances, the computer is appealing to information a biological system could not possess.

This point seems to be less straightforward than I imagined, so let me spell out the mistake. Starting from a random string, suppose the computer generates the sequence bndit disne sot sodiswn toswxmspw sso. Comparing the sequence with its target, it proposes to conserve the initial "B." But why? The string is gibberish. Plainly, the conservation of vagrant successes has been undertaken with the computer's eye fixed firmly on its future target, intermediates selected not for what they are (gibberish, after all), but for what they will be (an English sentence). This is a violation of the rule against deferred success. Without the rule, there is nothing remotely like Darwinian evolution. What the computer has in fact done is to match randomly selected items to a template, thus inevitably reintroducing the element of deliberate design that was banished from the Darwinian world.

KARL F. WESSEL believes that Marshall Horwitz and Lawrence Loeb have provided an experimental refutation of my main argument. He has misunderstood both their argument and mine. In "Promoters Selected from Random DNA Sequences" (Proceedings of the National Academy of Science, October 1986), Horwitz and Loeb reported that by substituting randomly derived DNA sequences for original promoters in certain bacterial plasmids, they were able to discover additional promoters that maintained or enhanced drug resistance. Promoter sequences, I should say, are not highly active; their function is to be bound by the proteins of the transcription complex. What is more, base pair sequences in promoters are very often low in specificity: only a few base pairs are crucial to the promoter's function.

What Horwitz and Loeb were after may be understood by a linguistic analogy. Consider an English sentence: the cat sleeps on the mat. Suppose one wished to discover which words might be substituted for the word "cat" while still preserving either the sense of the original sentence or a sense close to the original. One method might be to combine English letters at random to form three-letter words (rat, tar, aim, tic, etc.), the words inserted in turn into the original sentence. Sentences in which meaning is preserved would be counted as successes. But such an experiment -- similar in fact to experiments used in linguistics -- would be a device for generating substitution classes, not a demonstration that meaning is preserved under arbitrary substitutions. Ditto the experiment reported by Horwitz and Loeb.

In addition to missing the point of the experiment, Mr. Wessel conveys an erroneous impression of its quantitative structure. "Of the approximately 10¹¹ possible sequences of this type," he writes, referring to the total number of possible sequence substitutions, "it turns out that many promoted the function of the deleted natural sequence. . . ." In fact, Horwitz and Loeb report that ". . . very roughly, 2 percent of the 3 x 10¹¹ possible recognition sequences present in this construction may duplicate promoter recognition site activity." This, of course, is a statistical estimate, one based on a small sample. In the case of plasmids deficient in adenine, the relevant number of successes drops to 0.2 percent. The vast majority of random sequences thus failed to duplicate promoter recognition site activity.

Citing with satisfaction the work of John Koza ("Genetic Programming: A Paradigm for Genetically Breeding Populations of Computer Programs to Solve Problems," Technical Report STAN-CS-90-1314, Department of Computer Science, Stanford University), Mr. Wessel would argue that genetic algorithms embody the abstract properties of robustness in the face of randomness that I claim could not be a feature of living systems. In fact, I made no such claim. As their name suggests, genetic algorithms are structures designed to incorporate (or mimic) certain biological operations. Typically, strings or sets of strings are introduced as fundamental data structures and manipulated by operators that reproduce the effects of random variation, genetic crossing, and natural selection. Work in this area was initiated in the 1970's by John Holland in Adaptation in Natural and Artificial Systems.

Mr. Wessel's claim that "[m]any of these evolved programs perform their optimizing tasks better than the best intentionally designed ones" is surely not incontrovertible; some computer groups argue that genetic algorithms do not outperform hill-climbing algorithms (Proceedings of the Second International Conference on Genetic Algorithms and Their Applications, ed. J.J. Grefenstette, 1987). But let that pass. I do not for a moment deny the possibility that a controlled random search might be an effective way in which to explore a large space. In referring to the Face Print algorithm in my essay, I said as much. What is at issue is the nature of the controls. The Face Print algorithm may be a fine method for prompting a crime victim's memory; it may, in fact, be far superior to traditional methods in which the victim offers a police artist a description of the malefactor (Uh, let me see, big nose, yes, it was a big nose, I think, but no, not that big . . .); but the algorithm fails to capture an essential feature of a Darwinian mechanism, for fitness is evaluated in terms of an ever-progressing match between what the algorithm produces and what the crime victim remembers, with the crime victim's memory functioning as -- once again -- a forbidden design or template.

The larger question posed by genetic algorithms is whether they can reach any interesting lifelike structures. Although genetic algorithms are new, they make use of an old mathematical concept, a Markov chain. It is worth noting that mathematical models based on Markov chains cannot in principle generate the sentences of a natural language. This is something known since the 1950's; I offer it as an observation.

In speaking of redundancy, Mr. Wessel has appeared to misunderstand the basic technical facts. There are two theories loitering in the background. One is Shannon's theory of information, which was developed in the 1940's (see A.I. Khinchin, Mathematical Foundations of Information Theory); the other is the Kolmogorov-Chaitin theory of algorithmic complexity, which was developed in the 1960's and 1970's (see G. Chaitin, "Information-theoretic Computational Complexity," in IEEE Transactions on Information Theory, IT-20, 1974).

Information theory makes use of concepts drawn from the classical theory of probability; the theory of algorithmic complexity does not. When Mr. Wessel speaks of "maximally compressed programs, in the sense of algorithmic information theory," he is incoherently attributing to one theory the concepts of another. Code compression is an information-theoretic concept, not a concept of algorithmic information theory at all. What information theory reveals is that in order to determine the best code compression for a given text, all that need be considered is the entropy of the text -- the information it contains -- and the number of its symbols. The maximally compressed codes are, indeed, rare, but only because they are maximally compressed. (The fact that they are rare follows from the fundamental Shannon-Macmillan theorem, but it follows trivially.) Contrary to what Mr. Wessel imagines, the genetic code is "optimal," a fact demonstrated by Cullmann and La-bouygues ("Le Code genetique, instantane et absolument optimal," Compte re