subscribe to our mailing list:
Has Darwin met his match in Berlinski?
Posted December 31, 2002
Alexander Eterman: Extended commentary
Paul R. Gross: Berlinski vanquishes pseudoscience -- and science (again)
Mark Perakh: Are the Darwinian theory of evolution and intelligent design "theory" indeed equally unsubstantiated?
Jason Rosenhouse: No contribution to biology from Berlinski
Matt Young: Are Intelligent Designauts Functionally Illiterate?
By Alexander Eterman
December 31, 2002
In an article published in a 2002 issue of Commentary, David Berlinski sets out to analyze from every possible angle both the controversial "evolution vs. design" pair, and the more general opposition between "natural Universe and a universal tuning". Such an approach literally begs an extended commentary, quite impossible in the sparse framework of the reviewing realm. I am forced, therefore, to confine my reaction to an outline form.
- A logical discussion of the origins of life on earth can in no way refute the design argument. No formal reasoning is capable of disproving such theories as, for example, that life on our planet was deliberately created by, say, a highly advanced alien civilization, i.e. a designer, and not necessarily a supernatural one. How fruitful this argument is, and what empirical vistas it reveals, is another question.
- Those who take the design argument seriously have their work cut out for them. They must decide whether the designer monitored his creation for a long time - possibly to this day - intervening in its workings, or whether he detached himself immediately upon the "launch". If the presumed designer has long detached himself from his creation, it is crucial to determine whether he left it to its own devices, like a pet set free, or provided it with an in-built long-playing program that predetermined its operations, the way we leave the computer on over the weekend, having programmed it to perform certain calculations.
- A design theoretician, before he pits his strength against Darwin, should express his own narrative in an intelligible manner. Thus he should specify whether the evolution of biological species (regardless of its exact mechanism) took place at all, or whether the different species were produced by the designer gradually and independently of one another, or, perhaps, they all appeared simultaneously and went on in peaceful coexistence, until some of them gradually became extinct. If he accepts de facto an evolutionary process of any kind, he should decide whether blind natural selection is capable of creating anything new and sensible, or, on the contrary, that any biological change of slightest complexity is the outcome of enlightened design. Thus the proponent of the design theory will most likely reject the suggestion that the mammalian eye is the product of natural selection; yet he might concede that the hearing apparatus of mammals evolved in a natural fashion, with no outside help; or at least that the Indian and African elephants, so similar and yet so different, are two naturally divergent branches of the same tree, rather than the product of a sophisticated design.
- In a word, a hypothesis of biological design must be built not merely as a negation of the theory of natural selection, but rather as a series of diachronically expanded theoretical models of design, consistent in all of their links. It should be noted that building such a hypothesis is a thankless task, in the course of which a good half of "designers" empirically go over to the evolutionist side. In an aside, I would like to add that in and of itself, the design idea is logically consistent and almost harmless. In its more innocuous versions, however, it is largely devoid of substance and has no predictive power; and when it attempts to become a predictive theory, it runs into a tangle of knotty empirical problems. A rather complicated picture, though one that is far from unprecedented in the history of science.
- Of a totally different nature is the "argument from mathematical physics", simply known as the notorious anthropic principle, which argues that the laws of nature are designed in such a way as to bring certain complex phenomena into being, including the appearance of man. This is a classical scholastic hypothesis that would have done better had it never been born. Suffice it to point out that the validity of the "anthropic argument" cannot be empirically proven under any circumstances. Indeed, in order to examine the issue of the origins of known physical laws with an unbiased eye, it is crucial to have a thorough knowledge of the alternatives, in other words, by definition, to become as gods. How can a mere mortal, even one of Einstein's caliber, possibly know whether the presumed designer had the option of changing the universal law of gravity so that the force of gravity became proportionate to the cubic distance between masses? Another question: would Pythagoras' theorem thereby become invalid, and mathematics in general transformed beyond recognition? In fact, this state of affairs would make all surmises possible, all equally feasible -- and equally meaningless
- The anthropic principle is a seeming rather than a real argument, an argument from anthropomorphism. In its correct form, it asserts the following: let us assume that it is possible to create random physical versions of the world, yet in such a way that the mathematics would remain constant in every instance. In most of the versions, the emergence of man would be impossible; therefore the random creation of a world with man present is unlikely, and consequently has to be an outcome of design. This argument naively presupposes the feasibility of appealing to the theory of probability -- the only purely empirical mathematical theory - outside the boundaries of the empirical world, without building a model that incorporates the co-called probability space. What is more, this argument permits a logical discussion of phenomena that lie beyond the boundaries of logic!
- Mental experiments with a designer situated within the familiar realm of cause and effect yet outside nature, and discussing the latter, go back to ancient Greek philosophy and share most of its flaws. In days of old, when space and time were held to be absolute, creationist arguments included the claim that, if the world had existed forever, the total number of people that lived in all of time would have been infinite! The resemblance is uncanny, is it not?
- I would challenge the proponents of the anthropic principle to solve the following simple problem. Supposing we have before us a die, without any way of knowing whether it is genuine or fake. We are forbidden to subject it to any tests. Given the circumstances, can we venture any opinions as to the odds of getting three sixes out of ten throws? Definitely not. Yet if that is the case, how can we make any claims about the odds of an alternative version of the physical universe? After all, we know nothing whatsoever about the hypothetical rules of the game. It follows, therefore, that the theory of universal design is nothing but a fallacy that discredits the scientific integrity of its proponents.
And finally, for those who take issue with the theory of natural selection, my strong advice would be to leave Darwin alone. Combing through his works, it is no great feat to come up with all sorts of passages that are often contradictory or inconsistent. This is hardly of any importance. The name of Darwin, the discoverer of the theory of evolution, presently serves as merely its synonym. It would be ludicrous to look for answers to all the questions of modern biology in The Origin of Species. Darwin cannot be responsible for the claims of modern-day evolutionists; then again, they bear no responsibility for Darwin's own -- quite legitimate, in fact -- scientific doubts. There is a right time for everything.
By Paul R. Gross
December 16, 2002
Commentary's explorations of "Darwinism," as it is often called by those who don't know much evolutionary biology, arrive as essays of David Berlinski, whose latest is in the December, 2002 issue. The choice of subject is unremarkable. Evolution is the core of modern biological science and its ideas are increasingly a part of astrophysics, cosmology, and computer science. A distinguished intellectual journal should address it-expertly. After a century and a half, the science still means-for some but not all religious people-highly offensive notions on the history of life. What is remarkable is that Commentary, most of whose contributors are recognized experts in their subjects, should choose for this important scientific discipline one who contributes nothing to its professional literature.
Nevertheless congratulations are in order. He seems in this new essay to have reversed himself on the merits of the current effort to rehabilitate the ancient argument from design, now holiday-wrapped as "intelligent design theory." This change of mind is all the more praiseworthy because Berlinski is closely associated with the conservative Christian think tank that serves as primary promoter of "intelligent design." He is moreover the author of a flattering blurb for the book of another leading member of that organization. This is William Dembski's The Design Inference, whose argument and conclusions Berlinski here finds to be unsound.
The manner of dismissal, however, is less creditable. I take the liberty of paraphrasing Berlinski's argument: "'Intelligent design' proposals to rehabilitate Paley have failed, for good reasons, some of which I [Berlinski] give here. Evolutionary biology has failed, for good reasons which I also give here. Therefore nobody has a plausible scientific explanation for the diversity of life on Earth. This I, [Berlinski] unlike almost all the professional evolutionists and philosophers of science in the world, have now understood."
This seems to please Commentary's editors; but of course it is absurd. The reasons Berlinski gives for the failure of intelligent design, despite its massive public relations campaign to convince the laity that "Darwinism" is finished, have all been given before. The flaws of intelligent design theory have been explained, published widely in the professional literature, offered in the first biology courses of all decent colleges, in half a dozen very recent specialist books, on fifty flourishing scientific web sites. So, however well it is doing as politics, there is noting new about the abject failure to date of intelligent design apologetics as science.
On the other hand, none of the reasons Berlinski hints at for rejecting "Darwinism" stand up. They are the familiar creationist pabulum. All those supposed grand flaws and gaps in evolutionary biology, too, have been addressed in the professional literature, in tens of thousands of papers and also in dozens of excellent, best-selling popular science books. All those arguments, even when they really are arguments and not mere aspersions, have been soundly refuted. This includes, specifically, the published arguments of Berlinski's colleagues, the "design theorists," and Berlinski's own aspersions. It's all out there.
Unfortunately, a non-biologist reader of "Has Darwin Met His Match?," innocent of this vast body of knowledge, will have no notion of its content or even, perhaps, of its existence, and may therefore take Berlinski's assertions as true as well as deep, which they are not; as happy news that long-dead Darwin has met his match. Of course, this does accommodate the normal suspicion of science and all the normal discomforts with its implications for religion.
To deal fairly with Berlinski's aspersions would once again, as was the case for his earlier Commentary writings on evolution, need the same column-inches as his. That, I think, Commentary will not permit. But there is no reason why an honestly curious reader can't fully check my assertions here in about an hour, with a good search engine running on the internet. The data are all there, including those for such issues as the Cambrian "explosion," with commentary from the distinguished paleontologists who discovered the facts in the first place, and who have records of interpreting such facts successfully.
It is not that there are no problems in evolutionary biology. There are plenty. Not quite as many of course as in, say, high-energy physics, or, to go to the other end of the physical science spectrum, in weather forecasting. It is simply that Berlinski's problems are not problems; and the real problems do not lead where Berlinski's aspersions lead. The argument from design is dead: it was dead long before Paley. But "Darwinism" is very much alive and is about as sound, on evidentiary grounds, as the germ theory of disease. (Yes, I know: even there we have "therapeutic touch" and "quantum healing" flourishing politically. Facts can't beat politics or a good agent.)
This much, anyway, can be noted here, as example: Berlinski says that the (Paley's) argument from design has been taken up again in evolutionary biology and mathematical physics. This, if true, would give it some scientific seriousness. But: there is not one publication in recent biological journals, out of the tens of thousands of articles published annually, a huge subset of them devoted specifically to evolution, that undertakes to rehabilitate the argument from design. None of the intelligent design "theorists" mentioned in his essay has ever published the claim in a refereed, original article in a regular biological journal. Nor, of course, has Berlinski himself done so.
I do not know what Berlinski means by "mathematical physics," but if he is referring to Dembski's lucubrations on chance and probability-which Berlinski himself now finds flawed-then I know of no professional publications (other than Dembski's own books and his frenetic responses to critics) that treat them as of interest for mathematical physics. There are however a dozen point-by-point refutations of those claims by well-known physicists, philosophers, mathematicians, and biologists, available even on the internet. If Berlinski is, perhaps, referring to the so-called "anthropic principle," then a little search will take the reader to its primary sources, a few theoretical physicists, and to the recognition that it not only has nothing to do with the argument from design, but is hostile to its claim and implications.
By Mark Perakh
December 16, 2002
This country has the largest number of Nobel laureates, including physicists, chemists, and biologists, in the world. Strangely, it also has perhaps the largest percentage in the world of gullible people believing in astrology, the Bible code, communication with dead and all kinds of crank science. Crank science takes many forms but its unavoidable feature, which makes it recognizable through all of its often ingenious disguises, is the absence of evidence. One of the modern and very active versions of crank science is the so called Intelligent Design (ID) movement.
Commentary, although not a scientific publication, is a serious journal whose authors are usually experts in their fields. Therefore publication of an article by the mathematician David Berlinski (Has Darwin met his match? Commentary, New York, Dec 2002) which is the second such occasion (his previous opus on the same subject appeared in 1996) seems somehow surprising. Berlinski is far from being an expert in biology in general or the theory of evolution in particular. He has never conducted biological research, never published any material of biological character in any of the scientific journals. No wonder he so highly praises the opuses of another amateur in biology, lawyer Phillip Johnson whose attack on Darwinism was shown to display a lack of understanding of the subject he so brazenly set out to discuss and who has a penchant to misquote and to distort the views of his opponents (see, for example, the critique of Johnson in the Critique of Intelligent Design section on this site).
Berlinski's paper has other surprises as well. Perhaps the most amusing is his apparent change of mind compared with his earlier utterances. Berlinski supplied rave blurbs to the books by the prominent advocates of ID, William Dembski and Michael Behe. In this new paper, however, unexpectedly Berlinski casts doubts on the plausibility of the ID concepts so vigorously promoted by Dembski and Behe.
Berlinski's main thesis, expressed succinctly, is that both Darwinism and the ID alternative to it are equally unsubstantiated. If we formulate Berlinski thesis that way, perhaps the reasons for his new approach can be understood. So far, despite all the persistent efforts by the ID advocates, although they have been often successful politically, they have failed to get a foothold in science. A search through the scientific literature reveals absence of any discussion of the ID theory. In particular, though Dembski, Nelson, Wells, Meyer, all listed by Berlinski as having "respectable training," have indeed earned legitimate scientific degrees from respectable schools, they have not been publishing any scientific data which would have resulted from research in biology or other fields of science. Behe seems to be an exception as he authored a number of papers in biochemistry. His widely read book titled Darwin's Black Box where he introduced the concept of Irreducible Complexity (IC) has however no relation to his biochemical research. Dembski, who is a very prolific writer on ID, has apparently only one mathematical paper (on randomness) to his credit, published many years ago. Wells, by his own admission, set out to study biology on the advice of the Reverend Moon, with an explicit goal set in advance - to destroy Darwinism, hardly a proper motivation for an unbiased scientist.
Although the ID advocates rant incessantly about the supposed acceptance of their concepts by mainstream science, in fact almost all their publications appear only in their own outlets. They are anxious to be treated as equals in science. Asserting the equal status of Darwinism and ID, as Berlinski does in his paper, even if only by casting doubts on both equally, may seem to be a step forward from their standpoint.
Here are examples of specific points in Berlinski's paper which call for a rebuttal.
- Contrary to Berlinski's assertion, the fossil record, as incomplete as it is, admirably supports the tenets of Darwinism, all protestations by ID advocates notwithstanding.
- Contrary to Berlinski's assertion, the Cambrian "explosion" is not a major stumbling block for the evolution theory. Plausible explanations have been offered by biologists while the ID advocates have not provided reasonable arguments against these explanations.
- Contrary to Berlinski's assertion, biologists are not "standing in silence" regarding the explanations of "why the peacock tail?" and all other "whys" Berlinski suggests as allegedly insurmountable riddles for the evolution theory. In fact, biologists have suggested Darwinian answers to those "whys" long before Berlinski asked his "why" questions. He seems to be either unfamiliar with the pertinent literature or possibly suffering, using his own words, from an "inability to read the literature."
- A similar remark seems to be in order with regard to Berlinski's discussion of Nilsson-Pelger's (btw, not Pilger as Berlinski consistently misspells it) paper wherein the time necessary for the Darwinian evolution of a mammalian eye was estimated. Although in a footnote Berlinski accuses the physicist Matt Young in "an inability to read the literature," it looks like such an accusation can be rather properly made in regard to Berlinski's own misreading of Nilsson-Pelger's fine paper wherein he even did not notice the proper spelling of the author's name, not to mention his misinterpretation of Nilsson/Pelger's methodology (to which Young referred correctly).
- Berlinski's references to the anthropic principle reveal his insufficient familiarity with the literature on that subject (including such recent publications as those by Vic Stenger, Bill Jefferys and Michael Ikeda, and others: see the Anthropic Principle section on this site).
- A similar remark is in order regarding his discussion of chance and necessity. While his critique of Dembski's treatment of probabilities is often (although, in my view, not always) reasonable, it mostly repeats the critical comments offered before by other authors. Many articles containing such an earlier critique of Dembski's treatment of probabilities can be seen on this site.
Contrary to Berlinski's assertion, it is ID advocates like Behe and Dembski rather than scientists who have been obsessed with Behe's mousetrap example. A few scientists have made fun of that example by showing that Behe's mousetrap can be developed in small steps following an evolutionary mechanism. Contrary to Behe's insistence, his mousetrap was shown to be not "irreducibly complex," thus debunking Behe's and Dembski's claims based on that example.
Contrary to Berlinski's assertion, there is no penetration of ID into such areas as mathematical physics. In fact, all publications of the ID advocates are of "philosophical" type, lacking scientific data in support of their concepts. The most mathematically loaded material in the ID literature is in the books by Dembski, wherein he attempts (in my view, unsuccessfully - see the articles on Dembski in the Unintelligent Design section of this site) to apply concepts of information, probability, and complexity theories to substantiate the ID hypothesis, but nowhere it even remotely relates to mathematical physics.
The list of unsubstantiated assertions in Berlinski's paper can be easily extended. Overall, Berlinski's enviably eloquent paper offers almost nothing new or original, is largely misleading in his presentation of the situation in the biological science, and adds nothing of substance to the controversy in question beyond an effort aimed at putting the failed ID hypothesis on equal footing with genuine science.
My thanks to Randy Bennett, Pete Dunkelberg, Russell K. Durbin, Bill Jeffferys, Paul R. Gross, Jim Swan, and Matt Young for useful comments on the initial draft of this article.
By Jason Rosenhouse
December 16, 2002
It's nice to see David Berlinksi aim his rhetorical guns at a target more deserving than modern biology and cosmology ("Has Darwin Met his Match," Commentary, December 2002). The arguments he makes against intelligent-design theory (ID) suffer only from a lack of originality; critics of ID have been making the same points for years. In emphasizing the logical deficiencies and practical limitations of ID, Berlinski has given a good picture of why most scientists find it unpromising.
But his ongoing antipathy toward mainstream evolutionary biology is based on a caricature of the subject. For example, his statement that Darwinism "makes use of a fantastic extrapolation" through which "the development of antibiotic resistance in bacteria becomes the model for the development of such complex structures as the mammalian eye, or the immuno-deficiency system" is mistaken. Field studies of microevolution are certainly suggestive and important, but the evidence that natural selection is the primary shaper of evolution comes from many sources.
If we are constrained only by our imaginations it is not difficult to conceive of complex systems inaccessible to natural selection. Since selection manufactures complexity out of small alterations in simpler precursors, we conclude that multi-part systems formed from wholly unique parts require some other explanation. So it is telling that every complex biological system studied in detail gives the appearance of a Rube Goldberg machine. This was noted already by Darwin, whose work on orchids revealed that their wonderful insect-attracting contrivances were not pristine creations from nothing, but rather were cobbled together from simpler parts commonly found in flowers. The manifold proteins of the human blood clotting cascade have precisely the structure they should have if they were formed by familiar processes such as gene duplication and exon shuffling. And the evolution of the immune system, far from being the unfathomable mystery Berlinski describes, is actually the subject of a recent book (Origin and Evolution of the Vertebrate Immune System, L. Du Pasquier and G. W. Litman (eds.)).
Further evidence comes from animal behavior. Ethologists routinely use game theoretic techniques in their work. These are mathematical models based explicitly on the assumption that these behaviors were crafted by natural selection; the success of the models is evidence for the correctness of that assumption.
Berlinksi's other jabs also fall flat. The Cambrian explosion is a problem for evolutionists only in the sense that there are many possible explanations, but too little data for deciding between them. Blubbering about gaps in the fossil record can't change the fact that with millions of fossils collected and classified not one is out of place from a Darwinian standpoint. Also, Darwinism requires gradual change only at the level of the genotype, not the phenotype. And while it is true that information is independent of the medium containing it, there is nothing mysterious about the idea that changing the medium can alter the information. Genes do mutate, thereby coding for different proteins from before, and new functions are sometimes observed to arise as a result. The source of this marvelous code is indeed mysterious, a fact that would be troubling if Darwinism were a theory about the origins of life. Since it is not, Berlinski's hand-wringing on the subject is inappropriate.
The fruits of evolution are disseminated in tens of thousands of research articles in dozens of journals every year. The people charged with the responsibility of entering their labs and solving problems obviously find evolutionary theory useful. Numerous complex systems have been studied and the major steps of their evolution revealed. Where data is copious it is all in accord with Darwinian expectations; where mysteries remain the problem is a lack of data, not a lack of theoretical robustness. In response to all this, ID theorists fold their arms and shake their heads. That is their right. But for all their bloated claims and hyperbolic rhetoric, the ID community has made no contribution towards solving an actual biological problem. For that matter, neither has David Berlinksi.
By Matt Young
Department of Physics
Colorado School of Mines
December 16, 2002
Matt Young's Home Page
Someone named David Berlinski wrote in the December, 2002, issue of the magazine Commentary that, along with a lot of biologists, I am spectacularly unable to read the scientific literature (David Berlinski, "Has Darwin Met His Match?" Commentary, December, 2002, pp. 31-41). Specifically, Berlinski refers to a paper that details a simulation of the evolution of an eye and wrongly claims that the paper has been widely misread as using "nothing more than random variation and natural selection" (Dan-E. Nilsson and Susanne Pelger, "A Pessimistic Estimate of the Time Required for an Eye to Evolve," Proceedings of the Royal Society of London, Vol. 256, pp. 53-58, 1994). Berlinski writes in a footnote (p. 34),
The physicist Matt Young offers this inadvertently rich account of his own ability to read the literature: "Creationists used to argue ... that there was not enough time for an eye to develop. A computer simulation by Dan Nilsson and Susanne Pilger [sic] gave the lie to that claim: Nilsson and Pilger [sic] estimated conservatively that 500,000 years was enough time."
I plead guilty; except for the incorrect spelling, that is exactly what I wrote in my article published on the web page of the Pacific Coast Theological Society ("How to Evolve Specified Complexity by Natural Means", 2002). If anyone fails to understand the article, however, it is Berlinski, not I. Far from being able to read and understand the article, Berlinski seems not even to have noticed its title: "A Pessimistic Estimate of the Time Required for an Eye to Evolve" (my italics). The thrust of the paper was by no means to show that you can develop an eye in incremental steps (that is obvious), but rather to estimate how long it might take. The conclusion of the paper (p. 58) was stated clearly and concisely:
If we assume a generation time of one year, which is common for small and medium-sized aquatic animals, it would take less than 364 000 years for a camera eye to evolve from a light-sensitive patch.
Only someone who did not understand what he was reading or wanted to be deliberately deceptive could have overlooked that conclusion.
the number of apparently dyslexic biologists, I suppose I should be flattered
that Berlinski liked my writing so well that he singled me out for
criticism. Unfortunately, it is
Berlinski, not I nor the biologists, who appears functionally illiterate.
I wrote that Nilsson and Pelger (not Pilger, by the way) had performed a computer simulation of the development of the eye. I did not write that they used nothing more than random variation and natural selection, and I know of no reference that says they did.
Berlinski's carping that they described an eyeball, not an eye, is typical of those who tilt at neo-Darwinism. If Nilsson and Pelger had simulated the development of the light-sensitive patch with which they started, Berlinski would have asked where the original cells came from, and so on back to primordial ooze. Criticizing what we know about biology by harping on what we do not know is like criticizing a sturdy, durable, and functional concrete wall because there are chinks in it.
Nilsson and Pelger's is a sophisticated simulation that even includes quantum noise; it is not a back-of-the-envelope calculation. Their simulation begins with a flat, light-sensitive patch. They allow the patch to become concave
in increments of 1% and calculate the visual acuity. When some other mechanism will improve acuity faster, they allow, in various stages, the formation of a graded-index lens and an iris, and then optimize the focus. Unless Nilsson and Pelger performed the calculations in closed form or by hand, theirs was a computer simulation. Computer-aided simulation might have been a slightly better descriptor, but not enough to justify Berlinski's gratuitous sarcasm.
More importantly, had Berlinski understood what he supposedly read, he would have recognized that Nilsson and Pelger's accomplishment was not to "flog" a collection of cells up an adaptive peak, "a point never at issue because never in doubt." Rather, they showed that the required time was geologically short -- a point very much at issue among Berlinski and his neocreationist colleagues.
A recent article in Science has examined the evolution of a complex organ in fish and provided empirical
support for Nilsson and Pelger's time scale. (See news article, "Placentas May Nourish Complexity Studies," November 1, 2002, p. 945, and the accompanying technical article on pp.1018-1020; the authors studied the placentas of fish because they appear to display the intermediate stages of evolution, whereas the intermediate stages of eyes have been lost.)
I stand by my statement that the simulation by Nilsson and Pelger gives the lie to creationists' claims that an eye could not have developed within the available time.
A shorter version of Matt Young's note was submitted as a letter to the editor of Commentary
on December 9, 2002. I thank Erik for critically reading the letter and for alerting me to the Science article. Copyright © 2002 by Matt Young.
Matt Young is a professor of physics at the Colorado School of Mines, an associate at the National Institute of Standards and Technology, and the author of No Sense of Obligation: Science and Religion
in an Impersonal Universe,1stBooks Library, Bloomington, Indiana, 2001.