Monday, July 03, 2006

The Techno-Geek Speaks Greek

Discovery Institute co-founder and investment strategist George Gilder has written an article that will be appearing in the prominent conservative magazine National Review. The DI has the article available here, which is convenient for those of you who don’t have 5 subscriptions to the National Review like I do.

Would you believe that the article is terrible? In looking at reactions to Gilder’s previous articles, the most consistent criticism is that his writing is abstruse, incoherent, and filled with terminology that he either doesn’t understand or intentionally misuses (or worse, invents on his own). This piece continues that time-honored tradition.

First of all, very little of it has anything to do with evolution, whether by Darwinian means or any other. (He even spends several paragraphs plugging his own books, which have no clear relevance, but I guess the guy needs all the royalties he can get.) Staying true to the Discovery Institute’s tactics, he associates things with evolutionary biology that have little or no association at all, and in every case these just happen to be things that are disliked by right-wing ideologues such as George Gilder. People like him apparently need an all-purpose boogyman to make sense of the world, but it’s a poor substitute for genuine understanding. And in this case it has resulted in an article that consists mostly of disjointed ramblings with no coherent thesis. Secondly, Gilder has an bad habit of throwing in random quotes from noteworthy scientists, most of whom would probably have a very low opinion of George Gilder. In virtually no case do these quotes have any real relevance to whatever point, if there is one, that Gilder is trying to make. They appear to serve as the literary equivalent of name-dropping, lending a façade of authority to an otherwise nonsensical piece. And then there is Gilder’s favorite tactic, which is to wax profound about one scientific advance or another (with no indication that he knows what he’s talking about), and pretend as if this alone somehow constitutes an argument. There is just painfully little that rises up to the level of coherence.

Below the fold I will try to address the few claims that are on-topic and comprehensible enough to address. That’s not many, but it’s worth clearing a few things up.

Early on, Gilder repeats the old canard about natural selection being a tautology: “…at its root, Darwinian theory is tautological. What survives is fit; what is fit survives,” he writes. This is really bottom-of-the-barrel stuff, requiring a willful misunderstanding of evolutionary theory. Let’s clear this up: Natural selection is about the probability of an organism surviving and reproducing relative to the rest of the population. The theory requires that some features of living things are more conducive to survival and reproduction than are others; hence if these features are heritable, they will increase in frequency over successive generations. Since there is no a priori requirement that this be true of the world, it’s hardly a tautology, now is it? We could live in a world where all organisms, regardless of their traits, were equally likely to survive and reproduce. But a century of experiment and observation shows that this isn’t the case. In their famous work on Darwin’s Finches, Peter and Rosemary Grant found that a difference as small as 0.5 mm in beak size was enough to cause a measurable change in the likelihood of survival. Obviously, given that those features which improve survival can be detected empirically, Gilder’s blather about everything being equally good is nonsense.

Gilder also goes on and on about how we biologists are apparently unaware that genetic information can be understood as information rather than as chemistry. This seems to be a favorite hobby-horse of his, as he has repeated it several times elsewhere, claiming, quite amusingly, that biologists are going to be left behind in the new Information Age. This must come as a big surprise to those of us who use information technology on a regular basis, whether it be with sequence alignments, BLAST searches, or even phylogenetic trees. There is in fact a large and thriving field of biology known as bioinformatics that specifically focuses on using information theory to analyze the massive amount of sequence information that’s being produced. And like everything else in biology, evolutionary theory is a critical part of this field. But in spite of the fact that most research institutions have bioinformatics programs — and in some cases, whole departments — Gilder is apparently oblivious that this discipline even exists. Instead he lectures biologists about the need to incorporate this revolutionary new idea called information he’s just discovered. I hate to break it to you George, but you’re a half-century behind on this.

Gilder also makes a big deal about the idea that the information inherent in DNA is completely independent of the medium, seeming to base a large amount of his thesis (whatever it is) on this claim. Here are a couple of examples:

I came to see that the computer offers an insuperable obstacle to Darwinian materialism. In a computer, as information theory shows, the content is manifestly independent of its material substrate.
[…]
“Like a sheet of paper or a computer memory chip, DNA bears messages but its chemistry is irrelevant to its content.”

But anyone paying attention to molecular biology over the last 20 years knows that this is completely false. Many of the vital functions performed by nucleic acids are a direct result of their chemical properties. Consider, for example, the structure below:

This is a stem-loop, one of the most basic secondary structures that nucleic acids (in this case RNA) can adopt. These structures are necessary for the binding of proteins and the regulation of transcription, translation, and just about everything else that is important in how nucleic acids transmit their information. And while secondary and tertiary structures are determined by the primary sequence of DNA, and thus can be reduced to sequential information, they would not work if expressed in a different medium. In many cases, the information contained within DNA and RNA manifests itself functionally by means of the chemical properties inherent in the sequence itself — hence you cannot separate the medium from the message.

But even if Gilder weren’t totally wrong about this, it’s not even clear why it matters. Assuming the medium were completely independent of the information contained therein, what of it? What does that say about evolution and how does that make “intelligent design” any more plausible? As with everything else, Gilder doesn’t explain. He just hand-waves and pretends as if exuding profundities is a proper substitute for having a clue.

Gilder is at his most absurd when he attempts to discuss the Central Dogma of molecular biology, which postulates that information flows from DNA to RNA to proteins. Gilder is under the bizarre misapprehension that this means that DNA must have preceded protein during the origin of life. Of course this is not the case; the Central Dogma applies only to modern organisms and says nothing about alternative systems that may have existed prior to the last common ancestor. But assuming DNA did come before proteins, how is this relevant? Gilder doesn’t explain. And he seems completely unaware of the RNA World hypothesis, which holds that RNA containing both catalytic and information storage properties preceded both DNA and protein. But given that Gilder seems amazed by his recent discovery that DNA contains information, we’ll just have to take things one step at a time.

Having a hard time understanding what the Central Dogma actually means, Gilder goes on to say some pretty silly things about it. This is one of my favorites:

Over at NASA, U.S. government scientists make an analogous mistake in constantly searching for traces of protein as evidence of life on distant planets. Without a hierarchy of informative programming, proteins are mere matter, impotent to produce life. The Central Dogma dooms the NASA pursuit of proteins on the planets to be what we might call a “wild goo chase.”

Is he serious? Obviously, NASA wants its probes to detect chemicals that correlate with living things. It doesn’t much matter which chemicals they look for just so long as they’re easy to detect and can be reasonably assured to indicate the presence of life. The fact that they look for protein doesn’t mean that they don’t think space bugs would have nucleic acids as well. The Central Dogma isn’t the least bit relevant here. (By the way: Looking on the web, I can find no NASA program that looks for the presence of proteins on “distant planets”, or even on not-so-distant planets. What Gilder is referring to is anyone’s guess.)

But it gets worse. Combining his poor understanding of the Central Dogma with a equally poor knowledge of the history of science, Gilder says things that are, to be as polite as possible, weird:

Throughout the 20th century and on into the 21st, many scientists and politicians have followed Darwin in missing the significance of the “Central Dogma.” They have assumed that life is dominated by local chemistry rather than by abstract informative codes. Upholding the inheritability of acquired characteristics, Jean-Baptiste Lamarck, Trofim Lysenko, Aleksandr Oparin, Friedrich Engels, and Josef Stalin all espoused the primacy of proteins and thus of the environment over the genetic endowment.

It is truly a challenge to pack so much ignorance into such a small space, but Gilder is a techno-guru and capable of marvelous feats. Where to begin here? First of all, Darwin couldn’t have known anything about the Central Dogma of molecular biology because he lived a full century before it was worked out. The same is true of Lamarck, who lived in the 18th and early 19th centuries (Gilder is off by 200 years here). And neither one could have known much if anything about proteins, and certainly nothing about DNA.

The inheritance of acquired characteristics has a long history, but suffice it to say that Larmarck is usually credited with having come up with the idea in pre-Darwin times. In present-day textbooks, Lamarckian evolution is often contrasted with Darwinian evolution as a means of explaining how natural selection is very much different from Lamarck’s view. (Lamarck really gets a bad rep for this; the inheritance of acquired characteristics was a widely held view in his time, not unique to Lamarck at all, but it has sadly overshadowed his truly valuable contributions to evolutionary thought.) When Darwin came along, his biggest problem was that he couldn’t explain heredity. His theory required that heredity existed, which everyone accepted as true, but he couldn’t explain how it worked, and this remained a major thorn in his side. He incorporated so-called Lamarckian ideas into his work, but he never quite cracked the nut of heredity. Then in the early 20th century, Mendel’s laws of heredity were rediscovered and the modern science of genetics was born. It was precisely the melding of genetics with evolutionary theory that comprised the neo-Darwinian synthesis. The Central Dogma of molecular biology, by showing that heredity creates phenotype, and not the other way around, makes this even more explicit. How badly confused must Gilder be to think that this constitutes a valid argument against Darwinian evolution? He is apparently ignorant of the fact biologists do not and have not taken the theory of acquired characteristics seriously for nearly a century. The only major exception was T.D. Lysenko and his followers in the Soviet Union, who accepted a Lamarckian version of genetics — a thoroughly anti-Darwinian view — out of a devotion to ideology. It just goes to show that attacking science for ideological reasons, as Gilder does, is doomed to failure. But of course the irony is lost on him.

And finally, somewhere in the last half of the article, by which point the piece has descended into complete gobblety-gook, we finally get introduced to intelligent design. You know, the so-called theory which Gilder previously admitted has no content. For Gilder to know nothing about biology is understandable, but one would at least think that he kept abreast of the pseudo-theory that is championed by the institute that he co-founded. That’s not quite the case:

But intelligent design is merely a way of asserting a hierarchical cosmos. The writings of the leading exponents of the concept, such as the formidably learned Stephen Meyer and William Dembski (both of the Discovery Institute), steer clear of any assumption that the intelligence manifestly present in the universe is necessarily supernatural.

Um, no. Dembski at least has argued on many occasions that the “intelligence” responsible for whatever the heck it’s supposed to be responsible for must be supernatural. The Discovery Institute’s creationist wing mentions prominently that they are at war with naturalism, and says quite directly that science should incorporate that which is beyond the natural — in other words, the supernatural. That is their entire purpose for existing and wasting everyone’s time. And Gilder, with his own railings against “materialism”, is effectively saying the same thing. It’s pretty dishonest of him to pretend otherwise.

Beyond that there is little left in the article that isn’t muddled and impertinent. Gilder’s habit of listing break-through scientific theories seems to have little purpose other than to awe his readers by showing them that he’s heard of things like quantum mechanics. The National Review is perhaps the best known conservative magazine in America, and its readers deserve better than this. The Discovery Institute, on the other hand, deserves what it gets.

(Cross-posted to the Panda's Thumb.)