Thursday, January 26, 2006

An Alito Filibuster?

So, it looks like John Kerry might be leading the charge to filibuster Alito. Good for him. Alito is a creep. The whole "unitary executive" spiel freaks me out, but there's more to it than that.

A few months ago, I wrote that the discussion of Harriet Miers was bothering me because the main argument in favor of her was "She MIGHT succeed despite her complete lack of credentials." No, that's not the point. We shouldn't have to guess whether a nominee will make a good Supreme Court candidate or not. The point of having hearings is so that Congress can ask questions and the nominee can answer them.

I watched some of the Alito hearings and they pissed me off. I am sick of this stupid game that all the Bush nominees have played -- a game called "Guess how to make me talk." Whenever Congress asks a direct question like "How would you be likely to rule on this issue?" we get one of these canned responses:

  1. "I can't answer that because it might reveal what I think before I actually rule on it."
  2. "That's already been ruled on, so it's been settled and I can't answer it."
  3. "You can't read what I said I think as an indication of what I actually think."

You know how it goes...

Senator: "Judge Alito, you once wrote that you personally feel that abortion is the world's greatest evil and should be destroyed with extreme prejudice the moment we have the chance to do God's bidding."
Alito: "Well, first of all, I was only writing what my former bosses told me to write. And second of all, even if I ever did feel that way, it was a long time ago."
Senator: "Okay. How do you feel about it now?"
Alito: "Stare decisis. My opinion isn't relevant."

Can you imagine any other job interview where a candidate gets to act like this?

Interviewer: "How would you handle this hypothetical situation if we hired you for this position?"
Candidate: "I'm sorry, but I can't answer that question because it might have some relevance to how I would do this job."

For all the interviewer knows, this person MIGHT be supremely competent at handling the job. But there's no point in waiting to find out, because the candidate is an asshole, and the interviewer is completely justified in showing him the door.

And that's what I wish Congress would do. And keep doing it. And send a message to the president: "I'm sorry, but this idiot will not talk to us. If you're serious about getting a nominee approved, please send us someone who will answer the questions."

Alito should damn well SAY: "No, I changed my mind about this unitary executive stuff. I don't believe it, and if the president overstepped his boundaries, I would rule against him." And if he won't say it, after being pointedly asked, then there's no reason not to assume the worst.

If Bush and Co. want an anti-abortion candidate badly enough, can't they at least be honest about it? Let them send a guy who will say "I disapprove of Roe v Wade and I mean to do all I can to undermine it, if not overturn it." If they want somebody who will assign the absolute power of a monarch to the president, let them send a guy who will say that.

He'll lose, of course. But he deserves to lose if he can't get 60 votes. As I said about Miers, the vast majority of people have NO business being on the highest court in the land, and the nominee gets the burden of proof on whether he is one of those uniquely qualified individuals.

Unlike Harriet Miers, Judge Alito is not a clueless groupie. He knows the law. But that's not enough. I want the guy who will serve on the Supreme Court for most of the rest of my life to have at least two qualifications. I want him to be competent, AND not evil. One or the other doesn't cut it.

I don't think that's a high bar to set. Nor do I think it's unreasonable to expect the candidate to openly and honestly demonstrate that he meets those expectations.

Thursday, January 19, 2006

Classes start tomorrow

In case anyone is interested, I'm beginning classes towards my Master's degree tomorrow at the university of Texas. Check out my program at the Center for Lifelong Engineering.

My classes will meet once a month, all day Friday and Saturday. On both days, my schedule will be:

8am - 12pm
Data Mining EE 380L
Basic concepts of data mining, in parallel with a practical track involving hands-on experience with industrial strength software and a term project will be covered.
Instructor: Joydeep Ghosh

1pm - 5pm
Verification and Validation EE 382C.3
This course covers various traditional and state-of-the-art techniques for software validation, a process that includes reasoning about (the correctness of) programs and testing programs. The course content will include both techniques for dynamic analysis, such as glass-box and black-box testing, equivalence partitioning, test strategy and automation, regression testing and debugging, and techniques for static analysis, such as symbolic execution, and also techniques for software model checking including those that employ artificial intelligence based heuristics.
Instructor: Sarfraz Khurshid

Tuesday, January 10, 2006

Review of "Darwin's Black Box" by Michael Behe

Ten years after the publication of Darwin's Black Box, I finally took it upon myself to read the thing. I felt that I was pretty familiar with author Michael J. Behe already, having followed the intelligent design movement for years, read his arguments online, read detailed criticism on other sites, and listened to his testimony in front of the Texas State Board of Education in 2003.

Still, I'd never picked up his book until now. There's this Old Earth Creationist / Intelligent Designist on the message board I frequent, and he's always taunted me as being closed minded for not reading Behe. So I read the book, not so much because I thought I would actually get any new information, but because I wanted to stop having that be a part of our arguments.

Here were my preconceptions about the book. As I understood it, Behe is probably the smartest person in the Intelligent Design (ID) movement. He is a real, actual biochemist. He has published peer-reviewed scientific papers, although not on the subject of ID. He knows science and scientific language. He also, not coincidentally, differs hugely from self-professed creationists in the sense that he accepts evolution almost in its entirety. He believes, or at least doesn't contradict, that macro-evolution occurs and that the earth is billions of years old.

Although Behe thinks that evolution is reasonable, he disagrees with mainstream biologists in the sense that he denies that evolutionary processes alone -- random mutation combined with natural selection -- are enough to account for all the diversity of life on earth. He believes that certain biochemical systems exhibit what he refers to as "irreducible complexity". Irreducibly complex things cannot have evolved, proposes Behe, and that leaves the alternative that they were "designed". Behe picks several systems as examples of irreducible complexity, which should be well-known by most who have followed the ID political movement: blood clotting, the cilium, etc.

All this is what I knew before reading the book. Early reading bore me out in these impressions. In Behe's own words:

"For the record, I have no reason to doubt that the universe is the billions of years old that physicists say it is. Further, I find the idea of common descent (that all organisms share a common ancestor) fairly convincing, and have no particular reason to doubt it. I greatly respect the work of my colleagues who study the development and behavior of organisms within an evolutionary framework, and I think that evolutionary biologists have contributed enormously to our understanding of the world." (pg 5)

So as far as Behe's concerned, it's fine to think that everything is descended from a common ancestor; so humans do indeed share genes with chimpanzees, with apes, with all mammals, with all vertebrates, and so on.

Furthermore, evolution operates just fine on the macroscopic level, for most things. Here is Behe on the evolution of the famous Darwinian bugaboo, the eye:

"Somehow, for evolution to be believable, Darwin had to convince the public that complex organs could be formed in a step-by-step process.

He succeeded brilliantly. Cleverly, Darwin didn't try to discover a real pathway that evolution might have used to make the eye. Rather, he pointed to modern animals with different kinds of eyes (ranging from the simple to the complex) and suggested that the evolution of the human eye might have involved similar organs as intermediates.

[Behe then recaps Darwin's example.]

Using reasoning like this, Darwin convinced many of his readers that an evolutionary pathway leads from the simplest light-sensitive spot to the sophisticated camera-eye of man. But the question of how vision began remained unanswered." (pg 16)

After this, Behe gives a fairly technical biochemical explanation of vision in terms of photons and proteins and such. His point as I understand it is this: on a macroscopic scale, sure, evolution can account for the construction of complex machinery from chemical parts. But the existence of the chemical parts themselves are a mystery beyond the reach of blind natural processes.

This is what the "black box" means in the title. The cell is treated as a black box by evolutionists. In math and computer programming lingo, a "black box" is a routine that does a certain job reliably. You stick something in, and you get something out. You don't necessarily know HOW the routine does its job, because it's hidden inside the box. But as long as it gives you the right results, you don't care.

So Behe means that the cell might as well be magic as far as macrobiologists are concerned. And that's what he builds his case on. As a biochemist, Behe says he's uniquely qualified to see that, in a nutshell, the cell IS magic. You look inside the cell, you see all this intricate machinery that couldn't have evolved, so you marvel at the brilliance and foresight of an "intelligent designer" who must have planned the thing.

Before looking at his arguments, I'd like to take a moment to address the style of the book. It's a little bewildering. Much of the book is written in a highly patronizing simplistic manner. Behe illustrates his points with analogies to Calvin and Hobbes cartoons, Foghorn Leghorn predicaments, Rube Goldberg machines, and hugely tedious detailed descriptions of everyday activities. Don't get me wrong, often a book on a complicated subject can benefit from the occasional light-hearted cartoon or cutesy analogy. But Behe doesn't just throw out the cutesy analogy and move on; he spends page upon page explaining his analogies in insulting baby talk. For example, in chapter 3, he writes about the process of swimming:

"Suppose, on a summer day, you find yourself taking a trip to the neighborhood pool for a bit of exercise. After slathering on the sunblock, you lie on a towel reading the latest issue of Nucleic Acids Research and wait for the adult swim period to begin. When at long last the whistle blows and the overly energetic younger crowd clears the water, you gingerly dip your toes in. Slowly, painfully, you lower the rest of your body into the surprisingly cold water. Because it would not be dignified, you will not do any cannonballs or fancy dives from the diving board, nor play water volleyball with the younger adults. Rather, you will swim laps.

Pushing off from the side, you bring your right arm up over your head and plunge it into the water, completing one stroke. During the stroke, nerve impulses travel from your brain to your arm muscles, stimulating them to contract in a specific order..." (pgs 57-58)


Behe goes on for three pages like this. And then when he finishes this drivel, he takes another two pages to explain that what he meant to say was, "You may think that the act of swimming is simple, but it's not."

But interspersed with these obnoxiously simplistic passages, there are plenty of passages stuffed full of dense, unreadable technical language like this:

"Conversion of plasminogen to plasmin is catalyzed by a protein called t-PA. There are also other proteins that control clot dissolution, including α2-antiplasmin, which binds to plasmin, preventing it from destroying fibrin clots." (pg 88)

Or this:

"Enzyme I requires an ATP energy pellet to transform ribose-5-phosphate (the foundation) into Intermediate II. The enzyme has an area on its surface that can bind either ADP or GDP when there is an excess of those chemicals in the cell. The binding of ADP or GDP requires a valve, decreasing the activity of the enzyme and slowing the synthesis of AMP." (pgs 157-158)

Now, I'm not scientifically illiterate. It's just that biochemistry happens to be way outside my field. If you're going to go with a level of description that requires the use of Greek letters, then count on losing a significant chunk of your audience.

This leads me to wonder: who is Behe's intended audience? If this were a biochemical treatise meant for biochemists, then he could have submitted it to a scientific journal, but he didn't. I've never heard the book promoted anywhere except on daytime Christian talk shows and in right wing columns. No offense intended to those forms of media, but most of their audience is probably not going to have any more luck deciphering "ADP" and "GDP" and "alpha-sub-2" than I did. And anyway, if he were going after very scientifically literate readers, then he would have done well to scrap the lengthy explanations of Calvin and Hobbes, as biochemists who know the material would certainly find them as pointless as I do.

But if the readers don't already get biochemistry, then he clearly does not expect them to follow the scientific lingo. In fact, it looks to me like he actually wants most readers to skip over the technical stuff. Whenever he slips into technical mode, he sets off the entire section with little boxes at both ends. The first time he uses this technique, he makes the following comment:

"The following five paragraphs give a biochemical sketch of the eye's operation. (Note: These technical paragraphs are set off by [] at the beginning and end.) Don't be put off by the strange names of the components. They're just labels, no more esoteric than carburetor or differential are to someone reading a car manual for the first time. Readers with an appetite for detail can find more information in many biochemistry textbooks; others may wish to tread lightly, and/or refer to Figures 1-2 and 1-3 for the gist." (pg 18)

Also, at various points in the book he writes things like "I assume I've lost most readers in the labyrinth by now..." (pg 149) And at one point he actually ridicules those who attempt to give understandable explanations to a lay audience.

"I apologize in advance for the complexity of the material, but it is inherent in the point I wish to make. Richard Dawkins can simplify to his heart's content, because he wants to convince his readers that Darwinian evolution is 'a breeze.' In order to understand the barriers to evolution, however, we have to bite the bullet of complexity." (pg 48)

Balderdash. I've read a fair portion of what Dawkins has written, and at no point do I remember him saying that evolution is "a breeze". I have no doubt that Dawkins has written some equally dense papers for his colleagues. My subject is computer programming, which is very far off from the work that Behe does. If I were going to write a book on programming, I could easily lose casual readers with detailed explanations of recursive search algorithms, segmented CSB+-trees, and context-free grammars. I could even, if I chose, use Greek letters. But if I were trying to reach an audience of interested novices, I probably wouldn't do that.

What I'm saying is that there's a simple way to explain a subject and a complex way to explain the same. Behe intentionally chose to go the incomprehensible route. If he did it for the reason I think he did -- to prove that biochemistry is a hard subject that requires a lot of specialized schooling to follow -- then he was wasting his time. I already believe that biochemistry is hard. If I thought it was easy then I might have become a biochemist. Instead of filling up pages of diagrams that most people would only glance at, he could have devoted some of that page space to fleshing out his arguments better.

But I suspect that there is something else going on here. I think Behe really does NOT expect most people to read the parts in boxes, but merely to be impressed by the big words and fancy abbreviations. The average reader would just look at the technical parts and say "Gee whiz, you're smart, Dr. Behe!" Most of the main thrust of his arguments are by analogy, and the GeeWhiz passages are just meant to convey the impression that the analogies are valid because they were made by someone smart.

Back to the substance of the book. To make the case that "black boxes" (cells) require a designer, he fleshes out his notion of "Irreducible Complexity", with one example in each of chapters 3-6. Since I'm not a biochemist or anything close to one, I'm just not very qualified to speak about whether Behe is right that the evolution of these systems is really as big a mystery as he says they are. Luckily, people who are qualified have long since stepped up to the plate to write at length about Behe's examples, so I'll defer to their explanations.

Chapter 3: The cilium
Chapter 4: Blood clotting
Chapter 5: Vesicular transport (search the page for "vesicular")
Chapter 6: The immune system

Each link gives a biologist's response to the stated claim. Note that the link for chapter 5 is actually a pretty lengthy deconstruction of the whole book.

Since the "irreducibly complex" (IC) nature of these particular systems has already been addressed very capably on other sites, I want to address Behe's overall concept of IC as the test for design. In its simplest form, the argument runs like this: Consider a system X that has dependent parts A and B. If you remove part A, then X will cease to function. If you remove part B, then X will also cease to function. Since either A or B must have evolved first, it stands to reason that at some point, X must have existed without one or the other. The resulting system would be useless. Therefore, it cannot have evolved in steps. It must be designed.

To really emphasize how silly the argument is, let's suppose that "X" = "The human body", "A" = "head", and "B" = "torso". Logically, the IC argument means "If the human body evolved, then at some point in history it must have been either a blundering torso with no head, or a disembodied head with no torso. How preposterous! Neither of those could survive! Since there cannot be a head with no torso or a torso with no head, God must have planned the head AND the torso to work together from the beginning!"

In this form, it should be pretty obvious where the fallacy is. The head and body coevolved. At an early enough point, you don't find a body with no head; you find a body that performs the functions of the head with no clear separation between them. Earlier than that, you reach organisms that just don't do head-like things like seeing, hearing, or thinking, yet they survive just fine.

So it's not enough to say "You can't remove parts A or B." To prove the case of irreducible complexity, you also have to prove that there is no simpler form of A or B that does the same job, only a bit worse. And beyond that, there's the scaffolding issue, i.e., some body types had features that supported the adaptation of other features, but then went away.

Obviously no one would bother denying that there are a great many complex systems that exist in living organisms. That is exactly why the theory of evolution exists: because it explains the very complexity we observe, better and more thoroughly than any other concept proposed. Does evolution also explain complex chemical compounds such as those that make up the cell? Almost certainly, according to the theory of auto-catalytic cycles proposed by Stuart Kauffman and others. According to this principle, a large enough variety of chemicals may be almost guaranteed to produce complex systems that continue to create more copies of themselves. It also shouldn't be overlooked that prokaryotic (simple) cells had about a 2.8 billion year headstart on evolution before the first eukaryotic (complex) cells arrived on earth. That's more than twice as long as the rest of all evolutionary history.


There was one analogy in the book that particularly didn't sit right with me. Behe tries to explain that even though one complex system may appear to be a "descendant" of another system, there is no logical path to get from one to the other. For example, you could say that motorcycles are just juiced up bicycles. But there are no small, incremental changes that could be made to bicycles that would turn them into motorcycles.

Well, Behe is absolutely right to say this. Motorcycles didn't evolve from bicycles. For one thing, a motorcycle has a gasoline powered motor. The gas powered motor wasn't specifically made for motorcycles; it was invented separately and has been applied to all kinds of other inventions that share no intellectual ancestry with the motorcycle. Another example that Behe didn't mention would be the computer chip. Today, you find ridiculously powerful computer chips in appliances like thermostats and alarm clocks. The power of those chips is largely wasted in such appliances, but they are used there anyway because they can be cheaply mass-produced, now that the development has already been done for computers.

That's one way that motorcycles and alarm clocks are different from living organisms. Designers transfer parts from one invention to another easily, but that just doesn't happen in nature, as far as we've observed. Far from being a problem with evolution, this is one of the ways that evolution has been confirmed. The theory predicts that no such borrowing of spare parts will occur. Different animals can receive the same feature from a common ancestor, or they can separately evolve apparently similar body parts. But they cannot transfer precise information across the family tree if the common ancestor didn't have that information. If an animal had a feature that was clearly co-opted from an unrelated species -- such as, if an ostrich suddenly gained a perfect copy of human hands with opposable thumbs -- that would tend to discredit evolution. But we don't see that sort of thing happen. This is one of the ways that evolution is falsifiable, which is one of the reasons why it's a legitimate science.

So finally we get to the idea of Intelligent Design, in a chapter which is surely the precursor to a lot of the pro-ID arguments that we've heard in the last ten years. In one of the most famous passages in the book, Behe says:

"Imagine a room in which a body lies crushed, flat as a pancake. A dozen detectives crawl around, examining the floor with magnifying glasses for any clue to the identity of the perpetrator. In the middle of the room next to the body stands a large, gray elephant. The detectives carefully avoid bumping into the pachyderm's legs as they crawl, and never even glance at it. Over time the detectives get frustrated with their lack of progress but resolutely press on, looking even more closely at the floor. You see, textbooks say detectives must 'get their man,' so they never consider elephants.

There is an elephant in the roomful of scientists who are trying to explain the development of life. The elephant is labeled 'intelligent design.'" (pgs 192-193)

That is all well and good, except that it is a patently bogus bit of sleight-of-hand. At no point in Behe's book does he ever actually produce any elephants. Instead, he just insists "Now that I've ruled out the mainstream scientific explanation, the only alternative is elephants."

Worse, Behe never offers any reason to suspect that his elephant (the "intelligent designer") actually exists. A more appropriate analogy would be if the detectives were swarming around a 23rd story apartment in New York, with a mysterious murder but no visible signs of an elephant whatsoever. While the detectives are trying to do their job, Behe is saying "See, I told you that theory would hit a dead end. It must be elephants that did it!" And "THAT clue didn't pan out either, did it? Why don't you just admit that it's elephants?" When the detectives point out that no witnesses have seen an elephant, and that there is no clear way that an elephant could have gotten up the stairwell or elevator in the first place, he accuses them of anti-elephant bias.

In this situation, the burden of proof is clearly on Behe to give a reason why elephants should be even considered as a hypothesis. It's not that the detectives have ruled out elephants entirely; it's just that until there is compelling evidence to suggest that an elephant was there, "getting their man" is a much simpler approach to the crime.

In order to make a case for a designer, Behe has to do more than reject natural selection as an explanation; he has to actually provide a reason to think that a designer was available at the scene. In one passage, Behe writes about genetic engineering, saying,

"The fact that biochemical systems can be designed by intelligent agents for their own purposes is conceded by all scientists, even Richard Dawkins... Since Dawkins agrees that biochemical systems can be designed, and that people who did not see or hear about the designing can nonetheless detect it, then the question of whether a given biochemical system was designed boils down simply to adducing evidence to support design." (pg 203)

That is hardly a "concession" at all. I can't imagine anybody disagreeing that intelligent beings like us CAN use our intelligence and our current state of technology to alter a gene, or that this ability will be further enhanced in the future.

It is, however, a complete red herring. From "some genes can be designed" Behe makes the logical leap to "all genes were designed." This is like saying that because some vegetables are grown by farmers, it logically follows that ALL vegetables are grown by farmers, and none of them grown in the wild.

What is missing from Behe's argument is the fact that people can design genes, but only if there are any people around to do it. A hundred years ago, the capability to "intelligently design" genes did not exist, at least among humans. And obviously people could not have designed the genes of the first life. So the burden of proof is on Behe to show that there existed any "designer" back then who was capable of genetic engineering. He asserts that there was, but he's begging the question. Since Behe refuses to speculate on the identity of this designer, we're back to square one. Either present evidence that such a universal gene-tinkerer exists, or just acknowledge the fact that natural explanations are all we have to go on at this time.

Although Behe taunted scientists for letting the cell remain a "black box", Behe's solution to the matter is to propose that a particular kind of pre-human intelligence exists -- certainly not a trivial claim in any way. This intelligence is older than the oldest multi-celled life on earth, and is capable of performing genetic engineering on a scale far beyond any human ingenuity so far. How does this designer work? Where did it come from? How do we explain the inherent complexity involved in the designer's existence? We don't know, and it's not our business to ask questions about it. So to get rid of these tiny black boxes, Behe just creates the biggest black box of all out of nothing.

Wednesday, January 04, 2006

Classic Daily Show shot


I ripped this picture off the Comedy Central web site.

I have nothing in particular to say about it, just saving it for posterity.

Clarification for those who don't watch the Daily Show:

The picture is of Jon Stewart doing his usual very bad -- yet oddly hilarious -- impression of Bush.