27 January 2006

What a difference a day makes.

The Discovery Institute, over at their Media Complaints Division Blog, has posted yet another article castigating Judge Jones for ruling that Intelligent Design is unscientific.

This one, by a second-year law student, takes more or less the same tone as the others:
In this detailed analysis, I will take a close look at Judge Jones reasoning, and evaluate the potential legal basis for determining the scientific status of ID. Ultimately, I find that the Kitzmiller opinion has no legal basis to determine the scientific status of intelligent design, and as such, is merely the opinion of one man, not the law as proclaimed by a federal district court judge.
Ed Brayton, over at Dispatches, has already fisked the substance of that post. I'd like to take a second to look at something else: the Discovery Institute's pre-decision view of how the judge should rule.

Back on October 17th, in a press release titled, "Discovery Institute Tells Dover Judge Teaching About Intelligent Design is Constitutional," Robert Crowther had this to say:
Today, the Discovery Institute, the nation’s leading think tank researching intelligent design, filed an Amicus Curiae (i.e. “Friend of the Court”) brief in the Kitzmiller v. Dover Area School District case urging the judge to rule that it is not unconstitutional to teach about the scientific theory of intelligent design.
Let's be clear: they didn't ask the judge not to rule on the constitutionality of ID. They asked the judge to rule that ID was not unconstitutional.

That October release wasn't an anomoly; it was the pre-judgement norm. From a November article:
Although Discovery Institute does not support the particular policy adopted by Dover, it has been clear in supporting the principle of academic freedom when it comes to intelligent design. That is why the Institute supported filing a friend of the court brief on behalf of 85 scientists who sought protection of the freedom to research and write about intelligent design. That is also why the Institute itself filed its own brief defending the constitutionality of teaching about intelligent design.
Again, they weren't demanding that the judge stay away from the question of whether or not ID is constitutional. They were actively promoting the view that it is constitutional to teach ID. As recently as the day before the decision was released, Casey Luskin described the constitutionality of Intelligent Design as, "the big question at stake in the case."

The was the DI's view on December 19th was that the constitutionality of Intelligent Design was central to the Dover case. Starting on December 20th, they began to condemn Jones as, "an activist judge who has delusions of grandeur" for ruling on the constitutionality of Intelligent Design.

What a difference a day makes.

26 January 2006

Lonnig's "Dynamic Genomes" paper: A quick critique.

In the comments section of my most recent post on the Discovery Institute's publication track record, Spike made the following suggestion:
Here is the only scientific paper that one can link from the Discovery Institute’s list of “Peer-Reviewed, Peer-Edited, and other Scientific Publications Supporting the Theory of Intelligent Design (Annotated)” http://www.discovery.org/scripts/viewDB/index.ph… . (The rest you have to pay the publishers for, I suppose):


1. Can you, dear reader, understand it?
If so, could you explain it to us lay people?

2. Is it science?

Caveat Poster I have no special allegiance to “Darwinsists” (whatever those are), evolutionists, scientists or the people who feel they represent the Truth of Evolution. So don’t play into OSC’s hand and don’t use logical fallacies.

If you want to dismember this paper, do so on rational, scientific grounds. Por favor.
I started out intending to examine the entire paper, but it's taken me a while to thoroughly respond to (or dismember, if you prefer) just one of the claims. I do have other things to do, so I'm going to restrict my response to addressing his claims about the lack of differences seen between organisms. This doesn't mean I agree with the rest of the paper - it just means that I only have so much time available for this right now.

The bulk of Lonnig's paper, at least as I understand it, seems to center around the question of how, given the large number of different ways that it is possible for a genome to change, is it possible for any of the various features seen in organisms to remain the same. In other words, if mutations are common, then why don't we see more differences between different groups of organisms:
Becoming fully aware of the features specifying dynamic genomes as mentioned above, the overall impression most students of genetics inevitably have gained, could perhaps best be stated by the words assigned to the Greek philosopher Heracleitus of Ephesus (about 544 BC to ca. 475 BC), describing the essence of nature by his famous verdict: panta rhei, ouden menei (“all things flow, nothing abides”). For almost ‘everything’ in the plant and animal genomes seems to be in a permanent process of flux so that in the long run one should hardly expect any constant genomic (and corresponding morphological) characters at all. (p.104)
There is really no way to say this nicely, so I will be blunt. If Lonnig is trying to suggest that we should not see conserved genes-and as far as I can tell that is exactly what he is saying-then he has an abysmally appaling understanding of evolutionary biology.

Stabilizing selection (occasionally known as purifying selection)is a basic part of our modern understanding of evolution. In simple terms, stabilizing selection is nature's way of saying, "if it ain't broke, don't fix it." In slightly more technical terms, stabilizing selection is a type of natural selection that occurs when the version of a trait that is currently present in the population is the one that is associated with the highest fitness. The Wikipedia article linked to at the start of the paragraph cites human birth weight as one example of stabilizing selection.

Stabilizing selection is, at the absolute minimum, something that needs to be considered as a possible explanation for a lack of evolutionary change in a trait. Lonnig devotes something like half of that paper to discussing stasis, and why he thinks stasis is a problem for evolution, but he does not mention the concept of stabilizing selection anywhere in his paper. He is either unaware of the concept, or he deliberately decided to completely ignore the concept. Stabilizing selection is typically discussed in introductory biology classes, so it's difficult to believe that he didn't know about it.

Something else that Lonnig does not seem to mention is that we can actually predict, at least to a certain extent, how variable specific regions of DNA are likely to be within and between species. For example, the regions of DNA that code for proteins that are involved in basic cellular processes tend to be less variable than the regions of DNA that code for proteins involved in the organism's interaction with the environment.

But you really shouldn't take my word for it, especially when it's easy to look at examples. I'm going to try to err on the side of making the explanation too simple, so I apologize if I'm covering things that you already know.

Histones are a type of protein that is involved in packaging DNA. The shape of this protein is extremely critical to its function, and the sequence of amino acids in the protein determines its shape. Packaging DNA is a basic cellular function, and is critical to the function of the cell. This means that mutations that change the protein are likely to disturb the function of the protein. As a result, stabilizing selection will tend to weed out mutations that result in changes to the protein.

I'm going to list part of two DNA sequences. I'm going to give the first 30 letters of the sequences. Sequence 1 comes from a jellyfish (GenBank accession AY428830.1). Sequence 2 comes from a bivalve mollusk (GenBank accession AY654989.1). To make the differences between the sequences easier to spot, I'll use capital letters to mark the differences.
1: aga aaa tcA acC gga ggA aaa gcA cct CgT
2: aga aaa tcT acT gga ggC aaa gcC cca AgA
There are two things that you should notice about these sequences. The first is that there is actually a fair amount of variation between them - 6 of the 30 bases are different. The second is the way I've grouped the letters into sets of three. This is a portion of DNA that gets translated into a protein, and it takes three letters ("bases") to specify one amino acid. Next, I'm going to provide the protein translation for each of those sequences. Lower-case letters will be used to show differences.
As you can see, the DNA sequences are different, but the proteins that result are identical. (In fact, the entire protein sequences for this histone in these two species are identical, not just these ten amino acids.)

To an evolutionary biologist, this type of signal indicates that this particular gene is both very important to the proper function of the cell and very sensitive to change. That is certainly the case with histones.

Next, we'll look at a protein that is involved in interactions with the environment. The fly genus Drosophila has a protein called alcohol dehydrogenase (or "adh"), which it needs to live in areas where there is ethanol production, such as the rotting vegetation where many of these insects live.
The larger of the two flies in the picture is Drosophila differens. This is a species unique to Molokai, and is part of the Hawaiian Drosophila. The smaller fly is Drosophila melanogaster, the famous "fruit fly" used in genetics labs all over the world. I'm not going to go to the lengths that I did above to look at the genes, but a quick comparison of adh from these two species (GenBank accessions M63303.1 and M36580.1) showed that the DNA sequences were 78% similar, and that there were differences in the proteins. These two flies are much more closely related to each other than a jellyfish is to a clam, so it is clear that this gene is much more variable than the histone we looked at before.

When we see a protein that is more variable, it usually indicates that the protein can tolerate more change and still function properly and/or that the protein's primary function in some way involves the animal's environment. If the protein can function in a slightly changed form, then it becomes possible to pass on mutations that slightly change the protein. If the protein is involved in interacting with the environment, then it may need to be different in animals living in different environmental conditions.

Lonnig's list of possible ways for genomes to change is completely irrelevant to understanding why some proteins are evolutionarily conserved. The histone gene shows much less variability than the alcohol dehydrogenase did, but that doesn't mean that mutations occur any less often in the DNA that codes for the histone. It doesn't mean that any of the different ways that the gene could be mutated don't happen with histone genes. It just means that any mutations that do occur in the histone DNA can only get passed on to the next generation if they don't change the protein. Mutations that don't get passed on to the next generation are evolutionarily irrelevant.

Natural selection can be a force for change, but it can also be a stabilizing force. It all depends on the circumstances. In the case of basic cellular functions, it is usually a stabilizing force. We are separated from the first cells by an amount of time that is too vast to comprehend. The basic cellular processes were pretty much optimized a very, very long time ago. It should come as no surprise that selection usually acts to stabilize the genes responsible for basic cellular processes.

Looking up at this post, I see that I've written quite a bit more than I had intended to, and looking at the clock, I see that my "quick" critique has taken three hours, so I'm going to wrap things up. Even though I wasn't able to dismember the entire paper, I hope I've demonstrated two main things:

1. Even in cases where a protein is exactly the same in widely separated species, the DNA that codes for the protein may differ. In other words, the function may be static, but the genetics are not. In this sense, Lonnig's claim is somewhat misleading, if not just plain wrong.

2. A lack of divergence in a gene (or any genetic trait) between two different groups of organisms is not a problem for evolution. In fact, such similarities can (and do) often result from the stabilizing action of natural selection.

25 January 2006

How can you tell it isn't science?

Let's say that you are someone who is interested in science, knows a bit about it, but aren't an expert. You might be someone who reads a lot of popular science books, or who watches a lot of science programs on tv. You might read a lot of science fiction. It's even possible that you are a science fiction author.

You have heard a bit about the whole intelligent design thing, but you may not have been following it closely - particularly when it's not in the news. You are also at least a bit disposed to root for the underdog. It's a better story, and you know that it has been real sometimes. People really did laugh at Fulton and the Wright Brothers, and some scientific theories have faced opposition from entrenched opponents. So how do you know that this isn't the case with Intelligent Design? Why should you trust us when we tell you that the ID people aren't really doing science, and that their real motives are much, much more political than scientific. Why shouldn't you believe the DI's claims that we represent an entrenched "Darwinian orthodoxy?"

There are many different arguments that I could make right now, and many of them are valid. These range from the fairly basic ("Their arguments are just plain unscientific") to the completely obscure ("A simple Bayesian probability analysis can show that it is extremely unlikely that someone whose 'scientific theory' is being mocked is actually right"). Which argument you find to be the most convincing may vary. Personally, I think that the most damning argument can be made just by looking at what the Discovery Institute has published.

I've written about the Discovery Institute's list of documents that they claim as supporting ID before. The last time that I did, they were calling it a list of "peer-reviewed and peer-edited" publications. The last time I talked about their list, I pointed out that not only is "peer-edited" a term without scientific meaning, but that some of the publications on their list didn't even rise to that anaemic standard. The current title of the list is "peer-reviewed, peer-edited, and other scientific publications," but the items on the list are the same. I think that attitude says a little something right there - slap a new label on it, and everything will be fine.

But let's set that aside for a minute, and look at their claims. The list, as I have previously noted, contains some items that are trade press books. It contains some articles that are found in the philosophy literature. It lists a book as one item on the list, then goes on to use every chapter in the book as a separate entry. It even lists some articles both in a "featured articles" section at the start of the list, then lists them again later on. The total number of entries in their list, duplicates included, is thirty-four.

That's not a lot by scientific standards. Last semester, I wrote a review article for a class that discussed the geographic modes of speciation observed in Hawaiian insects and spiders. That's a limited group of organisms, living in a very limited area, and I was only looking at one aspect of evolution in the group. I still wound up citing 124 separate articles - almost four times as many as the DI lists as supporting their position. As a scientist, I do find the lack of publications to be a significant strike against them, but I can understand that a non-scientist might not see the significance as clearly.

So, instead of comparing the scientific output of the Discovery Institute to the scientific output of scientists, I'm going to compare it to something else. Let's see how their scientific output stacks up against their public relations machine.

In addition to containing a list of "scientific articles" supporting ID, the Discovery Institute lists favorable news articles. Some of these are written by reporters or op-ed columnists not affiliated with the Discovery Institute. Others are written by DI fellows. Many are press releases issued by the DI.

Let's see just how their PR output stacks up against their scientific output. To do this, I combed through the list of articles linked above, and counted only those articles that were both related to evolution or ID and that were written by someone affiliated with the Discovery Institute. What I wanted to see is how long it would take for me to reach a total of thirty-four of those articles - that's the same number as the number of items (duplicates included) on their list of "scientific" articles.

The first of the articles is dated today, and the 34th (working backward) is dated 10 November 2005. That's a period of 77 days. That works out to a rate of about 0.44 press releases per day. Now, let's look at the scientific output. The first article in the list of scientific articles is dated in 1985, but I'll be generous and round it to an even twenty years. If you do the math, that puts the scientific article production rate at 0.0046 per day.

Let's look at that again:
Press Releases: 0.44/day
'Scientific' pubs: 0.0046/day

To me, that's the comparison that shows the Intelligent Design Movement's priorities far more clearly than almost anything else. This is a group of people that are pumping out press releases and op-eds at about 100 times the rate that they are producing material that they claim is scientific.

One hundred times more PR than science. Still think Intelligent Design is the noble scientific underdog, fighting against the entrenched orthodoxy? Are they Fulton with a PR firm? Or are they just trying to conceal a political and religious agenda behind a (very) thin veneer of science?

Press releases/op-eds:
(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34)

24 January 2006

A non-update update

I'm trying to get back into the habit of posting something every day, so, without further ado...


Seriously, I'm working on a couple of actual science posts. I was hoping to get one of them up today, but circumstances conspired against it. I got tied up with a car repair, then got home to find my dog missing. He was eventually located in the custody of the Humane Society, but that kind of blew the evening. Then I got to grade papers.

Grading papers, for those of you who have not had the experience, can actually be a very interesting study in human psychology and miscommunication. It's also a very good way to kill off neurons, without all those pesky side-effects that you get using chemical methods of brain-cell elimination.

Tonight, I was grading lab notebooks. This was the first time that the students have had to turn them in, so I wasn't expecting too much from them this time - and I almost got it.

Here's the breakdown:
Notebooks expected: 15
Notebooks received: 14

# of Students who followed
all directions correctly: 2

#of different ways
directions misunderstood: 12


23 January 2006

The Koufax Awards

It would seem that I've picked a pretty bad time to take a few days away from the blog. I come back from the happy land of Amotivation to find that someone nominated me for Best New Blog over at the Koufax Awards.

I'm flattered by the nomination, but I won't be voting for me. I haven't decided who I will vote for yet, but there are quite a few blogs on the list that are much better than this one. Don't just take my word for it. There are something like 125 nominees right now. Browse the list and you'll probably find some really good blogs that you've never heard of before.