Monday, October 17, 2005

How Science works

One of the favorite quotations of the creationists comes from Lynn Margulis, who said:

It was like confessing a murder when I discovered I was not a neo-Darwinist. … I am definitely a Darwinist though. I think we are missing important information about the origins of variation. I differ from the neo-Darwinian bullies on this point.
They take this to be some sort of apostasy on the part of a prominent biologist, but it actually is much more sophisticated than that. To understand why, it's important to understand what Margulis did and how science works. Creationists tend to skip those steps, and misunderstand this quote. The Panda's Thumb looked a little more carefully at the context of the quote, and debunked the immediate meaning, but let's look at the broader issue.

Margulis achieved a measure of fame (at least among scientists) in the '80s, by proposing something called the Endosymbiotic Hypothesis. To understand it, we're going to have to go back into the depths of the week in 9th grade biology that covered cell biology.

MitochondrionAlmost all eukaryotes (organisms whose cells have membrane bound nuclei) have organelles called mitochondria, the so-called powerhouse of the cell. Each cell has a bunch of these, and they divide and replicate as the cell they're in grows and replicates. Each one has a single loop of DNA, and two membranes. With various fancy proteins, they maintain a proton gradient which they use to turn food and oxygen into useful energy (pedants will whine about that description because I just simplified something people spend a lifetime studying into two sentences. Read the Wikipedia or click on the illustration for the gory details.)

The point I want to emphasize is that these little buggers are complicated in lots of ways. It's hard to see how something like this would evolve in the middle of a cell. It meets most of the criteria for Irreducible Complexity, though as we'll see, the story is more complicated than that.

Lynn Margulis (and others before her) noticed a couple interesting things. First, that DNA loop. Other organisms with a single DNA loop are in the Domain Bacteria. Of course, they only have one membrane, but when one cell engulfs another cell, the engulfed cell is surrounded by two membrane. Indeed, it's possible to get even more membranes around a cell that way.

So, we have something that looks a lot like a bacterium that was engulfed by a different cell. Do some DNA analyses, and you learn that the DNA strand in all mitochondria is pretty similar to a particular group of Proteobacteria.

The complexity is not so irreducible. We just reduced one irreducible cell to two reducible cells, a host cell which derives from the Archaea and a symbiont that derives from the Bacteria.

The other problem now is that different species have different numbers of genes in their mitochondria. Mitochondrial genes sometimes shift out of the mitochondrion and into the nuclear genome. The existence of plasmids and other sorts of movable genes isn't really news, but the fact that the mitochondria are losing parts is intriguing, as it suggests that they aren't quite as irreducible as they might seem on their face.

Why did this inspire Dr. Margulis to say that she "differ[s] from the neo-Darwinian bullies"? The neo-Darwinian synthesis is the intersection between genetics and natural selection. As formulated in the 1940s, it states that there are 4 modes of evolution: mutation (which would include crossing over and other large-scale chromosomal rearrangements, including gene duplications), natural selection, genetic drift (the tendency of small populations to be more dominated by rare events), and gene flow (migration between populations). In this schema, the only source of novelty would be mutation, the other mechanisms work with the existing variation.

Margulis says she's a Darwinist because she accepts the scientific evidence that natural selection happens and the logical argument showing that under fairly simple conditions, it must happen. For natural selection to occur, there must be more individuals born than can survive and reproduce, there must be heritable differences between individuals, and individuals with certain heritable traits must be more likely to leave offspring. To the extent Darwinism is anything, it is a movement that holds this demonstrably true statement to be true. Darwin never claimed that nothing else would happen, just that this process has to take place.

Margulis denied being a neo-Darwinian because she was proposing a new source of variation in biological systems. As she looked at the cell, she saw numerous cases that could plausibly be explained by the same process of engulfment and integration that the mitochondria appear to have gone through. The chloroplast is a well documented case which also has a DNA loop, and appears to have originated as a cyanobacterium.

The talk she was delivering when she made the comments quoted above was about the possibility that the nucleus originated in the same way. She also has proposed that flagellae, cilia and peroxisomes may have endosymbiotic origins, though the evidence is shaky to non-existent for all but the mitochondria and the chloroplasts.

There are two important points here. First, contrary to the claims implicit in most creationist work (including IDC), it is possible to disagree with the neo-Darwinian synthesis without endorsing creationism. "Neo-Darwinism" and creationism could both be wrong, but you don't hear much about that from Dembski, who claims that he can eliminate all possible explanations based on Chance or Regularity in one fell swoop. That would mean that Margulis shouldn't have even bothered looking for natural explanations for biological complexity. Dembski could have eliminated her hypothesis without even knowing what it was, had he come along a few years earlier.

The second point is that this is how science works. Confronted with a real challenge, the origin of a very complex structure intimately linked to the functioning of a cell, scientists didn't toss up their hands, declare "Goddidit," and play cards for the rest of the day. They gathered evidence, studied the system, thought carefully about testable hypotheses, and gathered data. They evaluated their hypotheses, and if their predictions were right, they tested the hypothesis some more. If they were wrong, it was time for a new hypothesis. Most of my young bio majors were able to list those steps on their lab exams.

That's science. It's how the game is played. And one step at a time, you break something that seems irreducibly complex down into its components.

When I say that creationism is an argument from ignorance, this is what I'm talking about. It's not that creationists are necessarily more ignorant than anyone else, just that humans are imperfect, our knowledge will never be complete, and we're all ignorant on some issues.

On the other hand, creationists never talk about endosymbiosis. They talk about natural selection and mutation, ignoring the fact that endosymbiosis is a well-documented addition to the list of four (not two) evolutionary mechanisms that go back 60 years. That omission can only be ascribed to ignorance or malice, and Hanlon's Razor compels me to go with ignorance.