Col’s Rational World

Mis-Information Theory

After a few subtle hints, I’ve finally decided to fulfil the promise I made in an earlier post to talk a little bit about the subject of information in Evolution.

This all comes from the commonly-held creationist belief that evolution somehow disobeys the laws of thermodynamics or of information theory. In fact, that seemed to be the entire basis of the young-earth creationist beliefs of Answers in Genesis representative Paul Taylor (about whom I was writing in the post above). The claim is that we can’t get information from nowhere, so evolution can’t possibly have created complex creatures from simple beginnings. Creationists often cite the second law of thermodynamics which, they say, essentially encodes this problem.

Thermodynamics, for the record, is an area of physics dealing with how systems respond to things like adding and subtracting energy, or changing the pressure or temperature. The second law talks about how a value known as the ‘entropy’ of a closed, isolated system will always remain the same or increase over time.

In case you’re bothered, entropy is a measure of the complexity of an isolated physical system. It is a difficult and precise mathematical concept, but it basically corresponds to the amount of energy in a system that is ‘useless’ (that is, cannot be harnessed to do work). A system with high entropy is one, therefore, in which most of the energy is useless, in some precisely defined physical sense. Think about the room in which you are sitting – there’s lots of energy there just because the air is warm, but you can’t gather it up in any way in order to power a television. However, it came from a much lower entropy state like a lump of coal or a bucket of oil, where the energy is more useful. There’s the same amount of energy, but it’s just in a more usable form. Once you convert from low entropy to high entropy, you can’t get it back without putting even more energy in.

The creationist claim, needless to say, is false in every way. The most important way in which it is wrong is that it completely misrepresents the concept of entropy in thermodynamics. Entropy, so the creationists claim, is a measure of disorder. They claim that the second law of thermodynamics states that all systems tend towards greater disorder. Of course, this statement is untrue. The second law of thermodynamics actually states some complicated physics-y stuff that is utterly irrelevant to the plausibility of evolution, but which appears to have been hijacked by nutters.

So there you go, the second law of thermodynamics is utterly irrelevant to Evolution. And it’s irrelevant for two reasons: (1) it applies only to isolated systems – which the Earth isn’t, and (2) it’s not really talking about disorder at all.

But you can see that this argument doesn’t hold water for far simpler reasons – merely by applying common sense. For a start, we know plenty of physical systems that get more and more complex over time. Like, well… you. Humans start out as a microscopic cell, and end up as billlions of cells all operating in a beautifully choreographed whole.  Want another? How about crystals? Anyone who has ever seen crystals grow will realise that they form in a very organised manner from a chaotic soup of chemicals. Snowflakes are another regularly quoted example.

Kent Hovind tried to dodge the restrictions of the second law of thermodynamics by famously redefining it. He tried to claim that the scientific objection (that the Earth is not an isolated system because there’s energy flow into it, hence the second law doesn’t apply) was wrong. His argument was that adding energy to systems makes them even more disorganised. Take earthquakes, for example. But what Hovind is doing here is redefining the law itself. The second law states that the entropy of an isolated system will never decrease over time. But Hovind is saying that the ‘isolated system’ bit is just wrong; so what we have here is a totally different (and provably false) law that Hovind has invented – not the second law of thermodynamics. You can’t say “Clever physicists invented this amazing law of nature that disproves evolution. Except that they got the law slightly wrong and it actually should say something else.” That’s just massive dishonesty.

As an aside, we can show how adding energy to a system actually does increase order. Take a jar of grains of all sorts of sizes, add some energy by shaking it, and notice how the denser ones have sorted themselves to the bottom, with the lighter ones on top. Ah, physics – you gotta love it.

So on to the generation of new information in the genome. Let’s look at that closely.

The genetic code contains the instructions for building every cell in our body. You could say in a very real sense that the 3 billion-ish base pairs in the human DNA are like a recipe for generating a human being. If you’re new to genetics, Wikipedia has a fairly comprehensive introduction. It’s worth reading a little bit on this or it might not make much sense. There’s a few simple concepts that you need to grasp, and then the rest is just reading long words.

Every living thing uses DNA codes to specify how to make another individual. You can take, say, all the DNA from a mouse and lay it side-by-side with the DNA from an average human and add up all the differences. When you compare the genes, one against the other, then we’re about 85% identical. The other 15% would, therefore, encode the differences between a human and a mouse. OK, so this is slightly misleading – in reality there are other differences, not just the differences in genes – for example, we have extra non-coding sequences (‘junk DNA’) that are fairly different. But all the differences basically boil down to three things:

1. Mice have some chunks that we don’t have
2. We have some chunks that mice don’t have
3. We both have some chunks that are similar, but with some differences

So, all the difference in ‘information’ between mice and humans – or, indeed, any two species – boils down to differences in their DNA. And those differences are one of three things – either insertions, deletions or mutations. That is, either we add a bit, we remove a bit, or we change a bit. And the best fact is that we know exactly how all three of those can occur – through replication errors at the stage that DNA is being copied in a living cell. And, better than that, we have seen it happening in the lifetime of modern science. Here’s a simple method by which any of these could have happened:

1. Insertions: Endogenous retroviruses are chunks of DNA that were initially viruses but which managed to transcribe themselves into the DNA of the host that they infected. The human genome has very many of these, and we can compare them with our closest evolutionary relatives. In fact, this is an interesting aside because they brilliantly illustrate the evolutionary family tree. Another example is in so-called trisomy disorders, such as Down’s syndrome, where chromosomes are replicated too many times (three copies instead of two). We can look at domesticated crops and see many examples where extra copies of chromosomes have been inserted without harming the plant in any way.
2. Deletions: Bits of genetic code can be deleted rather easily simply by errors in the replication process, especially during chromosomal crossover. Williams syndrome and Jacobsen syndrome are both examples of genetic deletions in humans.  Deletions are generally bad, of course. In practice, deletion doesn’t play much of a part in the differences between humans and mice because one species has not evolved from the other. In reality, we have both evolved from a common ancestor, and the missing genes between the two DNA codes are simply due to chunks of DNA that have been added in the genome of one species, but not the other, after the two species diverged.
3. Mutations: Mutations are, of course, simply changes in the base pairs at certain locations of the genome. The genetic code consists of four letters, C, G, A and T, which correspond to four chemicals called Cytosine, Guanine, Adenine and Thymine. At any stage in the replication procedure, the copying process can simply accidentally switch one of these to any of the others. This is very rare, but when dealing with a few billion base pairs, then you’re going to see this happen a few times each time the DNA string is copied. As before, of course, this has been observed and documented in extraordinary detail.

Well, essentially I’ve just shown that there is absolutely no impossibility whatsoever with information theory applied to the genetics of evolution. I could just stop here and the case is closed, but just as a simple illustration, let’s look at how new information can arise.

Let’s take a string of four letters, “AAA”. Think of it like a genetic string. In this case, it codes for (‘translates into’) the amino acid Lysine. Amino acids are, as the oft-repeated phrase says, the building blocks of proteins which are the very basis of all life.

Now, let’s say there was a replication error which involves accidentally duplicating this string one time too many, so we get “AAAAAA”. Do we have any new information? Well, yes we do in a sense. This new string tells us not just to make one Lysine, but to make two.

But we don’t need to stop there. What happens when we accidentally get a genetic mutation in one point in this new string. What if the 5th ‘A’ swaps for a ‘G’? Now we get ‘AAAAGA’, which gives us Lysine and Arginine. So now we have a totally different string, which codes for two different amino acids. What we have here is completely new information which has been generated by well-known and totally understood evolutionary processes. Anyone who wants to claim that this is insufficient has only one way out – they have to show that there are some parts of animals that are not spceified by the genetic code. Good luck with that.

Job done.

Be Sociable, Share!

• 
•