Friday, July 2, 2010

Bob Dutko on Information and Evolution

When I did my in studio interview with Bob Dutko one question he posed related to a challenge regarding evolution. He says that the information content of a DNA molecule has never been observed to increase, and evolution requires that increase. This means evolution is wrong. This is at about the 50 minute mark.

I replied that certain bacteria had been observed to mutate and develop an ability to consume nylon, which is a recently invented substance. Why isn't that an increase in information?

Bob's answer was something like "I'm afraid you're incorrect. You do not have new genetic information you only have a reshuffling of existing information. That does not produce new genetic information." I admitted my ignorance and said I'd look into it.

And so I did and what do I discover? Bob is completely wrong. Mutations can and do increase information content. Also a reshuffling of existing DNA can increase the information content. It's pretty easy to see when when you understand what information is.

"Information" is a field of study related to computer science and electrical engineering. The definition can be somewhat technical, but here's a basic overview. Take a Word file as an example. Within your computer the file is represented as a string of 1's and 0's. Suppose a file is 1000 kilobits when saved. That can actually be reduced with compression. WinZip is a program that can compress files. Basically WinZip can recognize patterns and can represent repeating patterns with a reduced quantity of 1's and 0's. The fully compressed file might be 500 kilobits. That's the quantity of information for that Word file. The fully compressed # of bits required to describe the file.

What this means is that in fact something can be less coherent but have more information. A book with letters in a random sequence will have more information than an encyclopedia of the same length with English words. Real words form patterns that can be compressed into a reduced string of data. Random sequences of letters wouldn't have those patterns. So ultimately the number of bits required to describe the random book would be greater than the number of bits required to describe a book of the same length with real words.

So let's apply this to DNA. Suppose a mutation occurred and the new string produced a sequence that was more random. It could be that nothing but a reshuffling of existing information has occurred, but if it's more random it has more information. On the other hand it might produce a more repeatable pattern in which case the information content could reduce. But this really doesn't matter for evolution anyway. A more random sequence might express itself physically in a manner that produced a selective advantage or it might not. Natural selection doesn't care if the DNA sequence can be expressed with a large or small quantity of bits. All it cares about is whether the physical expression makes the organism more or less capable of reproducing successfully.

So now I'm annoyed. Bob Dutko knows that an average person doesn't know this stuff, so he can spring this on an unsuspecting person knowing that they'll just have to defer to what they think is his informed opinion. Of course I suspected he didn't understand what he was saying, but since I didn't either there wasn't much I could say.

Soon after my interview Bob Dutko interviewed Kenn Hamm where he made the same argument. I also had recalled that Bob makes the same argument in his Top 10 Proofs Evolution is Scientifically Impossible. It's obviously an argument he thinks is very important. So I thought I'd call in on open line Friday and discuss it.

So I did. I called in and told the screener Jan that I wanted to talk about Bob's argument related to evolution and information. The first time I tried I thought I was early in the queue, but he never got to my call. So I called again the following Friday and described my call the same way. This time Bob spoke with me during commercial. He wasn't going to take my call. He said something like "You probably got some obscure stuff from talkorigins and I'm not going to let you ambush me in this way." So I said to him "How about I do this. I'll send you my question in an email, send you any related links, you can look it over and I'll call you at a later date." He said OK.

So I sent him my question. There was some miscommunication as I tried to get the right contact information. Before Bob received my question he told me that he'd allow this kind of thing just this once, but he wasn't going to let me turn open line Friday into some sort of semi-regular debate between me and him where he has to run around and do a bunch of research. He's a busy guy and can't be doing that. He also told me that he wouldn't be able to discuss it for this upcoming Friday. Anyway, here is what part of what I wrote:

OK Bob. I'll tell you what I intended to say.

"You argue that evolution is not possible because it requires an increase in genetic information. I have two questions. How do you measure information? In other words how do you know if information has increased from one generation to the next. And second, how do you know that there has never been a generational increase in information?"

That is the totality of my question. I've looked in to what information is and as I understand it information content does increase with mutations. There is evidence here:

Contact me if you would like any of my further thoughts.

I like this question because it demands that Bob know what information is. And I don't think he does know what it is. This is obviously very bad for him. When you see what it is you see his argument that he's been pushing, and in fact selling in his CD series, is just completely wrong.

So I didn't hear back for a while and I also was swamped at work and couldn't even listen to his show. Finally this week things have eased up, so I emailed him to ask if it's a good time to call in about this.

Nope. He hasn't looked at it, doesn't know when he might have time to. And frankly he thinks I should just call in as a regular caller instead of expecting him to do a bunch of work preparing for my calls. I'm playing games that put him at an unfair advantage since I'm researching a topic he's not prepared for, and he's not going to allow me to come on and pull a fast one. If I'm all bent on talking about this particular topic and demanding he prep for a debate it will be a long wait.

These types of replies came over the course of a couple of emails. This obviously makes no sense whatsoever. I'm calling about an argument he makes routinely. Isn't this what open line Friday is for? And I'm submitting my question in advance. How is that pulling a fast one? He repeatedly criticized me for preparing with research. So what if I've prepared? Isn't that what reasonable people do? But it was clear to me that this is not getting through to Bob.

What's weird is I really think Bob believes what he's saying. He really thinks I'm being extra demanding of him. By submitting my question in advance so that he can have the opportunity to prepare if he so chooses this is somehow me being demanding. But obviously rationality is out the window here because the alternative is looking bad on the air by revealing the fact that he's basing an argument on a subject he doesn't understand.


Unknown said...

I find it interesting that Bob said that you were "... playing games that put him at an unfair advantage since I'm researching a topic he's not prepared for, and he's not going to allow me to come on and pull a fast one ..." I find it interesting because that is exactly what he does to people he debates with ...

ComfortForInjured said...

Perhaps you and Bob have been talking at cross purposes because your terms of reference are not the same.

I am not sure but could it be that you have researched "Shannon Information"? Shannon information is a limited to a mathematical and probabilistic definition that encompasses both meaningful and useless bit strings. Shannon information is the first level definition, and is indeed information, but not sufficient for many purposes.

For example, a garbled nonsense string of characters may qualify as good Shannon information, as will a meaningful string such as an English instruction to raise your hand.

To produce the kind of result expected (information gains in more useful DNA to the purpose of synthesis of new proteins), I would submit that we need semantics / meaning (which Shannon himself stressed was not what he intended).

A paper which supports your point of view and has been widely cited is the Long et al paper (Long et al., 'The origin of new genes: glimpses from the young and old.' (2003), Nature Reviews Genetics, 4, 865-875 (November 2003)), which assesses several research studies and concludes that DNA can indeed change via Darwinian pathways and produce meaningful information.

To evaluate the strength of Long's paper, one would have to research and evaluate the studies cited in the paper. I have the view that the jury is still out, but that Long could be correct... or not, (That's why we fund additional studies)

One critique of the Long paper I found is a chapter in Meyer's Darwin's Doubt (2013) that seems to cover this issue well, from a point of view that critiques the adequacy of the Darwinian mechanism to account for all of the necessary increase in genetic information. I believe it is not a question of my opinion, nor yours, nor Bob's - just a matter of evidence and evaluation -

As much as I would like to join you on several issues, I am afraid that I am not qualified to assess one way or another based solely on the evidence so far - just wanted to move the issue to a question of evidence, not stuck a possible confusion over a definition.

If Bob's young earth creationism and biblical references are not to your taste for exploring further, you could take a quick look at more formidable morsels at