Tuesday, October 02, 2007

The Crappiest Q&A of the Night

Okay, Im sick of typing out all of Dembskis stutters and erm and eeeeeh eeeeeeh eeeeeeehs and "I mean, you knows?" So this is an abbreviated Q&A. This was a softball question, and he still did it. Seriously the next time I hear someone insisting Creationists have a GREAT fr*me because they are dynamic speakers Im going to throw a CD of this garbage at them. But I vote this Q&A as the crappiest of the night. Its five minutes of two people talking and not saying anything. Awful.

Young gentleman-- I was wondering about the use of, probably DNA in terms of an information theory argument. Seems like a perfectly reasonable language, I don’t know if the complexities are too large, but are there people looking at DNA as an alphabet information carried in strings of DNA sort of in words and is there anything in that realm that is possible?

Dembski--Well sure! This is the idea of DNA being a code is well worn territory. I refer you to Hubert Yockey, who has written on this topic and is also critical of conventional evolutionary theory. He says the information in DNA is not something you can get from natural processes. So I mean, this merging of information theory and biology in the genetic code is something that is very much done. And in bioinformatics you look at sequences of DNA from different organisms and you do a lot of comparisons and data mining. This idea of DNA being a code is very much used.

You mathy folks will probably have more fun with this, but here is what I got out of this crappy Q&A:

  1. Dembski, given a softball opportunity to talk about his Blessed Information, could just gibber gobbledy-gook.
  2. Bioinformatics is impossible if Dembski is right. Bioinformatics means nothing without evolution. Even in the games we would play in undergrad in order to learn how to use all the online tools for bioinformatics, we had to use evolution. No idea why Dembski brought this field up. I mean honestly, Id really like to see him figure out what an unknown gene does in humans without using evolution. Knock-out babies, Dembski? Thats just about your only option.
  3. Who is Hubert Yockey? I Googled him and got a ton of Creationist pages. Im going to do more hunting...


Tyler DiPietro said...

I've never heard of this Yockey fellow either, but he doesn't seem to be a particularly important figure in CS theory. And if this quotation...

"He says the information in DNA is not something you can get from natural processes."

...is at all an accurate reflection of his views, then they are even more useless than Dembski's. Saying that you can't "get information" from natural processes is absolutely silly. If one conceives of DNA clusters based on functions as objects with Kolmogorov complexity, for instance, than there is no greater source "new information" than a process of shifting those functions through a (statistically) random process. That is because Kolmogorov complexity is equivalent to the compressibility of the string used to describe the object. Changing the regularities of the description is the equivalent to producing "information".

Tyler DiPietro said...

Sorry, after this:

"If one conceives of DNA clusters based on functions as objects with Kolmogorov complexity, for instance, than there is no greater source "new information" than a process of shifting those functions through a (statistically) random process."

I meant to add something to effect: "From what I understand, mutation fits the bill quite nicely for such a process."

Bob O'Hara said...

I bought one of Yockey's books a couple of years back. I only got to about Chapter 2 before I gave up - it's just crud.

You're not surprised at this, are you?

Dan Cardinale said...

You're probably getting tired of people asking, but any chance of the audio being made available?

SteveF said...

Yockey has produced a bunch of evolution/information themed papers. He's also a critic of most origin of life scenarios. A few papers:

Yockey, H.P. (2000) Origin of life on earth and Shannon's theory of communication. Computational Chemistry, 24, 105-123.

Yockey, H.P. (1977) A calculation of the probability of spontaneous biogenesis by information theory. Journal of Theoretical Biology, 67, 377-398.

Yockey, H.P. (1977) On the information content of cytochrome c. Journal of Theoretical Biology, 67, 345-376.

Yockey, H.P. (1979) Do overlapping genes violate molecular biology and the theory of evolution? Journal of Theoretical Biology, 80, 21-26.

Da Vinci said...

Review of Yockey's Information theory and molecular biology by Gert Korthof

Tyler DiPietro said...

Thanks Da Vinci, that review was interesting.

I did a little reading on Yockey, and it seems that Dembski has misrepresented him to an extent. My reading hasn't revealed any criticism on his part of evolution, but most theories of abiogenesis. Significantly different, AFAIK.

It also seems from the review, if it accurately represents him, that he's invoked a conceit common amongst computer scientists: attempting to reduce everything to questions of computability. You can apply information theoretic concepts in ad hoc ways to biology (I outline one above using KCS complexity), but it has yet to be demonstrated whether this particular notion illuminates anything about biology itself, even setting the conceptual difficulties pointed out by the reviewer aside.

I'll have to look into this some more, maybe some of his papers are in the public domain by now.

Shai said...

Um, I'd hate to do Dembski's work and do his own ad-hoc arguments for him, but I think he might be able to say (unwittingly, of course), that bioinformatics is possible because God, erm, I mean, "the designerwhoeverheis", created all of the living creatures from the same really smarty-pants building-blocks.

In short, Jebus did everything apparently just right so that it'd fit my claims.