Sunday, December 10, 2006

My overwhelming experience

I have, for the last month (I just looked back on the dates- it's been a month!), been engaged in a debate at Overwhelming Evidence about whether it's possible or not for evolution to cause increases in information and functions or if this would require intelligence. The ID stance (represented here by Mario, Patrick and TRoutMac) is, of course that this is impossible. Judging by the length of the discussion you would have thought that there would have been a lot of topics covered, but looks are, in this instance, deceiving. Most of my discussion with Mario is just trying to pin down exactly what he meant by increases in information and functions (and just by information and functions as such, whether new or not). I will not recount everything that was said in this debate since it can be read in full at the link above, but I do want to give a summary of what was stated (and small bits that were never said as the thread was closed down prematurely). When I write "ID claim" below, I am referring to what the corresponents at Overwhelming Evidence wrote.

ID claim: Only intelligence can create new information and new functions. Evolution can't. Mutations can only degrade information and functions.

After a LONG session with Mario I managed to get a fair understanding of what he actually meant by function (and what would constitute a new one). It's an interesting defintion he uses: At the level of a catalytic enzyme, two enzymes, one that ligates the molecules a and b to make c and another that ligates d and e to make f, would not be functionally different. That's right, even though two different enzymes perform different functions, they are not functionally different! Instead, the enzymes would have to DO something different like, for example, transport instead of ligate. Why this is an important distinction, I'm not sure.

Their defintion of information, it was finally mentioned after quite some time, seems to be William Dembski's CSI.

ID claim: "An increase of info. can only be the product of gene duplication..."

Mario was quite hung up on the idea that a gene would have to be duplicated before it mutates and creates new info/functions. His reasons seem to be that unless there is a prior duplication, the mutations would destroy the info/function of the old gene and thus there would be no NET increase in information. The horizontal transfer of said gene into an organism that retained to original gene would not count as this would only be information/function TRANSFER (not origin of). I have to regress a bit and state that there was, to the aforementioned, a highly related I point which I never pursued (because the thread was closed down). The point being how they use the terms creation (or new) versus destruction (or loss) when discussing information. In their minds, new information and a new function is created once. For example, let's say that we get a new enzyme (by their definition) with the ability to make x from y and z in organism A. When organism A reproduces or when the new enzyme horizontally transfers to another organism, this would not constitute the creation of a new function/information. Fair enough - under some circumstances, I would agree. However, when they talk about the destruction (loss) of information/function they do not mean the opposite of creation/new as one would expect. Destruction occurs when ANY organism loses it's info/fuction, EVEN THOUGH GAZILLIONS of organisms might still have them. So, if organism A (with it's new function) above multiplies to become 10^10 organisms, we still have a singular event of info/function creation. But if 10^9 of those organisms lose that info/function we have 10^9 events of info/function destruction (even though the information/function still exists. Even more bizarely, if an organism loses a gene it's a loss, but if it regains it through horizontal gene transfer, it is not a gain. Even though overall no change has taken place, there has still been a loss of something, they seem to think. And most bizarely perhaps, the creation of a new info/function gene by destruction of an old gene and then the subsequent transfer of that gene into another organism that still has the old gene, creating an organism that has a new gene with new info/function does not count as an increase in info/function). If these guys were to be consistent, the destruction of information/functions should only be counted when the LAST of it disappears. (All this is so strange that I wonder if I might have misunderstood what they wrote, but i don't think so. And we will never know, either).

Moreover, claiming that the horizontal transfer (and genetic recombination, leading to changes in genetic code) of genetic material can never be considered an increase in information whereas an actual duplication (with subsequent changes) apparently can, seems to me to be applying double standards. The rationale seems to be that horizontal transfer is only a transfer of existing information, but then so is a duplication. I would have liked to see a clarification on this point.

ID claim: It is IMPOSSIBLE for non-intelligence to create new information.

Unfortunately, I never made them back this claim up. Their conclusion seems to come from the fact that it has never been observed (as per their definition of information and what they constitute as evidence). Unfortunately for them, no intelligence has ever been seen to create the information necessary for the complexity of living organism from DE NOVO either. But that doesn't bother them (they wouldn't even call that little problem an extrapolation. Go figure).

Regarding evolutionary algorithms:
TRoutMac used an analogy between the destruction of biological information and the destruction of information if you were to mutate some text editing software. The analogy is pretty useless, so i simply responded that he might want to look up evolutionary algorithms instead. I got two responses to this. One from Patrick and one from TRoutMac. Patrick's was simply a link to some piece he had written earlier. Wrote he:

There are computer models that attempt to simulate biological evolution, and they are so vastly oversimplified and divorced from the biological reality they attempt to imitate that claims made on their behalf should be considered very carefully.
He wrote a lot actually, but the first paragraph included the above. Evolutionary algorithms are indeed oversimplified, but they are a heck of a lot better at modelling evolutionary processes than TRoutMac's hypothetical text editor is. I think this was lost on TRoutMac.


You're going to now argue against Intelligent Design by invoking Intelligent Design? Who devised those evolutionary algorithms, Hawks? Are you saying those people are stupid? That's not very niceā€¦ I suspect they would feel rather insulted.
One has to wonder why he wrote this. ANY sort of simulation will be programmed to some extent. At the very least, to simulate a biological system, even at the most basic level, you would at least have to have rules regarding the way atoms and molecules bond together. Would these rules in nature imply intelligence? (I would not be terribly surprised if TRoutMac answered yes, btw). If you wanted to simulate processes of information gain going on in organisms that are already "full of information", you would start of with a model using organisms that already are "full of information". This seems to be have gone straight over TRoutMac's head. TRoutMac, I would not call any of THOSE people stupid.

That was a long entry, and that's not even all. The rest will have to wait until a later date. Stay tuned.


Post a Comment

<< Home