Response to Criticisms of Behe's Edge of Evolution: Chapter 3
by GutsI havn't read Behe's entire book yet, but I did get passed chapter 3. Out of curiosity I wanted to take a look at the critisms that were already out there. Fortunately I found some that were at least trying to address the arguments in Chapter 3.
I wanted to address briefly what i found out there so far. I say briefly because I havn't really digested the book, having only obtained a few weeks ago. Nonetheless, the criticism that I was able to find so far with a few keywords in google was that of Chapter 3, where the critic stated that Behe:
1. Assumed that the fitness landscape is static
2. Assumed that all local maxima is an absolute trap.
There were more assumed "errors" but none of them can be found anywhere in Behe's book (as far as I can see). You can view the criticism I am referring to:
Mark C. Chu-Carroll's review of chapter 3 of Behe's new book
Others have taken these assumed assumptions as if Behe had actually said them. First, I don't see how Behe could have assumed that the fitness landscape is static. His very discussion of the evolution of malaria seems to contradict that assertion. Reading his discussion of the evolution of malaria it is easy to see how he envisions Darwinian evolution: a population tracks the optimum, it is specified by the nature of genetic variation in the population and the shape of the adaptive landscape. Whether or not there are distant, higher fitness maxima somewhere in trait space is not an important issue, because all of the dynamics of the population are driven by the nearby optimum. In this view, populations don't get "stuck" on low peaks, because the peaks can change, and peak shifts (i.e., moving through a trough to a higher adaptive peak) may not be necessary and may even be extremely difficult.
Mark goes on to criticize Behe's use of local maxima:
If the fitness landscape truly has many independent dimensions, then there are very few (if any) true local maxima. To assume that local maxima are common requires assuming that when moving through one dimension brings you to a local maximum, moving through any other dimension will also bring you to a local maximum at the same point - which is really another way of saying that the dimensions are not independent - they all reach maxima and minima at the same places.
It's easiest to think of the fitness landscape in two-dimensional space, but adding additional dimensions does not provide any important problem for the existence of local fitness maxima. Thus, even in high-dimensional trait space, there can (and probably will) be local fitness maxima. Moving to the maximum in one dimension does not imply or require that a maximum be reached in any other dimension. Thus, a population can be at a fitness maximum with respect to one trait, but off the maximum for another trait. In fact, this sort of scenario is to be expected when the optimum changes for a single, ecologically important trait. This kind of reasoning assumes that as I move in the second dimension the local optimum in the first dimension remains a local optimum - i.e. that the dimensions are independent. So independent dimensions, each of which contains local optima, retains local optima in the multi-dimensional space.
September 20th, 2007 at 1:52 pm
I think you are giving the critics too much credit. It is not that Behe doesn't acknowledge that fitness landscapes change it is that he doesn't allow for Super Monkey Fun Ball evolution. For example, Behe might argue that the distance between functional protein X and functional protein Y is too large to be crossed by a random walk, i.e. there are too many intermediate steps where function is lost. Behe's critics, however, counter that in the same way a player shifts the landscape to guide a monkey ball through a maze in Super Monkey Fun Ball, dynamic fitness landscapes guide random mutations through otherwise impassible fitness terrains. Behe's critics imagine that continual changes in the fitness landscape create a selective benefit for each of the intermediate steps necessary to cross otherwise functionless terrain, Super Monkey Fun Ball style.
The problem with Super Monkey Fun Ball evolution is that it deals with probabilities the way Enron dealt with money. A dynamic landscape does not improve the odds of a function evolving, it simply shifts the accounting of the probability from quantifiable random mutations to the relatively unquantifiable fitness landscape. In the end, Super Monkey Fun Ball evolution is no better than a random walk and, as Dembski has already shown, a random walk is incapable of producing the complexity of life.
Comment by Jehu — September 20, 2007 @ 1:52 pm
September 20th, 2007 at 9:57 pm
Also, it's worth noting that local optima today may actually still be local optima tomorrow, it is still a barrier if it remains for a timescale thats larger than the timescale necessary to reach the peak.
Comment by Guts — September 20, 2007 @ 9:57 pm
September 22nd, 2007 at 9:36 am
If this blog entry had appeared in the immediate aftermath of the book's publication there would have been a deluge of responses. It seems the critics were intent on a brief period of severe debunking followed by a strategy of ignoring the book.
Comment by Bradford — September 22, 2007 @ 9:36 am
September 22nd, 2007 at 3:07 pm
From the review by Chu-Carroll:
Considering that this sort of math is right up Chu-Carroll's alley, it doesn't look good for Behe.
Comment by steve — September 22, 2007 @ 3:07 pm
September 22nd, 2007 at 4:17 pm
Haha…
Precious.
Chu-Carroll's highly emotional response makes me think that there's more at play than simple concern for intellectual integrity. I have a hard time trusting someone that responds as oddly at this.
Comment by Doug — September 22, 2007 @ 4:17 pm