...and of course, being LBN, it sank like a stone. I was going to get around to posting it here but I'm happy to see you beat me to it, papau!
The common comparisons of genes to computer programs and blueprints can be quite deceptive.
It's becoming very clear that life on earth has evolved various means of evolving -- that there are mechanisms in life's toolkit that make the process of evolution more effective.
Populations of organisms that can evolve quickly when they are faced with some sort of environmental stress are more likely to survive than populations that cannot.
At first we thought of evolution as a process where essentially random mutations were simply selected for or against -- a process of "natural selection" that was easily explained using our genes-as-blueprints analogies. Mutant genes that are not benificial are weeded out of a population, and mutant genes that are benificial are reproduced and amplified within a population.
The next step in our understanding of the process was to model genes as chunks of code that could be duplicated and shuffled around into new, more competitive configurations. This model has led to a much more sophisticated understanding of natural selection. These so-called "genetic algorithms" are the subject of cutting-edge research in genetics, computing, and evolutionary biology. For examples my quick "I'm feeling lucky" google search took me here:
http://www.aic.nrl.navy.mil/galistBut such "genetic algorithm" models do not explain everything.
It seems this duplication and shuffling of genes, along with the sorting and collection of "random" mutations, is somehow a contolled process. Duplications, shufflings, and mutations of genes that are unlikely to be useful to an organism's descendents are somehow suppressed through a variety of unknown mechanisms, one of which may have been described above.
(Before anyone jumps in at this point claiming to see some "Invisible Hand of God" here, that's not at all what I'm talking about. The mechanism for these processes are unkown, but not supernatural.)
As always, we discover life's toolkit to be much more sophisticated than we thought it was...
I don't agree with the author that this is a "
radical addition to the widely embraced laws of Mendelian genetics," perhaps because I've been very keen on these sorts of research for the past twenty-five years. It's been clear for many years now that genes are not the static entities Mendel described.
The "radical" aspect of this -- which is not so radical when you actually think about it -- is that the capacity of a population to evolve or not to evolve is, in itself, an evolved trait.