[QUOTE=Frylock]
I understand that biological evolution by natural selection in fact procedes as it does in part because of the heritability of traits. But it is not my understanding that the concept of evolution requires that heritability of traits plays this role…Can you tell me where you got this information, that heritability is essential for evolution?
[/QUOTE]
Seriously? Inheritance of intrinsic (in the case of biological organisms, genetic) characteristics is fundamental to natural selection. If preferred characteristics aren’t passed on to the next generation in some fashion, there’s no “evolution”; all you have is drift and random combination of genes or instructions. Even in nonbiological concepts of evolution–for instance, the concept of “cultural evolution” or conceptual memetics–it’s intrinsic that information about past states and consequences is passed on, albeit in a fashion more akin to Lamarcks transmission of acquired characteristics than preferential selection of inherent qualities.
The dice analogy was a bias toward a statistical distribution of sixes in the Yatzee cup. There’s nothing carried on from one roll to the next–the dice neither know nor care what face they come up with–and unlike biological evolution, there is an ultimate, teleological goal (maximize the sixes) which doesn’t shift or show any competing influences once you’ve achieved the median. This make it, at best, a misleading analogue to natural selection, which is always a moving target. There is no ultimate predator, no top of the chain; everything eats everything else, and as the lion runs faster, so do the wildebeast, each only stopping when they come to an acceptible equilibrium of predation versus starvation.
[QUOTE=yelimS]
Again, I’ll argue that’s only a play on words, and that the only difference between evolution in the broader sense and genetic evolution is one of speed. All new changes are random, anyway. In the dice example, “good” traits survive, in the gene pool, good traits survive and multiply.
[/QUOTE]
With regard to biological organisms (save for humans and perhaps, to a very limited extent, a few other mammals) “genetic evolution”, i.e. natural selection is the only way in which evolution affects species. Changes are not “random” in any sense of the word; even if we assume that the input is a random spectrum (i.e. a population is subjected to predation in an entirely unbiased selection process), the result is decidedly non-random, with a bias going toward the animals who are best able to avoid being dinner, either by outrunning the predator or (in the case of herd animals) their herdmates. Even if we restrict the meaning of your statement to genetic mutations, these are not random, either. It’s well understood that some parts of the genome of any given animal are more prone to mutation than other parts, and that strongly entrenched genotypes are, generally speaking, highly resistant to mutation (most likely because they have a multifaceted purpose such that a defect or change would affect so many different developmental characteristics as to render the carrier nonviable).
In the dice example, coming up sixes is the characteristic selected for. But the dice aren’t punished or selected out for coming up some other number. They’ll roll again and again, regardless of which number comes up. The lack of consequences to either the individual die or to the system as a whole, nor any way to alter the rules means that there is no possibility of improvement in the direction of the selective pressure. What you have here is just a biased probability distribution (I think this is an exponential distribution, but someone will no doubt come 'round and correct my statistical misapprehension and/or laziness) rather than the asymmetric, non-zero-sum game that is characteristic of typical natural selection processes.
Stranger