[mdemare]

Evolution is essentially an algorithm. That's why we can use logic to predict that genes that regulate behavior for "less successful individuals to slink away and accept starvation" are unlikely to arise, since not a single one of the ancestors of that individual were less successful and slunk away to accept starvation.

[jballanc]

There's a mister Hardy and a mister Weinberg that would like to have a word with you about that logic...

http://en.wikipedia.org/wiki/Hardy-Weinberg

The other thing to keep in mind is that evolution works as a movement through nucleotide space over a fitness landscape. The fundamental flaw in much of Dawkins work is his reliance on a biologically unrealistic definition of a "gene". Sure, if you go with his definition of genes then logic can tell you all sorts of things about evolution that aren't necessarily true.