Heh... unfortunately, I've been very busy, and a thorough look into everything that's wrong with "Selfish Gene" would require more time than I have. It is well written, and that's one of the problems with the book, and Dawkins in general. That is, many people are willing to accept his views because they are well written and do not take the time (or have the background) to question their scientific validity.
For starters, very early on Dawkins acknowledges the current lack of consensus as to the exact definition of a "gene" (i.e. does it include promoter/enhancer/other regulatory sequences? should each polymorphism count as one gene? or do we group them together? and if so where, in the continuum of small differences between "genes" do we draw the line?). His proposed solution is to take as a gene any combination of genetic elements which gives rise to an expressed trait.
He then expands on this basic definition to describe how it leads to genes which are "selfish". That is, they are interested only in their own continuance, and so the traits they code for will be optimized around this goal. Unfortunately, this is dangerously close to circular logic. You see, the physical traits that appear in organisms rarely have a simple one-to-one correspondence with a particular subset of the genome. Effects such as epistasis, where the action of one gene masks the effect of another, nullify Dawkins' simplistic definition.
Or, if you would have it another way, genes which have epistatic effects would need to be included in with the original "gene" according to Dawkins' definition. Unfortunately, the network effects that give rise to higher plants and animals would eventually necessitate including the entire genome of an organism in one "Dawkins gene", making the unit of selection not the gene, as Dawkins would have it, but the individual.
As for topics which he over-complicates, the first that comes to mind is kin selection. Admittedly, I don't think his over-complication is intentional, but rather his exploration of the topic is hindered by his presumption that genes are selfish. In fact, kin selection can arise in evolving individuals without the need for the concept of a "gene" at all! Rather, any evolving individual which evolves in a mixed population will eventually acquire various kin selection traits.
As I said at the start, I'm not surprised that Dawkins' ideas are so commonly held as true. He is well written, and most of the work that would refute his claims is usually presented in language inaccessible to most laypeople.
For starters, very early on Dawkins acknowledges the current lack of consensus as to the exact definition of a "gene" (i.e. does it include promoter/enhancer/other regulatory sequences? should each polymorphism count as one gene? or do we group them together? and if so where, in the continuum of small differences between "genes" do we draw the line?). His proposed solution is to take as a gene any combination of genetic elements which gives rise to an expressed trait.
He then expands on this basic definition to describe how it leads to genes which are "selfish". That is, they are interested only in their own continuance, and so the traits they code for will be optimized around this goal. Unfortunately, this is dangerously close to circular logic. You see, the physical traits that appear in organisms rarely have a simple one-to-one correspondence with a particular subset of the genome. Effects such as epistasis, where the action of one gene masks the effect of another, nullify Dawkins' simplistic definition.
Or, if you would have it another way, genes which have epistatic effects would need to be included in with the original "gene" according to Dawkins' definition. Unfortunately, the network effects that give rise to higher plants and animals would eventually necessitate including the entire genome of an organism in one "Dawkins gene", making the unit of selection not the gene, as Dawkins would have it, but the individual.
As for topics which he over-complicates, the first that comes to mind is kin selection. Admittedly, I don't think his over-complication is intentional, but rather his exploration of the topic is hindered by his presumption that genes are selfish. In fact, kin selection can arise in evolving individuals without the need for the concept of a "gene" at all! Rather, any evolving individual which evolves in a mixed population will eventually acquire various kin selection traits.
As I said at the start, I'm not surprised that Dawkins' ideas are so commonly held as true. He is well written, and most of the work that would refute his claims is usually presented in language inaccessible to most laypeople.