[HilbertSpace]

Dawkins and Gould have never really gotten along - in part, it was due to a legitimate difference of opinion on some biological matters. In my opinion, it was also due to the fact that the most visible proponents in a field inevitably develop followers and schools of thought surrounding them, and wind up being identified with a simplified version of what are in fact a fairly complex suite of ideas.

For example, Dawkins used the phrase "lumbering robots" in Selfish Gene to poetically describe people (and all life) as being vehicles for propagation of genes. This phrase, because it is so visceral and even appalling to some models of humanity, was lifted from the wider context of the book and used to create a caricature Dawkins as a strict genetic determinist. Although he elaborated his position in The Extended Phenotype (a book he vastly preferred over Selfish Gene) and other writings, he became associated by his critics with a straw-man version of adaptationism. At the same time, his attacks on Gould's ideas for things like punctuated equilibrium were also mostly directed against a cartoon version of the concept, rather than against Gould's model particularly.

Punctuated equilibrium is a complex idea that is less simple, and less easily dismissed, than it might seem at first glance. The support is not only in the fossil record (which is the first bit that tipped of Gould, as a paleontologist), but also in some information-theoretic models of the genome and the organism. Specifically, the now well established models of neutral mutation combine with an information-theoretic and somewhat structuralist model of evolution to support the idea that punctuated equilibrium is highly probable.

Picture a geographical map of the US, and locate yourself within some state. If you were to move by taking steps in random directions, a lot of your movement would end with you not leaving your state. If you were close to a state border, especially in a corner, you might certainly transition into a new state, and more random steps might carry you further into that new state, with each step towards the center making it less likely that subsequent steps would take you back out quickly. This is one way of picturing the distinction between genotype (the DNA component of an organism) and phenotype (what the organism looks like, along with its physiology and behavior). Another supporting piece of evidence, in my mind, is that complex systems, being hierarchical, behave in a way that is non-linear. Incremental steps at a lower level (in this case the genes) accrue until they are sufficient to bring about a phase transition at a higher level of organization.

So, Dawkins' model of the selfish gene can be almost entirely correct, but needs to be seen alongside the larger picture of Gould's structuralism, including components of punctuated equilibrium and spandrels.