In the book Wonderful Life, Stephen Jay Gould made a subtle point about species diversity. While accepting that life over the past few million years "is more diverse than it has ever been" (Sterelny, p71), he characterises this as not the peak of "disparity" - diversity in a different sense, thus requiring application of a different term.He refers to disparity as the number of different types of body form or plan. So, for example, the large number of types of beetle contributes to the overall species diversity, but not disparity.
He attributes the peak of disparity to the time of the Cambrian explosion (530-odd million years ago), where a large number of multicellular animals with substantially different body types "came into being" so to speak. Specifically, it looks like he's referring to the the number of different phyla. (the animal kingdom is made up of 25 to 80 different phyla, depending on who you talk to. Most of the animals we are aware of come from just a handful of phyla, eg vertebrates, invertebrates, etc; most phyla are obscure, unnoticed, and with just a handful of representatives.)
The disparity of life forms never again matched that of the Cambrian era, which Gould says is one of the key issues of evolution that needs to be explained. Conversely, he says that the remaining lineages show - by comparison - "profound evolutionary conservatism".
As Sterelny (see below) points out, by analysing and comparing at the genetic level (DNA/RNA) two species, it's possible to make some observations on the length of time since they diverged from a common ancestor. Yet the technique is silent on the rate of divergence, for example whether they made key changes regularly, or all at once close to either the beginning of the divergence, or to the period contemporary to the species.
The Cambrian explosion presents little evidence of earlier formations, yet it is the time of (sudden) record of the greatest disparity of species ever.
Sterelny's book is called Dawkins Vs Gould: Survival Of The Fittest. Here's where he paints Richard Dawkins in opposition to Gould (and although I can't be sure how much of his dichotomisation is a straw man, it is clear there has been divergence of perspective - and opinion - at times). Sterelny portrays Dawkins and his adherents as "cladists". In this context, it means they are relatively blind to disparity (as opposed to diversity in general). Sterelny says the true cladist says morphology (the comparison of similar features between species) is unimportant compared to the genetic distance between two species. As mentioned before, in one sense this is true. But this well illustrates one definited divergence in perspective between Dawkins and Gould. It is one thing to say that taxonomy (and cladograms) should be wholly based on genetic proximity (and thus break down reptilia as a grouping). But genetic analysis only works on _our_ timescale, that is, what DNA can be extracted today. Morphology as paleontology is the only possible approach where the DNA evidence is unavailable, and is thus the only appropriate tool for analysis at large time scales. In any case, morphological taxonomy or cladistics must ultimately incorporate the findings of genetic analysis, so the two perspectives must converge.
So in Sterelny's terms, Dawkins would not regard the Cambrian explosion as particularly noteworthy. In the history of life on earth, those diverse phyla _must_ have been more closely related than many that exist today. However, this is where I see the straw man in Sterelny's argument, because I simply can't see someone of Dawkins' experience writing off the Cambrian experience thus.
If I've counted them correctly, Sterelny measures three chief points of convergence between the two evolutionary biologists, Richard Dawkins and Stephen Jay Gould. Foremost, Dawkins is a genetic biologist, and Gould is a paleontologist - that is, their analytical perspective begins at the micro versus macro level. Dawkins is further characterised as the arch-rationalist - as he has demonstrated - who believes that "the scientific description of outselves and our world... is true... beautiful and complete". Then Gould is the humanist who "does not think science is complete".
Being a rationalist and a humanist myself, I can vouch that the two are not mutually exclusive, despite that portrayal as diametric opposites. Science is science, and human sociology, per se, is not - it's humanity. Yet humanism can be brought into the sciences without in the least abrogating scientific method, as Gould readily demonstrates in all his books of essays.
(The third point of differentiation I will leave to another day. This relates to their divergent approach to ethical analysis, which Dawkins extrapolates from his work while Gould sees answers as outside the scientific paradigm.)
As a coda, I think it's important to remember that there is still _much_ to be discovered in both genetic analysis and paleontology. Sadly, both may be helped along by the greatest tragedy of our era, as climate change melts the Siberian permafrost, and unleashes both a torrent of field discoveries, and - if you credit some analyses - a vast reservoir of carbon that will superboost the climate changes already happening.