Here's the most important thing about The Ancestor's Tale that I couldn't fit in my review. I kept noticing how little Richard Dawkins mentioned the other celebrity evolutionary biologist of our time, Stephen Jay Gould. After all, Gould was a prominent character in many of Dawkins's previous books, cast as the brilliant paleontologist misled by leftist ideology.

Gould was famous for his attacks on adaptationism--the notion that the creative powers of natural selection are behind all sorts of fine points of nature, from jealousy to 11-year cicada cycles. Dawkins was an ultra-Darwinian fundamentalist in Gould's opinion. Gould thought that evolutionary biologists should widen their horizons. They should consider that things that look like adaptations might just be by-products of how organisms develop. They should consider how random catastrophes can override all of natural selection's work, wiping out fit and unfit alike. They should consider how selection may work on many levels--not just with selfish genes, but with populations, and even species. (This was why Gould thought punctuated equilibrium was so important.)

Dawkins would have none of this. He downplayed the importance of developmental constraints, of mass extinctions, and species selection. His attitude towards punctuated equilibrium has been, "Yeah, but so what?"

And then, in The Ancestor's Tale, the battle of Dawkins v Gould disappears. One possibility for the disappearance might be that Dawkins is respecting the dead. (Gould died in 2002.) Perhaps, but the silence is still weird. That's because in this book, Dawkins moves into the heart of Gould territory: the murky realm of evolutionary history. Dawkins has always been at his most eloquent and powerful when he ignores history. His arguments about selfish genes and the like are, at their heart, exquisitely organized reasoning. He did sometimes bring in actual details from biology to these arguments, but only as illustrations of his points. In The Ancestor's Tale, Dawkins takes on 4 billion years of evolution, in all its strange exuberance. The evidence--the fossil record, the relationships of living species revealed by DNA, and so on--dwarfs our explanations for it. We know there were giant scorpions in the oceans, and that they disappeared. But we don't know why. We know that birds survived mass extinctions 65 million years ago, but their close relatives--feathered, flightless dinosaurs--did not. But we don't know why. And so on. You'd imagine that this territory might make an adaptationist a bit anxious.

Dawkins handles himself very well as he moves across this terrain. He knows his natural history, his plate tectonics, and all the rest. He frequently throws up his hands about why the history of life took the turns that it did--although he remains confident that the best way to find the answer is to keep adaptationism first and foremost in mind. Gould shows up only in footnotes. Punctuated equilibrium remains an interesting empirical question but not a major principle. Species selection doesn't even show up in the index.

Yet I thought that sometimes Dawkins didn't acknowledge that some of the episodes in evolution he was writing about still raise some important questions about his selfish-gene centered view. I found this to be the case especially when he wrote about the origin of animals. Animals are multicellular organisms, in which trillions of cells come together as an individual, which then reproduces through just a few sex cells. Animals also descend from a single-celled ancestor. Making that transition isn't simple. A bunch of cells won't just come together and agree that a few of them will get to pass their own DNA on to the next generation. That doesn't make evolutionary sense. The only way to decipher this transition is to view evolution taking place at different levels--at the level of the genes, of the cell, and of the individual. Changes at one level may work against changes at the others, or they may all end up working together. I got interested myself in this subject a couple years ago while writing an essay for Natural History, focusing on the work of Richard Michod of the University of Arizona. It seems to me that the origin of animals is a case where Gould's multi-level selection may work well. Now, Dawkins might disagree, and yet he didn't even mention this challenge to his own views, let alone tear it apart as you'd expect from his previous books. In a 630 page long book, I find this omission puzzling.

It's always possible that Dawkins might eventually accept that in this case multi-level selection is important. He'd probably go on arguing that in most cases a gene-centered approach to life works best. I found it very interesting that he ends the book with a discussion of religion, saying that he suspects that many who call themselves religious would agree with Dawkins (an outspoken atheist) on many of the things he has to say about nature. He describes how "a distinguished elder statesman of my subject" was arguing for a long time with a colleague. The statesman said jokingly, "You know, we really do agree. It's just that you say it wrong."

I imagine Dawkins talking to Gould there.