The late Stephen Jay Gould more than once observed that, were it possible to roll back time and re-run evolutionary history, we would most likely end up with very different results. Minor differences in circumstances can lead to very different evolutionary pathways.
Others, most notably Simon Conway Morris, hold that evolution is far more predictable than Gould would have had us believe. As evidence, they cite the interesting phenomenon of convergent evolution, where different species evolve strikingly similar features in similar circumstances. A classic example is the similar body shapes of dolphins, sharks, ichthyosaurs, and (at more of a stretch) penguins: these predators’ ‘designs’ enable them to move quickly under water. If mammals, fish, ichthyosaurs, and birds evolved such similar shapes for moving at speed in the same environment, the argument goes, evolution must, to some extent, be predictable.
Those who maintain that evolution is more predictable than we might suppose sometimes go so far as to claim that upright, bipedal, intelligent life was almost inevitable on Earth. Had it not been for that pesky asteroid, they say, the world would now, quite possibly, be being ruled by dinosaurian, rather than mammalian, humanoids. This despite the fact that, as far as we know, upright, bipedal, intelligent dinosaurs failed to evolve in the 180-million years that dinosaurs actually did rule the earth.
Jonathon Losos's interesting book sets out to explore both the phenomenon of convergent evolution, and the possibility of performing experiments to assess evolutionary predictions. In the first part of the book, he describes many examples of convergent evolution. In subsequent sections, he describes experiments in the wild, and in more controlled environments, to determine whether the accuracy of various evolutionary predictions can be tested.
Although convergent evolution is a genuinely fascinating phenomenon, it is considerably less remarkable when the species in question are closely related. When presented with similar environmental challenges, is it really at all surprising when closely related species evolve similar solutions? Evolution can only tinker with what is already there; how many fundamentally different tweaks can be made to closely related lizards, for example, to help them evade a new predator? In fairness to Losos, he makes this point more than once, but, to this non-expert at least, it seemed as if more might have been made of it. There is a world of a difference between two species of stickleback, to cite another example, evolving brighter colours in the absence of predators, and dinosaurs evolving into intelligent humanoids. Even if small-scale convergent evolution of closely related species is common, extrapolating to claim that the evolution of intelligent humanoids is almost inevitable is another thing entirely.
Sensibly, Losos doesn't spend too much time examining arguments about putative humanoid dinosaurs—although he does eventually make his own position clear. This book is primarily about the experiments: how scientists have begun to test evolutionary predictions, and to assess how particular examples of convergent evolution come about. Both of which strike me as far more interesting and useful than coming up with untestable hypotheses about where dinosaurs might have gone next.
An entertaining book on an interesting subject.
Disclosure: I received a free review copy of this book from the publisher.