Comments on evolutionary convergence from Kenneth Miller’s Only a Theory, page 152-153
… Run the tape of life again, starting from the Cambrian or wherever one might choose, and it’s almost inconceivable that you’d get hairless bipedal primates with brains big enough to endow them with self-awareness, reflective thought, and calculus.
But, upon reflection, that’s not really the issue.
The issue is not whether the exact scenarios of this planet’s actual natural history would be repeated. They clearly would not. The genuine question is what sort of living world would emerge from a second or third running of the tape of life. Although we cannot predict the detailed outcome, this much we do know: Life vigorously explores adaptive space, and it finds its way to the same niches in that space again and again. One can, for example, study the ecological roles played by dinosaurs in various habitats—herbivores, scavengers, predators, keystone species—and discover that the exact same roles were quickly taken by mammals, birds, and reptiles when these great animals disappeared. Not even the most gifted naturalist could have looked at the world of the Cretaceous and predicted exactly how the balance of nature would settle in the postdinosaur world–but even the dullest would have been confident that settled it would be.
Turning our attention to the special case of our own species, we can be fairly confident, just as Gould tells us, that our peculiar natural history would not repeat, and that self-awareness would not emerge from the primates. Indeed, we would have no reason to suppose that primates, mammals, or even vertebrates would emerge in a second running of the tape. But as life reexplored adaptive space, could we be certain that our niche would not be occupied? I would argue that we could be almost certain that it would be—that eventually evolution would produce an intelligent, self-aware, reflective creature endowed with a nervous system large enough to solve the very same questions that we have and capable of discovering the events. It would be to maintain, for no particular reason that this corner of adaptive space was found once by the evolutionary process but could never be found again. Everything we know about evolution suggests that it would, sooner or later, get to that niche.
I’ll admit that there’s nothing to be gained by pretending that one can settle this question of repeatability with any certainty. So far as we know, nature has conducted the experiment just once, and the result was us (plus a few million other species). Science demands repeatability, and that’s not possible in this case. Perhaps at some point in our own development we will discover a second experiment, a planet with characteristics similar to our own, on which we can truly test the grand principle of evolutionary convergence. Maybe that data will even be good enough to satisfy a Steve Gould. But that’s a question for another book, and maybe even for another century. The point for today is that it’s perfectly reasonable to maintain that evolution as we know and understand it was almost certain to produce a species like ours under conditions that prevail on Planet Earth.