Jerry Coyne’s “Seeing and Believing” and convergent evolution

In “Seeing and Believing” (2009-02-04), evolutionary biologist Jerry Coyne reviews two books that discuss creationism/”intelligent design” and their history, and attempt to reconcile religion and evolution: Karl W. Giberson's Saving Darwin: How to be a Christian and Believe in Evolution and Kenneth R. Miller's Only A Theory: Evolution and the Battle for America’s Soul. I enjoyed it enormously and commend it to your attention.

One point in particular caught my eye.

Giberson and Miller assert that the evolution of humans, or something very like them, was inevitable. Given the way that evolution works, they claim, it was certain that the animal kingdom would eventually work its way up to a species that was conscious, highly intelligent, and above all, capable of apprehending and worshipping its creator.

[…] To support the inevitability of humans, Giberson and Miller invoke the notion of evolutionary convergence. This idea is simple: species often adapt to similar environments by independently evolving similar features. Ichthyosaurs (ancient marine reptiles), porpoises, and fish all evolved independently in the water, and through natural selection all three acquired fins and a similar streamlined shape. Complex “camera eyes” evolved in both vertebrates and squid. Arctic animals such as polar bears, arctic hares, and snowy owls either are white or turn white in the winter, hiding them from predators or prey. Perhaps the most astonishing example of convergence is the similarity between some species of marsupial mammals in Australia and unrelated placental mammals that live elsewhere. The marsupial flying phalanger looks and acts just like the flying squirrel of the New World. Marsupial moles, with their reduced eyes and big burrowing claws, are dead ringers for our placental moles. Until its extinction in 1936, the remarkable thylacine, or Tasmanian wolf, looked and hunted like a placental wolf.

[…] Miller and Giberson are forced to this view for a simple reason. If we cannot prove that humanoid evolution was inevitable, then the reconciliation of evolution and Christianity collapses. For if we really were the special object of God’s creation, our evolution could not have been left to chance. (It may not be irrelevant that although the Catholic Church accepts most of Darwinism, it makes an official exception for the evolution of Homo sapiens, whose soul is said to have been created by God and inserted at some point into the human lineage.)

[…] We recognize convergences because unrelated species evolve similar traits. In other words, the traits appear in more than one species. But sophisticated, self-aware intelligence is a singleton: it evolved just once, in a human ancestor. (Octopi and dolphins are also smart, but they do not have the stuff to reflect on their origins.) In contrast, eyes have evolved independently forty times, and white color in Arctic animals appeared several times. It is hard to make a convincing case for the evolutionary inevitability of a feature that arose only once.

It would be easy to take too much from this argument against convergence. It is not evidence against convergence of intelligence that it has evolved only once as such, because there hasn’t been time since the first instance for a second to arise — it would be the most extraordinary coincidence if it were to arise twice in distantly-related creatures within the space of a hundred thousand years, so all we know is that we are first. Rather, I think he is arguing that it fails to be evidence for it — that there’s no reason to suppose that every feature of life to be found in the wild, such as the trunk of an elephant, is one that would be “re-discovered” by convergence if a particular branch had failed to flourish. By Coyne’s account, Giberson and Miller seem to be positing exactly that sort of “universal convergence”: Miller writes
But as life re-explored 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 we have, and capable of discovering the very process that produced it, the process of evolution…. Everything we know about evolution suggests that it could, sooner or later, get to that niche.
I think I see another problem with universal convergence. If evolution naturally filled every possible niche that it was presented with over time, we wouldn’t see the sometimes dramatic successes of species we ourselves carried across the seas. Rabbits are a dramatic success in Australia, for example, where “universal convergence” should lead us to predict that a species more perfectly adapted to the niche they occupy should already have been native to Australia on their arrival.

I’d be very interested to hear from those with greater expertise whether this objection holds up.

Note also this longer blog post on the subject from Coyne and this paper discussing evidence against convergent evolution of intelligence (ht Black Belt Bayesian).

Add post to: Delicious Reddit Slashdot Digg Technorati Google
(already: 4) Comment post


6.01.2010 21:56 Alison Rowan

I’m not sure you can prove much from the success of an invasive species. Invasive species tend to succeed because they don’t ‘fit in the eco-system they’re introduced to. Research has shown that they’re healthier because they leave their parasites and diseases behind ( and rabbits, in particular, had few predators in Australia in part because the giant birds that might have had them for lunch were already extinct (plausibly because of people).

In contrast, in an island ecology, new settlers (it’s usually birds as they can get there easier) will evolve incredibly rapidly to fill every available niche: Galapagos finches are the canonical example. But filling niches just means learning to eat all the available food (and not get eaten in the process), it doesn’t have to mean convergent evolution. There’s more than one way to crack a nut.

7.01.2010 12:19 Liz W

Is he saying that all attempts to reconcile Christianity [1] and evolution fail if our evolution was not inevitable, or only that these specific ones do? He may be right on the latter — I haven’t read either book — but I’d disagree with the former. I probably don’t have time to unpack the reasons fully right now (but will happily do so over the beverage of your choice at some point). The short version is that I don’t think “inevitability” is a valid concept in the first place when it comes to God’s interaction with creation; it looks valid superficially, but only because we find it difficult to conceive of a God outside time and tend to speak of him as though he were inside it, even if we know perfectly well that classical Christianity teaches otherwise.

8.01.2010 12:23 Doug

There’s a huge difference between “had to evolve X” and “very likely to evolve X”. Most biologists would agree there exist many X which would be likely to crop up if you — to use Stephen Jay Gould’s phrase — re-ran the tape of evolution. Eyes are a very good example.

But to get to where Giberson and Miller seem to want to be, you need to have a determinism about evolution which I really don’t think is tenable. And certainly not where X=humans as we understand them.

It seems so unnecessary to me as well — it’s very ‘God-of-the-gaps’, trying to squeeze God in to spaces in the existing science. I reckon this is always a Bad Plan because science shifts over time, and it weakens your argument (rhetorically, at least, if not formally) if you have to keep moving your conception of God to fit the gaps in the latest theories.

Much better strategically to define your God well outside that frame of reference and have Him properly Transcendent and Omnipotent — which I think is the sort of thing Liz is doing, and others. It leaves you open to other challenges (e.g. Invisible Gardener, Omphalism) but at least avoids the embarrassing possibility of being incontrovertibly proved wrong about God.

(Just to be clear, I’m being slightly mischievous here. I am quite sure that almost all theists who hold a Transcendent view of God are doing so out of genuine conviction, not because they think it’s the strategically clever way to conceptualise God to ward off potential atheist critiques.)

8.01.2010 13:11 Paul Crowley

I am quite sure that almost all theists who hold a Transcendent view of God are doing so out of genuine conviction, not because they think it’s the strategically clever way to conceptualise God to ward off potential atheist critiques

Well, this is one reason I like the theory of memes. Religious thought has all sorts of features which seem designed to make it invulnerable to criticism, and I don’t think very many of them are deliberately designed by someone with that express intent — it’s just that a belief system with that feature is better at propagating.

8.01.2010 13:43 Liz W

Doug: I agree with you that God has to be defined out of the scientific frame of reference (and don’t worry, I appreciate the mischievous tone!) Omphalism is certainly one of the odder consequences of creationism, but surely doesn’t follow just from holding a transcendent view of God? As for the Invisible Gardener, I’m with those who think Flew’s philosophy of language was too restrictive; I go with a more Wittgensteinian theory of meaning.

All this reminds me that I have a book of essays in honour of John Wisdom sitting on my to-read pile. Really must get round to that Very Soon Now… right after the other things that I also need to read VSN

(For those who don’t know, it was Wisdom who originally came up with the Invisible Gardener parable, which Flew then developed further; helpfully, Flew’s (1968) article is available in the <a href=””>unofficial Stephen Jay Gould archive</a>.)

8.01.2010 13:45 Liz W

Ah, the <a> tag doesn’t work? I’ll need to take a closer look at this Markdown business. Also not sure why that comment hasn’t threaded correctly.

Comment post