Why there are no missing links

This topic came up earlier this week: creationists are always yammering about the "missing link" and how it's missing and therefore evolution is unsupported by the evidence. It's total nonsense, since evolution doesn't predict a "missing link", but it seemed worthwhile to explain why, since there was a recent publication of some exciting data that demonstrates the real complexity of the situation.

Jim Foley and John Hawks and Carl Zimmer have written up the story of the Denisovans. To summarize, another group of Pleistocene humans have been sequenced, called the Denisovans — their identity is murky, as they've only been recognized by a few bones, but the results show that they were genetically distinct from both modern humans and Neandertals, another Pleistocene group that has been sequenced. Like the Neandertal story, in which some Neandertal genes (less than 5%) were introduced into some modern human (European and Asian) populations, what we know about the Denisovans is that some of their genes, about 5%, also spread into a subset of modern human populations, in this case the Melanesians.

That's awesome stuff. There are all these splintered bits of ancestry that come together in complex ways to produce the human species, and that's why there is no missing link. Many people have this false notion that our evolution was a matter of a panmictic gemisch of people rolling fatefully down the smooth channel of history, everyone mingling, all of them tracing a common lineage back and back to a discrete ancestor. It wasn't. Our river of time looked more like this, a braided stream:

This is what we mean when we talk about populations having structure. Branches emerge, whether we call them Neandertals or Denisovans or modern humans, and they are distinct but there can still be genes flowing between them to some degree. Even within modern humans we have structure, where groups maintain a kind of genetic integrity over space and time; I can look at my own recent lineage and see how my mother's Scandinavian connections were maintained through several generations in America, or I can look at my father's pedigree that goes back about 400 years in the New World and see that even though they were constantly scudding along at the very edge of the American frontier, mingling with Native Americans and black slaves and freedmen and Chinese railroad workers and Japanese farmers, somehow in their marriages, nothing but Scots/Irish/Anglo-Saxon names turn up.

That's the nature of a species: many channels, many populations, not just one, separating and merging with circumstance. It's always been this way; when humans and chimps first diverged from their common ancestors, it wasn't like one tribe went left, one went right, and they never talked to each other again — it was many streams of ancient ape populations twisted about amongst each other, gradually disentangling to each form a spectrum of divergent channels for each separated species.

When a creationist demands to see the "missing link", it's like looking at the picture of a river above and asking for the one drop of water that started it all. There wasn't one. The question doesn't even make sense. It's why BioLogos looks so ridiculous when they worry over whether we can trace our ancestry back to two people, Adam and Eve — of course we can't, humanity has never been represented by just two unique individuals, and even considering the issue seriously reveals an absence of understanding of how populations evolve. It's so confused, it's not even wrong.

(I notice that Greg Laden comes to a similar conclusion, that using the term "missing link" should be avoided, but the nature of his argument looks about as tangled and discursive as the picture of the braided stream above…so maybe it's more true to the reality?)