If you look at human beings, it’s immediately obvious that there’s lots of variation. Much of that variation is accompanied by contributing genetic variation – on average, your DNA sequence differs from another person’s by about 0.1 percent (about 6 million letters of DNA).

Where did all this variation come from? From mutation, broadly construed. Whenever cells divide, copying errors can occur – from single-letter swaps, deletions, or insertions, to duplications of entire genes (or entire genomes – all the DNA!). Bits of DNA can be chopped out and moved around; in addition, viruses can insert their DNA sequences and sometimes these can stay put, like genetic squatters. All of these sorts of changes have been observed in the lab, and if any of them end up in sperm or egg cells, the resulting mutations are passed on to the next generation. A recent estimate for the single-letter mutations in humans suggests that you inherited about 30 of them from each of your parents! (That doesn’t include many of the other sorts of mutations, including e.g. gene duplications.) In other words: we’re all mutants.

A few of those mutations can have a large effect (sometimes good, but usually bad, causing a miscarriage), but most have a small or even zero effect. So variation accumulates – thus that 6 million letter difference that you have, on average, with someone else.

So far, young earth creationists agree.

Now, this variation is spread as evenly through the human population as all the sex allows. (Travel these days is making it more even.) But imagine that a group of people becomes isolated from the rest, so that nary a sperm makes it across the gap. What will happen? Well, both groups will continue to accumulate variation. But mutation is a random process; so the two groups will accumulate some different genetic variations. Maybe small differences that don’t matter at all, that are completely invisible – unless you sequence their DNA. Or maybe larger differences that still don’t matter, but that are visible (like racial differences).

Still, young earth creationists are nodding their heads.

Now what if those two groups stay isolated for, not thousands, but hundreds of thousands or millions of years. Those small differences will become larger, no? Like in Darwin’s finches, for example, or rock wallabies. Of course, there’s something else that can encourage change, and that’s natural selection. If one of the variants within a population has comparatively more offspring, that variant will become more common, and may even push out some other variants. For example, in a small population of brown bears that were isolated in arctic conditions, those with a light fur mutation would have been better off: precursors to the polar bear. This is a small but significant difference between the two groups. Other changes accumulated too, like skeletal and tooth changes, and we now have brown bears and polar bears as distinct species. (They can still interbreed in principle; they just tend not to because they don’t share a habitat any more.)

Still, John’s happy with all this. Because these changes are all within a “kind” – the human kind, the finch kind, the bear kind.

So where does he get off the boat? Actually, I have no idea. The longer the process goes on, the greater the variety that will be produced. He asks: how could mutation bring about large changes? I ask: how could it not bring about large changes!? It’s inevitable, given population separation and enough time. When the ball starts rolling down the hill, it keeps going unless you stop it. What stops it, John? (Maybe you’ll say: time, the young Earth. In which case the debate turns to the age of the Earth, and the impossible rate of evolution you’ll need to generate the observed variation within your “kinds”.)

How large a change is required for a change in a creationist “kind”? John is careful not to say. However, he is very clear that chimps and humans belong to different kinds. On the other hand, he thinks that all the rock wallabies belong to the same kind. Problem: the genetic distance between chimps and humans is the same as the genetic distance between the two most distantly related species of rock wallaby: 2%. If mutation and natural selection produced all the rock wallabies from a common ancestor, as John seems to accept, then how can he deny that it produced chimps and humans from a common ancestor too? For the wallabies, he accepts the mechanism (mutation and natural selection) and he accepts the common sense reasoning (the logic of common descent). But not in the chimp/human case, despite the fact that the mechanism would do the trick, the common sense reasoning applies, and the distance travelled would be the same.

Worse for him: the genetic distances within the kangaroo kind, the bear kind, the ape kind, the bird kind, and [gulp] the bacteria kind are much greater. And he indicates that they’re related by common descent too. (Though he won’t answer my direct questions to clarify that.) Well, if the range of genetic diversity within birds can be generated by common descent, so can the range of diversity within mammals. It’s roughly the same amount. John simply cannot have all the “kinds” he wants.

I’ve just shown that mutation and natural selection can produce large changes via common descent, and therefore changes in “kind”. Of course, John sees this problem. What he wants to say is this: God designed the “kinds” so that a certain amount of genetic variation was encouraged within them; they can mutate up to that point, but no further. (That’s ultimately what he means for the variation to be “built in” to the kind – see his finches point here.) The problem is there is absolutely no reason to suppose that mutation is restricted in this way. There’s nothing to stop the ball from rolling.

Worse for him, we know the ball didn’t stop rolling, because we have a record of the path it took in all the DNA sequences discussed in the “language analogy” thread, plus all the hundreds or thousands of similar DNA-based trees of descent that have been produced by biologists. These paths are consistent only with no-barrier mutation-based change. If there were barriers, they’d show up as separate, isolated trees of descent for each kind. But that simply isn’t what we find.