10 January 2009

On being created

Most online discussion of evolution gets into selection, transitional fossils, and suchlike, often from a viewpoint of attempting to explain why, yes, it really does work like that.

All of this—to my lasting annoyance, hence this post—is, from the one side, generally factual, but not generally effective, because it's missing the point.

Understanding descent with modification, speciation as process (and remember that this is the increase of diversity, really, because species are an artifact of taxonomy, not an actual substantial thing, even in tetrapod zoology and less so in botany or microbiology or the study of the great diversity of creatures that do not descend from a critter with four limbs), and genetics came later, sometimes much later. However fascinating the mechanics of all these things are, they're not the core insight behind the Darwinian view of life.

That core insight is this: "many more individuals of each species are born than can possibly survive".

This is an observation; a fact, independent of any one person's mind or understanding, verifiable, and verified, by many people in many times and places. (It is true of non-industrial human populations, too. Industrial and post-industrial human populations are something of a special case, one where most children don't die but where they don't reproduce at replacement levels, either, and not reproducing is not, from an evolutionary perspective, necessarily different from having died before you could reproduce.)

Once you start trying to explain that fact, to determine why this inescapably ubiquitous fact is so and what it means for living creatures, you wind up with two things; the definition of a Darwinian individual (born, dies, recognizably itself between those times, reproduces its kind, the reproduction includes the possibility of modification in its descent), and the idea of modified descent as a response to environmental conditions.

No one has come up with a better explanation of why many more living creatures of each species are born than can possibly themselves survive to breed; no one has come up with a better explanation of the filtering mechanism exerted by the not-all-survive being related to the environment producing descent-with-modification. Lots and lots of very smart people have tried very hard.

If you want the Darwinian view of life to be wrong, you have to come up with a better explanation, one with more predictive and explanatory power, of why many more are born than can possibly survive. (Pointing at post-industrial humans won't do it; there will be a lot of people pointing back at fish and pine trees and ducklings, and more pointing at "co-operation in groups" and third-order consequences of being specialized for that.)

The folks who assert that the Darwinian view of life must be wrong are generally emotionally motivated; they're also generally trying really hard to not acknowledge that this is, at core, a theory explaining differential odds of survival.[1]

Until there's an acknowledgement that explaining this simple, narrow, factual thing—many more are born than survive—is the point and power of Darwin's theory of evolution, all of the clever reasoning and really impressive understanding (HOX genes and fig reproduction, just for starters, are well into ow-my-head territory, and someone figured that out) doesn't carry any possibility of emotional conviction, and that emotional conviction is what's required.

[1] It's one fearsomely elegant system; all the diversity of life comes out of five properties (the Darwinian individual definition, above) and two rules. Rule one is that differential reproductive success—of a population of genes—is affected by environmental fitness, so organisms become better fitted to their environment over time, and rule two is that change cannot happen that reduces differential reproductive success with respect to the remainder of the population. That's it. That's all it takes to go, in time, from RNA to butterflies and beluga.

EDIT, because from Brook's comment I was clearly too terse with rule two. Let this be a lesson to me to avoid footnotes.

"change", there, is standing in for "a change in the distribution of traits (or genes) in a population of individuals, where the population is sufficiently large to function as a statistical universe with respect to the distribution of traits, the selective process is constrained so that it will always approach the local maximum for fitness in a continuous way."

The location of that local maximum may be moving (the climate is getting wetter or drier, say), and it can go away entirely (post-glacial flooding seems to have drowned the nesting areas of the north Atlantic albatross population...) but the selection process is continuous and approaches the fitness local maximum as determined by the environment. (Which is not obliged to be either continuous or consistent, which is where a lot of the interesting stuff gets into things.)


brooksmoses said...

I'm confused about your second rule, in part because I'm not sure I'm understanding you. In particular, I'm not sure whether you're talking about change at the level of the individual or at the level of the population; both seem wrong.

There are many individuals born which have a chance of reproductive success that's lower both than their parents and than the remainder of the population; mutations are after all random and are generally unhelpful to greater or lesser extents. So, for an individual, change (with respect to the parents) that reduces the differential reproductive success is certainly possible, and the rule seems erroneous at that level.

For the population as a whole, the fact that change that reduces the reproductive success does not happen is an emergent property of the fact that individuals with low reproductive success have fewer descendants, so at that level the rule seems redundant.

Graydon said...

I've edited the post to be clearer, and while you'd think that at the level of populations the rule would be redundant (because it's certainly implied strongly by the idea of differential dying), the continuous approximation of the local maximum for fitness in a particular environment part is important, because it's a tool for understanding what's going in with respect to environmental change (climate, animal movement, etc.). It's also in some sense counter-intuitive, and the folks who actually teach this stuff have to keep teaching that bit.