Irreducible complexity and endosymbiosis

At Billy Dembski's blog, Granville Sewell wonders What if we DID find irreducibly complex biological features?:

In any debate on Intelligent Design, there is a question I have long wished to see posed to ID opponents: “If we DID discover some biological feature that was irreducibly complex, to your satisfication [sic] and to the satisfaction of all reasonable observers, would that justify the design inference?”

No, at least not necessarily. It isn't clear why we should make that immediate leap.

There are two related problems with that jump. First, the fact that something meets the definition of Behe's irreducible complexity doesn't mean it couldn't have evolved (Behe has acknowledged this point). Second, the fact that something is unevolvable by known means (which is what Sewell and Behe usually seem to mean by IC) doesn't mean that the only available explanation is design.

A simple example comes from natural arches in the desert Southwest. Just like human made arches, they are structures which are currently irreducible (remove the wrong bit of rock and they'll tumble) but which unlike the Golden Gate Bridge, no one invokes a designer to explain structures like Rainbow Bridge.

Why not? One reason is that, when Rainbow Bridge formed a few million years ago, there wasn't anyone around capable of conceiving of or constructing such a thing. If there was no designer and no construction crew, there was no design. There's also no particular evidence that anyone would want to have built a bridge at that site, nor any evidence of mechanical scraping or other external interference with the rock, so there's no positive case to be made that the design and construction actually happened. Perhaps more importantly, there are good explanations for natural arches already. The geological theory behind those explanations makes testable predictions, and we can verify those predictions against measurements in the field. "Irreducible" doesn't mean something isn't accessible by gradual processes, just that, having reached a certain point, you can't remove parts.

On top of that, there already are objects which biologists would gladly agree seem to meet Michael Behe's strict definition of irreducible complexity, but which no one (not even Behe, so far as I know) thinks require any sort of "design inference."

There are tiny things called mitochondria in each of our cells. Those organelles process oxygen, oxygen which would otherwise kill the cells by interacting with important molecules. Each of those mitochondria in each of our cells consists of various structures, including two membranes surrounding a circular chromosome and the necessary molecular entourage to produce the proteins with which it processes oxygen and releases molecular energy (ATP) into the cell.

Remove the mitochondria, and the cell dies. Remove parts of the mitochondria, and they die and take the cells with them. Half a mitochondrion would be unquestionably useless. So where did it come from, and why don't we invoke design to explain it?

Part of the reason is that there's no evidence that there was anything at all capable of designing, let alone manufacturing, cells with mitochondria in them when the first mitochondrial cells existed. The exact date is still a bit fuzzy, but those first cells date back at least 1.2 billion years, and probably further back. All multicellular life has mitochondria, and the equipment and knowledge necessary to manufacture such a thing from scratch in a laboratory didn't exist a decade ago, and may not exist today. Invoking design seems a little silly under the circumstances.

But there's a better reason not to fuss with such an implausible idea. That circular chromosome found in mitochondria is very similar to the circular chromosome found in some bacteria. And those two membranes are exactly what you'd find if one cell engulfed another into a vacuole (as they do before digesting food particles). What if some ancient cell engulfed a bacterium which could process oxygen, but didn't manage to digest it? That engulfed cell would soak up oxygen from the engulfer, and the engulfee would digest partially processed molecules from the cell, and emit excess ATP for the engulfer to use. Win-win. It's a testable hypothesis that dates back at least to the 1950s, and which was championed most prominently by Lynn Margulis.

Margulis cited similar evidence to argue that various other organelles in the cell, and other forms of biological novelty above the cellular level, can be attributed to the effects of symbiotic cooperation. The response to this proposal has been mixed. With mitochondria (and chloroplasts) there's good reason to think that endosymbiosis can explain how you'd get what looks like one cell living inside another. With structures like peroxisomes or cilia, the evidence is less compelling.

Margulis rightly points out that her mechanism differs from those which form the core of the neo-Darwinian synthesis: mutation, selection, genetic drift and gene flow. Because she proposed a new evolutionary mechanism, she does not describe herself as a neo-Darwinist. This isn't a challenge to the basic practice of science nor to common descent. She simply recognized that existing evolutionary mechanisms were not adequate to explain certain biological phenomena, and developed a new idea to help explain that.

As I've point out before, the sort of cellular endosymbiosis we infer took place with mitochondria can be seen in transitional form in the world around us. Researchers at the University of Rochester and the Craig Venter Institute produced yet another example. Fruit flies infected with a bacterium which enters the host cells – Wolbachia – were found to have imported part of the bacterial genome into their own. Mitochondrial genes have also been found gradually migrating into the genome of host cells, and one of the authors of the Wolbachia paper explained "In a way, Wolbachia could be the next mitochondria. A hundred million years from now, everyone may have a Wolbachia organelle." Again, the discovery of large swaths of DNA suddenly appearing in the fruit fly genome required an explanation beyond those developed for the Modern Synthesis (it is, as I put it above, "unevolvable by known means"). Scientists needed an explanation beyond those which existed in traditional biology. But there was no rush to invoke invisible intelligent genetic designers or engineers. A novel natural processes was more than adequate, and yields new testable results and important new avenues of research (which a design inference wouldn't do).

(I've written about Wolbachia before, if you're interested).