Sunday, October 10, 2004

Is Wright's Argument Any Good?

Before continuing on with the discussion of what Dennett said vs. what Wright said he said, let's pause to consider whether Wright's argument is actually any good.

We have already seen that Wright's argument depends critically on the analogy between embryonic development and the evolutionary process generally. The analogy is roughly that in both cases we go from simple beginnings to functional complexity and diversity. If there is anything more to the analogy than that I'm not sure what it is.

The trouble is that there are obvious differences between ontogeny and evolution. In ontogeny we see the unfolding of a genetic program. Thus, the embryo goes from uniform simplicity to diversified complexity because of the instructions encoded in the genes and the cellular machinery that exists for decoding those instructions.

The process by which natural selection crafts complexity is entirely different. Natural selection does not represent the unfolding of a preordained program. Rather, it represents the end result of many years of trial and error. There is no guarantee that his process will lead to anything more complex than you started with. And even if you do, there is no guarantee that complexity will not increase for a while, and then plateau, having reached an evolutionary cul-de-sac. There is definitely no guarantee that anything like higher intelligence will emerge.

Now, elsewhere in the interview Wright expresses his belief that intelligence was indeed inevitable as a result of the evolutionary process. This assertion imparts a clear directionality to evolution. Dennett, by contrast, was far more circumspect about the sense in which evolution leads to complexity. He points out that over the “long, long, long haul” we can say that evolution will probably lead to human-like intelligence but that it is not guaranteed to do so in any finite length of time. He also describes the evolutionary process as being like a sawtooth function. Complexity will steadily increase for a while, but then the whole thing might come crashing down, say as the result of some natural catastrophe. But then it will start building up again.

So I think the analogy breaks down completely.

There is another point to be made. In embryonic development we do not have a single natural law or process causing a particular effect from a given cause. It is not comparable to saying that a rock released close to the surface of the Earth falls to the ground because of the effect of gravity. Rather, the process of development is itself so complicated that it needs to be explained in terms of something else. Most biologists would answer that the processes of development are themselves the product of natural selection.

For Wright's analogy to be valid we would now have to argue that evolution by natural selection is so complicated a process that we can not think of it as a simple natural law in the same sense that the law of gravitation is such a law. Rather, it had to emerge as the result of some sort of design process.

But this doesn't seem right. Evolution by natural selection is a simple consequence of having imperfect replicators competing for resources. That this process can, in some cases, lead to great complexity does not mean the process itself is incredibly complex. I think this is another point where Wright's analogy breaks down.


At 9:54 PM, Anonymous Anonymous said...


I think there is something important that you are missing, and it is to be found in Wright's book, "Nonzero."

In this book, he uses the logic of game theory to make an enormously compelling argument for directionality in both biological and cultural evolution, and before you write off his analogy, you should read the book and think it through. In his argument (as opposed to Gould's, for example) there are never plateaus (for various reasons, they are logically/mathematically impossible); there may be enormous setbacks (e.g. natural catastophes) but if life continues to go on afterwards,, there's an extremely good chance that complexity will steadily grow and that some form of intelligent life will eventually emerge. (It didn't have to be us, in other words, but it would have be something).

The directional view really does stand up to scrutiny, so the analogy doesn't break down as easily as you seem to think.

At 10:14 AM, Blogger Jason said...

Thanks for the comment. I have read Nonzero, but I don't find Wright's argument there convincing. Game theory has nothing to do with whether an organism will evolve itself into a corner. When I talked about complexity increasing until it reaches a plateau I was talking about engineering constraints.

For example, let us suppose that dinosaurs had remained the dominant form of life on Earth. Would evolution have produced more complex, more intelligent dinosaurs? Or had dinosaurs reached the maximum level of complexity possible given a reptilian body plan? Certainly there's nothing in the fossil record to suggest that over the course of 200 million years dinosaurs were getting brainier.

By contrast, brain growth in hominid species happened very quickly. This suggests that producing a species with human-like intelligence is not just a matter of species playing an increasing number of non zero-sum games. It is also necessary to start from a body plan that will allow such size increase to occur. I see nothing inevitable about producing such a body plan, at least not in any finite amount of time.

At 9:32 PM, Anonymous Anonymous said...

I would say that there's absolutely no way to know where the development of dinosaurs would have gone had it not been interrupted by an enormous chunk of rock from the sky.

From what little I understand about the history of dinosaurs--A. there was a branch of dinosaur development that clearly was the precursor to birds, which did indeed get more complex. B. there is *some* evidence that some dinosaur species were becoming social animals. As Wright demonstrates, the possibilites for steadily increasing intelligence go way up once animals form social cooperatives.

Now the question of whether or not a species of social dinosaurs could have evolved self-awareness and began the process of cultural evolution in a world with so many huge animals roaming around is open. But it's definitely not closed.

At 6:17 PM, Blogger Jason said...

I didn't mean to suggest that it was closed, but the burden of proof is definitely on Wright and I don't think he has met that burden. I'm not sure what you mean in suggesting that birds are more complex than dinosaurs. Also, the fact that there is some evidence of social behavior among dinosaurs does not answer my objection about engineering constraints.

I agree that we can't know what would have happened had the dinosaurs remained the dominant form of life on Earth. But Wright's argument requires that we do make some statement about what would have happened to them.


Post a Comment

<< Home