Saturday, May 26, 2007

The Map: Probability and Evolution

Many people, including myself, have failed to appreciate the degree to which the basic evidence for evolution undermines a theory of design. Common descent is a near fatal blow to any theory of design.

In neo-Darwinian Evolution (NDE), every life form can trace its lineage back to the very first, most primitive life forms. The tree of life has no floating branches that begin in mid air. Everything is attached to the same trunk. More generally, evolution could consist of two intertwined trees of life, but even then, there are no floating branches.

In contrast, a designed tree of life can have floating branches. Indeed, every branch could be floating, or just one, or any number in between. For example, a designer could have chosen to make man (or gazelles) a separate floating branch with no ancestors on the tree of life. Or the designer could have chosen to make each family a different branch. Or each species. Or any combination of the above. There are millions of species, so there are at least trillions of topologies of descent that one can imagine given generic design.

After Darwin made his prediction about common descent, we did not know as much about the fossil record. Nor did we know about DNA and genomes. It was possible that we might have found that humans only coincidentally look like apes, and that we might not have even had DNA. Or maybe apes would have had a triple helix instead of a double helix for their DNA.

That's not what we found. We found that all life appears to trace its line back to a common ancestor. Thousands of research projects have shown this to be the case. We share most of our genome with apes, and we even share accidental mutations, in a pattern that proves we are related by a common ancestor.

Bayes' Theorem demands that we grant NDE a huge boost factor in our confidence, and a corresponding huge suppression factor for generic design. Simply put, there's only one way to do descent with NDE, and (more than) trillions of ways to do it with generic design. And after we look at life, we find life did it like NDE. We should therefore favor NDE over generic design by factors of trillions.

Here's an analogy. I have two boxes. Each box is filled with shuffled decks of cards. In the first box are normal decks of playing cards. Each deck in the second box is special, and contains 52 copies of the Queen of Spades. You don't know which box is which, but you go to the box closest to you. You draw a deck, then turn over the top card. It's the Queen of Spades. What are the odds you're looking at the box of ordinary card decks? Well, though it's unlikely, it's still possible that you went to the box containing ordinary decks, but you're 52 times more likely to be looking at the box of special decks. You cast aside the rest of the first deck and pick up another. It too has the Queen of Spades as the top card. Again, it is possible that this is a box of ordinary decks and that you've just been unlucky. Very unlucky, but it could happen. Eventually, the probability that you went to the box of ordinary decks becomes so small that you have to conclude that you're looking at the box of special decks.

This is our situation. Not saying anything about a designer, we must say that each of the designer's design decisions is equally likely. And yet the evidence clearly shows that the generic designer would have had to make the one design (against trillions of others) that evolution would make.

(I mention a "generic designer" because advocates of Intelligent Design are loathe to say anything about the designer, despite the fact that they have no one in mind but God.)

So, the bottom line is that a generic designer is effectively ruled out by common descent.


The objection to this argument was that the assumption of equal probability for each possible descent design decision in not scientific, but theological, because what God would do is not something to which one can apply probability.

It's really rather perverse that theists first obscure God by talking about a generic designer in scientific terms, but when we try to reason about a generic designer scientifically, we are told we cannot do that without getting all theological. I was also told that the probability of the designer choosing a particular course of descent was "I don't know" and that "I don't know" cannot be plugged into an equation.

When I raised the counter-objection that probability is specifically designed to apply to unknowns, I was taken on a ride wherein the entire field of probability analysis was brought into question.

Stumbling Blocks

Were it to be shown that evolution alone could explain life and our mental faculties, the concept of objective morality would be at risk. This is confirmed by the kinds of discussions that arise when NDE is discussed in conjunction with morality. Therein, Christian theists blame the atrocities of Hitler and Stalin on evolutionary biology, as if the way we developed historically implies a murderous political philosophy.

I hate to bring it back to psychoanalysis, but theists have no answer to the probability argument. They are simply ignoring it because it is inconvenient.

No comments: