I'm not, at this point, a huge fan of certain parts of "intelligent design" -- the main proposition of which is positive and religious. I think such impressions are subjective (what are the criteria?), and not something you can "prove". I believe the movement has a valid concern, but it should have focused on raising objections rather that inserting positive assertions.

On the other hand, others -- particularly some vocal atheistic scientists -- have no more nor less business proposing that things happened without intention, or as a result only of "blind" processes. (An equally religious proposition.) If you can't detect intention, and it's wrong to discuss intention, then its absence must be equally impossible to detect and wrong to discuss. And if it's fine to teach one, then it's equally fine to teach another.

But on the subjective criteria demanded for "evidence of design", I think many will stand self-convicted.

Consider these reactions:

In 2001, Alexei Erchak at the Massachusetts Institute of Technology, in Cambridge, unveiled a groundbreaking light-emitting diode that suffered none of the light losses that can plague ordinary LEDs. Erchak's device was modified with a special reflective layer and an optical structure, known as a photonic crystal, that together captured light that would otherwise have been lost and channeled it into a useful beam. The design funneled six times as much light into its beam as an unmodified one, an improvement that astounded LED researchers at the time.

Now a UK team says the structure that makes Erchak's LEDs so special is not unique after all. Look hard enough and you can find it in the fluorescent wings of male African swallowtail butterflies of the Princeps nireus species. What is more, the structure of these wings is subtly different from the MIT design in a way that may offer clues for improving LEDs further....

"Who knows how much time could have been saved if we'd seen this butterfly structure 10 years ago," says Vukusic.

P. nireus may have more to teach. In Erchak's LED, with its perfectly periodic crystal, light is better transmitted at some angles than others. But there may be a work-around in the quasiperiodic structure of P. nireus's photonic crystals.

In other words, when observers thought the idea originated with a scientist, they felt it was an "astounding" feat. But, well, not so much when we learn that somehow the swallowtail butterfly does the same thing, but even more efficiently than the researcher conceived. What would be a brilliant design from a scientist is not at all indicative of "intelligence" when found in an insect.

I agree that criteria for evidence of design (or for evidence of anything, really) is subjective. Some may require a higher threshold of proof than others, or may vary depending on subject matter. But certainly one should avoid double standards, no?

But there is actually a way of addressing such subjective questions. In court, we appeal to the idea of the reasonable man. And here we have a panel of optical researchers who feel, yes, that design X, considered on it's own, is astounding, and even more efficient (e.g. intelligent) than their own best design efforts.

Dawkins would be quick to point out that said appearance is an illusion. But when we say it is an illusion, we are also admitting, inadvertently: Yes, it does have the appearance of design. We are also appealing to an objective (not subjective) criteria for knowing what "design" looks like. So oddly enough, Dawkins and the ID'ers are in the same court: both substantially agree (even if they protest otherwise) there is an absolute, objective criteria for what "design" should look like.

But they differ regarding the history of how the apparent design came about. Dawkins responds by saying life was created in a series of small steps, which sometimes go down a blind alley, with no apparent end goal in sight.

Natural selection is the blind watchmaker, blind because it does not see ahead, does not plan consequences, has no purpose in view. Yet the living results of natural selection overwhelmingly impress us with the appearance of design as if by a master watchmaker, impress us with the illusion of design and planning.

There are several problems here:

Dawkins transfers the argument about "design" from the appearance of the object itself to the process by which it was created. In effect, Dawkins seems to be conceding the appearance is consistent with design.

So the question is no longer Paley's question about finding a watch sitting in a field, and inferring a watchmaker. Now the question is, in essence, about Dawkins' "magic box", as I will call it. Instead of finding a watch, fully-assembled, we find a "magic box" which, if you put some simple watch parts into it and shake it for a while, out will come a very simple, working timepiece. Put the timepiece back in, toss in a few more spare parts, shake for a while longer (maybe toss it in your car's trunk) and out comes a more complex chronograph. Do it a while longer, and you get a computer or human being.

So Dawkins answers the question about the watch by proposing a "magic box" -- but doesn't notice that the box in itself seems to connote design. The complexity of the watch has been transfered cleverly to the complexity of the box's behavior. And such a box would indeed require an explanation!

Dawkins' weakness here can be seen most clearly in his resort to computer software. Other than life itself (the appeal to which would be circular), there are no examples of such "magic boxes" in nature -- so Dawkins must resort to using computer programs to illustrate how a random process can apparently produce apparently intelligent behavior over time.

Yet that just goes to prove the opposite point: the software "works" because it was made to work by a developer. (I've written genetic software: it's not easy to design) The "goals" apparent in the output were first entered as constraints in the software itself -- and the random-number generator is merely a magician's stage prop to distract you from what's really going on.

To illustrate, let's say I wanted to prove that random processes can, over time, generate valid math equations. So I write software which substitutes random values into "X + Y = Z" and then only prints the valid equations. Did the random number generator choose the valid equations? Did it create new "information"? No, my software, which already knew how to do math sums, and was already programmed with the goal of selecting valid equations, did the choosing. So the "information" we see coming out does not arise spontaneously from the random number generator (which is just a distraction -- a series of loops would have done the same thing), but actually from the selection criteria the developer entered in the first place. Information in => information out.

The other tact Dawkins takes is to distribute the watch's assembly over time. Yet even the simplest self-sustaining organism is already far more complex than Paley's watch. (And indeed, the probability numbers seem to indicate it should never have happened.) If we had a simple self-maintaining robot, we might not be surprised if, even by chance, it happened upon a few useful adaptations. But Dawkins' story starts with finding, not merely a watch in a field, but a self-maintaining and self-modifying robot which is far more complex than Paley's watch.

But where did the robot (and it's initial programming and behavior) come from in the first place? Again, this is so much slight of hand: Dawkins' simply shuffles the complexity and information away from your view, like an untidy housekeeper shoving the mess into a closet before company arrives. (To be clear, I think he deluded himself too, because he didn't think clearly about what he was doing. You can't get something for nothing, and you can't make the "information" go away by shoving it around.)

Mapping backwards, just as we'd ask who wrote that wonderful genetic computer program which could replicate information with random changes, and prefer some outcomes over others, so also we'd ask where the information in the natural equivalent came from. Not the randomness (which is readily available), but the programming. It's as if Paley stumbled on a computer with a genetic algorithm loaded onto it. Surely, you'd infer at least Intel and a developer.

This is an argument Dawkins has soundly lost, but I suspect it has done fine at achieving its main purpose, which is polemic, not rational.

There's another problem here, but I don't want to bore you further, and it's not my main point, so I'll dispatch it as quickly as possible: By offering a definition of a "blind" watchmaker, Dawkins also saying that a "seeing" watchmaker (i.e. God) would have, in contrast, done other things. Dawkins is thus indirectly offering a definition of God, and saying his particular view of God is the only God who can exist -- a God whose design style would be Danish Modern (minimalist, with smooth lines everywhere), not baroque (seemingly pointless curlicues and ornaments galore).

But where did Dawkins discover that only one particular kind of God, with one particular "design style" could exist? This is, after all, a theological statement he's making. When I look at the content of the bible, I don't see a neat progression from Abraham to Messiah. I see a story of a group of people who sometimes go forward, but often go backwards. There are stories with ambiguous implications. The royal dynastic line splits, and both halves seem to go nowhere. The Christian Messiah doesn't conquer, but instead, unexpectedly, shows mercy. The God depicted therein is certainly not a god of neat, linear progressions.

Arguments of this sort thus inevitably boil down to: "Well, I don't like your conception of God, and since mine clearly doesn't exist, yours can't either." That's a fine emotional, subjective rant, but it's certainly not an example of ineluctable logic.