I just spent about an hour googling "Recognizing Intelligent Design" (and a number of variations thereof), and this appears to be one of those "I know it when I see it" things.
The most interesting thing that I found was the difference between the published (Nature, Science, etc.) discussions and the underground (I suppose I can include FreeRepublic in that category, but there are LOTS of sites on the internet discussing this topic).
In the published articles, Dembski, Behe, and other proponents of ID harp on complexity, (either irreducible or specified) as the key indicator of design, and the (published) opponents of ID nitpick the concepts of irreducibilty and specificity.
In the underground category, examples of design focus on simplicity rather than complexity. A good example of that category is the picture of the Nazca spider that AndrewC posted -- very simple and obviously produced by an intelligence (and derived from a very complex design from nature).
Even seemingly complex designs by humans (an automobile, for example) are less complex than what they serve to replace (a horse). An airplane is less complex than a bird. A computer is less complex than a brain.
Complexity would seem to be a contra-indication of design...
Thanks for a good post. I agree with you as would any good programmer.
Exactly. The IDers are not consistent in this. The Great Designer would have designed everything. Supposedly, He left His signature behind in the form of "specified complexity". The claim is that natural laws don't give rise to specified complexity. And, in selection, only goal-directed selection can give rise to specified complexity. Thus, highly developed Leghorn chickens are designed by humans.
Natural selection is not goal-directed because the selection criteria themselves evolve in a non-deterministic manner. However, we have many examples of natural selection creating specified complexity. On one scale we see bacteria evolving methods for dealing with nutrient starvation or poisoning. On another scale we see termites building mounds.
Thus, the ID argument that natural laws do not lead to specified complexity utterly fails.
At this point, the IDers will perform a bait and switch. They'll pretend that the argument still holds for macro-scale processes. Why, you ask? They "know it when they see it".