By my own reckoning, intelligent design is a study concerned with information and complexity analysis of empirical data. Specifically, one of its functions is to seek algorithmic explanations for sets of empirical data. At a certain statistical threshold, it is assumed that an explanatory algorithm must be preferred to explanation by appeal to random actions. At this point, the algorithmic explanation is categorized based on whether it is 1. a suitable description for the observed data and directly consequent of other known algorithmic patterns (viz. physical laws) or 2. a suitable description for the observed data and not the product of other known algorithmic patterns (i.e. the algorithmic pattern is not an emergent property of simpler rules, but rather exists unto itself). Any phenomena that falls strictly into the latter case demonstrates intelligence (or "design").

Examples:
1. Condensing of gas into stars. A star has certain simplifying properties that distinguish it from a more randomly disbursed cloud of hot gas. This can be explained by more fundamental principles such as gravitation and the strong nuclear force. It thus falls into the first category. (This is preferred over the equally valid explanation that the aggregation of molecules is simply an extremely improbable anomaly of their random movements.)
2. Organization of stones into a monument. The combination of the stones follows simple geometric patterns. However, this geometric organization is not an emergent property of such other physical properties as gravitation. It thus falls into the second category. (This is preferred over the equally valid explanation that the aggregation of bricks is simply an extremely improbable anomaly of quantum tunneling and whatnot.)

I hope it is immediately evident, first, that this method of categorization is reasonable and, second, that it easily extends to all phenomena: chemical, biological, etc. I am often rather taken aback by accusations that ID (intelligent design) cannot produce evidence of design. Obviously, my second example gives an instance of where it can, and there are innumerable other instances. There is no reason that I can see to divorce science from analysis of intelligent action. That is the specific concern of forensic science--to determine how complex patterns correspond to intelligent action. The scientists employed at SETI certainly think that they can differentiate between random data and data produced by an intelligence. And, just so, ID is a perfectly sound way to analyze collected data.

Something I really need to emphasize is that ID is not an "alternative" to evolution. ID is an analytic tool that can be applied to evolutionary theory as well as any other natural phenomenon. With ID we are simply examining whether evolution is an emergent property of other well-known universal principles, or whether it exhibits statistical anomalies which would best place it in the second enumerated category (meaning that evolution should probably be explained--at least in part--by appeal to some kind of intelligent intervention).

That being said, ID does also have the potential to prefer certain models over others, as any good scientific method should, and it is possible that evolutionary models may be found to be inferior to other models.

As such, it is my opinion that ID is indispensable to modern evolutionary theory, at least if we want evolutionary theory to be counted as a real science. Unfortunately, many scientists of modern day submit to the philosophy of mechanistic naturalism. That is unfortunate because, bounded by their philosophy, these scientists must believe that a suitable evolutionary pathway will always exist. If a simple pathway does not exist, a more complicated one must be chosen, or one more complicated still. Think epicycles. If it is your founding belief that the sun rotates about the earth, then instead of rejecting the theory when explanations become complicated, you are bound to simply introduce more complications into your explanation.

I think particle physicists give a much better example of "good" science (though I imagine many physicists are also mechanistic naturalists, for them it is not damaging to their scientific inquiry). Physicists may posit, for example, new particles, but these are not fact until they are actually observed, and if an entirely new particle is observed that was not predicted, well, back to the drawing board. In fact, I really do not think anyone in physics would be too terribly astonished if it turned out the proper theory was something altogether different from anything presently anticipated. There are many competing theories in modern physics about how reality works--traditional quantum theory, string theory, loop quantum gravity, etc. Physicists are content to go about their way without worrying too much that a competing theory might have too much public support or some such nonsense. Neither do I think there should be all this upset about postulations that evolution may not be as pan-explanatory as it is sometimes made out to be.

So that is my view on the purpose and use of ID. It looks for patterns of design. If it finds them, great. If it doesn't, great. The important thing is to keep looking, and to keep writing about it--both pro and con--and to not get all boxed in on what we are willing to believe about physical phenomena based on our personal philosophies.

I do, of course, have a position on what the data indicates. I'll give some substantiation of that in a subsequent post.