One thing that I did learn, though, is that the ID community has backed away from the “irreducible complexity” argument (since it’s been shot to pieces so many times, I guess) and brought forth its red-headed step child: “specified complexity”.
“Specified complexity” is the notion that if a structure is both unlikely and fits a recognizable pattern, something more than natural forces must be at work. A common example is Mount Rushmore. Every mountain is unique, so any given mountain feature is unlikely, and all mountains are therefore complex. Mount Rushmore, however, also fits a recognizable pattern of four specific human faces. It therefore has “specified complexity” that indicates intelligent design.
For simpler examples, they’ll use a string of letters. In the example on one of mike’s reference sites, you draw Scrabble tiles from a bag at random and line them up. To keep the math simple, we’ll assume that the bag has one of each letter and that you replace each tile you draw with another just like it. Suppose you draw this string of twenty-eight letters…
WETHEPEOPLEOFTHEUNITEDSTATESWhat are the odds of drawing that particular string of letters in that order? Something like 1 in 4*10^39 (four followed by thirty-nine zeroes). It’s extraordinarily unlikely that you would randomly draw that particular string of letters. Now let’s draw another one.
OMECIHUATLACEUCUCYOTICIHUATIWhat are the odds of drawing that string at random? Exactly the same. You might be fooled into thinking that the second string is more likely, though, because it looks more random. It doesn’t fit a recognizable pattern (English words that form a phrase that most Americans know). In fact, it resembles a sequence we would expect to see when drawing letters randomly from a bag or banging randomly at a keyboard; a string of letters with no recognizable meaning.
The problem with identifying “specified complexity” is that you have to have a context in which to recognize it. We recognize Mount Rushmore as “specifically complex” because we’ve seen many mountains, and we have expectations of what mountains shaped by the natural forces of continental drift and erosion will look like. We also have expectations of what man-made sculptures look like. On Mount Rushmore, we see shapes which are inconsistent with our expecations about natural mountains and consistent with our expectations about man-made sculptures, so we recognize a deliberate design.
But how do you recognize “specified complexity” without a context? If you were not familiar with mountains and sculptures and human faces, would you have any reason to think Mount Rushmore was the result of design instead of natural forces? It may not be as clear then.
Let’s use the simpler example of our letter strings. Suppose you were from a very obscure community in the remotest part of China. You don’t speak English, and you haven’t even seen the English alphabet. Under those circumstances, would the first string of letters look any more “specific” to you than the second? If you answered yes, ask yourself if the second string looks any more specific if you know that Omecihuatl and Acuecucyoticihuati are the names of goddesses from Aztec mythology.
This is the problem when ID advocates claim that they recognize the signs of design in something like the structure of the human eye or the bacterial flagellum: they have no basis for comparison. They don’t have a context in which to differentiate a designed structure from a natural structure. They can’t say, “We recognize that this bacterial flagellum is designed, because we’ve seen numerous examples of natural flagella and artificial flagella, and this one is consistent with the artificial ones.”
“Specific complexity” is a smoke screen, because there is no way to tell a designed biological structure from a natural biological structure. Consequently, there is no way to test a biological structure to determine if it is the product of design or natural forces. A claim that is not testable is not scientific.
I have to throw out a big credit to Jason Rosenhouse, the author of Evolutionblog, who wrote an excellent article that addresses the same subject for CSICOP.