since any precursor to an irreducibly complex system is by definition nonfunctional.
Nonfunctional as a mousetrap
maybe, but hardly non-functional period
. Behe's analogy fails because it predetermines a highly specific function
to evaluate success or failure against. Let's consider his humble mousetrap.
(1) a flat wooden platform
(2) a metal hammer
(3) a wire spring
(4) a sensitive catch
(5) a metal bar
And each of those piece would certainly make for a poor mousetrap, or, as Behe would put it, a non-functional mousetrap. But what if you started with needing to just get start congregating mice to a specific location? You have a mouse infestation, but for purposes of the party you're throwing tonight, you really just need to make sure that there are fewer mice running around your living room in random directions. A wooden platform would serve as an adequate landing point to put bait, say, cheese, without messing up the carpet. So you put a wooden platform with a few hunks of cheese in another room, in hopes that the mice will have a greater tendency to hang out in the room with cheese so that your guests don't see so many of them bumming around the living room.
The night of the party...
Success!!! There were only, like, 20 mice running around the living room as opposed to the normal 50 or so that typically happen. But, as it happens, some people are still turned off by the presence of mice, and you've noticed that attendance to your shindig is still steadily going down. Well, hell, in order to keep people coming to the party, perhaps you should do something to not just lure mice away but to keep them away
. So an easy solution is, when the mice are congregating around the wooden platform with the cheese, you just kill them. Those bastards aren't going to be bugging people in the living room after that! So you grab a metal hammer and put it in the room so that, when you're in there, you can kill the little buggers without messing up your shoes.
And that works, but you're kinda tired of the whole polling process of having to periodically pop into the room and see if there were any mice to kill. A-ha! You'll just put a little sensitive trigger there that will make a noise whenever a mouse grabs himself some cheese. That way, you'll know
when you need to pop into the room to kill some mice. Easy-peasy.
Unfortunately, while a few weeks of this has improved party attendance, you realize that if attendance doesn't significantly improve anytime soon you're going to have to cancel these parties. The mice numbers are down, but it's apparently not enough. You've got to seriously
reduce these mice numbers so that they just plain don't show up at the party. But, you just don't have enough time to go mouse killin'. Besides, it's not like you're home 24 hours a day and always within earshot of the room. You need a way to have this wooden platform kill the little buggers for you, so you don't have to be such an active participant.
So you start experimenting, trying out a few things here and there. Maybe you can give the mice a heart attack or something by scaring them with a really loud noise! You attach a spring to your sensitive trigger to make a much louder "Thwap!" sound. You try this for a few weeks, but the mice population hasn't really been positively or negatively affected by it (upon observation, you notice that the mice just don't pay any attention to the noise). Well, that little spring isn't really reducing the efficacy of your device, and you're sorta lazy so you don't bother to get rid of it.
Then, one day, quite by accident, you leave your hammer on the end of the spring. Providence has shined down upon you! Much to your amazement, you witness a greedy mouse eat some cheese and "Bam!" the catch triggers, the spring disengages, taking the hammer with it and smacking the mouse on the head! Of course, your repeatability of this result is...lacking, to say the least. About 99% of the time the spring just goes "Thwap!" without taking the hammer with it. So you attach a string to the spring and hammer via hooks so that the hammer goes with it every time.
98.5% failure rate. That sucks. The hammer just lands at random locations, occasionally making contact with the head of a mouse. You fiddle around with string length, attaching strings to random locations on the wooden platform, spring, and hammer until the hammer lands roughly in the same place every time, i.e. the vicinity of a mouse's head. Ah...success. And then you notice that, with all of the hook placements you've put on that wooden platform, you now have a pretty easy place to put a metal bar for the hammer to rest against that is in close enough proximity to the spring to allow for direct coupling! You slap a metal bar on there, get rid of the strings, and now you're good to go!
So, my little storytelling exercise is not without fault, and many, many aspects of that can be picked at, or possibly refined to better match with an actual reality. But I think I've shown the idea that an irreducibly complex system like a mouse trap cannot
be created through small, iterative, successive implementation is simply false. Maybe some of those intermediate steps couldn't be considered a mouse trap (e.g. wooden platform with a hunk of cheese on it), and so some of those intermediate steps could be considered as non-functional as a mousetrap
, but hardly non-functional, period.
I didn't even have to talk about scaffolding.
P.S. Holy crap-in-a-hat, 18 new replies while I was typing this one out. I spent way
too much time typing this drivel not to hit 'Post', but now I'm afraid that bunches of what I typed are just invalid
. Apologies all around.