Saturday, December 06, 2008

Specified prolixity

PZ Myers cited in vain

We nonbelievers grew weary a long time ago from the efforts of proselytizing Christians to “explain” Christianity to us and “share the good news” with us. Sorry, guys. We've heard the so-called good news, but it's like watching a news report on Fox: We simply don't believe it. The recruiters for the cult of Christianity assume we don't know what the Bible says, but a lot of us do know. We just reject it.

We find it insulting to be assumed ignorant.

Creationists are trying to borrow this page from the nonbeliever's book. You see, they understand evolution and natural selection. They just reject it.

Oh, really?

We probably need to make some allowances for the conscious liars who use discredited arguments because they're effective rhetorical devices for advancing their holy cause. (Does anyone really think that Duane Gish doesn't know that his argument about the second law of thermodynamics is invalid? He was refuted in detail time and time again, but he never abandoned it in his public presentations because people continued to fall for it.) They know their arguments are bogus, but the “noble lie” is a pragmatically mendacious approach to redeeming souls for Christ. (Apparently the ends justify the means.) But many of the anti-evolution polemicists are likely to be entirely sincere. Their arguments reveal their profound ignorance rather than their willful dishonesty.

A case in point is this month's issue of Acts & Facts from the Institute for Creation Research. The December 2008 edition contains an essay titled More Than Just “Complex.” The author is Brian Thomas, who is described as a “science writer.” Thomas cites a remark by PZ Myers that criticizes creationists:
The lesson of Darwin is that unguided natural processes have the ability to generate complex functionality, so it takes more than just showing complexity and function to demonstrate purpose. Creationists don't understand that at all, so they keep whining “it's complex!” as if they have brought up an irrefutable argument for design, when they've done no such thing.
In typical creationist fashion, Thomas goofs up the quotation a bit. He truncates PZ's final sentence by deleting his conclusion and moves the exclamation point outside the quotes (“it's complex”!), as if PZ were not imputing the exclamation to the creationists. Sloppy, but not surprising.

After the Myers quote, Thomas rolls out the we-know-but-reject argument:
The reason that “the lesson of Darwin” is rejected by creationists is not because they don't understand it. Rather, it is because they rightly observe that “unguided natural processes” cannot generate both complexity and functionality...
(By the way, I truncated the final sentence in my quote from Thomas's article. Go check out the original to see if I did violence to it.) Thomas goes on to complain that natural selection is incapable of generating “specified complexity,” but then falls promptly into a creationist's classic post-hoc misconstruction of the problem. He does this by rhapsodizing about the protein complexes known as chaperonins:
Chaperonins have a precisely-placed enzymatic active site, detachable caps, flexible gated entryways, a timed sequence of chemical events, and precise expansion and flexion capacities. Each of the parameters—size, shape, strength, hydrophobicity distribution, timing, and sequence—represents a specification. With each additional specification, the likelihood of a chance-based assembly of these parts diminishes…to miracle status.
A miracle! God is obviously required.

Let's try a simple explanation of this inane interpretation of complexity. If I draw a hearts royal flush from a standard deck of cards, that's pretty close to a miracle. (No, this has never happened to me.) There's only 1 chance out of 2,598,960 of drawing the ace, king, queen, jack, and ten of hearts in a random selection from a standard deck. (That big number is 52C5, the number of ways of choosing 5 items from a set of 52 distinct things.) We could choose even less likely things for our illustration, but this will do as an example.

We understand, of course, that any hand of five cards has one chance in 2,598,960 of being drawn, but most of those hands are nondescript and insufficiently interesting to catch our attention. A hearts royal flush, however, is a killer winning hand. If you were to predict (specify) in advance of a deal that you were going to get a hearts royal flush, it would be a pretty miraculous example of meeting your specification if the dealer were then to deal out those cards to you. (People might call for an investigation.)

But a hearts royal flush is not the only “killer winning hand” that the dealer could have given you. There are two royal flushes made up of red cards. Would you be unhappier with a diamonds royal flush? I think not. It would serve your purposes quite as adequately. It would look a little odd if you had “specified” hearts and then drew diamonds, but you're still a winner. And your chances of a red royal flush are twice that of a hearts royal flush alone. You know what you did? You over-specified.

Did you grasp the lesson? Even if we don't go on to talk about equally satisfactory outcomes with black royal flushes and results that are nearly as good with other straight flushes and four of a kind?

While Thomas waxes eloquently over all the details of chaperonins (I wonder how much he really knows about them?), listing all the specific features they have as they exist today, he completely misses the possibility (likelihood, actually) that there are millions (billions?) of other formulations that would have created alternate-universe chaperonins of equal functionality.

Try this little experiment the next time you're playing poker: When you pick up your five cards, recoil in amazement at the hand you're holding (no matter what it is, strong or weak), throw the cards down on the table face up so that your companions can see them, and scream, “Oh, my God! It's a miracle! This hand occurred in the face of odds greater than one in two million! Surely an Intelligent Dealer was involved in specifying this hand!”

Depending on the intelligence of the dealer in question, he may lean over and punch you on the shoulder.

12 comments:

PZ Myers said...

Wait -- it should work. In principle, the exact opposite of a poker face should be just as successful as being expressionless. It's just a lot more annoying.

Anonymous said...

Outstanding. Now, a non-statistics-trained-person question:

Is it misleading for your hypothetical pokerp layer to exclain, upon getting his five cards, "Wow, I just beat the odds by better than 1 in 2 million"?

Isn't there a 100% chance he was going to receive SOME combination of 5 cards, since (your scenario assumes) he was going to be dealt a hand? And should that somehow be worked into the 1-in-2-million odds?

It really is a ingenuous question, even if it sounds persnickity.

Zeno said...

It's a perfectly reasonable question, Anonymous. You're right: The player will inevitably get some combination of 5 cards. It's the probability of any one specific hand that has only one chance in 2,598,960. If you had predicted that particular hand in advance, it would be amazing (or some kind of sleight-of-hand parlor trick, more likely). The Acts & Facts article implies that the way chaperonins are is the way they must be, therefore being specified in incredibly exact detail. But just like there are many different winning hands in a card game, there are many different protein combinations that perform identical functions. The "specified complexity" is largely an illusion.

Anonymous said...

I'm with you on this. My favorite example is more familiar to software people since that is my field. This is only an approximation, but did you know that a team of 1,000 monkeys typing on PC keyboards produces a Microsoft Vista release every 78 minutes.

llewelly said...


... did you know that a team of 1,000 monkeys typing on PC keyboards produces a Microsoft Vista release every 78 minutes.

hm. 1000 monkeys. Let's say 180 keys pressed per monkey per minute. Say 7 bits per key press. That's 98,280,000 bits in 78 minutes. That's only about 12 megabytes of total data. Assuming the monkeys type as fast as the fastest human typists - 120 wpm or 720 keys per minute only gets us to 48 megabytes. Now - I haven't run vista, and I don't know how big it is, but I do know XP was much, much larger than 48MB - hell, it wouldn't even run with 48MB of ram. I think there's something seriously wrong with your numbers.


n

llewelly said...

By the way. If your hand wins the pot, hang on to it for the next round of poker. That's what evolution does.

wstachour said...

I appreciate the solidity of this argument, though I don't expect the mystically-inclined to be swayed.

But we never get to it: "God did it" is no argument at all, since the "God" is surely much more miraculous that whatever He is invoked to explain; ergo, "God" only constitutes a bigger problem, not a solution to our current one.

(Not that I think you doubt this...)

Anonymous said...

The reason argumentation won't work is because all apologetics do not stem from a quest for learning, but out of a somewhat exhausting effort to preserve a very strong emotion that belief in a god emits. This means that people who try to sell god either as an intelligent designer or as the magic-sky-fairy of the bible aren't really listening to a word you say. They're interested in feeling safe with what they're told, and the only "discussion" you have with them is emotional. I'm saying this because die-hard creationists often reject even stark contradictions to their claims, and their continual to do so is so irrational, I can only assume it's a result of some emotional tie to their preheld beliefs.

Perhaps the only way to actually change that is not with clever reasoning (which is always a treat for me, though), but with clever bullshit-tactics. Bullshit-tactics do a lot to sway the gullible particularly *because* they address a person's feelings and not his intellect. That would be dishonest and a bad precedent, but I'm willing to bank some money that it will be effective, at least in the short-run.

Oh, and I'm not endorsing that in any way, nor would I ever do something like that. I'm into science because it attempts to be sans bullshit.

Shygetz said...

llewelly has the right of it; it's even better than you implied. Which is why most evolutionary solutions you find are more analogous to four-of-a-kind than a royal flush; they are REALLY REALLY good, but not optimal as they had to be built up from other pretty good but not optimal precursors. So, you're more likely to end up with four-of-a-kind (which you can build from a pair, which is a rather common hand) than the optimal royal flush.

Anonymous said...

did you know that a team of 1,000 monkeys typing on PC keyboards produces a Microsoft Vista release every 78 minutes.

I knew there would be a good explanation for Vista! Who knew!

shrimplate said...

I'm saying this because die-hard creationists often reject even stark contradictions to their claims, and their continual to do so is so irrational, I can only assume it's a result of some emotional tie to their preheld beliefs.

Might I suggest that it's a dysfunctional tie to preheld beliefs, in the sense that there's probably some emotional "mistake" also being made by the sort of people under discussion.

Their hide-bound adherence to irrational and ultimately emotionally-dissatisfying creationist bullshit serves some kind of function, but it's so fucked up it's just hard to deal with.

I guess all I mean to say is that holding creationist beliefs is not just irrational, it's emotionally and psychologically suboptimal, too.

Anonymous said...

Feynman made a comment on odds like that. He would start off a lecture by saying something along the lines of, "You know, on the way in here today I saw a car with a license plate that read AQ1RW7. What are the chances!"

I'd say the much bigger problem with his argument has nothing to do with the probabilities, it has to do with age-old creationist misconception that chaperonins developed all at once. We're not talking about one draw of the cards, so the analogy is seriously flawed to begin with. Chaperonins have an evolutionary history.