Sunday, October 03, 2004

Verification bias

My old buddy, Stumbling Tongue, sent me this good link on verification bias (I know, I know, Rathergate is an old issue). It seems that folks have been trying to decipher this thing called the Voynich manuscript for 400 years, but then this guy came up and proved it was a hoax--there was nothing to decipher. The reason this new guy was successful was because he actually considered the hoax possibility and investigated it.

The article talks in terms of an "expertise gap"--where the expertise needed to solve a problem actually exists *between* well developed expertises--and outlines an approach that systematically works to bridge this gap and produce a solution. And while I agree that this is very real and there is much value in solving that problem, the only verification bias I could find was at the beginning, ie. "hey, maybe this thing is a hoax, let's investigate that instead."

Deliberate contrariness is very valuable, very simple, and remarkably difficult to do in practice. The next time you feel convinced about something, try to come up with four or five reasons why that position is wrong. Feel your brain pretzeling, but once you are done you will have a more, dare I say it, nuanced understanding of the situation. And when I say "reasons", I mean really good, well thought out reasons, not stupid, lazy, or generic reasons. When I ask people now to critique work I've done, I ask them for a list of five reasons why my argument is wrong. I really ought to be doing this myself, but there you go.

I had dinner with some friends last night. These folks are good people, smart people, close buddies, and they invited me to a presidential debate party they are having whenever the next debate is on. Ye Gads. I had not read this bleat at the time, but the images flashing through my mind were frighteningly similar. Here's quote:
I could talk about the blogger party tonight where the luminaries of the Northern Alliance gathered to watch the debate, and peck out snark and insight.... I hate the debates. I have a vision of 65 million undecided Americans tuning in and making a snap judgment for all the wrong reasons. Wow, he pounded the podium to emphasize each word - but the other guy pounded each syllable. What’s this about sealing Fallujer? Is it leaking? Did they have a flood?

But mostly I hate the debates because I simply cannot abide hearing certain statements I’ve been hearing over, and over, and over again. I can’t take any more talk about bringing allies to the table. Which ones? Brazil? Mynmar? Microfrickin’nesia? Are there some incredibly important and powerful nations out there whose existence has hitherto escaped me? Fermany? Gerance? The Galactic Order of the Belgian Dominion? Did we piss off the Vulcans? Who? If we mean “France and Germany,” then please explain to me why the reluctant participation of these two countries somehow bestows the magic kiss of legitimacy.
Now I understand the value of people getting together and confirming their bias and prejudice. If shared bias and prejudicial isn't the basis of friendship and loyalty, I don't know what is. And if I create an image of chimps picking ticks out of each others' hides in a big monkey-tick-eating session, it's not because I look down on such events (which I don't) but because I acknowledge their tremendous social value and contribution to group cohesion and feel no need to give them airs beyond that. But other things generate that kind of group warm fuzzies too (autumn bbqs, shared drinks, hell, even apple picking expeditions) without being, you know, annoying to others. Whatever.

Go read this Slate article on why it's rational not to vote. Then read this Times piece on why you should vote anyway. (Here are the reasons: 1) your vote will not make the difference, even in close races, and 2) vote out of Duty, especially if you vote for the Superior candidate. Look, I know 2 is lame, but what do you expect out of the Times?)

On the futility of debating at all, read this excellent Stumbling Tongue article on what works, what doesn't, and what does "work" really mean anyway? An excerpt:
It seems obvious that “debate-style” ways of arguing only antagonize the majority of people, who are offended by all direct contradiction. And when someone is antagonized, they are not receptive to anything new. So, except in very rare cases, debates shut down communication. Witness the blogosphere: except for such paragons as Eddie Thomas, it’s just a mass of people shouting past each other. (The sad thing is, people logon to find like-minded people to shout with, not even oppositely-minded people to shout at — never mind anything as radical as civilized disagreement.)

But what is the alternative to “debate-style” arguments? There is simply something dishonest about spending your energy trying to seduce someone into agreeing with you. For one thing, you are usually hiding the fact that you’re not explaining yourself directly. Seduction isn’t communicating an idea with a person; it’s trying to insert an idea into a foreign mind. Seduction targets the man, not the idea, and so to that degree it is sneaky and manipulative.
Well said.

The closest I got to honesty around debates was in the historic Owl Bar in Baltimore, where a friend agreed that her "not-Bush" vote was based on the belief that there was a way of solving terrorism that did not involve turning the Middle East to glass or replacing totalitarian despotisms with more open, tolerant societies. She did not know what that was, she did not think Kerry knew what that was (or that he was even thinking about the problem) but she felt that circumstances would push him towards that anyway. Personally, I am OK with blind faith being a reasonable motivation, just don't confuse it with anything else.

While we are on the subject (a little tangentially) of cognitive function and dysfunction, I don't think this philosophic take on cognitive fallacy is useful, although I think the fallacies themselves are right on (via AL Daily. Yes it is true that people confuse authority with knowledge/expertise, care too much about motivation rather than argument, etc. etc. I don't want a laundry list, I want a systematic model that probably has its roots in anatomy, is filled with behavioral evolution, and has equations cooked up by economists. I honestly don't think it's complex, I just think that we are too close to the subject to be able to figure it out well.

One last thing. A friend of mine who has been reading political blogs and seeing them become increasingly polarized cites this as the value of traditional "mainstream media": "Yes, they are often wrong, and generally biased, but they don't get so hysterical that they become unreadable by half the country, and having a centralized source of fairly good information is valuable." I acknowledge that the centralization instinct is a natural one in humans, but I also beleive that the chaos produced by shouting (not this informed debate carnard) does more to attentuate bad decisions than anything else.


Post a Comment

Subscribe to Post Comments [Atom]

<< Home