Saturday, August 21, 2004

The Limits of Certainty

Amidst the onslaught of Swift Vets charges and counter-charges, it's worth considering: 1) What is evidence?; 2) What kind of evidence should we demand from people making politically charged accusations against others? From a column written by Michael Kinsley last June:
______

According to a Harris poll out Wednesday, a majority of Americans still think the Bush administration was telling the truth before the war when it said it had hard evidence of WMD. A Knight Ridder poll released last weekend reports that a third of the populace believes that the weapons have been discovered. A Fox News poll last week found that almost half of Americans feel that the administration was "intentionally misleading" about Iraq's weapons, but more than two-thirds think the war was justified anyway. A Gallup Poll released Wednesday concludes that almost 9 out of 10 Americans still think Saddam had or was close to having WMD.

By now, WMD have taken on a mythic role in which fact doesn't play much of a part. The phrase itself "weapons of mass destruction" is more like an incantation than a description of anything in particular. The term is a new one to almost everybody, and the concern it officially embodies was on almost no one's radar screen until recently. Unofficially, "weapons of mass destruction" are to George W. Bush what fairies were to Peter Pan. He wants us to say, "We DO believe in weapons of mass destruction. We DO believe. We DO." If we all believe hard enough, they will be there. And it's working.

The most striking thing about polls like these isn't how many people believe or disbelieve some unproven factual assertion or prediction, but how few give the only correct answer, which is "don't know." In the Fox News poll, vast majorities expressed certitude one way or the other about the existence of WMD in Iraq, the likelihood of peace in the Middle East, and so on. Those who voted "not sure" (an even more tempting cop-out than the pollsters' usual "don't know") rarely broke 20 percent and usually hovered around 10. Four-fifths or more were sure about everything.

As someone who manufactures opinions for a living, it is my job to be sure. And my standards for the ingredients of an opinion are necessarily low. There may be a few ancient pundits such as George Will who still follow the traditional guild practices: days in the library making notes on 3-by-5 cards, half a dozen lunches at the club with key sources, an hour spent alone in silence with a martini and one's thoughts and only then does a perfectly modulated opinion take its lovely shape. Most of us have no time for that anymore. It's a quick surf around the Net, a flip of the coin, and out pops an opinion, ready-to-go except perhaps for a bit of extra last-minute coarsening.

Still, even the most modern major generalist among the professional commentariat likes to have a little something in the way of knowledge as he or she scatters opinions like bird seed. The general public, or at least the part of it that deals with pollsters, is not so cowardly. Most people, it seems, will happily state a belief on a question of fact that nobody knows the answer to, and then just as happily do a double back flip from that shaky platform into a pool of opinions about which they are "sure."

Pollsters themselves, and the media who report their findings deadpan, are partly responsible for this. Every news report about a poll result reinforces the impression that opinion untethered to reality is valid or even patriotic (and to be "not sure" is shameful). The modern pundit culture is also partly to blame, I suppose, with its emphasis on televised argumentation. Viewers do not always grasp the difference between low standards and no standards at all.

Are there weapons of mass destruction in Iraq? Sure there are in every sense that matters, reality not being one of them.
Link