Bollettino

There's an excellent little book by Italian researcher Massimo Piattelli-Palmarini , "Inevitable Illusions." P.P contends that our usual cognitive mechanism suffers from certain mental "tunnels," especially when it comes to probability, causal inference, and what I would call the narrative urge -- the drive to create, out of events, stories that are consonant with the pattern of stories we like. P.P's section on Predictability in Hindsight seems particularly apposite given the state of the War. The War evolved in two stages: resistance in the South, and less resistance in the center, followed by a wholly unpredicted collapse in Baghdad. The fedayeen, who nobody mentioned in the press pre-war, fought as well as they could; in contrast, the Republican guard, who accrued tons of print, were terrible fighters. The Republican guard fought the American war -- conventional confrontation between two armed forces -- and were wiped.

P.P reports an interesting experiment, comparing two cases. In one case, a real result, and real prior data leading up to the result, was given to the subjects of the experiment, who were then asked if they could have predicted the result from the prior data. In a second case, they gave the same data, but an opposite result (in other words, they lied). In both cases, the subjects were confident, from the data, that they could have predicted the result. As long as we think we have a certain result, we immediately create a plausible backstory; and in the creation of that backstory we become confident of our power to correctly appraise each piece of evidence.

It is this quality that Jack Shafer makes fun of in a recent hit on Johnny Apple, the NYT journalist. You'll remember that Jack Shafer's first hit on Apple made fun of his prediction that Afghanistan would be difficult to govern. You'll rememember we commented that Shafer's remark -- that Afghanistan is comparable to San Francisco in governability -- was the acme of dumbness. Given the firefights this last week in Afghanistan, and the government's own reports on the return of the Taliban, one would think that Shafer's newest piece on Apple would be tempered by the humility induced by his own rashness in pronouncing Afghanistan pacified territory. In fact, Shafer makes the odd assumption that results consequent to American victories are historically, and thus militarily, irrelevant. So for him, the taking of Kabul by the Americans has closed the book on Afghanistan -- an assumption that the soldiers of the British Empire could probably have told him something about. He makes fun of Apple for asking the common sense question about what defines the end of the War. Here is Shafer thinking himself a real cock of the walk:

"By April 6�a whole day later [from the first article Shafer analyzed]�Apple constructs new victory benchmarks for the coalition in "Allies' New Test: How To Define Victory." It's not enough that the Americans and Brits have encircled Baghdad and subdued Basra in less than three weeks of fighting and eviscerated the Iraqi army and its irregulars. His impatient lede asks, "How and when, it seems worth asking, will the United States and its allies know they have won the Iraqi war?"

Apple doesn't answer his own question directly but implies that the allies' recipe for victory pie would have to include a new, democratic government in Iraq; the elimination of Saddam Hussein; the uncovering of his weapons of mass destruction; and the departure of U.S. troops�sooner rather than later.

By defining victory "up," Apple subtly retreats to his original, March 27 position that nothing but quagmire, quagmire, quagmire awaits the United States in Iraq."

You'll notice that Shafer is accusing Apple of doing exactly what the subjects in PPs' experiment did. And you will notice he is making the accusation by ignoring evidence that Apple's original predictions about Afghanistan are coming true, since Shafer has decided that the defeat of the Taliban in 2001 was the definite end of the Taliban -- he's anchored his certainty there. Meaning that he's protecting himself from the Predictability in Hindsight problem by hemming and hawing on his own predictions, and editing facts to reflect badly on Apple.

This will happen a lot for as long as celebrations of the Meat Machine's demise are broadcast on tv and the radio.