US:Winning a war - it's not how many tanks you have but how much pain you will take, writes Shankar Vedantamin Washington
Let's say you and I are generals of opposing armies. You have 10 divisions and I have four. If your troops are as good as mine, when would our hypothetical war end?
The simplest answer is that the war will be over once four divisions of yours neutralise my four divisions; you have six divisions left, so you win.
The second answer, which is better than the first, is that the war should be over before it begins, because we both know what the outcome would be.
If we agree on the outcome without fighting the war, neither of us would have to lose four divisions.
It's the third answer that has particular relevance to real life as the United States finds itself enmeshed in a struggle in Iraq that has already claimed the lives of more than 3,000 US troops.
Before we get to it, however, we need to unpack the implications of the second answer. Given that the opposing sides in wars are often mismatched - the US and Saddam Hussein's Iraqi army, for example - the question that bothers many political scientists is why wars occur at all.
War imposes non-refundable costs - four divisions each in our hypothetical example - on both the winner and the loser.
If the weaker side is going to surrender anyway, doesn't it makes sense to cry surrender before the war starts? And shouldn't the stronger side be willing to accept a little less than total surrender if it can get most of what it wants without the loss of blood?
"The puzzle is if both sides knew what the outcome of the war would be, then both sides would be better off not fighting the war," says Dan Reiter, a political scientist at Emory University.
What this suggests is that wars are not primarily about conflicts between two sides, because most conflicts can be settled to the benefit of both sides without actual fighting.
War, Reiter and others argue, primarily reflects a lack of information: when two sides cannot agree on how much damage they can inflict on one another - and how much damage they can sustain - war offers a mechanism to provide that information.
"If both sides think they are going to win, they will not be able to reach a bargain," Reiter says. "You fight wars to find out who is stronger and who is willing to take more punishment. When enough information has been provided, the war ends."
Let's apply his reasoning to the war in Iraq, which will then provide us with the third answer to our hypothetical question.
Everyone knows that the US military is the strongest player in the conflict. Nearly four years into a war that has morphed into a complicated struggle with insurgents, what information remains to be provided?
The missing information is not about each side's ability to inflict damage, but the other component that Reiter described - the ability of each side to withstand losses.
The third answer to when the hypothetical war will end is this: it depends. You have an advantage of six divisions and, if we were both willing to fight to the last man, you would win. But no one is ever really willing to fight to the very last man.
What if you were really willing to lose only one division, while I was willing to lose three out of my four? Even though you outnumber me by more than two to one, I would win. In fact, I would win even if your troops were nearly three times as good as mine.
Knowledge about how much pain each side is willing to absorb is usually the most difficult information to obtain in war.
Both sides always exaggerate their capacity for pain, because if your opponent knew you were willing to lose only one division, he would just hang on until you reached that breaking point.
You need to deceive your opponent into believing you can tolerate huge losses.
In fact, says John Mueller, a political scientist at Ohio State University, you probably have to deceive yourself about it, too.
During the Vietnam War, Mueller says, the US assumed that beyond a certain point of losses, the North Vietnamese would break. "Both sides in Vietnam talked about staying the course - the issue is whether they were really ready to do it."
As it turned out, the North Vietnamese were willing to accept casualties on a scale virtually unprecedented in the history of combat. Vietnam war-era secretary of state Dean Rusk once calculated that North Vietnamese losses in the war, when measured as a proportion of population, were the equivalent of the US losing about 10 million lives.
American losses in Vietnam, by that same measure, were 175 times smaller - but even that was too high. "There was a breaking point," Mueller concludes. "It was the Americans who broke."