Three lessons here. The first: Virtually every post mortem that relies on the blame game has the same result. There is always someone who asked for more resources and warned of the risk before the event. There are always enough intelligence indicators so that once you go back — knowing the pattern of actual events — it becomes possible to predict the past with 20-20 hindsight.

The problem is that the post mortems and hearings tend to be useless. Every prudent security officer has always asked for more; the indicators that could provide warning with 20-20 hindsight will still be buried in a flood of other reporting that warns of crises that don’t take place; U.S. officials will still have to deal with what intelligence experts call “noise” — the vast amount of reporting and other data that make it impossible to sort out the right information until the event actually occurs and the patterns are known. All of this makes it hard to know what request or warning ever matters. …

The second lesson is that we cannot deal with crises like the political upheavals in the Arab world, or the more direct threats that countries like Iran and North Korea can pose, unless our diplomats and military advisers take risks — and more casualties — in the process. …