Many things went wrong in Iraq War II, and many things went right. That makes it like every other war ever fought. War is unpredictable by its nature, and often won by the side that adapts most quickly to the unexpected. For this reason, our military is already deeply engaged in evaluating lessons learned so far and figuring out how to change. Last month I had the opportunity to hear a presentation by military analysts about these issues. Here is my take on several of their more salient observations.

Too much too soon. Sometimes, what goes wrong is everything going right. Training and technology led to an execution of the war and a speed of advance so rapid that it surprised everyone and outstripped planning. The military expected to take seven weeks to reach Baghdad, but it took only two. The rapid sweep left a vacuum in its wake, and politics abhors a vacuum. Backfill put people into power who were sometimes no better than those deposed.

Yet no one would argue that we should have purposely gone slower. The surprising speed was a great help in many ways, but we hadnt adequately prepared for it. Unexpected success brings unique problems, as well as easily missed opportunities. We were not prepared to transition so early into a peacekeeping mode.

Reality TV. Before the war, many military leaders opposed wartime teleconferencing. They feared it would encourage premature decisions and their promulgation before careful review. But now most have changed their minds. Face-to-face discussions convey information that can get lost in carefully composed memos. Remote commanders get a better sense of the battlefield, and troops get a better sense of what the commanders want and expect. So far, teleconferencing has led not only to quicker decisions, but to better ones.

Cities are jungles. Iraq is mostly desert, but that proved mostly irrelevant. Virtually all fighting took place in or near cities, where visibility is low, and the greatest dangers are ambush, snipers, and booby traps more akin to the Vietnam experience than to Iraq War I. Over the past two decades, about 70 percent of U.S. military engagements have been urban for example, in Mogadishu and Panama City so we should have been better prepared. But we have grossly inadequate facilities for urban training, and our soldiers spend little time doing it. That must change.

The city environment also neutralizes many of our high-tech advantages. GPS doesnt work indoors, and often fails outdoors in narrow alleys. Our high tech communications also have problems. Some of our radios use frequency hopping (rapid changes in frequency) to avoid detection and location. But these systems work only when there is good signal propagation at all frequencies, a condition often not met in cities. So after a few weeks of urban fighting, some soldiers (and officers) had their families send them citizen band walkie-talkies from Radio Shack. When you are under fire, it may be more important to be able to call for help immediately than to maintain covert communications. This experience is reminiscent of Gulf War I, when families sent soldiers cheap GPS receivers.

Problems of precision. On D-Day in World War II, we dropped leaflets warning all French citizen who lived within 50 kilometers of the coast to evacuate. Our bombers and artillery demolished entire towns because it was the quickest way to eliminate a handful of entrenched Nazis. We now put a much higher priority on protecting civilians. Minimizing collateral damage has become a major constraint in modern war fighting. Our precision weapons are still not perfect, but they are getting much better; they reduced the number of noncombatant deaths to a much lower level than many predicted. As a result, most Iraqi civilians chose not to evacuate cities, and the massive refugee problem that many feared never materialized. But an unfortunate consequence of precision is that U.S. troops had to fight battles in the midst of innocents the people they were there to save.

The military describes the current situation as a three block war. In block one we are feeding and giving medical care to the Iraqi people. In block two we are patrolling, acting as peacekeepers and policemen. In block three we are engaged in full combat. In Iraq all three blocks are sometimes adjacent and coincident in time. Follow a suspected sniper, but be careful; if you throw a hand grenade into his hiding place, you may kill innocent civilians. Even a flash bang stun grenade is impermissible, because that could hurt a baby. This kind of fighting is so new that abstract planning is of little help. We are learning as we go.

Insufficient psyops. Psyops, for psychological operations, is the modern version of propaganda war. The important aspects of current doctrine include: talk the local language, know the local culture, and speak the truth. This last requirement surprises some people, but the military wisely makes the assumption that truth is our ally and the enemy of our enemies. If you never lie, you have hope of winning the trust of the civilians. Psyops worked remarkably well in Afghanistan. Our Special Operations Forces could speak local languages, and they could leverage the help of local people. The remarkable result: Afghans saw themselves liberated by fellow Muslims – their own countrymen.

But skill at psyops is largely a specialty of the Army Special Operation Forces, and their role in Iraq was important but limited. With the much larger force deployed in Iraq, psyops failed. The average Army soldier has virtually no knowledge of Arabic, and only superficial understanding of local culture. The Marines and the other forces have even less preparation in psyops.

Knowledge of culture goes well beyond being familiar with Muslim customs such as not passing food with your left hand, or not showing the bottoms of your feet. For example, if a soldier chases a terrorist into a building, he is expected to knock before entering. I found this requirement astonishing, but I spoke to several soldiers who had recently been in Iraq, and they confirmed this practice. It sounds ludicrous, but if you dont knock, and as a result you see a woman uncovered (maybe just her face) you could capture your terrorist but create several new ones. A husband or brother or both may feel obliged to take revenge for the insult in order to restore family honorregardless of their political beliefs.

Decentralized intelligence. In the continuing conflict, a surprise success is “Dragon Eyes,” a GPS-guided, remotely piloted vehicle that can be carried in a backpack. It looks like a model airplane, with a wingspan of only 1.2 meters and weight of 2.5 kilograms. It can be launched with a toss, or with a bungee cord. It flies quietly at an altitude of 150 meters using a zinc-air battery for power, and can transmit 18 frames per second of visible or infrared video from a range up to 10 kilometers. If spotted by the enemy, it is easily mistaken for a bird.

What makes Dragon Eyes so valuable is that it is easy to use (training takes less than a week), and it provides actionable intelligence information needed immediately. Soldiers deploy it when they need to know what lies behind that building, or near that bridge. Its cost is so low (soon to come down to $50,000) that it can be owned at the platoon level. (Generals dont waste time with things that cheap.) In the next two years, the Marines are slated to get 342 of these little marvels.

Despite the rise of the dragon, the most important source of actionable intelligence remains humint, short for human intelligence. Humint exploitation teams (HETs) get reports from sympathetic Iraqis, not only for big news items (where is Saddam Hussein?), but more frequently for less newsworthy (yet crucial) information such as the location of a roadside bomb. We now find a large number of these before they are set off.

Good news. It is important to learn from success too. The oil fields in Iraq were saved, even though Saddam had loaded them with explosives. His troops arrived at the huge Mosul dam to blow it up but our military (with decisive help from local Iraqis) prevented them from doing so. A great sandstorm, the kind that had foiled President Carters hostage rescue attempt in Iran in 1980, was endured without major problems. Most of the Iraqi infrastructure was preserved, so the post war recovery could proceed at a slow but measurable pace.

Some say the military is always fighting the last war. That is not my impression. We are far from mastering the new kind of urban war in which we do battle in the midst of innocents and demand extremely low collateral damage. But our armed services do a better job of learning from their mistakes than any other large organization I know. I wish that the rest of government, and scientific establishments, could learn with similar speed. Our strength is enormous – but, just as in biological evolution, it is often more important to be adaptable than to be strong.