Dirty war

Though many idealists will welcome this insistence on "limited" rather than "total" war, Carr makes it clear that his advocacy of protection of civilians from armed conflict is based on sheer pragmatism. His is a quintessentially Hobbesian view; people will support the regime that best ensures "the safety of ordinary citizens by establishing a professional, politically disinterested and highly disciplined army." Break from that guideline and sooner or later some consequence of your mistake will jump up and bite you on the ass and your empire will fall apart. Hence, "terrorism" always eventually fails.

While this is a reassuring scenario, it turns out to be tricky to apply both to history and to our current crisis, despite the fact that "The Lessons of Terror" seems specifically intended to do so. If "terrorism" consists only of attacks on civilians, then what do we call al-Qaida's early actions in Saudi Arabia or the 2000 bombing of the USS Cole in Yemen, all of which selected military targets? Did the group only become "terrorists" when they bombed two American embassies in Africa in 1998? It's also difficult to compare the attacks of Sept. 11 with the oppression inflicted by occupying Roman armies in Germania or Louis XIV's brutal creation of a cordon sanitaire on France's frontiers in the 17th century, both of which Carr characterizes as "tactics of terror."

First of all, an attacking army doesn't conceal its identity or its goals. In the case of the Sept. 11 attacks, the ultimate culprits and their motives remain facts we can only infer. Second, ruthless imperial armies force their opponents to choose between fighting back, perhaps to the death, and a life of subjugation that might be even worse than death. Islamist terrorists confront Americans with a choice between living under the constant threat of sudden violence and such alternatives as pulling U.S. troops out of Saudi Arabia or ending aid to the secular government of Egypt, foreign policy commitments most Americans were probably unaware of to begin with. Most of the other contemporary terrorist situations Carr touches on -- in such places as Israel and Ireland -- involve far more intimate and tangled relations between attackers and victims. (On the explosive subject of Palestinian-Israeli violence, Carr blames both sides. He sees the region's bloody history as deriving from the fact that Israel was founded with the use of terror -- he cites the bombing of the King David Hotel in the waning days of the British Mandate by Jewish terrorists -- but blames the Palestinians, too, for perpetuating the cycle of violence.)

Furthermore, Carr's recommended response to the threat is to opt for "preemptive military offensives aimed at making not only terrorists but the states that harbor, supply, and otherwise assist them experience the same perpetual insecurity that they attempt to make their victims feel." And yet, while "war can only be answered with war," we must avoid any actions or policies that punish civilians. Carr advocates the abolition of the CIA, which he sees as bearing "an uncomfortable resemblance to an organ of state terror" and as "focusing enormous amounts of attention and funds on covert operations, while missing calls on larger, truly vital world developments" and just generally failing at its putative mission of gathering intelligence. He thinks the kind of military action required is best performed by new unmanned aircraft like the RQ-1A Predator and highly trained special operations forces who can strike surgically, but he disdains long-range air strikes.

Exactly who, though, should we sic these formidable assets on if we want to slap down al-Qaida? We can't find most of the group's members, and no one seems to know which, if any, states are sponsoring them. Rich Saudis are said to have funded anti-American Islamist militants for years, but we can no more bomb our ally Saudi Arabia than the British can bomb us because some Irish-Americans send money and guns to the IRA.

Carr's unpredictable opinions make "The Lessons of Terror" a provocative read -- he may hate the CIA, but he likes Donald Rumsfeld; he condemns America for historically pursuing a bloodthirsty strategy of "senseless battles of attrition" in most of its wars, but says that Native Americans first set that tone with "their own vicious techniques of destructive war." But in the end, the confidence with which he teaches his lessons seems premature.

Exceptions to Carr's rules tend to creep into the reader's mind, undermining the whole enterprise, even when he takes the trouble to refute them. For example, the U.S. strategic bombing of Japan during World War II, both conventional and nuclear, intentionally killed huge numbers of civilians, yet Japan continues to be one of America's most valued allies. Carr claims that the exceptional generosity that the victors of that war showed toward the defeated nations was an example of "unprecedented decency" that drew both Germany and Japan back into the "community of constructive, civilized nations" and "undid a great deal of the cruel stupidity of the Allied civilian-bombing campaigns." But rather than supporting the idea that "terrorism" as Carr defines it always backfires, this suggests that whatever damage may be done by brutal military tactics can then be undone by political means. The war was won, after all, and so was the peace, and, as much as I'd like to believe otherwise, the amount of civilian carnage involved doesn't seem to have affected either that much.

Thinking hard about "The Lessons of Terror" tends to lead the reader into this sort of cul de sac. Perhaps that's an inevitable hazard when very, very general principles are so forcefully asserted. It would be comforting to believe that somebody somewhere understands absolutely, positively how to handle terrorism, but I suspect that we're better off listening to the kind of experts who grasp that they, like the rest of us, are still learning.

• View blog reactions