Year Zero

Gulfari and the Magic Button

by Jonathan Wallace

February 18, 2002

Year Zero is a series of essays mixing my personal account of September 11 and its aftermath with reflections on ethical, legal, political, religious and other implications. The essays are all collected here. You can also subscribe to the Year Zero mailing list here.

Last October, U.S. planes carried out an air strike against a tractor pulling a trailer on an isolated road in Afghanistan, within a few miles of a Taliban stronghold. My source for the the following information is an article on page 1 of the February 13, 2002 New York Times. The article does not specify whether the vehicle was a "tractor-trailer" truck or an agricultural tractor, a detail which is not terribly important. In any event, U.S. planes killed 21 civilians who were fleeing the vicinity of a strike on the Taliban compound, including seventeen children and three women. The survivors of the hit on the trailer were brought into a nearby house, which itself was struck in a follow-up sortie just minutes later. The victims ranged from an eighteen month old infant to a twenty-five year old woman. A sixteen year old girl named Gulfari related how she was splattered with the blood and body parts of her relatives. "When I put my hand up I felt blood. It was like meat and I threw it away."

Lt Col Jim Yonts, speaking classic Pentagon-speak: "We verified the target and on the night of the 21st, we dropped some precision-guided munitions on the target and destroyed that target. All the munitions were accounted for--on the target."

The attack on the Afghan trailer-full of children made me think of the parable of the magic button, familiar from your introductory college ethics class. A genie appears and tells you that in order to eliminate lung cancer from the world, you need only push the red button in front of you. However, a side effect of pushing the button is that one innocent child, half the world away, will die horribly. Do you push it?

You probably will discuss the button story when you get to John Mill's idea of utilitarianism, and the issue on the table will be whether killing the distant child does not, in fact, accomplish the greatest good for the greatest number. Under this approach, you may in fact have a distinct duty to push the button, unless you can derive a reason why saving the life of one child (whom you don't even know personally) is more important than the lives of millions of lung cancer patients.

Libertarian Wendy McElroy finds the button story particularly offensive. In a 1997 essay she says:

Such puzzles are not profound moral problems at all: they are philosophical sleights of hand. To take the button-pushing scenario seriously, you would have to inhabit a universe so vastly different than ours that your moral code would not even remotely resemble the one you hold today. After all, your morality has been constructed upon the realities as you know them. You have derived a code of behavior based on certain assumptions about the nature of the universe and your own nature as a human being. These assumptions did not and do not include a magic button that causes both miracle cures on a global level and instant unexplainable deaths. If the universe ran along principles that included magic buttons, you would undoubtedly have an entirely different code of behavior than you have now.

I respect Wendy McElroy tremendously. But she's wrong. Magic buttons may not appear every day, but that's not the point of the story. Every day, the decision-makers running what we loosely call our "civilization" make decisions which result in the deaths of one or more children or adults to achieve some perceived "greater good". Even issues which have nothing to do with war and aggression have this kind of impact; decisions about laws pertaining to health, homelessness, and traffic safety can be traced directly and convincingly to human mortality. While we are taught in public school and Sunday school to think uncritically that every human life is sacred, we have never actually operated our civilization that way.

The strike on Gulfari's trailer seems to me a perfect real life implementation of the button story, including an actual button. Here, instead of saving lung cancer sufferers, we are protecting Americans from terrorism. The genie pops up and says that "If you press that button and drop some of Lt. Col. Yonts' nice precision guided munitions, you will protect 260 million American citizens from terrorism. The kicker is that you will also kill 21 Afghan civilians, including an eighteen month old baby, and what's more, you will splatter a sixteen year old girl named Gulfari with the blood and body parts of her relatives."

In order to carry on from here, we need to solve the button story. This is a task I don't find at all difficult. God, natural rights, and ideas of "progress" or "civilization" don't unambiguously tell us why we shouldn't (or should) kill babies or splatter Gulfari with gore. In Professor Mothersill's introductory ethics class at Columbia in 1975, we spent most of the semester searching the fabric of the universe for a solid cornerstone for human ethics. I don't remember what we concluded.

The whole problem goes away if we recognize that ethics is human legislation, the rulebooks we adopt to live by, and that legislation has a reality and a life somewhat independent of the reasons which motivated it. If we find what happened to Gulfari and her family repugnant, rather than wasting irreplacable years of our lives debating whether God or natural law planted in us the seed which makes it repugnant, we can write a human code to protect Gulfari against splattered gore. While also protecting 260 million Americans against further iterations on the World Trade Center attack.

In fact (notice how neatly I set this up) that rulebook already exists. It is called the international law of war, as embodied in the various Geneva conventions, to which both the United States and Afghanistan are signatories. Several principles of international law in particular apply, those of "proportionality", "noncombatant immunity", and "discrimination". For a neat summary of these and related concepts, see James L. Carney, Is it Ever Moral to Push the Button?, which, as the title indicates, jumps off from its own nuclear variation on the button story. Carney is a Lieutenant Colonel (apparently a more articulate one than Yonts) and also a graduate of Harvard Law School. He defines "proportionality" as follows:

[P]roportionality means having a reasonable relationship between the goals and objectives to be achieved and the war means being used to achieve them.

"Noncombatant immunity" is the rule which says that it is not appropriate in a "just war" to kill "infants and small children (normally all children), the infirm, aged, wounded, or sick, and those otherwise helpless to protect themselves." The concept of "discrimination" adds to this that we must work hard to discriminate between combatants and civilians. Carney quotes Army Chaplain Donald Davidson:

Virtually every moral commentary on war since World War II, whether focused on the air battle or ground combat, has discussed the problem of noncombatant immunity. The issue is not whether noncombatants should be immune to attack; there has been general agreement on this point since classical times. Rather, the problem is deciding "who" is a noncombatant; that is, the problem of discrimination. The difficulty of differentiating between combatants and noncombatants has escalated with each stage in the development of modern warfare: the advent of conscript armies and large standing armies in Napoleon's era, new weaponry developed in the industrial revolution, the mobilization of whole societies in major wars, the large-scale employment of guerrilla or insurgency war and terrorism, and the invention of weapons of mass destruction.

It is hard to imagine a murkier environment in which to try to discriminate between civilians and fighters than Afghanistan. The Taliban and Al Quaeda did not wear uniforms. Extensive factions in the population switched sides with some regularity based on bribes or convenience. In another incident last fall, the U.S. bombed what it thought was a Taliban convoy, with numerous fatalities. Locals said the group consisted of village elders and leaders travelling to Hamid Karzai's inauguration. It is possible that both sides were right, that among the dead were ex-Taliban, now converted to the side of the new government, and traveling to Kabul to honor Mr. Karzai.

While to many readers the general murkiness of the situation may excuse the civilian casualties, I believe the opposite is true. The rules of proportionality and of discrimination read together would demand that we work particularly hard in a murky surrounding to find the combatants. In some cases, as with the ex-Taliban convoy, it may be hard even for a human eye up close to make this distinction. In others, as with Gulfari and the trailer full of children, it will not be difficult at all.

There are things we take for granted that once examined are pregnant with injustice. It is typical of the human mind that we tend to throw a comforting blanket over horrors that are convenient or even just familiar. Not doing so would invite distress and require effort, both of which cost us energy. I have written before that certain words are stop-signs: they do not contain any meaning other than the statement, "Stop thinking now".

In the quote from Lt. Col. Jim Yonts given above, the stop-sign word is "precision-guided munitions". Since the Gulf War we have had the comforting but quite unjustified idea that "smart bombs" exist which are capable of flying into a particular apartment building and up the stairs, and killing only the occupant of apartment 4D.

There are two problems with this. The first is that human inventions never tend to work as well as we pretend they do. In many fields of endeavor, such as architecture and engineering, we account for this by overdesigning our works, and building in redundancies. A bridge is typically built to withstand forces much greater than we anticipate it ever will, and it is also designed so that if some of its supports or cables are knocked out, the load can be shifted to the others so that the structure does not fall. By definition, in tools that are intended to kill people or destroy property it is hard to design for redundancy; you cannot design a munition which will kill Gulfari but simultaneously keep her alive, just in case. And it is incontestable--even if you believe only the vanishingly small number of cases not denied by our government--that precision-guided munitions have been known to stray and hit targets other than the ones they were being "precision-guided" to.

The second problem is that, in a fuzzy and uncertain world in which truth is hard to pin down (raise your hand if you don't think we live in one), "precision-guided munitions" may perform their job admirably, hitting the designated target, but which turns out to be the wrong one based on human error and bad intelligence. The Iraqi bunker with hundreds of civilians that we thought was an intelligence center, the Sudan pharmaceutical factory we thought was making chemical weapons, or the Chinese embassy in the former Yugoslavia are three notorious examples. There have been many more cases in Afghanistan, including the killing of Gulfari's family and the attack on the convoy of village elders. There is a troubling theory making the rounds that factions jockeying for control in the new Afghanistan are lying to U.S. agents in order to call down airstrikes on one another.

"Precision-guided" bombs are really not capable of deciding who is a combatant before choosing to explode, nor would most of us (especially if we were on the receiving end) want a machine making this decision. So it should seem clear that the phrase "precision-guided munitions" should really mean nothing more than "munitions which hit the designated target a larger percentage of the time than old fashioned munitions, assuming we have designated the right target". But, as our experience in Afghanistan shows, such fuzzy munitions still aren't enough to comply with the rules of proportionality and noncombatant immunity, which demand more precision still.

The crucial question about any kind of bombing--and one which has been ignored by most of the public since World War II--is what the ethical implications are of doing something at a distance that one could presumably do up close with greater accuracy. The practical implications are usually much clearer. For example, "We don't have troops in Dresden now, and we can't get them there anytime soon, and we know that there will be massive loss of Allied lives if we try. But we can fly planes over Dresden with an acceptable level of losses and destroy the city with bombs." Of course, I am picking a particularly stark example, but the point is that the prior two sentences completely evade the question of why it is important to bomb Dresden at all. Dresden contained no important military targets and was targeted to inflict a huge and demoralizing loss of civilian lives--a clear violation of the rules of proportionality and noncombatant immunity.

For an interesting discussion of the practicalities (not the ethics) of bombing civilian populations, see Terror Targeting: The Morale of the Story, by Lt Col Eric Ash, USAF (yet another lieutenant colonel). Ash speaks of a "seesaw" relationship between enemy morale and ours. The purpose of large scale attacks on civilian centers such as Dresden was to raise the morale of our own aircrews (who would feel like they were accomplishing something tangible in an uncertain and dangerous time) while destroying enemy morale. Ash writes that this was not an effective tactic:

[T]he terror bombing of civilians was not very successful. As a strategy, it caused negative morale among bomber crews, and it failed to target the Schwerpunkt of German morale, just as firebombing Japanese cities failed to break the Japanese will to resist. Why then did Allied decision makers go for the terror-bombing option? There are many plausible reasons: desire for revenge and "eye-for-an-eye" retribution, inability to do anything else while facing a daunting enemy and a very uncertain future, perceived opportunity to prove the raison d'Ítre of the air forces, avoidance of friendly ground casualties, and belief that it would break enemy will.

Ash concludes that short of "the morally questionable extremes of killing most of the people or completely destroying their ability to survive," the terror bombing of civilian populations is not a good way to break their morale or bolster our own. We are better off, he says, concentrating on bombing the enemy's command and control structures. Ash's conclusion reminds me of Talleyrand's dictum that the assassination of the Duc d'Enghien by Napoleon was "not only a crime, but a mistake".

When you largely remove ethics from the discussion, and focus only on practicality, as Ash does, than an American essay can dovetail alarmingly with an Al Quaeda manual. The destruction of the World Trade Center, a target of no military significance, clearly was intended to confuse and frighten the U.S. public and unify the radical Islamic world. As Ziad Jarrah, the lead hijacker of the flight which crashed in Pennsylvania, wrote in a letter to his Turkish girlfriend in Germany, "It is a great honour and you will see the result, and everyone will be celebrating." An Al Qaeda manual found in Britain and introduced in evidence in the African embassy bombings trial in New York begins with a remarkable statement:

The confrontation that we are calling for with the apostate regime does not know Socratic debates, Platonic ideals, or Aristotelian diplomacy. But it knows the dialog of bullets, the ideals of assassination, bombing and destruction, and the diplomacy of the cannon and machine gun.

The young came to prepare themselves for Jihad, commanded by the majestic Allah's order in the holy Koran. "Against them make ready your strength to the utmost of your power, including steeds of war, to strike terror (into the hearts of) the enemies of Allah and your enemies...."

One view of the global conflict between Al Qaeda and the U.S., popular in Europe, is that it is a face-off between two powers with similar philosophies of force, utilizing similar approaches. This view, repellent to Americans, violates our idea that we are the ones wearing the white hats. But our view of our own virtue is often based on nothing more that a tautology, that we are good because we are Americans, and you need more than that to win a moral argument. Only movie super-villains believe they are evil, and glory in it. Real humans, even quite horrible ones, tend to believe, like Americans do, that they are in the right and are actually quite nice people. Al Qaeda operatives feel this quite strongly, and so did Himmler when he made his famous petulant speech to the SS:

Most of you must know what it means when a hundred corpses are lying side by side, or five hundred, or a thousand. To have stuck it out and at the same time remained decent fellows, that is what has made us so hard.

If we believe we are good, and others evil, it seems logical that we must be able to indicate the differences, which must be more than "We are Americans and they are Arabs" or "We terrorize for democracy and they do for Islam". A good strong distinction would be as follows: We believe in the rule of noncombatant immunity and they do not. They deliberately killed three thousand American civilians in a nonmilitary target and we would never do such a thing.

Secretary of Defense Donald Rumsfeld has made this very point relentlessly since September 11. For example, he said in a press conference on October 29: "We didn't start the war, the terrorists started it when they attacked the United States, murdering more than 5,000 innocent Americans..."

The attack on Gulfari's family and numerous other examples reported in the press in recent months are an embarrassing contradiction of the premise that we do not kill noncombatants. While we have not engaged, this time around, in the kind of mass terror-bombing of nonmilitary targets that Dresden represented, we have killed civilians steadily both with straying munitions and targeting based on bad information. How do we reconcile this with the premise, "We are good"? While a single act of negligence may not be enough to make us bad, certainly a persistent repetition of negligent acts, coupled with resolute denial they are occuring or that they are wrong, starts to look intentional after a while.

The pathology of our various responses to the killing of Afghan civilians is complex and interesting. Sometimes we deny that it happened, or that it could have happened, as Lt. Col. Yonts did; after all, the munitions are "precision-guided" so they couldn't possibly hit the wrong people. Donald Rumsfeld in his daily press briefings looks pained, stares down the questioning reporter, and says with steely resolution that it never happened, as he did in October when asked whether the U.S. had accidentally bombed a hospital in Herat: "We have absolutely no evidence at all that would suggest that that allegation ... is correct. I'm sure it's not." Sometimes he takes the approach that since Al Qaeda and the Taliban started it, AND the death of innocent people inevitably results from war, Al Qaeda and the Taliban are responsible when we kill civilians: "Responsibility for every casualty in this war, be they innocent Afghans or innocent Americans, rests at the feet of the Taleban and Al Qaeda." He suggested that the foe deliberately puts civilians at risk by using them as "human shields by placing their armor and artillery in close proximity to civilian schools, hospitals and the like." ( October 29 press conference).

However, the most pathological answer of all, based on a pervasive philosophy of irresponsibility, is: "We didn't do it; our bombs did." To find an example of an uncritical adoption of this view by an otherwise intelligent U.S. journalist, see Karen DeYoung's February 12 "news analysis" in the Washington Post, War's Black and White Phase Turns to Gray. DeYoung is indignant about a recent U.S. commando raid, apparently based on bad intelligence, in which we killed some Afghan government soldiers, and captured and beat some others. DeYoung says:

But what allegedly happened on Jan. 23 in the south-central Afghan province of Uruzgan is widely seen as something qualitatively different than anonymous bombs dropped from thousands of feet in the sky. In Uruzgan, according to the Afghans, their American friends stood right in front of them, looked straight at them and fired.

And she quotes James B. Steinberg, President Bill Clinton's deputy national security adviser, as saying that "'this beating-up thing is going to hurt [the Bush administration] much more' than any anonymous bombing mistake."

What DeYoung is saying, without apparently having given it any real thought, is that mistakes which are egregious and horrifying if we make them face-to-face are somehow more acceptable if we make them at a distance by means of "precision-guided munitions".

The unquestioned and entirely false backdrop to this assertion is a philosophy called technological determinism, which holds that technological developments drive social developments, including ethics, rather than the other way around. Professor Daniel Chandler of the University of Wales, Aberystwyth, writes of "the frequent assumption or implication that technological developments, once under way, are unstoppable: their 'progress' is inevitable, unavoidable and irreversible." Chandler points out that technological determinists tend to argue that every technological development is morally neutral, and can be used for good or evil, the way a scalpel can be used to kill or to cure. Chandler counters this proposition by quoting Jerry Mander, from his influential Four Arguments For the Elimination of Television:

Many technologies determine their own use, their own effects, and even the kind of people who control them. We have not yet learned to think of technology as having ideology built into its very form.

Mander is correct. As a technology, "precision-guided munitions" contain and conceal a very precise ideology, which includes the idea, unwittingly expressed by DeYoung, that "anonymous" mistakes committed at a distance are not morally as blameworthy as the same mistakes committed when looking into the victim's eyes.

But why should this be so? Peace to Henry David Thoreau, who came up with a catchy phrase when he said that we have become the tools of our tools. Thoreau was surprisingly resigned to the deterministic idea that our tools shape us, even though we are their creators. It is in fact a poor workman who blames his tools. If under our moral rulebook, it is wrong for me to kill Gulfari with a bullet, it is not a defense that I did it with a bomb instead. The fact that people as smart as DeYoung fall into the trap of believing that a moral distinction exists here, only illustrates the pervasiveness in American society of the deterministic viewpoint: namely, that we need not examine the morality of tool use because the tools themselves somehow reside in a zone above morality.

This brings us to a variation on the genie's words about the magic button. "If you press this button, you will kill our enemies without risking American lives to do so. The kicker is that you will also kill 17 children and four women, and splatter sixteen year old Gulfari with blood and body parts."

In the past 20 years, Americans have formed a remarkable belief that military actions should (and can) be carried out without the loss of a single American life. Bloody set-backs such as the killing of 241 Marines by a car-bomb in Beirut in 1983 and the death of 18 soldiers in an action in Somalia in 1993 led us immediately to withdraw from those countries without punishing the malefactors. (Our passiveness on these and other similar occasions certainly emboldened Al Qaeda to commit the September 11 attacks.) On the other hand, we have shown that we can carry out actions like the Gulf War and now, the war in Afghanistan, without the loss of a significant number of U.S. troops in combat. We do this mainly by allowing proxy troops and precision-guided munitions to do most of the fighting.

I am about to argue that we have an obligation under certain circumstances to risk our own lives rather than use precision-guided munitions, in order to protect noncombatants. I imagine correspondents of mine like Bob Wilson popping from their chairs, eyes flashing and hair standing on end, to protest. As Bob, who is a Vietnam veteran, once wrote here:

I personally favor our military having the nastiest, most menacing and crippling hardware that technology is capable of producing...Might I remind you that the only purpose a military has is to kill people and break things.

(It is interesting to note that Bob was talking about land mines, another type of so-called "smart munition" favored by the United States which kills numerous civilians.)

Perhaps it seems self-evident that we should go to almost any length to protect the lives of our soldiers and citizens--but it is really not. Imagine a hostage situation in which a terrorist has taken over a nursery school classroom and is threatening to kill one child an hour until certain demands are met. Since it is impossible for us to meet these demands, our only other option is an armed raid on the classroom at a significant risk of loss of life, both the children and the soldiers or police who go in. Let's say that in a worst case estimate we expect to lose all forty children in the classroom, plus ten soldiers or law enforcement personnel. We discover that the terrorist has a four year old son whom he adores. We could certainly produce the child, stand him up within sight of his father, and announce that we will cut off one of his fingers or toes every hour (without stanching the flow of blood) until all of the hostages are released. Assuming this response would be effective (it might not with an Al Qaeda terrorist, who would assume he would meet his son in heaven), why do we never deal with dangerous hostage situations this way? We don't because it is repugnant to torture and kill children, even when the result will be a significant savings in American lives. This is of course, one more iteration of the magic button scenario. We will not push the button to kill one child, even the child of a murderer, in order to save fifty American lives.

There is no moral distinction between this scenario and one in which, in order to avoid risking U.S. lives, we use precision-guided munitions that we know will go astray or hit poorly-selected targets and kill innocent people. A "smart bomb" is only a substitute for a U.S. soldier on the ground with a gun. U.S. soldiers surrounding Gulfari's trailer would most likely have chosen not to machine-gun seventeen children and four women. A precision-guided munition, once targeted, is incapable of making such a choice. Therefore, we should not use precision-guided munitions when it is reasonable to put soldiers on the ground, even though we are taking a heightened risk of loss of U.S. life. Today, in Afghanistan, we control the airways and our allies control the ground, except for small pockets of disorganized resistance. Therefore, we could send in special forces in almost any of the situations in which we have used precision-guided munitions instead. Since these munitions are currently hitting civilians an unacceptably high percentage of the time, our refusal to send in troops is the wrong magic button choice, the same one we wouldn't make in the hostage situation. We are able to make it only because it is shrouded in the fog created by technological determinism and denial.

I am only arguing, by the way, that troops on the ground would do a better job, not that they are a panacea. In her article, Karen DeYoung was complaining of a special forces raid on an Afghan compound, in which we killed some allies (some were apparently shot to death after being handcuffed). But, as DeYoung pointed out, in that scenario we are much more accustomed to holding people and command structures responsible for errors than we are in an "anonymous bombing" from thousands of feet up.

A story reported in yesterday's Times (February 17, 2002) illustrates the extreme immorality promoted by the use of remote weapons. A pilotless Predator drone relayed back video of three men examining scrap metal in a former Al Qaeda camp in a remote area. One of the men was taller than the two others and seemed to be treated with "deference" by them. Based on wishful thinking that this man might be Osama bin Laden, one or more anonymous CIA officials decided to launch a missile from the Predator, killing all three men. The dead turned out to be desperately impoverished local villagers who planned to sell the scrap metal for a few cents. Since Predators can stay up in the air monitoring a target for many hours, it probably would have been possible to track the three men until a special forces team could have been inserted to capture and interrogate them. Such a team would probably not have chosen to execute the three men with a bullet to the back of the head--morally exactly the choice made by the CIA handlers who launched the missile. In the comforting cloud of technologically determined nonresponsibility, when a pilotless drone launches a precision-guided munition, it is convenient to forget that there is a human decision-maker anywhere in the picture.

I do not know if there is an international law doctrine of "equivalency" but if not I would like to propose one as follows. The life of the tall Afghan scrap metal collector, who was the sole support of his desperately poor family, is not worth any less than the life of a broker, investment banker, secretary, maintenance worker, firefighter or homeless person killed at the World Trade Center. Nor is the intensity of his family's grief any less than that of the families bereaved here on September 11.

How would you feel if the genie said, "Press this button and you can redress the murder of 3,000 innocent Americans on September 11. However, in order to do it you will have to kill 3,000 innocent Afghans, including small children." It seems we have made our choice and have already pressed that button. One of the ways we cope with it is by not keeping track of the numbers. University of New Hampshire economics professor Marc W. Herold has made an estimate though, based on reports in the world press and government sources: more than 3,00 noncombatant Afghans dead by last November 23.

The punchline to the magic button story may be a poll taken by Psychology Today in 1983. The magazine asked, "If you could secretly push a button and thereby eliminate any person with no repercussions to yourself, would you press that button?" 69% of responding males said yes, 56% of women.