The Battle of the Casbah:
Terrorism and Counter-Terrorism in Algeria, 1955-1957
by General Paul Aussaresses (Enigma Books, 250 pp., $25)
Suppose you have credible information that a bomb is about to go off, and you have detained suspects who may be able to tell you when and where it will explode. If they won't talk, do you torture them? This "ticking-bomb test" is intended to put human rights and other forms of Kantian good intentions under pressure, to see whether you will stand by the principle never to use human beings merely as a means even when hurting one person might save thousands of others.
The question is not merely a philosophical one. In recent weeks Alan Dershowitz and Jonathan Alter have speculated publicly about the necessity of torturing terrorist suspects--under judicial warrant, of course--if torture is the only way to obtain information that may prevent future attacks. "If we want to prevent the death of hundreds of innocent people by subjecting one guilty person to non-lethal pain," Dershowitz recently wrote, "then we must find a way to justify this exception to the otherwise blanket prohibition against torture." If, in August 2001, we had used torture on Zacarias Moussaoui, the man who is accused of having been the twentieth member of Mohammed Atta's team of hijackers, who is now awaiting trial in Virginia federal court, could we have averted the attacks? Should Abu Zubaydah, Al Qaeda's chief of operations, now in custody, be tortured to disclose information on future operations, or on the whereabouts of Osama bin Laden?
On the face of it, hurting a single person in order to save the lives of thousands of people sounds like the kind of hard choice that is beloved by the tough-minded realist. It may seem this way until it is remembered that the ticking-bomb experiment turns out to be the rationale of torturers everywhere. In French Algeria, for example, the ticking-bomb test was always used by the old hands to overcome the scruples of the new recruits in the torture cells of the French prefectures and barracks. When Army Captain Paul Aussaresses first arrived in Algeria in 1954 and discovered that the French police were regularly torturing suspects, one policeman took him out for a pastis at a local bar and told him that moral scruple about inflicting pain was not the real issue. The real choice was this: "Torture a suspected terrorist or tell the parents of the victims that it is better to let scores of innocent people be killed rather than make a single accomplice suffer."
Even people who detest torture might be prepared to do it once, rather than have to explain themselves to the grieving victims of an atrocity. Yet single examples, taken in isolation, falsify the problem. In the ticking-bomb example, you are asked to decide only once in order to avert a single incident of clear and present danger. But that is almost never the real issue. In the real life of institutions, once you have made the first decision to torture, it is but a short step until you are asked to do so again.
What may have begun as an exceptional instance, justified by exceptional circumstances, will rapidly become a policy. The new recruit who rationalized torture at the beginning as a single act of cruelty designed to protect his loved ones soon finds that he is complicit in a banal routine of everyday horror. This was precisely what happened in Algeria: initial justifications of the exception ended up justifying a rule, and those who overcame their scruples just once found themselves in an infernal machine, unable to escape.
American policemen regularly interrogate suspects, and use intimidation, physical pressure, and threats to break down alibis, excuses, or lies. This is not pretty, but it usually stays just short of torture. If we were to change the rules and allow torture for a category of suspected terrorists, how would we prevent torture from being applied to break down suspects in cases--say, the murder of children--where the police have strong feelings of detestation for the suspect in question? If we deal with this problem by erecting institutional firewalls around torture, handing terrorist suspects over to special interrogators and special agencies, how do we keep these special units under executive and legal control?
Alan Dershowitz believes that requiring police to seek a "torture warrant" every time they need to torture a suspect would not only keep the practice under control but even reduce the use of torture. A torture warrant is supposed to act like a search warrant. The problem is that torture is harder to control than searches. How can a judge define in advance how much torture is enough? The heart of the problem of torture is institutional: how do you keep control of it once you have let it out of the box? As reports by the Israeli human rights group B'Tselem show, in Israel it has proved difficult to keep the use of physical force in interrogation under effective judicial and executive control.
The thought experiment also falsely assumes that torture will unlock the necessary knowledge in a timely fashion. But if you have suspects who will not talk in normal interrogations, it is likely that they will resist torture as long as possible. If they hold out long enough, you will have the worst of all worlds: a suspect tortured to death and a stadium full of dead bodies. That, of course, may be precisely what terrorists intend. In this way torture may play into the hands of terrorists.
Torture is fundamentally a political strategy. Generally, it is not used, as the thought experiment supposes, to save lives, but to break the will of political opponents. Whenever a state is locked in an ongoing battle of wills with a terrorist group, the purpose of torture soon ceases to be the extraction of information alone. Its objective becomes to spread fear, to reduce the number of people willing to work for terror organizations. That is the basic reason why torture metastasizes so rapidly from the exception to the rule. And the natural strategy of terrorists is to respond in kind. If you want to make more terrorists, one sure way to do so--the French in Algeria would be the example here--is to torture more suspects.
Thinking that torture will help us in a war against terror also falsifies what our problem is. We think that our problem is information, and so we need torture to get the truth. In reality, before September 11 there was plenty of information in the possession of the American authorities (noise, but no signal). No, our problem is not a problem of knowledge. It is a problem of belief. It is not what terrorists know that makes them dangerous; it is what they believe. And beliefs cannot be changed by physical duress. Indeed, they may be reinforced. Those who survive torture become living monuments to the brutality that has been inflicted upon them. If they die under torture, they become martyrs to their cause.
Any counter-terror campaign is a battle to persuade as well as to dissuade. Terrorists do need to know that what they believe about us is false. They believe that we are weak and will not fight; and so we should prove them wrong. They believe that we are hypocrites; and so they need to know that we actually believe in the constitutional prohibition against cruel and unusual punishment. They need to know all this if we are to win. Winning is about not losing our nerve, about not losing control in the face of provocation. The military logic of terror is to provoke us into reciprocal atrocity that will lose us the war for legitimacy and the war for opinion.
The barbarians who kidnapped Daniel Pearl undoubtedly tortured him. He was subjected to indecent abuse, followed by horrifying death, because he was an American and a Jew. It is hard not to want to do the same in return, but it would be a mistake. Torturing his captors would set in motion an escalation of reprisals that would probably end up jeopardizing the life of every American in Pakistan. The people who killed Pearl may have violated all humane norms, but we have strong prudential reasons for holding on to these norms, even when our enemies do not.
Controlling the impulse to escalate in a counter-terror campaign is not easy, but other countries have shown that it can be done. British interrogation techniques in Northern Ireland in the early years of the Troubles did fall foul of the European Convention on Human Rights. Then the British realized that their methods were losing them important friends abroad, not to mention the support of the Catholic population in Belfast. Over time, they shifted from interrogation under duress to signal interception and infiltration, and managed to gain the upper hand. Information was never effective enough to prevent all bombings: mistakes and tragedies occurred, but each bombing ended up with the terrorists slowly losing public support.
Still, if terror has to be fought with military means, then why is it acceptable to kill terrorists and not to torture them? First, it has to be said that it is not always acceptable to kill terrorist suspects. It all depends on the circumstances. It is one thing to send Special Forces units to Afghanistan to hunt down and to kill terrorist fighters, such as those of Al Qaeda. It would be quite another to send Special Forces to eliminate suspects in some hideout in Barcelona, Hamburg, or London. Shoot-to-kill policies would violate a suspect's rights to the presumption of innocence. (Yes, even terrorists have rights.)
Britain learned the cost of shoot-to-kill strategies when it dispatched an assassination squad to pick off terrorist suspects in Gibraltar in 1988. The squad shot some suspects in broad daylight, but a military success turned into a political setback when the European Court ruled that the shootings violated the European Convention on Human Rights. Besides, in countries with adequate legal and police systems and extradition arrangements with the United States, a shoot-to-kill policy would be unnecessary. On the other hand, in Afghanistan, where terrorists have formed themselves into military units and have already struck against the United States, Americans have little choice but to take them on with military force and to kill them in combat.
If we can kill warrior terrorists in combat, why can't we torture sleeper terrorists when we capture them in our midst? But the situation is not so simple. If we are at war, then the ethics of war apply. The basic principle of those ethics is: kill if you have to kill, but do not inflict unnecessary suffering. A warrior kills his enemy as efficiently as possible. A barbarian seeks to inflict suffering. This moral distinction is not just a warrior's code of honor, dependent on reciprocity among fellow warriors. It is meant to apply also in the conduct of warriors toward civilians, and even toward barbarians. A soldier who would not hesitate to kill an enemy has good reason to have additional scruples about torturing him.
There is something about inflicting suffering, even for the extraction of useful information, that exposes those who act in our name to unnecessary moral harm. We may not feel any tenderness toward terrorist suspects, but we should feel special concern for our own agents. Torture is a face-to-face activity. As Wislawa Szymborska has written, it is highly dependent on the human capacity for empathy:
The body is painful,
it must eat, breathe air, and sleep,
it has thin skin, with blood right beneath,
it has a goodly supply of teeth and nails,
its bones are brittle, its joints extensible.
In torture, all this is taken into account.
If a good torturer must be able to imagine, with some precision, what pain must be like for another human being, then we can only imagine what inner violence the torturer must do to deafen his own instincts to the screams that he causes. Over time, when torture becomes routine, it creates dead souls. To the degree that we care about causing moral hazard, we should be concerned about exposing our own citizens to the face-to-face infliction of pain. To advocate torture, in other words, we must be willing to ask another person to do it in our name and to be scarred by it. This is the deontological ditch--not just the harm done to the victim, but the harm done to the perpetrator--that I, for one, would prefer not to cross.
All of these preliminary thoughts about torture, it might be said, do not escape the frame of the thought experiment. They, too, remain abstract and unreal and high-minded. We need the test of relevant historical contexts, which might offer credible examples where, despite what liberals may think, torture actually works. The central case here is Algeria.
Although successive French governments from the 1950s right up to the present administration have denied that torture was anything but an exceptional and sadistic abuse by particular individuals, the evidence is overwhelming that torture was a systemic and central tool in the French struggle against Algerian independence. One obvious reason for this practice was racism, linked to fear. The belief that those whom you torture do not quite belong to the same human race certainly helps when you are applying the electrodes. It is a general rule that torture anchors itself in situations--apartheid South Africa, for example--in which one race or ethnic group is trying to keep another group down by force.
From a strategic point of view, the rationale for torture in these racially charged situations is rather the same as the rationale for terror. It becomes rational to torture when the group that you are trying to control outnumbers your forces and when that group appears to have you on the run, just as it can seem rational to resort to terror when the state you are opposing has such a preponderant advantage of force that only asymmetrical assault will make a difference. Both torture and terror are the weapons of the weak.
The French might have decided that a better response to Algerian demands for self-determination would be to win Algerians over to some continuing association with France, just as the Algerian FLN movement might have decided that non-violent public demonstrations, strikes, and sit-ins would build a better case for freedom than the random butchery of settlers. There was nothing inevitable about the descent into the inferno. As Alistair Horne's exemplary history of the Algerian war, A Savage War of Peace, makes clear, even as late as 1958 thousands of Muslim Algerians greeted Charles de Gaulle's historic visit to Algiers with joy, believing that some kind of association between France and a free Algeria could be achieved without more violence. Indeed, one of the key lessons of the Algerian war is that the state counter-terror actions have to maintain control of the escalation gradient. Terrorists deliberately seek to provoke ever more extreme reactions in order to foreclose peaceful alternatives. The state party that succumbs to this provocation rapidly loses control of the capacity to control its own violence or the violence of the other side.
By launching attacks on French settlers on May 9, 1945, the Algerian independence movement set in motion the brutal reprisals of Setif that confirmed the FLN in its determination to resort to terror. Reprisal, torture, and killing by both sides became the inexorable consequences of the initial choice of violence. The violence that the FLN used in its freedom struggle provoked French retaliation, and the French response provoked violence from the Algerians. In the end, the side that prevailed was the one that had the capacity to absorb the most punishment. This turned out to be (as could have been predicted from the beginning) the side that was fighting for its freedom on its own native soil.
In the years since Algerian independence in 1962, it has become conventional in the center of French public opinion to believe that torture helped to lose Algeria. It disgraced the country, hardened Algerian hatreds, and did not stop the independence movement. This view was always contested at the margins by Jean-Marie Le Pen's right-wing political followers, and by the Algerian pieds noirs who went into exile in France after 1962. They have always believed that France lost Algeria not because it was too brutal, but because it was not brutal enough.
This view found a recent champion in the person of a retired French general named Paul Aussaresses, who had been a captain in the French special services attached to the French army in Algeria between 1955 and 1957. In a memoir published in France last year and now translated into English, Aussaresses wrote that he routinely tortured suspects and killed them afterwards, and also ordered and supervised reprisals of civilians after FLN terrorist attacks. His military and police superiors in Algiers explicitly approved all of these actions, and the French government--including Francois Mitterrand, the French minister of the interior at the time and later president of France--knew what he was doing.
It was not the revelations themselves that caused the scandal. The use of torture was widely denounced by many French intellectuals. What was distinctive about Aussaresses was his jaunty impudence about the whole issue, his insistence that he had no regrets and that he would do it again in the context of the contemporary war on terrorism. When his American publisher got him onto 60 Minutes and Mike Wallace asked the inevitable question--should the United States torture Zacarias Moussaoui and other suspected terrorist detainees to force them to disclose future plots?--Aussaresses replied: "It seems to me that it's obvious."
Aussaresses's perky refusal to apologize for torture outraged human rights groups in France. They launched a successful prosecution to have him convicted of "justifying war crimes." While it was ludicrous for human rights groups to be on the side of the restraint of free speech, their campaign was driven by the belief that with Le Pen still a factor in French electoral politics, it was essential to deny reactionary revisionism any further prestige with the public. Le Pen fought in Algeria, and Aussaresses mentions his service record there with approval.
As the historian William Cohen has shown in a recent study, while President Chirac has apologized publicly for Vichy and for French complicity in the dispatch of French Jews to concentration camps, both he and the rest of the French political class remain much less willing to admit the stain on French honor represented by the Algerian war. The reasons for this are complicated, but they include the electoral weight of the aging pieds noirs and their descendants, together with large numbers of aging Frenchmen who did their military service in Algeria. In publishing his memoir, Aussaresses perhaps gambled on this equivocal attitude on the part of French authorities. He gambled wrong. He was convicted of defending war crimes in public, ordered to pay a fine, and stripped of his Legion d'Honneur. And instead of being bowed or broken by the backlash against him, Aussaresses seemed to bask in his obloquy.
It is easy to see why. As he tells it, the same tender liberal hypocrites who judge him now were responsible for France's losing Algeria. Most Frenchmen wanted to keep Algeria French, but they did not have the stomach to will the necessary means. Aussaresses prides himself on being one of the few moral realists with the inner toughness to do what had to be done. "What I did in Algeria was undertaken for my country in good faith, even though I didn't enjoy it. One must never regret anything accomplished in the line of a duty one believes in." Instead of bowing before liberal opinion, he taunts it:
The methods I used were always the same: beatings, electric
shocks, and in particular water torture, which was the most
dangerous technique for the prisoner. It never lasted for
more than one hour and the suspects would speak in the hope
of saving their lives. They would therefore either talk
quickly or never.
Aussaresses's most challenging claim is that all this routinized brutality actually worked. According to his account, the systematic torture of the FLN cadre inside the Algiers casbah between 1955 and 1957 broke the cells of the organization one by one and gradually freed the city of terror attacks. By the end of the campaign against the casbah, where torture was used to break the entire network of Ali la Pointe and his gang, peace had returned to Algiers.
Aussaresses writes as if the French won in Algeria. Actually, all he is entitled to claim is that torture momentarily broke the back of the FLN cells in one city, Algiers, for a period of a few years. He seems to forget that the French ultimately lost, decisively and irrevocably. Torture may have won a battle, but it lost the war. The victory for torture that he celebrates in 1957 was undone by 1962, when France granted Algeria independence after the Evian accords.
Torture divided and weakened the French determination to keep Algeria. Aussaresses's memoir confirms that it was opposed not only by intellectuals. Prefects and policemen and military leaders asked to be transferred. A few military doctors moved tortured prisoners into civilian hospitals to save their lives. Journalists wrote against it: the means corrupted the end itself, destroyed the very justification for staying. Aussaresses regards all these forms of dissent and moral scruple as hypocritical weakness. In fact they were a strength that saved France. It had granted independence to Tunisia and Morocco and thereafter to its other French colonies. It was impossible, over the longer term, to keep Algeria, whether by torture or by any other means. De Gaulle's moral realism--that no amount of repression could succeed against a people bent on independence--displayed a far greater grasp on reality than the hard-boiled Aussaresses.
The real picture of torture in Algeria comes not from Aussaresses's smug and disagreeable memoir, but from Frantz Fanon's recollections of the asylum in Blida, Algeria, where until his flight in 1956 he cared for both victims and perpetrators of torture. Fanon was wrong about many things, especially the fatuous idea, expressed in The Wretched of the Earth, that violence in the service of liberation actually heals and transforms the oppressed; but he was not wrong about the special horror into which societies sink when they meet terror with torture. In his psychiatric clinic at Blida, Fanon treated a twenty-eight-year-old French policeman who was so disturbed by memories of the screams of the men whom he had tortured that he stopped up his windows, put cotton in his ears, and still could not stop the screaming inside his head.
At the same time, Fanon was treating a "patriot" (an Algerian militant) who had been tortured and was still recovering from the use of electric shock. One day the tormented policeman, while waiting to see Fanon, happened to spot the "patriot" in the hospital garden. When Fanon came upon the policeman shortly afterwards, he was "leaning against a tree, looking overcome, trembling and drenched with sweat." Fanon took him home, where the policeman, lying on the sofa, confessed that he had just met one of his victims. After Fanon had sedated the policeman, he went back to the hospital and found that the policeman's victim, the patriot, had tried to hang himself in a toilet. The staff of the clinic could only calm him down by lying: they persuaded him that he had never actually seen his torturer. As for the torturer, he secured a sick-leave transfer back to France.
Torture in Algeria was not only a crime. It was also a mistake. It corrupted victim and perpetrator, French and Algerian alike. Paul Aussaresses's memoir is written in a style of rugged tough-mindedness, which calls itself realism, and which likes to mock liberal scruple for its sentimentality and inability to face hard choices. But this supposed realism and tough-mindedness is nothing of the sort. It is the kind of realism that wins battles and loses wars, a nihilistic cynicism that betrays and eventually destroys the causes it claims to defend. Americans, sorely tried by terror, should take note.
Michael Ignatieff is Carr Professor of Human Rights Policy at Harvard University's Kennedy School of Government.
By Michael Ignatieff