John Ip: Torture justification breaks down

In a Herald article last week Professor Mirko Bagaric argued torture for interrogation is justifiable in some cases, and accordingly that the absolute legal prohibition on torture should be reconsidered.

Bagaric's claim that torture is justifiable essentially boils down to three arguments: the ticking bomb scenario, an analogy to self-defence, and that torture is effective. All three arguments are flawed.

The first argument is overtly utilitarian: "live-saving compassionate torture" is justifiable if the benefit ("lives saved") outweigh the cost ("pain inflicted on the wrongdoer"). How might such a situation arise? Cue the ticking bomb scenario. Imagine authorities have captured a terrorist who has planted a nuclear bomb that is due to detonate in several hours in a large city.

The terrorist refuses to talk. Can the interrogator use torture to extract the critical information?

The answer that one is supposed to come to is, of course, "yes". That's the whole point of the ticking bomb scenario. It exposes us all as closet utilitarians.

Our commitment not to torture is contingent - all that is left is to haggle over the price. But notice the number of assumptions that are conveniently built into the ticking bomb scenario. The authorities just happen to know that the bomb will detonate in several hours; there's no way of making the terrorist talk other than with torture; the authorities have captured the culpable terrorist (the "wrongdoer" as Bagaric calls him), who is the only person who knows how to defuse the bomb.

For this reason, the ticking bomb scenario is more appropriate for stumping first year university students than for assessing the legal prohibition on torture.

Notice also that Bagaric's proposal has no logical stopping point. Although he posits that the nefarious terrorist be tortured for the critical information, the person to be tortured need not be culpable at all.

Suppose that the terrorist (because of nerve damage and intense training) is impervious to physical pain. However, authorities have captured the terrorist's beloved 3-year-old child. If the child is tortured in front of the terrorist, then the terrorist will reveal the critical information. Should authorities torture the innocent child?

If one follows Bagaric's utilitarian logic, the answer is yes, assuming that the benefits (lives saved) outweigh the cost (pain inflicted on child plus mental pain inflicted on terrorist).

Bagaric's second argument is an analogy to self-defence: "If we can kill to protect others, it is nonsense to suggest that we can't inflict lesser forms of harm, including torture, to achieve the same result."

The difficulty with this argument is that torture for interrogation is not directly analogous to self-defence.

It is true that the law allows the use of deadly force in self-defence, which may include the defence of another. Suppose a knife-wielding assailant attacks me. Fearing for my life, I kill my attacker. My attacker is morally culpable in threatening harm to my life, and my act of killing him is an effective means of saving the life of the victim (me).

Compare this with the terrorist who is strapped down - incapacitated and helpless - in preparation for his torture. It could be said that the terrorist is still in a sense a threat because he refuses to divulge the key to diffusing the bomb. But recall that Bagaric's argument logically justifies the torture of persons other than the terrorist, such as the terrorist's 3-year-old son. This is where the analogy to self-defence fails.

Indeed, advocates of torture implicitly concede this by requiring the number of lives saved to be large. Self-defence, by contrast, does not require this: it is enough that just my life is saved by my killing of the knife-wielding attacker.

The third argument that Bagaric makes is that torture really does work, meaning that it is an effective way of eliciting the truth. Despite Bagaric's claims there is a wealth of evidence that torture works, the number of examples put forward by advocates of interrogational torture is limited.

And some of those examples do not stand up to scrutiny. For instance, Bagaric cites an account of the CIA's interrogation of an al Qaeda member, Abu Zubaydah, which supposedly broke Zubaydah "in less than 35 seconds" with a form of torture known as waterboarding. This is hard to square with the recently disclosed report of the CIA Inspector General, which states that Zubaydah was waterboarded 83 times. Moreover, much of the valuable information Zubaydah provided was actually obtained by the FBI without torture or coercion.

This last point shows that the debate is not really whether torture works or doesn't work. The real question is torture's effectiveness in eliciting the truth relative to other interrogation techniques. The choice is not, as torture advocates often suggest, between torture and sitting on one's hands.

* John Ip is at the University of Auckland Faculty of Law.