It’s a useful discipline of thought to occasionally play devil’s advocate, and attempt to make a convincing argument on the opposing side to that with which you agree. This was brought home to me recently when I re-read Michael Levin’s famous essay The Case for Torture. Anyone who’s familiar with this piece will know that it is a disarmingly frank, simple and logical espousal of a very unpopular position: that torture by the ‘good guys’ is sometimes, in extreme circumstances, justified.
I had no idea who Professor Levin was when I was first introduced to the essay a few years ago. My subsequent discovery that his politics are somewhat to the right of Genghis Khan gives it some context other than that of an intellectual exercise. Even so, it’s accepted by most objective authorities that torture, quite apart from its barbarism and cruelty (which is beside the point of the essay anyway), tends to yield information which is dangerously unreliable at best. But can Levin convince us that its use is ever acceptable? To answer that question, we had better be aware that there are holes in his argument so big you could lose a medium-sized town down one of them. So let’s take a closer look and see if we can’t help him out.
From our post-9/11 perspective, it's tempting to see Michael Levin as some sort of visionary. Writing in 1982, he had no obvious real-life examples which were extreme enough to illustrate his thesis. So he hypothesizes two scenarios: a nuclear bomb set to go off imminently, with the terrorist in custody but refusing to give up the bomb’s location; and a bomb on a plane, with the terrorist (again, presumably, in custody) being the only person who can disarm it. These sound compelling, but we must be cautious. Not only do both these scenarios have no parallel in the real world, but also the particular sets of circumstances would be highly unlikely to come together in the way Levin describes.
Levin moves on to an ‘informal poll’ he conducted, asking four mothers whether torture was justified to retrieve a kidnapped baby. All responded that it was. This is a classic example of the argumentum ad populum fallacy: basically, Levin is saying, ‘These people agree with me, so I must be right.’ Because personal opinion is irrelevant to whether torture is morally justifiable, he would have done better to leave out this little anecdote – especially as he later refutes the ‘personal opinions’ of those who feel that torture is barbaric.