Some people who assert expertise in the subject offer a "utilitarian argument" – rough treatment need never be used in any interrogations because it never works. In other words, there is no need to debate the morality of any sort of coercive methods, because, these folks declare, they only yield bad information. And since they declare themselves experts, we're expected to take their unsupported word for it.
The problem with a utilitarian argument like this is that it is exquisitely vulnerable to new data. As soon as any data shows up that indicates that rough treatment does, in fact, work, the utilitarian argument goes out the window. (But not for them. Anyone who presents this data is either stupid or lying. Maybe both.)
The piece I linked above mentions cases where rough treatment, up to and including water-boarding, did in fact yield useful information, and in much less time than the comfy-chair would have required. One quote that I find of interest:
In a 2004 book titled The Interrogators, for example, co-author Chris Mackey, who conducted Army interrogations in Afghanistan, condemned torture but detailed how "the harsher the methods we used -- though they never contravened the [Geneva] Conventions, let alone crossed over into torture -- the better the information we got and the sooner we got it."
Now, I don't have a copy of the book lying around, but I did find it at Amazon. Using the "search inside" feature, I found this in the epilog:
In all of the soul-searching over the [Abu-Ghraib] scandal and the effort to understand what interrogators do, there has been a familiar refrain –
the adage that harsh treatment of prisoners only produces bad intelligence, that a tortured prisoner will say anything to stop the pain. That line has been recited for years by schoolhouse instructors and has now gained currency among those rightly condemning the abuses at Abu Ghraib. I know many experienced and fine interrogators who believe that tenet of interrogation doctrine wholeheartedly. But I don't find it particularly persuasive. If a prisoner will say anything to stop the pain, my guess is he will start with the truth. Our experience in Afthanistan showed that the harsher the methods we used -- though they never contravened the [Geneva] Conventions, let alone crossed over into torture -- the better the information we got and the sooner we got it. Other agencies seem to have learned the same lesson. In its interrogation of high-ranking Al Qaeda figures, the CIA has obtained secret legal rulings from the Justice Department to use certain coercive methods, including one called water-boarding in which a prisoner is strapped to a board and submerged in water until he is sure he will drown. If coercion doesn't work, why would the agency go to the trouble?
I think the source of the "utilitarian argument" against torture is laziness. If you can declare that torture – or even harsh methods that fall short of torture – just plain don't work, you don't have to stand up and defend the moral case, which Mackey offers in the very next paragraph:
The reason the United States should not torture prisoners is not because it doesn't work. It is simply because it is wrong. It dehumanizes us, undermines our cause, and, over the long term, breeds more enemies of the United States than coercive interrogation methods will ever allow us to capture.
If you convince yourself that water-boarding doesn't work – can never, ever work – you can face the weeping relatives of the victims of a terrorist attack free of any worry over whether those lives were worth the principle you upheld. You can rest secure in the knowledge that no trade-off of principle for lives was made, or ever will be made.
Good for you.