Since it is, indeed, a rhetorical question, he is not expecting a substantive answer, but instead offers it to make his point – that the CIA is hiding torture and believes the use of torture can't be justified.
My response is
A few years ago, one of my co-workers was in a head-on collision with a cop. The collision was on a one-way street, and my co-worker happened to have been the one going the right direction on that street. The cop stated that he'd gotten an emergency call and was rushing to the scene when he ran into the other car.
OK, fair enough.
The cop also refused to let anyone take pictures of him at the accident scene. It seems he was an undercover cop, and having a photograph of him, in circumstances where he was known to be a cop, would have impaired his ability to keep his identity secret. I suppose it might have put his life in danger. If the wrong people had gotten hold of that photo, it would make it a lot easier to target him for reprisals, or to pre-emptively take him out.
And the way he drives, I guess he doesn't need any more risk factors in his life.
So I could see an intelligence agency destroying tapes of interrogations to keep the identity of interrogators secret. Given the fits people threw over the disclosure of Valerie Plame's identity, when she hadn't been a covert operator for years, and was not likely to be ever again, I can understand the desire to protect identities of agents who might be working under cover right now.
A second reason for destroying the tapes is that the CIA might not want details of interrogation techniques to leak out.
Indeed, this especially might be the case if they don't use torture.
I offer this premise: Non-coercive methods of interrogation may work better (or at all) largely because the subjects of these methods are convinced, somewhere in the back of their minds, that if they don't cooperate, eventually the gloves will come off and the "rough methods" will come out. If the CIA tapes leaked out and it became obvious torture was never used in interrogations, the non-torture methods would quit working.
Update: There have been comments, including Terry's advice:
As for Karl, it's best to ignore him, on this subject(you may make your own decision on other topics). He believes torture works, that those who are opposed to tortures he doesn't think are tortures are being willfully blind; because despite the people like myself who say it doesn't work, the CIA has a couple of apologists (with decidedly mixed motives, as they say they know, and imply they have performed, tortures being used) keep saying that torture as means of collecting useful information is useless, he still avers that it's quite the handy tool, and ought to be used more often.
My reply to him is
Let's be clear here. You write:
As for Karl, it's best to ignore him, on this subject(you may make your own decision on other topics). He believes torture works...
Let's be clear here.
I believe torture sometimes works.
You believe torture never works.
As I pointed out in my open letter to you here, the truth is a bit more complicated than you like to present it.
According to the sources you choose to believe, torture never works. And since you define torture as any physical or mental coercion, you can't even detain a prisoner without torturing him, since absent coercion, the prisoner would not remain confined. (Now's your chance to clarify your definition, by the way.)
As evidence that forceful interrogation (again emphasizing that I draw the line between "torture" and "not-torture" in a slightly different place than you do) works, I cite sources you choose to disbelieve.
These include Ed Morrissey's friend here
However, there are many different scenarios for interrogations, including time-critical emergencies, such as hostage rescue or impending attacks. "Effective interrogators need every range of options in these cases, including methods that use coercion to elicit information, for the different situations that our forces not only might face, but have faced. He used the examples of rescuing captured American soldiers from terrorists when we know they will be brutally tortured and murdered if not found immediately and rescued. "I'm guessing that the vast majority of Americans who vote would not have a problem with us using coercive tactics to get that kind of information from a terrorist."
Asked about interrogation guidelines, and whether more assertive methods would ever be authorized, Colonel Stuart Herrington, an expert on interrogation retired from the United States Army,
Yeah, well I think the answer to that is that you know, the type of information you’re trying to get is obviously situation dependent, and sometimes the situation is more critical than others, but there’s got to be, and that’s what’s going on now, a healthy deliberation, and a laying down of here are the procedures…and this has been done already, here are the procedures that are authorized, here are some more aggressive procedures that are not authorized without the approval of so and so, and here are procedures that you will never do, and so that everyone knows basically what the ground rules are, so there’s no room for hot doggery, you know?
My point is, different situations need different techniques. Some of these techniques will be coercive.
Since you define anything the least bit coercive as "torture", then either torture works, or your definition needs some work.
I draw the line above that threshold, though I don't know exactly where it should go.
My main point is, I see the debate over where that line should be as an open one, and reasonable people can disagree.
To you, it is closed, and there are no reasonable or moral excuses for disagreeing with you.
To me, it has become obvious that Terry does not care to build or defend his case. He finds it easier to impugn the character of those who dare to disagree with him.