About that David Rose piece for Vanity Fair I mentioned: really, give it a read. Rose patiently tracks how the collapse of Jose Padilla’s prosecution begins with the detention and torture of a recovering heroin addict named Binyam Mohammed. When Rose asks FBI Director Robert Mueller if any terrorist attacks on the U.S. were disrupted because of torture, Mueller replies, “I don’t believe that has been the case.” And there’s a special guest appearance by Libby-and-Blagojevich-hunter Pat Fitzgerald, who accompanies an FBI agent in 1998 to Morocco to interrogate a terrorist. (Interestingly nicknamed “Joe The Moroccan.”) They secure his testimony — and use it to convict his accomplices — without laying a finger on him.

And in light of the controversy over John Brennan and finding a new leadership for the intelligence community untainted by torture, consider carefully what this anonymous CIA official tells Rose:

“We were done a tremendous disservice by the administration,” one official says. “We had no background in this; it’s not something we do. They stuck us with a totally unwelcome job and left us hanging out to dry. I’m worried that the next administration is going to prosecute the guys who got involved, and there won’t be any presidential pardons at the end of it. It would be O.K. if it were John Ashcroft or Alberto Gonzales. But it won’t be. It’ll be some poor G.S.-13 [bureaucrat] who was just trying to do his job.”

You wouldn’t approve of treating Lynndie England or Charles Graner as the architects of torture at Abu Ghraib, would you?

Finally, here’s something that rarely gets discussed in all the talk about intelligence failures from 9/11 to Iraq. Best intelligence practices say torture is an unreliable method for yielding useful information. Yet intelligence analysts who had to craft assessments from the debriefs of captured Al Qaeda operatives were left without any way of knowing that the basis for their reports was unreliable:

“We didn’t know he’d been waterboarded and tortured when we did that analysis, and the reports were marked as credible as they could be,” the former Pentagon analyst tells me. “The White House knew he’d been tortured. I didn’t, though I was supposed to be evaluating that intelligence.” To draw conclusions about the importance of what Abu Zubaydah said without knowing this crucial piece of the background nullified the value of his work. “It seems to me they were using torture to achieve a political objective. I cannot believe that the president and vice president did not know who was being waterboarded, and what was being given up.”

That’s pretty much designing failure into the architecture of intelligence work. If the point of the torture is, as Dick Cheney says, to yield reliable information — leave aside the actual relationship between torture and unreliable information — it’s nonsensical to exclude such a crucial data point from intelligence analysts. The only reason to do so is to protect the torture program from outraged CIA or Pentagon analysts who might blow the whistle to an inspector-general, or to Congress, or to the press. And at that point it’s clear that good intelligence work is no longer what’s at issue here.