The Washington Post has a fascinating article by Laura Blumenfeld featuring interviews with three interrogators—one American, one British, one Israeli. Much of the focus is on the American, Tony Lagouranis, a 37-year-old military intelligence specialist who served in Iraq in 2004 and who has a new memoir out. He says he is anguished by his service, wracked by guilt over having to “torture” suspects.

Well, everyone is against torture in principle—at least everyone who is not a sadist. But what constitutes torture? That’s the nub of the problem. Blumenfeld’s article sheds interesting light on this vexatious issue by juxtaposing Lagouranis’s comments with those of his far more experienced Israeli and British counterparts.

First up is a man identified only by his first name, now living on an unidentified Mediterranean island because of death threats from the IRA. James, 65, was one of Britain’s most experienced interrogators in Northern Ireland. Starting in 1971, James said, he worked for the Special Branch of the Royal Ulster Constabulary (RUC), interrogating Irish nationalists Gerry Adams, Bobby Sands, and others whom the British government suspected of being terrorists. Blumenfeld’s article offers these vignettes of how James operated:

Once, IRA leader Brendan Hughes claimed that James had cocked a gun to his head. James does not deny it. “You fight fire with fire,” he said, the memory igniting his blue eyes. Another anecdote: “My friend once saw a guy planting a bomb,” he said. He laughed. “My friend tied a rope around the guy’s ankle, and made him defuse it. Now that’s how to deal with a ticking bomb.” Yet James denies being guilty of torture: “Yes, a bloke would get a cuff in the ear or he might brace against the wall. Yes, they had sleep deprivation,” he said. “But we did not torture.”

Then there is “Sheriff,” the code name of the recently retired chief of interrogations for the Shin Bet, Israel’s internal security service. The article describes his technique as follows:

For Sheriff, interrogation was more psychological than physical. He used flattery on Palestinians who put bombs under playground benches: “You say, ‘Hey! Wow! How did you connect these wires? Did you manufacture this explosive? This is good!” He played good cop, and bad: “One day I was good. Next day I was bad. The prisoner said, ‘Yesterday you were good. What happened today?’ I told him we were short on manpower.”

Presumably those kinds of psychological ploys are exactly what opponents of “torture,” broadly defined, think we should use to extract information. Yet even for someone as skilled as Sheriff, they weren’t always enough:

But when the pressure mounted for intelligence, Sheriff said, the best method was “a very little violence.” Enough to scare people but not so much that they’d collapse. Agents tried it on themselves. “Not torture.”

James’s and Sheriff’s justifications won’t convince those who consider everything from the good cop/bad cop routine to “a little violence” as torture. That’s a comforting, consistent position to take. In fact, it is basically the policy laid out in the new Army Field Manual on Interrogations, a policy that prohibits many of the more coercive techniques employed in the days after 9/11.

But is this new policy sufficient to keep us safe? Sheriff doesn’t think so:

“You have to play by different rules,” the Israeli interrogator told an American visitor. “The terrorists want to use your own system to destroy you. What your President is doing is right.”