Hacker News Reader

Comment by overcast | Hacker News Reader

Well it didn't deliver results, because none of it was true. So how could the guy give up any information that didn't exist?

View on HN

Imagine you had a scale, onto which you put a weight labeled "100kg". You know it weights 100kg because it says so right there. The scale however, only registers 85kg. You take it off, you put it back on, the scale always says the same thing. You try different scales, they say 85kg, but you keep going because it's 100kg and you're going to find a scale that's accurate! Eventually you find a scale that says what you want.

This straw man illustrates that at some point, it isn't the tool, but the process that's flawed. Even if your tool is accurate, if you don't believe it, you can't get at the truth, and if you work hard enough you can find a broken tool to tell you what you want.

So perhaps there is some truth to the idea that torture can extract true information from people, but if the whole process operates like the above straw man, it totally invalidates the use of the tool.

It's more like, you will extract information, but only in hindsight can you ever know if it was worthwhile, accurate, a trap, etc. You will also kill some people, some people won't break easily, and some will babble anything and everything. You won't know which is which in the moment, and usually you're torturing for time-sensitive information.

What you get is what happened during WWII, when the resistances knew they just had to withstand torture for 24 hours, 48 at most. It's amazing what motivated people can do when they have a goal, and that goes for the tortured as well as the torturers.

Part of the result of any interrogation is to figure out if the subject has the information you want.

So it didn't deliver at all.

Right, but the point of the article is that they were doing it to the wrong person. That doesn't mean it does not work. Causation does not imply correlation.

If it takes you more than 83 separate instances of waterboarding somebody to decide that they don't have useful information to share with you, then waterboarding is not an efficient way to get reliable information.

If waterboarding is not an efficient way to get reliable information, then I'd be very curious to hear about what definition of "work" you're using where waterboarding might satisfy, but fail in this particular instance.

For all the times it didn't work, how many times did it work for others? One person, with no information, doesn't mean anything.

"Here are eight cases cited in the report where the C.I.A. made the case that its tactics thwarted plots and led to the capture of terrorists, and how the committee's report undercut those accounts."

http://www.nytimes.com/interactive/2014/12/08/world/does-tor...

Full report:

http://www.feinstein.senate.gov/public/index.cfm/senate-inte...

Torture is illegal under the Eight Amendment. The Supreme Court has affirmed this multiple times.

If we're going to go against legal precedent as well as international convention, the burden of proof for torture's effectiveness is on torture advocates. The evidence provided does not stand up to scrutiny.

Your question is poorly formed because the basis of my comment was asking how you defined "work" in the context of waterboarding. You didn't provide any sort of definition, you just said "Well, maybe it worked for other people".

But if it doesn't work reliably, for whatever "work" means, then it doesn't matter if it's successful sometimes and unsuccessful other times because if you don't have a way to distinguish the success from the failure, it's impossible to measure.

That's why I asked you to clarify what you mean when you say it might "work".

well, either it means they were just gratuitously torturing him for fun, or it means they can't tell whether they were getting any useful information or if he had any to give

neither option suggests the technique is useful

it's also about as immoral as you can get

> Right, but the point of the article is that they were doing it to the wrong person. That doesn't mean it does not work. Causation does not imply correlation.

I think this case make it obvious that it's impossible to tell if the person you're torturing has the information you're looking for and is withholding it, or simply doesn't have the information. That's a textbook definition of "ineffective".

What's the alternative? I'm generally curious as to what would be considered more effective.

The alternative is to operate within the law and international convention.

"We got more information out of a German general with a game of chess or Ping-Pong than they do today, with their torture," said Kolm, 90, an MIT physicist who had been assigned to play chess in Germany with Hitler's deputy, Rudolf Hess [1].

Also: Why Torture Doesn't Work: The Neuroscience of Interrogation [2]

[1] http://www.washingtonpost.com/wp-dyn/content/article/2007/10...

[2] http://www.amazon.com/Why-Torture-Doesnt-Work-Interrogation/...

From the sound of it, "do nothing" would be no less effective, and a bit cheaper.

Have you quantified the times when it has been effective though?

If you can't tell the difference between when its effective and when its ineffective, then the technique is worthless.

Reply on news.ycombinator.com