“Torture” is a label you attached to things, and then when you ask if something is torture you’re making a disguised query but you can’t get out more than what you put in. Strong arguments against anything anyone affixes the label “torture” to don’t exist.
If one has a way of carving up reality such that yields (set of activities 1), and another that yields a strongly overlapping (set of activities 2), one doesn’t make the sets synonymous by acting as if there is only one mental bin as if there was only one set. An argument against each member of one set will always look like an argument against the members of the other if one makes this error.
This is assuming the cluster structure of thingspace doesn’t make the argument against everything in (set of activities 1) valid or invalid, which it usually does if the set’s boundaries aren’t arbitrary and sharp.
What I was thinking was that if it’s true that pain is much more likely to elicit answers that the pain-giver wants to hear than anything else (an argument against torture for the purpose of getting information), then it’s worth establishing which sorts of treatment supply sufficient pain to get that sort of reaction.
If you have strong arguments about whether torture is worth doing, than knowing what should be categorized as torture would be useful.
“Torture” is a label you attached to things, and then when you ask if something is torture you’re making a disguised query but you can’t get out more than what you put in. Strong arguments against anything anyone affixes the label “torture” to don’t exist.
If one has a way of carving up reality such that yields (set of activities 1), and another that yields a strongly overlapping (set of activities 2), one doesn’t make the sets synonymous by acting as if there is only one mental bin as if there was only one set. An argument against each member of one set will always look like an argument against the members of the other if one makes this error.
This is assuming the cluster structure of thingspace doesn’t make the argument against everything in (set of activities 1) valid or invalid, which it usually does if the set’s boundaries aren’t arbitrary and sharp.
What I was thinking was that if it’s true that pain is much more likely to elicit answers that the pain-giver wants to hear than anything else (an argument against torture for the purpose of getting information), then it’s worth establishing which sorts of treatment supply sufficient pain to get that sort of reaction.