On Torture

On Torture:

This came from a discussion elsewhere which concerned the wrongness of torture.

Torture cannot exist without creatures capable of being tortured.

Furthermore, torture necessarily is something that thwarts the strong desires of the creature being tortured. Which means that it is necessarily the case that it is something the tortured creature has reason to avoid.

Consequently, it is necessarily the case, in all possible worlds in which torture can exist, that there are reasons to cause others to refrain from torturing. There are, in fact, very powerful reasons to do so.

However, there is no reason to perform any action that is not effective at preventing torture. Such actions are a waste of time. The reasons to avoid torture provides agents with no reason to snap their fingers three times, unless snapping their fingers three times is an effective way of avoiding torture.

But, if prayer and snapping one's fingers ARE effective ways of avoiding torture, then there are reasons to do these things.

The reasons for avoiding torture are reasons to promise incentives or threaten punishment of potential sources of torture that have desires and aversions. However, it provides no reason to promise incentives and threaten punishments to . . . for example . . . inanimate objects such as a campfire (that can set fire to one's clothes). Campfires can cause significant harm, but campfires are not moral agents - moral terms do not apply.

But let us assume that we are talking about a source of torture that has desires and aversions - but not a reward system. Let us assume that you are dealing with an entity (creature, machine, it does not matter) which you discover will inflict a severe and prolonged electric shock (torture) unless doing so causes a light to come on. You will have a reason to turn on a bright light every time the creature inflicts a shock.

However, it makes no more sense to morally blame the entity for delivering the shocks than to blame the campfire for catching your clothes on fire. Moral concepts simply are not applicable here. You simply threaten to turn on the light if the entity should shock you, and the entity quits shocking you.

Now, give the entity a reward system. You can now make it the case that the entity has an aversion to delivering shocks by the use of praise and condemnation. Previously, where you had a reason to turn on the light switch, now you have a reason to praise and condemn. You have a reason to call the entity that inflicts shocks "bad" and those that refrain "good". You have a reason to call the act of delivering the shocks "wrong".

The overall point is that it makes no sense to even use moral terms - to call an act right or wrong - unless we are talking about agents that have a reward system. Campfires can cause severe burns, but are not moral agents. Entities that merely have desires and aversions that cannot change are not moral agents. It makes as little sense to praise and blame them as it does to praise or blame a machine in order to get the desired output. Praise or blame only makes sense where praise and blame itself can change behavior.