> punishment is strongly disliked. the only evidence for punishment > being good that we are aware of being cited for me is a single > religious reference that is overinterpreted for its number of > rewritings. i do see that _reduction_ can be good, as well as > _effective communication with resulting behavior_. i also see that > people need to _know that harm is acted on_. > > _none of those imply punishment_, which contains harm which can spread.
what we eventually found was that we need our model of reduction to be analogous in scale to the severity of the problem. things that are extremely dangerous get extremely reduced so that the danger is not caused. we still have a lot of copying behavior. it was hard to derive this; we have bigger problems than we could experience "punishment" for if causing them, and we don't model a norm of holding that (e.g. hiroshima and nagasaki). apes don't fight -- they thump their chests instead. as humans, we try to take that model farther and refrain from violence. but we seem too smart to catch up with ourselves. long story short: _there is waaaaaaaaaaaaaaaaaaaay too much suffering going on to punish anybody or hold that as a norm_ _we do still need to model the reasons for punishment, such as making it very clear that very bad things were bad and are being stopped_
