I agree with the general ideas that morality is useful-maybe-necessary to civilization, and that large-scale cooperation provides a huge advantage.
I *disagree* that civilization requires culpability. If the game is set up to weed out defectors, culpability is irrelevant. The question isn't "did you mean to do it and if so why," the question is "are you likely to do it again."
Culpability is a pretty good starting place for that, but it's not the be-all-end-all.
Yes, I agree that functionally, it shouldn't matter if an individual is culpable, just whether they are likely to do it again. Yet the culpability concept seems so widespread in civilization, I have to wonder if it turns out that having most people believe in culpability results in better individual and social controls. Or maybe it's just the simplest solution, rather than the most powerful solution, so a lot of people stumble on it. Or conceivably, it's just a very sticky or spready idea.
I think deterrence is the major thing missing from that analysis.
We don't just want to wait until people do something wrong the first time and then decide whether they look likely to do it again. If we did that, there'd be a general perception that you could get away with a crime if you could argue convincingly in court that it had been a one-off. (I don't know - for example, perhaps someone who murdered their father could argue that it didn't imply anything very much about their general safety for anyone else to be around, because only their father had ever had the opportunity to commit whatever acts of egregious mis-parenting in their childhood had inspired their murderous inclination
( ... )
I think the only thing I'd add to that is that, y'know, it can rapidly become difficult to figure out whether someone *meant* to do a thing or not, and if the system leans on "did you have control over this" to decide which choices to "disincentivize", people will start finding ways to claim that they didn't have a choice.
Your example in the second paragraph reads like a "burning bed" defense, which is accepted in some courts. Essentially, it's a form of premeditation which is also considered self defense.
I'd like to point out that murderers, especially burning bed types, have extremely low rates of recidivism.
Also, I don't think the punishment paradigm is necessarily an effective deterrent. Murder rates per 100,000 of population tend to be higher in states with the death penalty, for instance. Those burning bed cases? If there weren't social deterrents in place against less extreme solutions, then murder wouldn't have been on the table.
Essentially, there are a whole group of violent criminals who accept that they'll be punished, sometimes to an extreme degree, but still think it's better than continuing to be under the thumb of an abuser.
@londo, it's usually easy to tell when someone has a legitimate "let's kill the abuser" defense. Unlike most criminals, they tend to be nearby the crime scene awaiting the arrival of the police or waiting
( ... )
I *disagree* that civilization requires culpability. If the game is set up to weed out defectors, culpability is irrelevant. The question isn't "did you mean to do it and if so why," the question is "are you likely to do it again."
Culpability is a pretty good starting place for that, but it's not the be-all-end-all.
Reply
Reply
Reply
We don't just want to wait until people do something wrong the first time and then decide whether they look likely to do it again. If we did that, there'd be a general perception that you could get away with a crime if you could argue convincingly in court that it had been a one-off. (I don't know - for example, perhaps someone who murdered their father could argue that it didn't imply anything very much about their general safety for anyone else to be around, because only their father had ever had the opportunity to commit whatever acts of egregious mis-parenting in their childhood had inspired their murderous inclination ( ... )
Reply
I think the only thing I'd add to that is that, y'know, it can rapidly become difficult to figure out whether someone *meant* to do a thing or not, and if the system leans on "did you have control over this" to decide which choices to "disincentivize", people will start finding ways to claim that they didn't have a choice.
But, yeah. Good call.
Reply
I'd like to point out that murderers, especially burning bed types, have extremely low rates of recidivism.
Also, I don't think the punishment paradigm is necessarily an effective deterrent. Murder rates per 100,000 of population tend to be higher in states with the death penalty, for instance. Those burning bed cases? If there weren't social deterrents in place against less extreme solutions, then murder wouldn't have been on the table.
Essentially, there are a whole group of violent criminals who accept that they'll be punished, sometimes to an extreme degree, but still think it's better than continuing to be under the thumb of an abuser.
@londo, it's usually easy to tell when someone has a legitimate "let's kill the abuser" defense. Unlike most criminals, they tend to be nearby the crime scene awaiting the arrival of the police or waiting ( ... )
Reply
Leave a comment