Fear is frequently fun—ask any carnival promoter, or fans of Silent Hill. (That’s small-f fun; from a big-F standpoint, we’d be looking at fear as an aspect of sensual engagement or emotional involvement, but I think the argument still holds.) Without taking into account secondary effects like grief, it’s not at all clear to me that an environment containing a suitably calibrated level of lethal interpersonal threats would be less fun or less (instantaneously) Fun than one that didn’t, and this holds whether or not the subject is adult.
I do think those secondary effects would end up tipping the balance in favor of adults, though, once we do take them into account. There’s also a fairly obvious preference-utilitarian solution to this problem.
But the fear you get from Silent Hill is fear you can walk away from and know you’re not going to be attacked by zombies and nor will your loved ones. You choose when to feel it. You choose whether to feel it at all, and how often. Making fear that is known to be unfounded available on demand to those who choose it is not even in the same ballpark as making everyone worry that they’re going to be killed.
True enough, and I’m not going to rule out the existence of people calibrated to enjoy low or zero levels of simulated threat (I’m pretty sure they’re common, actually). It’s also pretty obvious that there are levels of fear which are unFun without qualification, hence the “suitably calibrated” that I edited into the grandparent. But—and forgive me for the sketchy evopsych tone of what I’m about to say—the response is there, and I find it unlikely that for some reason we’ve evolved to respond positively to simulated threats and negatively to real ones.
Being a participant in one of the safer societies ever to exist, I don’t have a huge data set to draw on. But I have been exposed to a few genuinely life-threatening experiences without intending to (mostly while free climbing), and while they were terrifying at the time I think the final fun-theoretic balance came out positive. My best guess, and bear in mind that this is even more speculative, is that levels of risk typical to contemporary life would have been suboptimal in the EEA.
How would you feel about a society otherwise similar to our own which included some designated spaces with, essentially, a sign on the door saying “by entering this room, you waive all criminal and civil liability for violent acts committed against you by other people in this room” and had a subculture of people who hung out in such places, intermittently mutilating and murdering each other?
I think I’d be okay with it in principle, in the absence of some well-established psychology showing strong negative externalities and in the presence of some relatively equitable system for mitigating the obvious physical externalities (loss of employment due to disability, etc.), preferably without recourse to the broader society’s resources. I probably wouldn’t participate in the subculture, though—my own level of fun calibration relative to threat isn’t that high.
Well, keep in mind, even inside such a room social norms would rapidly evolve against letting things get too exciting, it’s just that there wouldn’t be any recourse to a larger legal system to resolve the finer points.
Maybe a big guy sits down in the corner with a tattoo across his bare chest saying “I am the lawgiver, if anyone in the room I watch is injured or killed without appropriate permission I will break the aggressor’s arms” and mostly follows through on that. When somebody kicks the lawgiver’s ass without taking over the job, everybody else votes with their feet.
a few genuinely life-threatening experiences [..] I think the final fun-theoretic balance came out positive
Death represents pretty significant disutility; if the experience was significantly life-threatening, you’re attributing some correspondingly significant utility to the experience of surviving. How confident are you?
Ah. I probably should have been clearer about that. Above I haven’t been talking about expected utilities (which are likely negative, although I’d need a clearer picture of the risks than I have to do the math); in the last paragraph of the grandparent I was discussing the sum of fun-theoretic effects applying to me in the local Everett branch, and previously I’d been talking about what I assumed to be the utilitarianism of Bakkot’s hypothetical (which seemed to make the most sense as an average-utilitarian framework with little or no attention given to future preferences).
My preferences do contain a large negative term for death (and I don’t free climb anymore, incidentally). I’m not that reckless.
Okay, yes. However, I’m almost certain that having killers running around unchecked will not produce the optimal level and type of fear in the greatest possible number of people.
I find it unlikely that for some reason we’ve evolved to respond positively to simulated threats and negatively to real ones.
Why? A simulated threat prompts an immediate response, but killers on the loose prompts a lot of worrying over a long period of time. While fighting off a murderer might spike your adrenaline, that’s not what killers on the loose will do. Instead people will lock their doors. They’ll fear for their safety. They’ll be afraid to let strangers into their home. They’ll worry about what happens if they have a fight with their friend—because the friend can commit murder with impunity. They’ll look over their shoulders. Parents will spend every second worrying about their children. The children will have little or no freedom, because the parents won’t leave them alone and may just keep them inside all the time, which is NOT optimal. People will have a lot of cortisol, depressing immune systems and promoting obesity.
That’s NOT THE SAME as a single burst of adrenaline, whether from falling while climbing or from watching a movie or even from fighting for your life. So I guess you’re right that it’s not about whether it’s real or not (though if it’s a game, then when it gets too intense, you can just turn it off, and you can’t turn off real life), but about the type of threat. However, the simulated threat doesn’t actually make you less likely to continue living, whereas a real threat does.
Well, of course I don’t think that allowing murder without restriction is going to make everyone fun-theoretically better off, let alone maximally satisfy their preferences over the utilitarian criteria I actually believe in. My original claim was a lot narrower than that, and in any case I’m mostly playing devil’s advocate at this point; although I really do think that fun-theoretic optimization is best approached without reflexively minimizing things like fear or pain on grounds of our preexisting heuristics. That said, I’m not sure this is always going to be true:
A simulated threat prompts an immediate response, but killers on the loose prompts a lot of worrying over a long period of time. While fighting off a murderer might spike your adrenaline, that’s not what killers on the loose will do. Instead people will lock their doors [...] People will have a lot of cortisol, depressing immune systems and promoting obesity.
We know about a lot of societies with a lot of different accepted levels of violence. The most violent that I know of present up to about a 30% chance of premature death, so much higher than anything Western society presents that it’s scarcely conceivable (even front-line soldiers don’t have those death rates, although front-line service is more dangerous per unit time). But there’s very much not a monotonic relationship between level of violence and cultural paranoia, or trust of strangers, or freedom given to children. Early medieval Iceland, for example, had murder rates orders of magnitude higher than what we see now (implicit in textual sources, and confirmed by skeletal evidence); but children worked and traveled independently there, and hospitality to strangers was enshrined in law and custom. The same seems to go for more contemporary societies if the murder rates I’ve seen are at all accurate, although I don’t have as rich a picture for most of them. Our cultural fears of violence are very poorly correlated with actual expectations, as even a cursory glance over the most recent child molestation scare should show.
If studies of relative cortisol levels have ever been performed, I don’t know about them; but the cultures themselves don’t seem to show evidence of that kind of stress. I’d expect to see more paranoia following a recent uptick in violence, but I wouldn’t expect to see it well correlated with the base rate.
Early medieval Iceland, for example, had murder rates orders of magnitude higher than what we see now (implicit in textual sources, and confirmed by skeletal evidence); but children worked and traveled independently there, and hospitality to strangers was enshrined in law and custom. The same seems to go for more contemporary societies if the murder rates I’ve seen are at all accurate, although I don’t have as rich a picture for most of them.
Okay. What kind of murder are we talking about? What made up most of the extra—was it all sorts of things or was it duels? And was it accepted or was it frowned on? Were murderers prosecuted? Did victims’ families avenge them?
I’d expect to see more paranoia following a recent uptick in violence, but I wouldn’t expect to see it well correlated with the base rate.
Okay. What kind of murder are we talking about? What made up most of the extra—was it all sorts of things or was it duels? And was it accepted or was it frowned on? Were murderers prosecuted? Did victims’ families avenge them?
I’m not historian enough to say for sure, unfortunately. Judicial duels were part of the culture there, but the textual sources indicate that informal feuds were common, as were robbery and various other forms of informal violence. You could bring suit upon a murderer or other criminal in order to compel them to pay blood money or suffer in kind, but there was much less central authority than we’re used to, and nothing resembling a police force.
Yes. Don’t get too hung up on the specific example, though; I chose it only because it’s a time and place that I’ve actually studied. The pattern (or, really, lack of a pattern) I’m trying to point to is much more general, and includes many cultures that don’t have a strong emphasis on honor.
Fear is frequently fun—ask any carnival promoter, or fans of Silent Hill. (That’s small-f fun; from a big-F standpoint, we’d be looking at fear as an aspect of sensual engagement or emotional involvement, but I think the argument still holds.) Without taking into account secondary effects like grief, it’s not at all clear to me that an environment containing a suitably calibrated level of lethal interpersonal threats would be less fun or less (instantaneously) Fun than one that didn’t, and this holds whether or not the subject is adult.
I do think those secondary effects would end up tipping the balance in favor of adults, though, once we do take them into account. There’s also a fairly obvious preference-utilitarian solution to this problem.
But the fear you get from Silent Hill is fear you can walk away from and know you’re not going to be attacked by zombies and nor will your loved ones. You choose when to feel it. You choose whether to feel it at all, and how often. Making fear that is known to be unfounded available on demand to those who choose it is not even in the same ballpark as making everyone worry that they’re going to be killed.
True enough, and I’m not going to rule out the existence of people calibrated to enjoy low or zero levels of simulated threat (I’m pretty sure they’re common, actually). It’s also pretty obvious that there are levels of fear which are unFun without qualification, hence the “suitably calibrated” that I edited into the grandparent. But—and forgive me for the sketchy evopsych tone of what I’m about to say—the response is there, and I find it unlikely that for some reason we’ve evolved to respond positively to simulated threats and negatively to real ones.
Being a participant in one of the safer societies ever to exist, I don’t have a huge data set to draw on. But I have been exposed to a few genuinely life-threatening experiences without intending to (mostly while free climbing), and while they were terrifying at the time I think the final fun-theoretic balance came out positive. My best guess, and bear in mind that this is even more speculative, is that levels of risk typical to contemporary life would have been suboptimal in the EEA.
How would you feel about a society otherwise similar to our own which included some designated spaces with, essentially, a sign on the door saying “by entering this room, you waive all criminal and civil liability for violent acts committed against you by other people in this room” and had a subculture of people who hung out in such places, intermittently mutilating and murdering each other?
I think I’d be okay with it in principle, in the absence of some well-established psychology showing strong negative externalities and in the presence of some relatively equitable system for mitigating the obvious physical externalities (loss of employment due to disability, etc.), preferably without recourse to the broader society’s resources. I probably wouldn’t participate in the subculture, though—my own level of fun calibration relative to threat isn’t that high.
Well, keep in mind, even inside such a room social norms would rapidly evolve against letting things get too exciting, it’s just that there wouldn’t be any recourse to a larger legal system to resolve the finer points.
Maybe a big guy sits down in the corner with a tattoo across his bare chest saying “I am the lawgiver, if anyone in the room I watch is injured or killed without appropriate permission I will break the aggressor’s arms” and mostly follows through on that. When somebody kicks the lawgiver’s ass without taking over the job, everybody else votes with their feet.
Death represents pretty significant disutility; if the experience was significantly life-threatening, you’re attributing some correspondingly significant utility to the experience of surviving. How confident are you?
Ah. I probably should have been clearer about that. Above I haven’t been talking about expected utilities (which are likely negative, although I’d need a clearer picture of the risks than I have to do the math); in the last paragraph of the grandparent I was discussing the sum of fun-theoretic effects applying to me in the local Everett branch, and previously I’d been talking about what I assumed to be the utilitarianism of Bakkot’s hypothetical (which seemed to make the most sense as an average-utilitarian framework with little or no attention given to future preferences).
My preferences do contain a large negative term for death (and I don’t free climb anymore, incidentally). I’m not that reckless.
Okay, yes. However, I’m almost certain that having killers running around unchecked will not produce the optimal level and type of fear in the greatest possible number of people.
Why? A simulated threat prompts an immediate response, but killers on the loose prompts a lot of worrying over a long period of time. While fighting off a murderer might spike your adrenaline, that’s not what killers on the loose will do. Instead people will lock their doors. They’ll fear for their safety. They’ll be afraid to let strangers into their home. They’ll worry about what happens if they have a fight with their friend—because the friend can commit murder with impunity. They’ll look over their shoulders. Parents will spend every second worrying about their children. The children will have little or no freedom, because the parents won’t leave them alone and may just keep them inside all the time, which is NOT optimal. People will have a lot of cortisol, depressing immune systems and promoting obesity.
That’s NOT THE SAME as a single burst of adrenaline, whether from falling while climbing or from watching a movie or even from fighting for your life. So I guess you’re right that it’s not about whether it’s real or not (though if it’s a game, then when it gets too intense, you can just turn it off, and you can’t turn off real life), but about the type of threat. However, the simulated threat doesn’t actually make you less likely to continue living, whereas a real threat does.
Well, of course I don’t think that allowing murder without restriction is going to make everyone fun-theoretically better off, let alone maximally satisfy their preferences over the utilitarian criteria I actually believe in. My original claim was a lot narrower than that, and in any case I’m mostly playing devil’s advocate at this point; although I really do think that fun-theoretic optimization is best approached without reflexively minimizing things like fear or pain on grounds of our preexisting heuristics. That said, I’m not sure this is always going to be true:
We know about a lot of societies with a lot of different accepted levels of violence. The most violent that I know of present up to about a 30% chance of premature death, so much higher than anything Western society presents that it’s scarcely conceivable (even front-line soldiers don’t have those death rates, although front-line service is more dangerous per unit time). But there’s very much not a monotonic relationship between level of violence and cultural paranoia, or trust of strangers, or freedom given to children. Early medieval Iceland, for example, had murder rates orders of magnitude higher than what we see now (implicit in textual sources, and confirmed by skeletal evidence); but children worked and traveled independently there, and hospitality to strangers was enshrined in law and custom. The same seems to go for more contemporary societies if the murder rates I’ve seen are at all accurate, although I don’t have as rich a picture for most of them. Our cultural fears of violence are very poorly correlated with actual expectations, as even a cursory glance over the most recent child molestation scare should show.
If studies of relative cortisol levels have ever been performed, I don’t know about them; but the cultures themselves don’t seem to show evidence of that kind of stress. I’d expect to see more paranoia following a recent uptick in violence, but I wouldn’t expect to see it well correlated with the base rate.
Okay. What kind of murder are we talking about? What made up most of the extra—was it all sorts of things or was it duels? And was it accepted or was it frowned on? Were murderers prosecuted? Did victims’ families avenge them?
Good point.
I’m not historian enough to say for sure, unfortunately. Judicial duels were part of the culture there, but the textual sources indicate that informal feuds were common, as were robbery and various other forms of informal violence. You could bring suit upon a murderer or other criminal in order to compel them to pay blood money or suffer in kind, but there was much less central authority than we’re used to, and nothing resembling a police force.
Was it by any chance a culture of honor?
Yes. Don’t get too hung up on the specific example, though; I chose it only because it’s a time and place that I’ve actually studied. The pattern (or, really, lack of a pattern) I’m trying to point to is much more general, and includes many cultures that don’t have a strong emphasis on honor.
Okay.