Morality for sociopaths, 101

[madd0ct0r]

Hypothesis: using game theory, it is possible to construct a purely selfish moral framework that has identical outward results as other social or religous models.

[Simon_Jester]

Almost certainly not, although that doesn't stop people from trying to invent rationalizations so they can pretend that every ethical qualm they have is purely rational... for some reason or other.

Rule utilitarianism probably comes closest to emulating our actual moral systems of any major kind of moral philosophy, because "follow the rules that give the best outcomes in general, even if they don't do so in this particular case" allows us to encompass such complicated values as 'fairness,' 'mercy,' and 'prudence.'

But rule utilitarianism cannot be justified in purely selfish terms, because it requires an individual person to knowingly accept less than the best for themselves

[madd0ct0r]

only in one off transactions.

For instance, instead of paying for my groceries I could steal them - shoplifting.
In the short term, that's an improvement for me, no tedious queuing or having to spend money. If I am never returning to the shop again, and consider the risk of being caught this time small enough, why not do it? It benefits me.

In the long term it's risky as the cost of a court case may wipe out the value gained from many 'shopping' trips. You don't know if they've got you on cctv, so returning imposes a certain risk of consequence on you, meaning you can't shop there any more, with a further cost of convenience. Factor in the fact I know I have imperfect knowledge, and therefore there's a fuzziness on those probabilities and I need a safety factor, and the lowest risk option (paying) looks pretty damn good.

If i worked this out every time I go shopping, that's a significant thinking cost, so it's cheaper in the long run to just pay and miss out on the rare opportunities to shoplift £3 of veg.

[Cykeisme]

Not sure, but I think fear of getting caught and the legal reprisal/social ostracism you may face is extraneous when we're considering morality alone.

The question is, if you knew for sure you'd get away with the theft without anyone knowing, would you (SHOULD you) nick those veggies?

Correct me if I'm wrong and tell me why, but I don't think any framework constructed solely from game theory, or rational (but amoral) thought, would answer "no". In fact, amorally, I'd say you should steal the veggies if you want/need them.

[Purple]

Not sure, but I think fear of getting caught and the legal reprisal/social ostracism you may face is extraneous when we're considering morality alone.

The question is, if you knew for sure you'd get away with the theft without anyone knowing, would you (SHOULD you) nick those veggies?

Correct me if I'm wrong and tell me why, but I don't think any framework constructed solely from game theory, or rational (but amoral) thought, would answer "no". In fact, amorally, I'd say you should steal the veggies if you want/need them.

You forgot one thing. A rational actor recognizes that if I can do something so can everyone else. And if everyone was to start stealing shops would simply stop keeping stock as it would make no sense for them to do so. So it is in my interest not only to restrain my self but to restrain others from stealing in order to ensure that stores keep the stock I want them to.

If you want to construct any form of utility based morality system you simply have to step away from just observing cases of isolated risk-reward calculations and translate to a much wider view.

[madd0ct0r]

Not sure, but I think fear of getting caught and the legal reprisal/social ostracism you may face is extraneous when we're considering morality alone.

The question is, if you knew for sure you'd get away with the theft without anyone knowing, would you (SHOULD you) nick those veggies?

Correct me if I'm wrong and tell me why, but I don't think any framework constructed solely from game theory, or rational (but amoral) thought, would answer "no". In fact, amorally, I'd say you should steal the veggies if you want/need them.

Yes. It would be a net benefit to you, and that's all that matters. I'm not sure purple's argument holds water since by definition you are operating outside of group morality. There is effectively zero probability that everyone else will also steal. There is, however, always a non-zero chance you get caught. It's less of an issue at £3 of veg, but for something like murder it's very hard to predict what forensics might be able to recover 30 years from now. Look at all the cold cases being settled on dna now. The key to being a successful psychopath is to know the limits of your ability to predict the future.

[Luke Skywalker]

But rule utilitarianism cannot be justified in purely selfish terms, because it requires an individual person to knowingly accept less than the best for themselves

IMHO: If the rules in this society are adequately enforced, then disobeying them would not be very conducive to your self interest, and you would agree to pay into the enforcement of those rules for fear of others otherwise hurting you. This is especially prudent where anarchy would result in MAD, not where particularly powerful and immunized individuals would have otherwise benefitted from a lack of such rules (ie, anti-trust laws are good for most people but bad for the people who would have profited from the trusts, but rules against murder are a plus for everyone, because even the most badass killer knows he'd otherwise not live for very long.)

Of course, in Real Life, there are plenty of instances where you could break the rules knowing full well you would get away with it, or receive only a minor punishment, or where a proposed rule would "even" the playing field rather than be universally beneficial, and therefore be opposed (in purely sociopathic terms) by those "above" in the playing field, etc. In an idealized society where the rules were perfect and the enforcement was inescapable, sociopaths would be especially obedient.

You forgot one thing. A rational actor recognizes that if I can do something so can everyone else. And if everyone was to start stealing shops would simply stop keeping stock as it would make no sense for them to do so. So it is in my interest not only to restrain my self but to restrain others from stealing in order to ensure that stores keep the stock I want them to.

That seems to be based on the fallacy that your choice to engage or not engage in theft would somehow have an effect on others' choices to engage or not engage in theft, even if you are never caught or detected. Likewise, in purely selfish terms, you would not want to go through significant financial burden and inconveniences to make sure that your home and transportation are environmentally friendly. Even though your self interest would benefit from a cleaner Earth, your individual efforts would be diluted into the entire ecosystem and have a negligible effect on your own wellbeing, while taking a massively greater proportion from your wallet. Your argument (unless if I am horribly misunderstanding it) comes in to point out that, if everybody else had this mindset, everybody else and yourself would be fucked. But that's irrelevant; your choices do not magically effect others' choices. The only way they would is if there were some sort of public accountability system. From a sociopath's perspective, it would be logical to agree to this system, but it would only be logical to actually follow it insofar as you could get caught.

[Luke Skywalker]

Apologies for the double post, edit time ran out:


I suppose the key to a system for sociopaths outwardly identical to normal ethics is the enforcement. You need the Rules to serve, on net, the benefits of a large and powerful enough portion of the society to suppress the dissidents, to a significant enough extent that the sociopaths find that they will net-benefit from the implementation even factoring in the money needed for enforcement. So, to 99% of the population, taking wealth from the upper 1% would be great for them, and they're numerous and powerful enough to enforce this regardless of the minority dissenters.

I think one of the reasons, beyond the obvious, that enforcement has to be so perfect would be that, to a sociopath, the argument "if we don't do X, everyone will suffer, including me" doesn't really mean anything. Your paying your taxes, for example, might seem logical given the obvious benefit of government revenue. But, your individual taxes will be so diluted that they'll hardly benefit yourself at all, so while you'd love it if everyone else paid, you yourself would have no intrinsic desire to, and, in absence of adequate enforcement, whether or not you pay would not at all impact whether or not everyone else pays. So the net result is that nobody pays their taxes, and everybody starves.

Not to turn this political, but doesn't this sort of reveal a pretty fatal flaw in libertarian philosophy? There are certain public issues such as climate change or even taxation that, from a self-interest game theory perspective, invariably result in disaster. The only workarounds are reliance on individuals' consciences, which is not exactly the most universal feature, or to have a public accountability system, ie, government enforcement. So you can't just rely on the invisible hand of the market, because the invisible hand assumes that your actions will have immediate and karmic repercussions on yourself, rather than the diluted "I don't notice a thing" consequences of shoplifting or littering, that have collective consequences.

[Simon_Jester]

An externally imposed legal system that punishes a hypothetical sociopath for doing things that everyone else considers wrong is one thing.

That is not the same as what is proposed in the original post: a "purely selfish" system which produces identical results to a recognizable moral code.

Because if you're only doing what society deems right because otherwise you get in trouble for breaking the law, you're not doing it because you're following a purely selfish ethical system. You're doing it because other people are following an unselfish ethical system. And they went out of their way to compel you to act in line with their values, against your will.

only in one off transactions.

For instance, instead of paying for my groceries I could steal them - shoplifting.
In the short term, that's an improvement for me, no tedious queuing or having to spend money. If I am never returning to the shop again, and consider the risk of being caught this time small enough, why not do it? It benefits me.

In the long term it's risky as the cost of a court case may wipe out the value gained from many 'shopping' trips. You don't know if they've got you on cctv, so returning imposes a certain risk of consequence on you, meaning you can't shop there any more, with a further cost of convenience. Factor in the fact I know I have imperfect knowledge, and therefore there's a fuzziness on those probabilities and I need a safety factor, and the lowest risk option (paying) looks pretty damn good.

If i worked this out every time I go shopping, that's a significant thinking cost, so it's cheaper in the long run to just pay and miss out on the rare opportunities to shoplift £3 of veg.

This argument breaks down when it comes to unethical and obnoxious and immoral actions that are not actually illegal. There are a lot of ways to be immoral that are legal.

[Luke Skywalker]

An externally imposed legal system that punishes a hypothetical sociopath for doing things that everyone else considers wrong is one thing.

That is not the same as what is proposed in the original post: a "purely selfish" system which produces identical results to a recognizable moral code.

By "purely selfish moral framework", I was thinking about one that was adopted by an entire society, designed to function without reliance on any altruistic morality. ie, how a community of sociopaths would interact with one another. Corporations and, to a lesser extent, nation-states are virtually sociopathic by human standards, and they have created codes of conduct with one another to varying degrees of success.

If we're just talking about adopting a moral system for an individual, then we would have to figure out what self interests this individual prioritizes. If it's happiness and life satisfaction, and the individual does have the emotional feedback loop from altruism and selfishness, then wouldn't "I don't want to feel guilty" and "I want to feel happy about myself" justify charity and good deeds? If the individual prioritizes wealth or power and/or is a literal sociopath, then it doesn't seem to be possible.

Because if you're only doing what society deems right because otherwise you get in trouble for breaking the law, you're not doing it because you're following a purely selfish ethical system. You're doing it because other people are following an unselfish ethical system. And they went out of their way to compel you to act in line with their values, against your will.

Yes, so? If you're trying to develop this moral system for the real world, you'll have to take into account realities such as the existence of external legal and ethical systems beyond your ability to alter. This is sort of like saying that, "don't shoot people because you'll cause severe injury or death", isn't something you're doing by your own choice, but because the laws of physics and biology are imposing rules on your against your will. Either way, they're realities your moral system will have to deal with.

only in one off transactions.

For instance, instead of paying for my groceries I could steal them - shoplifting.
In the short term, that's an improvement for me, no tedious queuing or having to spend money. If I am never returning to the shop again, and consider the risk of being caught this time small enough, why not do it? It benefits me.

In the long term it's risky as the cost of a court case may wipe out the value gained from many 'shopping' trips. You don't know if they've got you on cctv, so returning imposes a certain risk of consequence on you, meaning you can't shop there any more, with a further cost of convenience. Factor in the fact I know I have imperfect knowledge, and therefore there's a fuzziness on those probabilities and I need a safety factor, and the lowest risk option (paying) looks pretty damn good.

If i worked this out every time I go shopping, that's a significant thinking cost, so it's cheaper in the long run to just pay and miss out on the rare opportunities to shoplift £3 of veg.

This argument breaks down when it comes to unethical and obnoxious and immoral actions that are not actually illegal. There are a lot of ways to be immoral that are legal.

The law isn't the only consequence one might have to deal with from immoral behavior. One could face social alienation, retaliation, guilt/negative emotions, etc. You'd have a hard time networking with others if you're a consistent jackass. Even if you are just doing this to strangers you'll never hear from again, at least in some situations you could argue that the negative habit-forming, mental effort and potential risks would still have a negative payoff. You would still feel guilt and perhaps embarrassment unless if you are by nature incapable of such emotions.

[madd0ct0r]

The law isn't the only consequence one might have to deal with from immoral behavior. One could face social alienation, retaliation, guilt/negative emotions, etc. You'd have a hard time networking with others if you're a consistent jackass. Even if you are just doing this to strangers you'll never hear from again, at least in some situations you could argue that the negative habit-forming, mental effort and potential risks would still have a negative payoff.

Exactly - for a trite example, someone who avoids the time and cost of preparing his own lunch by stealing other people's from the office fridge is probably going to frozen out of the office very quickly. The legal system is really just this sort of effect writ large and with a long memory.

Game theory suggests that for a single game of prisoner's dilemma, cheating is the default choice, but for long repeated games, it gets much more complicated, with cautious cooperation generally being a good strategy. It also suggests a certain proportion of a population can be assholes. More then that and the assholes start to hurt each other and it's less good to be an asshole. Below that proportion, there's room for another asshole to invade the population.

[Simon_Jester]

By "purely selfish moral framework", I was thinking about one that was adopted by an entire society, designed to function without reliance on any altruistic morality. ie, how a community of sociopaths would interact with one another. Corporations and, to a lesser extent, nation-states are virtually sociopathic by human standards, and they have created codes of conduct with one another to varying degrees of success.

Yes, which mostly works because the actual decisions are still being made by actual humans who at least reference concepts like 'fairness,' 'justice,' 'mercy,' and so on. The laws are made, or at least are supposed to be made, by people who believe in 'fair' for some meaningful standard of 'fair.'

I don't think it'd actually work in the complete absence of altruistic morality; there is a reason our ancestors evolved to instinctively follow such morality among their own fellows.

It's like, think about where the word "sociopath" comes from: 'society-suffering' or 'social-sickness.' It really is hard for me to imagine widespread sociopathy not being a handicap when trying to design a functional social system.

Because if you're only doing what society deems right because otherwise you get in trouble for breaking the law, you're not doing it because you're following a purely selfish ethical system. You're doing it because other people are following an unselfish ethical system. And they went out of their way to compel you to act in line with their values, against your will.

Yes, so? If you're trying to develop this moral system for the real world, you'll have to take into account realities such as the existence of external legal and ethical systems beyond your ability to alter. This is sort of like saying that, "don't shoot people because you'll cause severe injury or death", isn't something you're doing by your own choice, but because the laws of physics and biology are imposing rules on your against your will. Either way, they're realities your moral system will have to deal with.

What I mean is that if your system of ethics relies on external legal constraints, it cannot be said to be 'functionally equivalent' to a system that doesn't.

And if we look at the original question, he seems to be asking for a functional equivalent.

The law isn't the only consequence one might have to deal with from immoral behavior. One could face social alienation, retaliation, guilt/negative emotions, etc. You'd have a hard time networking with others if you're a consistent jackass. Even if you are just doing this to strangers you'll never hear from again, at least in some situations you could argue that the negative habit-forming, mental effort and potential risks would still have a negative payoff. You would still feel guilt and perhaps embarrassment unless if you are by nature incapable of such emotions.

Then you're relying on your own instinctive responsiveness to moral rules that, by officially being "purely selfish," you're supposed to not care about.

I mean, I might as well just copy-paste from Hillel: "That which is hateful unto you, do not do to your neighbor... The rest is commentary. Go forth and study." But that relies on your own sense of what is hateful, what is unfair, what is unjust... and if you already have such a sense then you are by default operating on some non-selfish moral system.

[madd0ct0r]

I was thinking on this a little more.

It seems reasonable to me a intelligent sociopath would recognise the harm to their future self due to a tragedy of the commons. They would not act with noble restraint to prevent the field being overgrazed since not acting harms them directly and the amount they suffer from the overgrazing would be scaled down, but they could selfishly support legislation that would control everyone's grazing rights, including their own.

Sort of leviathan writ large.

[Simon_Jester]

At that point you're envisioning a hyper-intelligent sociopath, and one who doesn't expect to be able to use personal gain to shield themselves from harm.

I mean, if you really want to see how a bunch of high-functioning and successful sociopathic types behave, look no farther than the hedge fund managers. And they showed no sign of wanting to regulate and restrain their own industry. Why bother, when they personally would be made so wealthy by all the piles of fraud and nonsense and financial three-card monte? The sheer mountains of cash involved would make them practically immune from prosecution in the regulatory environment of the time, they knew it, and they liked it that way.

[madd0ct0r]

hyper intelligent? really? It's not that advanced, especially for someone already relies on game theory to model people. Maybe they've read around the topic?

Good point with the 1) immunity from wealth and 2) hedge fund managers.

I was actually thinking about climate change, where the cost of change needed per person is quite high and the benefits less certain. It's not far off a calculation that most people make:

On your way to work, you see a kid drowing in a pond, and the cost to you to rescue him would be $75 cleaning bill for your suit. Would you rescue him?

On your way to work, you see an advert for the famine in Ethiopia. They are asking for $75 to feed a child for the crisis duration. Would you send them $75?

[PKRudeBoy]

Yes to the first and no to the second, for the simple reason that even assuming that the charity is not a fraud or poorly managed, the food will most likely be seized by a warlord to sell on the black market anyways.

[Simon_Jester]

hyper intelligent? really? It's not that advanced, especially for someone already relies on game theory to model people. Maybe they've read around the topic?

Our proposed game theorist sociopath... hm. I doubt there are many, because mathematical systems like game theory are bad at modeling if you don't have some understanding of what you're modeling, and sociopaths don't. But if he exists, it might be comparatively trivial to duplicate one and only one of these 'enlightened self-interest' decisions. Then again, that's a far cry from somehow deriving a consistent system of 'best practices' that he follows rigorously and that precisely mimic a real system of ethics.

It's not far off a calculation that most people make:

On your way to work, you see a kid drowing in a pond, and the cost to you to rescue him would be $75 cleaning bill for your suit. Would you rescue him?

On your way to work, you see an advert for the famine in Ethiopia. They are asking for $75 to feed a child for the crisis duration. Would you send them $75?

That has to do with social instincts like diffusion of responsibility. Humans feel more responsible for an emergency that occurs right in front of them, where at least implicitly they are one of the few people in a position to act. The hypothetical drowning boy is presumably not drowning in front of a small army of spectators including lifeguards... while the Ethiopian famine is no more proximate to our hypothetical commuter than it is to hundreds of millions of other people and numerous national governments.

But a sociopath is not going to share this feeling, they're going to be quite unsympathetic to disasters right in front of them, not just disasters that occur far away. Moreover, they will be unsympathetic, in a way the commuter probably isn't, because the commuter is likely to respond to the emotional appeal on some level even if they ultimately decide "hey, I need the money, and if I went around donating 75$ every time I heard about someone in trouble I'd go bankrupt in a few weeks."

[madd0ct0r]

and PKRudeboy proves my point on people estimating probability as part of the 'social instinct'.

I see your challenge though - I can arbitrarily create a method for dealing with any example, but it's not a consistent system of best practice, just an algorithm for generating one. That's rather more difficult

[Simon_Jester]

Well, the thing is, I don't think any (quasi-simple) mathematical algorithm can replicate the evolved system of 'fuzzy logic' that humans actually use for morals. This isn't because the system is somehow transcendent; it's because the system is messy, inconsistent, and often outright illogical or even antilogical. There are situations where you will come across as a pure bastard-coated bastard with bastard filling* just by logically pursuing your own interests. Or even by suggesting that your own in-group logically pursue its own interests.

So replicating it by precise, logical, analytical calculations is never going to be easy. At best you might be able to come up with a system of logical actions that can be justified as ethical to another logician. Which is not the same thing.


*Thanks to Dark_Hellion for inventing that phrase...

[madd0ct0r]

hmm. Messy, illogical, limited agreeement upon outputs? Normally faced with that I'd add a random walk term

Today you die Mr Bond, becuase it's Tuesday.

[Simon_Jester]

Except it's not random, it's actually fairly predictable... it's just different. People are very consistently illogical in certain ways.