Fun post! I like Parfit's argument from analogy, but don't think either horn of your first argument really works:
1. God could reward irrationality. In that case, it would be rational to *acquire* this irrational state. But it doesn't make the state itself rational.
2. I don't think it's "clearly irrational" to have extra degreed concern in proportion to degrees of similarity (possibly in combination with the right kind of causal connectedness). When I think about Parfitian identity-spectrum cases, for example, it seems entirely intuitive to think that my degree of prudential concern for the resulting person should be proportional to how much of "me" survives in them. It also seems that I have more self-interested reason to care about the myself of next year than the (more different) myself of decades hence.
I have to agree with point (2) above. I found it bizarre that section 2 was accompanied by a note saying that the argument is taken from Parfit even though it doesn’t seem faithful to Parfit’s work.
Parfit does not use the brittle-ness of personal identity to claim that there is no basis for egoistic concern whatsoever. He instead argues that egoistic concern is accounted for by psychological continuity.
On a more tangential note, I am curious to know how believers in an immaterial soul explain Parfit’s division case (i.e. the left and right hemispheres being transplanted in different bodies). This case seems at least as hard for dualists to deal with as reductionists.
Very interesting post, I'm glad I read this! But this all relies on moral realism being true (I think?): If morality is subjective, then your moral goals (e.g minimising suffering) are just things that you desire rather than objective moral obligations. So if you buy a coffee for $5 instead of donating it, even if you consider yourself utilitarian, then that is selfish (according to your own subjective moral axioms/beliefs) - But it can't be irrational because you're just doing what you desire in the moment, which might be all that morality is. I believe you're a moral realist anyways, but do you think this logic holds if we assume moral anti-realism for a second?
"So then if atheism is true, personal identity is sufficiently unreal not to be the source of reasons."
I think this proves too much, and not for reasons that specifically have to do with personal identity. Personal identity isn't *unusually* vague or indeterminate compared to lots of other stuff we care about. The logical endpoint of this style of argument is that the only thing we can rationally care about is the distribution of fundamental particles, or what the state of the universal wave function is, or some other type of fact for which we have some kind of strong guarantee that no vagueness/indeterminacy could arise. But none of our actual concerns are like that.
Alternatively, the atheism horn of the argument, if successful, hits a lot more than just selfishness!
With the other things, I think there's some potentially important joint-carving specification. E.g. doesn't seem bad to assign some modest intrinsic value to friendship given some specific way of defending it.
This is one of the better arguments for hedonism imo--that pleasure is most plausibly joint-carving.
Friendship will be just as vague and indeterminate--more so, no?--than identity. The same spectrum/sorites style arguments Parfit uses to argue that identity can be vague/indeterminate will work for friendship too. Start with a case where you're friends with someone, modify it bit by bit to end up with a case where you're enemies, hard to say just when you stop being friends. (Of course, you can think of friendship as a matter of degree, which is very natural, but you can make the same sort of move to save self-interested concern, as Richard suggests above.)
As for hedonism, I agree, with the important caveat that this sort of argument only works for a very particular sort of non-materialist about the mind. You need to think, not just that mental facts are fundamental (so all forms of materialism are out), but pleasure facts are among the metaphysically fundamental mental facts. Maybe that's not a big bullet to bite if you already like dualism/idealism, I'm not sure.
I am a philosophical altruist, but I will make a devil's advocate argument for egoism based on personal identity. While it might be true that we are in some sense a "different person" than we were when we were 8, and will also be different by the time we're 80, there is something past and future selves have relative to us that truly other people don't. That is a continuous stream of first-person perspective. My 8 year old self took part in a stream of qualia that I still have memory of now, and when I am 80 I will likewise have memory of present qualia. Meanwhile, qualia of truly other people are completely off-limits to me. They will never be part of my stream of consciousness nor will I remember them. So long as we recognize that there is some first-person experiential subject that experiences "its own" qualia over time, but never those of other people, there is a sense in which it is sensible to privilege the past and future self over true others. Simply put, I have felt and will feel my share of the world's experience, and will never feel that of anyone else, so it makes sense to privilege benefiting my own experience over that which I will never feel.
I worry about thought experiments which don’t justify their relevance. You can’t turn me gradually into Barbara. You can turn 4 year old me gradually (a day at a time) into 40 year old me. They aren’t equivalent.
But this is a discussion about gradual change. The point I’m trying to make about thought experiments is you need to show they are actually relevant for the argument.
If I cut off your leg and Barbara's leg and grafted Barbara's leg onto your body, the overlap between your new atoms and pre-amputation-Barbara's atoms would be much greater than the overlap between 4-year-old you's atoms and 40-year-old you's atoms.
Point 1 doesn't work, because either are unselfish out of gratitude, you love because he first loved us, or you are unselfish out of fear, and that cannot work long term because the fear of eternal punishment or missing out on external rewards just isn't strong enough. Even then this is stupid because a lot of people come to God because they are broken or flawed and need him; they aren't full to begin with to be able to act unselfishly, and it may be a sign of God that they change even a little over time.
Point 2, ugh. seriously, this is stupid. Yes, people change over time, but they don't change into an entirely different person. That's just a figure of speech meaning they believe different things. Even if your brother changes, he is still recognizable as your brother and still has the history and memories that link you to him. A seed for an oak tree will not grow into a cat.
Point 3 is kind of dumb also, if people can't manage even to take care of themselves, who they do know, how can they do so with someone they barely know in part? They would need to fix themselves then worry about others, because it's harder to help people you don't know than people you do.
You need to be careful about trying to solve everything intellectually. Effective Altruism is turning into ineffective altruism because it's doing stuff like this argument; shearing things like unselfishness from the real world and turning them into intellectual concepts divorced from existence. You can try to be a little better, you can't just be something even Jesus Christ wasn't.
I read through this piece twice to make sure that it didn’t mention shared memory or consciousness, which I think are the most obvious reasons we think of our future selves as the same as our present selves. Isn’t that a glaring omission?
Im a christian but still a little bit puzzled by your 'if God exists' scenario. Morality (to me at least) seems tobe obviously decided by intentions. Setting out to be less selfish because God will reward you for it does not seem like a good intention because you still have selfish intentions, you are just being optimally selfish by helping others.
Does helping others only for your selfinterest make you better than a person acting selfishly only for their selfinterest?
True but that does not really answer my question. Of course it would be better for the world if a person did a good thing with purely selfish motivations but it would not make him a better person right?
Your argument depends on what you mean by "selfishness." If you mean valuing yourself at 100% and everyone else at 0%, then it works for these and lots of other reasons. If you mean valuing yourself at 100% and everyone else at 100% then this gives no guidance at all. For example, giving all your money (or all your money above the world average) to others would be equivalent to moving your money from one pocket to another (or one bank to another). That seems silly.
I think a better model would be that others should simply enter into your utility function. For example, your utility function would be U(X,P) = U(X) + sum of U(p sub i)*alpha sub i, where X is a large vector of the state of the world for you (all goods and services you consume, now and in the future) and P is a vector of every other person (p sub 1, p sub 2, ... p sub n) multiplied by alpha sub i (0 <= alpha sub i, and probably <= 1 for almost everyone but perhaps > 1 for someone you love). That is, you include everyone else's utility function in yours.
The strong version of your argument is that alpha = 1 for everyone. This seems incorrect, certainly descriptively and probably normatively. I don't think God will punish you or fail to reward you in the afterlife for not giving away everything above the world average. And I don't think people are unable to distinguish meaningfully between themselves and others, despite the fact that there are an infinite number of small changes between you and Barbara and you and your 4-year-old self.
But setting alpha = 0 for everyone is pure selfishness and irrational.
One interesting question is whether alpha can have different values for different people. Can we value our children, parent, family, friends, community more than distant strangers? I think so. How much more? Tricky question.
Separate question: can alpha be negative for someone? That is, can we value the disutility of an enemy?
All of these are arguments for the second thing. "I don't think God will punish you or fail to reward you in the afterlife for not giving away everything above the world average"
There's some reasonable chance of getting reward proportional to the effort put in.
I am incredibly sympathetic to the arguments here, and am grateful for the post, but I have trouble resolving some counterarguments to the "atheism" points (having not read Parfit).
I can think of one distinguishing property that seems to be a "genuine reason" to care more about your future self over others: your continuous stream of subjective experience. "I care about me more because I know that I will have that person's experience and personally suffer the consequences!"
A counter to this might then be: "but why does it matter that you will experience it personally rather than someone else?," in which case one might reply "because I care about my future self more due to the continuous experience stream", and it's not clear how this is resolved.
A defense for the kind of partial selfishness (particularism), that it is better to help people close to us than kids in Malawi: we have more information and influence.
1) we know who deserves help, that is, who will not use that help for something evil, or squander it
2) we know their needs better, know better how exactly to help
3) we can compel them to pay it forward
We just don't know whether mosquito net in Malawi will cause overpopulation and a civil war over resources, for instance. But we know our nephew who wants to be a doctor and needs money for education is a reasonably decent person, actually capable of becoming a good doctor, and will honor a promise to pay it forward.
Fun post! I like Parfit's argument from analogy, but don't think either horn of your first argument really works:
1. God could reward irrationality. In that case, it would be rational to *acquire* this irrational state. But it doesn't make the state itself rational.
2. I don't think it's "clearly irrational" to have extra degreed concern in proportion to degrees of similarity (possibly in combination with the right kind of causal connectedness). When I think about Parfitian identity-spectrum cases, for example, it seems entirely intuitive to think that my degree of prudential concern for the resulting person should be proportional to how much of "me" survives in them. It also seems that I have more self-interested reason to care about the myself of next year than the (more different) myself of decades hence.
I have to agree with point (2) above. I found it bizarre that section 2 was accompanied by a note saying that the argument is taken from Parfit even though it doesn’t seem faithful to Parfit’s work.
Parfit does not use the brittle-ness of personal identity to claim that there is no basis for egoistic concern whatsoever. He instead argues that egoistic concern is accounted for by psychological continuity.
On a more tangential note, I am curious to know how believers in an immaterial soul explain Parfit’s division case (i.e. the left and right hemispheres being transplanted in different bodies). This case seems at least as hard for dualists to deal with as reductionists.
Very interesting post, I'm glad I read this! But this all relies on moral realism being true (I think?): If morality is subjective, then your moral goals (e.g minimising suffering) are just things that you desire rather than objective moral obligations. So if you buy a coffee for $5 instead of donating it, even if you consider yourself utilitarian, then that is selfish (according to your own subjective moral axioms/beliefs) - But it can't be irrational because you're just doing what you desire in the moment, which might be all that morality is. I believe you're a moral realist anyways, but do you think this logic holds if we assume moral anti-realism for a second?
Agree the arguments work better on realism. But they also give some reason to think an anti-realist's reflective values would be selfless.
"So then if atheism is true, personal identity is sufficiently unreal not to be the source of reasons."
I think this proves too much, and not for reasons that specifically have to do with personal identity. Personal identity isn't *unusually* vague or indeterminate compared to lots of other stuff we care about. The logical endpoint of this style of argument is that the only thing we can rationally care about is the distribution of fundamental particles, or what the state of the universal wave function is, or some other type of fact for which we have some kind of strong guarantee that no vagueness/indeterminacy could arise. But none of our actual concerns are like that.
Alternatively, the atheism horn of the argument, if successful, hits a lot more than just selfishness!
With the other things, I think there's some potentially important joint-carving specification. E.g. doesn't seem bad to assign some modest intrinsic value to friendship given some specific way of defending it.
This is one of the better arguments for hedonism imo--that pleasure is most plausibly joint-carving.
Friendship will be just as vague and indeterminate--more so, no?--than identity. The same spectrum/sorites style arguments Parfit uses to argue that identity can be vague/indeterminate will work for friendship too. Start with a case where you're friends with someone, modify it bit by bit to end up with a case where you're enemies, hard to say just when you stop being friends. (Of course, you can think of friendship as a matter of degree, which is very natural, but you can make the same sort of move to save self-interested concern, as Richard suggests above.)
As for hedonism, I agree, with the important caveat that this sort of argument only works for a very particular sort of non-materialist about the mind. You need to think, not just that mental facts are fundamental (so all forms of materialism are out), but pleasure facts are among the metaphysically fundamental mental facts. Maybe that's not a big bullet to bite if you already like dualism/idealism, I'm not sure.
Self-interest being a matter of degree seems a lot weirder than friendship being a matter of degree!
I am a philosophical altruist, but I will make a devil's advocate argument for egoism based on personal identity. While it might be true that we are in some sense a "different person" than we were when we were 8, and will also be different by the time we're 80, there is something past and future selves have relative to us that truly other people don't. That is a continuous stream of first-person perspective. My 8 year old self took part in a stream of qualia that I still have memory of now, and when I am 80 I will likewise have memory of present qualia. Meanwhile, qualia of truly other people are completely off-limits to me. They will never be part of my stream of consciousness nor will I remember them. So long as we recognize that there is some first-person experiential subject that experiences "its own" qualia over time, but never those of other people, there is a sense in which it is sensible to privilege the past and future self over true others. Simply put, I have felt and will feel my share of the world's experience, and will never feel that of anyone else, so it makes sense to privilege benefiting my own experience over that which I will never feel.
I worry about thought experiments which don’t justify their relevance. You can’t turn me gradually into Barbara. You can turn 4 year old me gradually (a day at a time) into 40 year old me. They aren’t equivalent.
But this is a discussion about gradual change. The point I’m trying to make about thought experiments is you need to show they are actually relevant for the argument.
If I cut off your leg and Barbara's leg and grafted Barbara's leg onto your body, the overlap between your new atoms and pre-amputation-Barbara's atoms would be much greater than the overlap between 4-year-old you's atoms and 40-year-old you's atoms.
Sounds like a very strong argument against materialism, since it is obvious he will not resemble Barbara more after that grafting.
However, I recall that people that receive heart transplants do gain traits from the donor.
Point 1 doesn't work, because either are unselfish out of gratitude, you love because he first loved us, or you are unselfish out of fear, and that cannot work long term because the fear of eternal punishment or missing out on external rewards just isn't strong enough. Even then this is stupid because a lot of people come to God because they are broken or flawed and need him; they aren't full to begin with to be able to act unselfishly, and it may be a sign of God that they change even a little over time.
Point 2, ugh. seriously, this is stupid. Yes, people change over time, but they don't change into an entirely different person. That's just a figure of speech meaning they believe different things. Even if your brother changes, he is still recognizable as your brother and still has the history and memories that link you to him. A seed for an oak tree will not grow into a cat.
Point 3 is kind of dumb also, if people can't manage even to take care of themselves, who they do know, how can they do so with someone they barely know in part? They would need to fix themselves then worry about others, because it's harder to help people you don't know than people you do.
You need to be careful about trying to solve everything intellectually. Effective Altruism is turning into ineffective altruism because it's doing stuff like this argument; shearing things like unselfishness from the real world and turning them into intellectual concepts divorced from existence. You can try to be a little better, you can't just be something even Jesus Christ wasn't.
I don’t care
I read through this piece twice to make sure that it didn’t mention shared memory or consciousness, which I think are the most obvious reasons we think of our future selves as the same as our present selves. Isn’t that a glaring omission?
Im a christian but still a little bit puzzled by your 'if God exists' scenario. Morality (to me at least) seems tobe obviously decided by intentions. Setting out to be less selfish because God will reward you for it does not seem like a good intention because you still have selfish intentions, you are just being optimally selfish by helping others.
Does helping others only for your selfinterest make you better than a person acting selfishly only for their selfinterest?
I agree that it would reflect a bad disposition. But it would be better to help others for selfish reasons than not to help them at all.
True but that does not really answer my question. Of course it would be better for the world if a person did a good thing with purely selfish motivations but it would not make him a better person right?
Your argument depends on what you mean by "selfishness." If you mean valuing yourself at 100% and everyone else at 0%, then it works for these and lots of other reasons. If you mean valuing yourself at 100% and everyone else at 100% then this gives no guidance at all. For example, giving all your money (or all your money above the world average) to others would be equivalent to moving your money from one pocket to another (or one bank to another). That seems silly.
I think a better model would be that others should simply enter into your utility function. For example, your utility function would be U(X,P) = U(X) + sum of U(p sub i)*alpha sub i, where X is a large vector of the state of the world for you (all goods and services you consume, now and in the future) and P is a vector of every other person (p sub 1, p sub 2, ... p sub n) multiplied by alpha sub i (0 <= alpha sub i, and probably <= 1 for almost everyone but perhaps > 1 for someone you love). That is, you include everyone else's utility function in yours.
The strong version of your argument is that alpha = 1 for everyone. This seems incorrect, certainly descriptively and probably normatively. I don't think God will punish you or fail to reward you in the afterlife for not giving away everything above the world average. And I don't think people are unable to distinguish meaningfully between themselves and others, despite the fact that there are an infinite number of small changes between you and Barbara and you and your 4-year-old self.
But setting alpha = 0 for everyone is pure selfishness and irrational.
One interesting question is whether alpha can have different values for different people. Can we value our children, parent, family, friends, community more than distant strangers? I think so. How much more? Tricky question.
Separate question: can alpha be negative for someone? That is, can we value the disutility of an enemy?
All of these are arguments for the second thing. "I don't think God will punish you or fail to reward you in the afterlife for not giving away everything above the world average"
There's some reasonable chance of getting reward proportional to the effort put in.
And you you're a two-boxer...maybe if you look at Parfit's "Traveling Hitchhiker" rather than Newcomb's transparent box?
I am incredibly sympathetic to the arguments here, and am grateful for the post, but I have trouble resolving some counterarguments to the "atheism" points (having not read Parfit).
I can think of one distinguishing property that seems to be a "genuine reason" to care more about your future self over others: your continuous stream of subjective experience. "I care about me more because I know that I will have that person's experience and personally suffer the consequences!"
A counter to this might then be: "but why does it matter that you will experience it personally rather than someone else?," in which case one might reply "because I care about my future self more due to the continuous experience stream", and it's not clear how this is resolved.
Jesus and Buddha were selfish, their selves just happened to include everybody else.
A defense for the kind of partial selfishness (particularism), that it is better to help people close to us than kids in Malawi: we have more information and influence.
1) we know who deserves help, that is, who will not use that help for something evil, or squander it
2) we know their needs better, know better how exactly to help
3) we can compel them to pay it forward
We just don't know whether mosquito net in Malawi will cause overpopulation and a civil war over resources, for instance. But we know our nephew who wants to be a doctor and needs money for education is a reasonably decent person, actually capable of becoming a good doctor, and will honor a promise to pay it forward.
The "no god" argument is basically Buddha's argument
> For this reason, and others, I’d suggest that at the very least you should take the morally low-hanging fruit and ...
I was expecting that to end with "...sign up for a paid subscription"! :-)
Pleasantly surprised by your nobleness.