You may be aware of this already, but I think there is a clear difference between saving an existing person who would otherwise have died—and in the process reducing suffering by also preventing non-fatal illnesses—and starting a pregnancy because before starting a pregnancy the person doesn’t exist yet.
What is that difference, from a consequentialist perspective?
(For the purpose of comparing apples to apples, let’s ignore the suffering reduced by preventing an illness. What’s the difference in outcome for a child between poofing them away at a young age, and preventing their birth?)
An existing child may have particular interest in their continued existence that is or has been realized physically while a nonexistent child doesn’t. Often actual (sometimes explicit) preferences to continue to exist, or future-oriented desires, even short-term ones, that would be frustrated if they die. Or maybe they have implicit interests in joy or the things they like, desire satisfaction or what they desire, because their brain is already arranged to like or desire these things, or some things in general, e.g. they have pleasure, desire and generally motivational/valuing subnetworks in their brain. Those interests would be less satisfied than otherwise if they die.
Am I understanding correctly that the distinction you outlined exists in preference utilitarianism, but not in hedonic utilitarianism? For example, if I were poofed away right now, from a hedonic utilitarian perspective, the only downside seems to be from the prevention of happy experiences I would have later had.
Also, does your argument work symmetrically? For example, if I could choose between ending the torture of an existing person, and preventing the creation of a person who would have been tortured, would your argument give strong reason to choose the former?
Within the preference-oriented perspective of your comment, has there been any exploration on how strong the trade-off should be between the preferences of existing moral patients and the future preferences of future moral patients?
For a simplistic example, when choosing between saving an existing happy person’s life, or creating 10 happy people, many consequentialists would prefer the latter. (The latter creates 10x the amount of happy life-years, which plausibly dominates the existing person’s preferences.) But even a 10-to-1 tradeoff would mean that preventing a happy person’s existence is 10% as bad as killing them—pretty bad!
Am I understanding correctly that the distinction you outlined exists in preference utilitarianism, but not in hedonic utilitarianism?
I gave two kinds of distinctions. Future-oriented preferences don’t seem to count in themselves under hedonic utilitarianism (the first distinction), but the fact that a being already exists with the circuitry and tendencies to experience pleasure (and/or suffering) could matter on some hedonic utilitarian views (the second distinction).
Also, does your argument work symmetrically? For example, if I could choose between ending the torture of an existing person, and preventing the creation of a person who would have been tortured, would your argument give strong reason to choose the former?
I think the arguments I gave don’t really say much either way about this. I think views where future people don’t (really) matter in themselves, like presentism or necessitarianism, are compatible (although I don’t give much weight to such views). On the other hand, you could defend similarly strong reasons and the procreation asymmetry based on actualism, Frick’s conditional reasons, or harm-minimization views. See this discussion. I think at least actualism is asymmetric in a non-question-begging way, and maybe Frick’s views, too, as I argue in the linked discussion.
Within the preference-oriented perspective of your comment, has there been any exploration on how strong the trade-off should be between the preferences of existing moral patients and the future preferences of future moral patients?
Within a view, I can’t really imagine how you would ground any particular tradeoff ratio, other than basically purely subjectively, i.e. your own preferences about these tradeoffs. You could get something like this across moral views under maximizing expected choiceworthiness over moral uncertainty with the right kind of intertheoretic comparisons between person-affecting views and total views (Greaves and Ord, 2017, another version), but it’s not clear what would ground such intertheoretic comparisons, because the views disagree about what makes something valuable. Maybe also something similar under other approaches to moral uncertainty, but the ratio could depend on the choices available to you.
From a consequentialist utilitarian perspective, there is also the impact on those that know the existing person.
The death of an existing person often causes suffering and loss for those that know them. Whereas preventing the existence of a future person typically does not cause this wider suffering (except perhaps in some cases, e.g. when parents strongly desire a child but are unable to conceive).
You’re right that that’s a source of consequentialist difference. However, out of the consequences of a person’s death, their death’s harsh effect on those who love them seems unlikely to be worse than the deprivation of decades of future happy life for the person who died.
To see this, let’s run with the premise that most of the badness of a person’s death comes from the suffering of their close friends/family members. For simplicity, let’s say whenever a person dies, 10 of their close friends/family suffer considerably. Let’s further say that every person lives through exactly 10 of their close friends/family dying.
Now imagine a baby who’s “choosing” whether or not they’d like to be born. If they choose not to be born, then they lose out on an entire happy life’s experience. BUT that experience would include living through 10 of their close friends/family members’ deaths, which our premise stated is worse than losing an entire life’s experience! So the baby should prefer to not be born.
So if most of the badness death of an existing person comes from the suffering the death causes on the person’s close friends/family, then ignoring other considerations, we should advocate for human extinction. Being born and having to suffer through 10 close friends/family dying is worse than not being born at all.
To me, this conclusion seems absurd. Even though I’ll suffer over life from the deaths of the people I love, on the whole, I’m happy to be alive.
In conclusion, while one’s death is worse because of the suffering of their close friends/family, it still seems that most of the badness of death comes from the person losing future happy life. So we’re still left with the conclusion that preventing a person’s existence is close to as bad as killing them.
I agree that ‘most of the badness of death comes from the person losing future happy life’.
However, there are also other factors that are relevant to whether ‘preventing a person’s existence is close to as bad as killing them’ (this obviously also depends what is meant by ‘close to’).
The claim seems to imply that we are doing something almost as bad as murder if we are failing to have as many children as possible. But a society where legislation reflected this position would reduce the quality of life of people who don’t want many (or any) children, would force women into being baby-making machines, and would plausibly result in a situation where individual children aren’t viewed as particularly important (if one dies, well, no time to grieve—another one is on the way). This strikes me as more absurd and dystopian than advocating for human extinction.
For me, this conversation is analogous to that surrounding Peter Singer’s book The Life You Can Save (TLYCS).
In TLYCS, Peter Singer argues, in my opinion quite convincingly, that we have a moral obligation to give up everything we have to help those in extreme poverty. Singer argues that every 5000 USD we spend on ourselves and not donate is equivalent to condemning a person whose life we could have saved. He then follows up with a far more modest ask: That we donate 1% of our income to effective charities.
There are many people who balk at Singer’s conclusion that we have a moral duty to donate everything above our bare survival needs to effective charities, and then reject his comparatively modest 1% ask. They might reply:
[Singer’s] claim seems to imply that we are doing something almost as bad as murder if we are failing to [donate as much money as possible]. But a society where legislation reflected this position would reduce the quality of life of people [in developed countries to that of those in extreme poverty], would force [people in developed countries] into being [money]-making machines, and would plausibly result in a situation where individual children [in extreme poverty] aren’t viewed as particularly important (if one dies, well, no time to grieve—another one is [waiting to be saved]). This strikes me as more absurd and dystopian than advocating for human extinction.
This isn’t what EAs actually advocate for. Singer’s conclusion is far too much to ask of most people, and even the most ardent EAs would balk at legislating it. However, many EAs, myself included, would agree that Singer’s philosophical conclusion really is correct.
Similarly, I’ve made the philosophical argument that there’s little moral difference between preventing a person’s existence and killing them. Given that conclusion, there are many compelling criticisms of what personal or legislative changes should follow. However, I haven’t found any convincing rebuttal to the philosophical argument.
There are many considerations which lessen the magnitude of the conclusion. Preventing the suffering of the close friends and family of a person who dies matters. One might have a high credence in a person-affecting view, endorse the procreation asymmetry, or place substantial credence on non-consequentialist theories.
But in my opinion, if you’re a consequentialist who holds even mild credence (say ~10%) in the non-person-affecting view, then preventing a person’s existence is on the order of badness of (say, ~10% as bad as) killing them. If you disagree, then I’d love to understand your perspective further, and see if there’s some crucial consideration I may be missing.
As with Singer’s arguments in TLYCS, I don’t think the truth/falsity of a philosophical argument is contingent on how radical its conclusions are. I also don’t think the existence of radical conclusions precludes the implementation of common-sense conclusions, like donating 1% of one’s income to effective charities.
This conceptually seems similar to the meat eater problem argument against global health interventions.
You may be aware of this already, but I think there is a clear difference between saving an existing person who would otherwise have died—and in the process reducing suffering by also preventing non-fatal illnesses—and starting a pregnancy because before starting a pregnancy the person doesn’t exist yet.
What is that difference, from a consequentialist perspective?
(For the purpose of comparing apples to apples, let’s ignore the suffering reduced by preventing an illness. What’s the difference in outcome for a child between poofing them away at a young age, and preventing their birth?)
An existing child may have particular interest in their continued existence that is or has been realized physically while a nonexistent child doesn’t. Often actual (sometimes explicit) preferences to continue to exist, or future-oriented desires, even short-term ones, that would be frustrated if they die. Or maybe they have implicit interests in joy or the things they like, desire satisfaction or what they desire, because their brain is already arranged to like or desire these things, or some things in general, e.g. they have pleasure, desire and generally motivational/valuing subnetworks in their brain. Those interests would be less satisfied than otherwise if they die.
Thanks for this perspective!
Am I understanding correctly that the distinction you outlined exists in preference utilitarianism, but not in hedonic utilitarianism? For example, if I were poofed away right now, from a hedonic utilitarian perspective, the only downside seems to be from the prevention of happy experiences I would have later had.
Also, does your argument work symmetrically? For example, if I could choose between ending the torture of an existing person, and preventing the creation of a person who would have been tortured, would your argument give strong reason to choose the former?
Within the preference-oriented perspective of your comment, has there been any exploration on how strong the trade-off should be between the preferences of existing moral patients and the future preferences of future moral patients?
For a simplistic example, when choosing between saving an existing happy person’s life, or creating 10 happy people, many consequentialists would prefer the latter. (The latter creates 10x the amount of happy life-years, which plausibly dominates the existing person’s preferences.) But even a 10-to-1 tradeoff would mean that preventing a happy person’s existence is 10% as bad as killing them—pretty bad!
I gave two kinds of distinctions. Future-oriented preferences don’t seem to count in themselves under hedonic utilitarianism (the first distinction), but the fact that a being already exists with the circuitry and tendencies to experience pleasure (and/or suffering) could matter on some hedonic utilitarian views (the second distinction).
I think the arguments I gave don’t really say much either way about this. I think views where future people don’t (really) matter in themselves, like presentism or necessitarianism, are compatible (although I don’t give much weight to such views). On the other hand, you could defend similarly strong reasons and the procreation asymmetry based on actualism, Frick’s conditional reasons, or harm-minimization views. See this discussion. I think at least actualism is asymmetric in a non-question-begging way, and maybe Frick’s views, too, as I argue in the linked discussion.
Within a view, I can’t really imagine how you would ground any particular tradeoff ratio, other than basically purely subjectively, i.e. your own preferences about these tradeoffs. You could get something like this across moral views under maximizing expected choiceworthiness over moral uncertainty with the right kind of intertheoretic comparisons between person-affecting views and total views (Greaves and Ord, 2017, another version), but it’s not clear what would ground such intertheoretic comparisons, because the views disagree about what makes something valuable. Maybe also something similar under other approaches to moral uncertainty, but the ratio could depend on the choices available to you.
From a consequentialist utilitarian perspective, there is also the impact on those that know the existing person.
The death of an existing person often causes suffering and loss for those that know them. Whereas preventing the existence of a future person typically does not cause this wider suffering (except perhaps in some cases, e.g. when parents strongly desire a child but are unable to conceive).
You’re right that that’s a source of consequentialist difference. However, out of the consequences of a person’s death, their death’s harsh effect on those who love them seems unlikely to be worse than the deprivation of decades of future happy life for the person who died.
To see this, let’s run with the premise that most of the badness of a person’s death comes from the suffering of their close friends/family members. For simplicity, let’s say whenever a person dies, 10 of their close friends/family suffer considerably. Let’s further say that every person lives through exactly 10 of their close friends/family dying.
Now imagine a baby who’s “choosing” whether or not they’d like to be born. If they choose not to be born, then they lose out on an entire happy life’s experience. BUT that experience would include living through 10 of their close friends/family members’ deaths, which our premise stated is worse than losing an entire life’s experience! So the baby should prefer to not be born.
So if most of the badness death of an existing person comes from the suffering the death causes on the person’s close friends/family, then ignoring other considerations, we should advocate for human extinction. Being born and having to suffer through 10 close friends/family dying is worse than not being born at all.
To me, this conclusion seems absurd. Even though I’ll suffer over life from the deaths of the people I love, on the whole, I’m happy to be alive.
In conclusion, while one’s death is worse because of the suffering of their close friends/family, it still seems that most of the badness of death comes from the person losing future happy life. So we’re still left with the conclusion that preventing a person’s existence is close to as bad as killing them.
I agree that ‘most of the badness of death comes from the person losing future happy life’.
However, there are also other factors that are relevant to whether ‘preventing a person’s existence is close to as bad as killing them’ (this obviously also depends what is meant by ‘close to’).
The claim seems to imply that we are doing something almost as bad as murder if we are failing to have as many children as possible. But a society where legislation reflected this position would reduce the quality of life of people who don’t want many (or any) children, would force women into being baby-making machines, and would plausibly result in a situation where individual children aren’t viewed as particularly important (if one dies, well, no time to grieve—another one is on the way). This strikes me as more absurd and dystopian than advocating for human extinction.
For me, this conversation is analogous to that surrounding Peter Singer’s book The Life You Can Save (TLYCS).
In TLYCS, Peter Singer argues, in my opinion quite convincingly, that we have a moral obligation to give up everything we have to help those in extreme poverty. Singer argues that every 5000 USD we spend on ourselves and not donate is equivalent to condemning a person whose life we could have saved. He then follows up with a far more modest ask: That we donate 1% of our income to effective charities.
There are many people who balk at Singer’s conclusion that we have a moral duty to donate everything above our bare survival needs to effective charities, and then reject his comparatively modest 1% ask. They might reply:
This isn’t what EAs actually advocate for. Singer’s conclusion is far too much to ask of most people, and even the most ardent EAs would balk at legislating it. However, many EAs, myself included, would agree that Singer’s philosophical conclusion really is correct.
Similarly, I’ve made the philosophical argument that there’s little moral difference between preventing a person’s existence and killing them. Given that conclusion, there are many compelling criticisms of what personal or legislative changes should follow. However, I haven’t found any convincing rebuttal to the philosophical argument.
There are many considerations which lessen the magnitude of the conclusion. Preventing the suffering of the close friends and family of a person who dies matters. One might have a high credence in a person-affecting view, endorse the procreation asymmetry, or place substantial credence on non-consequentialist theories.
But in my opinion, if you’re a consequentialist who holds even mild credence (say ~10%) in the non-person-affecting view, then preventing a person’s existence is on the order of badness of (say, ~10% as bad as) killing them. If you disagree, then I’d love to understand your perspective further, and see if there’s some crucial consideration I may be missing.
As with Singer’s arguments in TLYCS, I don’t think the truth/falsity of a philosophical argument is contingent on how radical its conclusions are. I also don’t think the existence of radical conclusions precludes the implementation of common-sense conclusions, like donating 1% of one’s income to effective charities.