Johann Frick’s ‘Conditional Reasons and the Procreation Asymmetry’ is widely regarded as the best existing defense of the asymmetry. It’s a thought-provoking paper, and I enjoy teaching it in advanced ethics classes. But in this post, I’ll diagnose three conceptual errors that I take to undermine the paper’s central arguments.
1. Conflating Individual-directed and Existence-conditional reasons
Frick contrasts “state-regarding reasons” with “bearer-regarding” ones. But he seems to equivocate between two very different distinctions.
(1) He first introduces “state-regarding reasons” as teleological, implying reasons to bring valuable entities into existence:
for Moore and other teleologists, our moral reasons are state-regarding reasons, since they are reasons to cause what is valuable to exist and what is disvaluable not to exist.
(2) He later treats “state-regarding” reasons as ones that treat the value of a state as normatively prior to that of the valuable entities it contains (such that the latter are treated as fungible constitutive means to realizing state-value):
a teleological view which regards our welfare-related reasons as purely state-regarding… views [people] as fungible receptacles for well-being, not as mattering qua individuals.
The problem is that these are two different distinctions. As I explain in ‘Value Receptacles’, fungibility is a consequence of a token-monistic axiology on which the only ultimately valuable “thing” is an abstract aggregate. But consequentialists can easily avoid such fungibility simply by accepting a token-pluralistic axiology, assigning fundamental value to each concrete person’s well-being, separately. (Past philosophers didn’t clearly mark this distinction, but I argued that it’s more charitable to read past teleologists as implicit token-pluralists, since the token-monistic view has literally nothing going for it in comparison. Frick even cites Moore as asking what things “ought to exist for their own sakes?” which is surely suggestive of ascribing ultimate value to the concrete particulars in question.)
So, if “bearer-regarding reasons” are just reasons that fundamentally stem from valuable concrete individuals (rather than from the abstraction of aggregate value), then even utilitarianism (charitably understood) posits bearer-regarding reasons rather than state-regarding ones. Alternatively, if “bearer-regarding” reasons are essentially ineligible to be promoted via creation, then objecting to fungibility does not yet provide any reason whatsoever to prefer bearer-regarding reasons over (suitably non-fungible) state-regarding ones.
In short, Frick’s central objection to teleology (and apparent motivation for taking reasons of beneficence to exclude existential benefits)[1] rests upon a conceptual error and a straw man.
2. Considering reasons against but not for a world
Frick later considers this “wide person-affecting view” on which we have reason to bring happy people into existence because this would benefit them, but hastily dismisses it:
Reasons that derive from the existence of a particular person cannot, at the same time, be reasons to make it the case that this person exists… If our moral reasons to be concerned with S’s happiness derive, not from the contribution it makes to a valuable state of affairs, but rather from S herself, then, in a world where S herself is absent, there is no moral reason to lament the absence of S’s potential happiness.
Here Frick seems to assume that reasons of lamentation are the only kinds of reasons that there are. For if there can also be reasons for (and not just against) states of affairs, his reasoning would immediately fall apart. Consider the valence-inverted parody:
“Reasons that derive from the existence of a particular person cannot, at the same time, be reasons to make it the case that this person does not exist… If our moral reasons to be concerned with S’s misery derive, not from the contribution it makes to a (dis)valuable state of affairs, but rather from S herself, then, in a world where S herself is absent, there is no moral reason to approve the absence of S’s potential misery.”
It would be foolish to deny the moral significance of existential harms on these grounds. In footnote 38, Frick writes:
[I]t is consistent with the view I am defending that I have moral reason to avoid a state of affairs in which S exists and is miserable even if S does not currently exist (and will never exist, if I act rightly). We must distinguish “reasons against a world (or state of affairs)” from the reasons we have “in a world”. According to my view
(1) I can have S-regarding reasons against bringing about a world w only if S exists in w.
Specifically, it is no S-regarding objection against bringing about a world w in which S never exists that in a different possible world, S would have been happy. But (1) is compatible with
(2) I can have S-regarding reasons in a world w1 against bringing about a world w2, even if S does not exist in w1 but only in w2.
That is, even if S does not presently exist, the fact that S would be miserable if I created her gives me S-regarding reason not to create [her].
Certainly, there can be reasons to oppose a world in virtue of S’s suffering in that world, even if S does not actually exist. But equally, there can be reasons to favor a world in virtue of S’s flourishing in that world, even if S does not actually exist.
There is no argument for asymmetry here. Frick is building in an asymmetry by only considering individual welfare-based reasons to oppose worlds and not even consideringthe possibility of individual welfare-based reasons to favor worlds. Even though the very logic he appeals to in order to make sense of our reasons to avoid existential harms (i.e., that we can have S-dependent reasons in a world even if S does not exist there, so long as the reasons speak to, and stem from, a world where S does exist) clearly undermines his basis for claiming that existential benefits can’t ground moral reasons of beneficence. Thus:
(3) We can have S-regarding reasons in a world w1 to bring about a world w3 where S exists happily, even if S does not exist in w1 but only in w3.
There is absolutely no principled reason to deny this, once you grant that we can have S-regarding reasons in w1 to avoid a world w2 where S exists miserably. Again, Frick offers no argument to the contrary. Instead, he assumes (without argument, or even acknowledgment) that person-regarding reasons can only stem from laments. But this is to beg the question on a massive scale.
2.1 Existence-conditional or Negative reasons?
Something this discussion brings out is that existence-conditionality may not, after all, be what’s most distinctive about Frick’s account of our reasons of beneficence. (S-dependent reasons don’t depend on S’s actual existence, and being conditional on S’s existing in the world where S has a morally significant welfare level obviously doesn’t restrict matters at all—that condition is trivially satisfied.) Rather, the distinctive feature of the account is that it is purely negative: we have reasons to avoid a certain outcome in which S exists (in a bad state, or a worse state than was otherwise securable), but no non-instrumental reasons to positively want any state in which S exists (over non-existence).
One way to see this is to consider that if one’s goal is a material conditional (if p then q), that’s logically equivalent to taking as one’s goal the negated conjunction: NOT-(p and not-q).
And this seems to mesh with our understanding of promissory reasons. As Frick notes, we’ve no reason to make extra promises simply in order to keep them. He suggests that the reason to keep a promise is conditional on having made it in the first place, but it may be clearer to just say that (i) breaking promises is bad, whereas (ii) keeping promises is neutral. Our promissory reasons are just to avoid broken promises. If you’ve made a promise, the best you can hope for is to maintain moral neutrality. So that’s why there’s no reason to make promises merely in order to keep them. It’s because promises have no value in themselves, no intrinsic moral upside. (They may be useful instruments for social coordination, but considered in themselves their only valenced [non-neutral] potential is negative.)
Frick’s analogy to promises thus helps to bring out what an appallingly nihilistic view this is of the value of humanity. For the parallel claim is that people have no value in themselves. (We may be useful instruments, but considered in ourselves our only valenced potential is negative: we’re all moral downside, no intrinsic upside.) So to view people as morally akin to promises in this way is, I believe, deeply offensive and disrespectful to the value of humanity.
[Image caption: Shall I compare thee to a barren rock?]
As I’ve argued elsewhere, proper respect for the value of persons instead requires positively desiring/appreciating that they have good lives, not just that they avoid bad lives (which is equally well achieved by non-existence as by happy existence). Our happy existence is not a matter of moral indifference. And so our reasons of beneficence are not conditional—or purely negative—in the way that Frick’s account implies.
3. Better satisfied life-standards ≠ Better lives
The main positive contribution of Frick’s paper is to show how one might reconcile the Procreation Asymmetry with the non-identity intuition that we should prefer to create a better life (C) over a distinct merely good one (B). Frick appeals to the
Selection Requirement: In a choice between creating two possible persons, I have contrastive moral reason to create that person for whom I can better satisfy the moral standard that will obtain if I create that person.
Frick assumes that we better satisfy the C-standard when C has a great life, than we do the B-standard when B has a merely good life. If so, the Selection Requirement can explain why we have contrastive reason to create C rather than B, without implying any unconditional reason to create either.
It’s a neat move. But I’m a bit suspicious about whether it will work as Frick desires, for the two standards might diverge in ways that undermine the hope of tracking greater comparative existential benefits. For example, suppose that B essentially has a lower “ceiling” on their possible well-being. (You might suppose that B has a severe congenital condition, and further accept the genetic essentialist view that anyone born with different genes would be a different person.) It might then turn out that B’s good life is in fact the best possible life that B could have. You would then maximally satisfy the B-standards by creating B, while C’s much better life less-than-maximally satisfies the C-standards. In that case, the standards-relative view would seem (on a natural interpretation) to incorrectly imply greater reason to create B than C.
(If we extend the principle beyond just persons, it becomes even more troubling: should we prefer to create worms at 100% of their meagre capacity for well-being over 99%-happy people?)
To save the view, we need all life-standards to use the same (unbounded) scale. But that seems an awkward fit with the conditional, laments-based view. Is there reason to regard a happy chicken’s life as lamentable, just because it lacks the richness of a human life? There’s certainly no reason to lament this for the chicken’s sake. I think we have more reason to want a human than a chicken to exist, but I don’t see any essentially comparative reason here; rather, the reason to want this is derivative of our strong absolute reasons of beneficence to want the human to exist (and the comparative fact that our absolute reasons of beneficence to want the chicken to exist are much weaker). There’s no failure in the chicken’s existence. Just a lost opportunity in the absence of the human’s. But such judgments seem contrary to Frick’s whole approach.
Perhaps the best move is to limit his solution to cases in which the same scale naturally applies. This would be to give up on fully capturing the non-identity intuition. But it would still be an interesting result for the asymmetry to be rendered compatible with the non-identity intuition in at least some cases.
(Though I still have some residual suspicion that the Selection Requirement fits ill with the standards-relative approach: I don’t really see the internal motivation for preferring to “better satisfy” one standard rather than another, distinct standard. It would seem more natural to see the two as incommensurable, rather than pursuing a hegemonic meta-standard of maximally satisfying sub-standards regardless of their differing contents. Doesn’t the meta-standard seem to treat the distinct standards as too… fungible?)
Conclusion
Despite my criticisms, I should reiterate that there’s a lot to like about Frick’s paper: it’s thought-provoking, well-written, and makes a number of philosophically sophisticated moves that make it great to discuss in a seminar. It may indeed be the best existing defense of the asymmetry! But it’s not a defense that should convince anyone.
A second reason he mentions is to avoid the repugnant conclusion. This is a common mistake. In ‘Puzzles for Everyone’, I explain why denying value to good lives doesn’t actually resolve the deeper issues underlying quantity-quality tradeoffs, and so cannot provide any such reason.
Against Conditional Beneficence
Link post
Johann Frick’s ‘Conditional Reasons and the Procreation Asymmetry’ is widely regarded as the best existing defense of the asymmetry. It’s a thought-provoking paper, and I enjoy teaching it in advanced ethics classes. But in this post, I’ll diagnose three conceptual errors that I take to undermine the paper’s central arguments.
1. Conflating Individual-directed and Existence-conditional reasons
Frick contrasts “state-regarding reasons” with “bearer-regarding” ones. But he seems to equivocate between two very different distinctions.
(1) He first introduces “state-regarding reasons” as teleological, implying reasons to bring valuable entities into existence:
(2) He later treats “state-regarding” reasons as ones that treat the value of a state as normatively prior to that of the valuable entities it contains (such that the latter are treated as fungible constitutive means to realizing state-value):
The problem is that these are two different distinctions. As I explain in ‘Value Receptacles’, fungibility is a consequence of a token-monistic axiology on which the only ultimately valuable “thing” is an abstract aggregate. But consequentialists can easily avoid such fungibility simply by accepting a token-pluralistic axiology, assigning fundamental value to each concrete person’s well-being, separately. (Past philosophers didn’t clearly mark this distinction, but I argued that it’s more charitable to read past teleologists as implicit token-pluralists, since the token-monistic view has literally nothing going for it in comparison. Frick even cites Moore as asking what things “ought to exist for their own sakes?” which is surely suggestive of ascribing ultimate value to the concrete particulars in question.)
So, if “bearer-regarding reasons” are just reasons that fundamentally stem from valuable concrete individuals (rather than from the abstraction of aggregate value), then even utilitarianism (charitably understood) posits bearer-regarding reasons rather than state-regarding ones. Alternatively, if “bearer-regarding” reasons are essentially ineligible to be promoted via creation, then objecting to fungibility does not yet provide any reason whatsoever to prefer bearer-regarding reasons over (suitably non-fungible) state-regarding ones.
In short, Frick’s central objection to teleology (and apparent motivation for taking reasons of beneficence to exclude existential benefits)[1] rests upon a conceptual error and a straw man.
2. Considering reasons against but not for a world
Frick later considers this “wide person-affecting view” on which we have reason to bring happy people into existence because this would benefit them, but hastily dismisses it:
Here Frick seems to assume that reasons of lamentation are the only kinds of reasons that there are. For if there can also be reasons for (and not just against) states of affairs, his reasoning would immediately fall apart. Consider the valence-inverted parody:
“Reasons that derive from the existence of a particular person cannot, at the same time, be reasons to make it the case that this person does not exist… If our moral reasons to be concerned with S’s misery derive, not from the contribution it makes to a (dis)valuable state of affairs, but rather from S herself, then, in a world where S herself is absent, there is no moral reason to approve the absence of S’s potential misery.”
It would be foolish to deny the moral significance of existential harms on these grounds. In footnote 38, Frick writes:
Certainly, there can be reasons to oppose a world in virtue of S’s suffering in that world, even if S does not actually exist. But equally, there can be reasons to favor a world in virtue of S’s flourishing in that world, even if S does not actually exist.
There is no argument for asymmetry here. Frick is building in an asymmetry by only considering individual welfare-based reasons to oppose worlds and not even considering the possibility of individual welfare-based reasons to favor worlds. Even though the very logic he appeals to in order to make sense of our reasons to avoid existential harms (i.e., that we can have S-dependent reasons in a world even if S does not exist there, so long as the reasons speak to, and stem from, a world where S does exist) clearly undermines his basis for claiming that existential benefits can’t ground moral reasons of beneficence. Thus:
(3) We can have S-regarding reasons in a world w1 to bring about a world w3 where S exists happily, even if S does not exist in w1 but only in w3.
There is absolutely no principled reason to deny this, once you grant that we can have S-regarding reasons in w1 to avoid a world w2 where S exists miserably. Again, Frick offers no argument to the contrary. Instead, he assumes (without argument, or even acknowledgment) that person-regarding reasons can only stem from laments. But this is to beg the question on a massive scale.
2.1 Existence-conditional or Negative reasons?
Something this discussion brings out is that existence-conditionality may not, after all, be what’s most distinctive about Frick’s account of our reasons of beneficence. (S-dependent reasons don’t depend on S’s actual existence, and being conditional on S’s existing in the world where S has a morally significant welfare level obviously doesn’t restrict matters at all—that condition is trivially satisfied.) Rather, the distinctive feature of the account is that it is purely negative: we have reasons to avoid a certain outcome in which S exists (in a bad state, or a worse state than was otherwise securable), but no non-instrumental reasons to positively want any state in which S exists (over non-existence).
One way to see this is to consider that if one’s goal is a material conditional (if p then q), that’s logically equivalent to taking as one’s goal the negated conjunction: NOT-(p and not-q).
And this seems to mesh with our understanding of promissory reasons. As Frick notes, we’ve no reason to make extra promises simply in order to keep them. He suggests that the reason to keep a promise is conditional on having made it in the first place, but it may be clearer to just say that (i) breaking promises is bad, whereas (ii) keeping promises is neutral. Our promissory reasons are just to avoid broken promises. If you’ve made a promise, the best you can hope for is to maintain moral neutrality. So that’s why there’s no reason to make promises merely in order to keep them. It’s because promises have no value in themselves, no intrinsic moral upside. (They may be useful instruments for social coordination, but considered in themselves their only valenced [non-neutral] potential is negative.)
Frick’s analogy to promises thus helps to bring out what an appallingly nihilistic view this is of the value of humanity. For the parallel claim is that people have no value in themselves. (We may be useful instruments, but considered in ourselves our only valenced potential is negative: we’re all moral downside, no intrinsic upside.) So to view people as morally akin to promises in this way is, I believe, deeply offensive and disrespectful to the value of humanity.
[Image caption: Shall I compare thee to a barren rock?]
As I’ve argued elsewhere, proper respect for the value of persons instead requires positively desiring/appreciating that they have good lives, not just that they avoid bad lives (which is equally well achieved by non-existence as by happy existence). Our happy existence is not a matter of moral indifference. And so our reasons of beneficence are not conditional—or purely negative—in the way that Frick’s account implies.
3. Better satisfied life-standards ≠ Better lives
The main positive contribution of Frick’s paper is to show how one might reconcile the Procreation Asymmetry with the non-identity intuition that we should prefer to create a better life (C) over a distinct merely good one (B). Frick appeals to the
Frick assumes that we better satisfy the C-standard when C has a great life, than we do the B-standard when B has a merely good life. If so, the Selection Requirement can explain why we have contrastive reason to create C rather than B, without implying any unconditional reason to create either.
It’s a neat move. But I’m a bit suspicious about whether it will work as Frick desires, for the two standards might diverge in ways that undermine the hope of tracking greater comparative existential benefits. For example, suppose that B essentially has a lower “ceiling” on their possible well-being. (You might suppose that B has a severe congenital condition, and further accept the genetic essentialist view that anyone born with different genes would be a different person.) It might then turn out that B’s good life is in fact the best possible life that B could have. You would then maximally satisfy the B-standards by creating B, while C’s much better life less-than-maximally satisfies the C-standards. In that case, the standards-relative view would seem (on a natural interpretation) to incorrectly imply greater reason to create B than C.
(If we extend the principle beyond just persons, it becomes even more troubling: should we prefer to create worms at 100% of their meagre capacity for well-being over 99%-happy people?)
To save the view, we need all life-standards to use the same (unbounded) scale. But that seems an awkward fit with the conditional, laments-based view. Is there reason to regard a happy chicken’s life as lamentable, just because it lacks the richness of a human life? There’s certainly no reason to lament this for the chicken’s sake. I think we have more reason to want a human than a chicken to exist, but I don’t see any essentially comparative reason here; rather, the reason to want this is derivative of our strong absolute reasons of beneficence to want the human to exist (and the comparative fact that our absolute reasons of beneficence to want the chicken to exist are much weaker). There’s no failure in the chicken’s existence. Just a lost opportunity in the absence of the human’s. But such judgments seem contrary to Frick’s whole approach.
Perhaps the best move is to limit his solution to cases in which the same scale naturally applies. This would be to give up on fully capturing the non-identity intuition. But it would still be an interesting result for the asymmetry to be rendered compatible with the non-identity intuition in at least some cases.
(Though I still have some residual suspicion that the Selection Requirement fits ill with the standards-relative approach: I don’t really see the internal motivation for preferring to “better satisfy” one standard rather than another, distinct standard. It would seem more natural to see the two as incommensurable, rather than pursuing a hegemonic meta-standard of maximally satisfying sub-standards regardless of their differing contents. Doesn’t the meta-standard seem to treat the distinct standards as too… fungible?)
Conclusion
Despite my criticisms, I should reiterate that there’s a lot to like about Frick’s paper: it’s thought-provoking, well-written, and makes a number of philosophically sophisticated moves that make it great to discuss in a seminar. It may indeed be the best existing defense of the asymmetry! But it’s not a defense that should convince anyone.
A second reason he mentions is to avoid the repugnant conclusion. This is a common mistake. In ‘Puzzles for Everyone’, I explain why denying value to good lives doesn’t actually resolve the deeper issues underlying quantity-quality tradeoffs, and so cannot provide any such reason.