Mar 21, 2023·edited Mar 21, 2023Liked by Richard Y Chappell
Actualism and Frick's views imply the procreation asymmetry, but I wouldn't consider them absolute views or standard Existence Anticomparativism. Rather, I'd still consider them merely comparative views (although we could argue about definitions). They allow certain kinds of comparisons, but for principled reasons and without directly assuming asymmetry between apparent goods and apparent bads or asserting the existence of absolute harms while denying absolute benefits:
1. Actualism allows actual beings to compare their lives to nonexistence (or us to do so on their behalf), but not non-actual beings to compare their lives with anything (or us to do so on their behalf). That seems intuitive to me, because the actual and only the actual can have objects that exist in the universe we can point to in order to ground their interests (e.g. brains with interest-generating and identity-establishing structures*).
2. Frick's conditional reasons are of the form "I have reason to (if I do p, do q)" (or Bykvist and Campbell, 2021's contrastive version "I have reason to (if I do p, do q) rather than not (if I do p, do q)").
If all person-regarding reasons are actualist or conditional (as I think is assumed in actualism and Frick's views, respectively), then we get the procreation asymmetry. You also concede (or at least don't strongly object to) actualist reasons and so at least weak asymmetry in Rethinking the Asymmetry, as "partiality towards the actual". The existence of absolute reasons seems to require more than the existence of actualist reasons or Frick's conditional reasons, at least with respect to apparently good and bad lives, because actualist and conditional reasons (at least with respect to individual welfare) each have absolute counterparts that make stronger claims, and there are more absolute reasons. So, it seems that there's an extra burden of proof to establish absolute reasons.
I think we can also use Frick's conditional reasons (and only conditional reasons), along with transitivity and the independence of irrelevant alternatives, to establish absolute harms while denying absolute goods, basically antifrustrationism (e.g. Fehige, 1998).
*This is a bit subtle, because we want to consider the whole of the future, including future actual beings, not just presently actual beings. This depends on specific views of the nature of time, although I wouldn't be surprised if the differences turned out to be merely normative, anyway. It also seems like you're sympathetic to something similar, with "timeless difference" and 4-dimensionalism, as you discuss in https://www.philosophyetc.net/2009/08/must-harms-be-temporally-located.html. Another possibility is Bader's account.
(1) I take it that 'actualist reasons' = what I call 'person-directed' reasons in the OP. The quick argument for absolute (or undirected) reasons is that we need them to explain our reasons *not* to create Misery (a merely possibly person, who *would* be miserable were she to exist).
One could instead appeal to some kind of structural reasons of ratifiability: if we created Misery, we'd then have (actualist) reasons to regret it. But this just pushes the asymmetry back a step, for what is the justification for adding structural reasons of regret-avoidance but not structural reasons of gladness-enticement (we would have actualist reasons to be glad of Joy's existence, were we to create her)?
(2) I worry that Frick's conditional approach presupposes rather than justifies the asymmetry. After all, it seems a general feature of "conditional obligations" (like promises) that violating them is bad, whereas satisfying them is merely neutral rather than good. [Cf. https://www.philosophyetc.net/2021/10/puzzling-conditional-obligations.html ] But that's an abhorrent view to take of people: that our existence is possibly bad, and neutral at best!
This brings me to my central argument for absolute value/reasons, which is just that its denial seems to express an abhorrent/disrespectful attitude towards human life, as I explain in the OP. We should view each other more positively than that.
(You suggest that "there's an extra burden of proof to establish absolute reasons." I reject that picture of moral methodology. Occam's razor is no reason at all to ascribe less value to people, or to smaller numbers of people. Moral parsimony is not a virtue. We should simply endorse the moral views that are best-warranted on their first-order merits.)
"(2) I worry that Frick's conditional approach presupposes rather than justifies the asymmetry. After all, it seems a general feature of "conditional obligations" (like promises) that violating them is bad, whereas satisfying them is merely neutral rather than good. [Cf. https://www.philosophyetc.net/2021/10/puzzling-conditional-obligations.html ]"
I'm not sure either way if I would say it presupposes ethical asymmetry, but that falls out basically immediately, so I guess it's at least "close" to presupposing it. Like with actualism in my other reply, if you tried to reverse Frick's conditional reasons, what you get wouldn't capture any kind of view of existence being prior to welfare, so it's hard to interpret it in person-affecting or non-absolute terms: "I have reason to do p and do q". p and q can be switched in that statement without affecting its truth value, but you can't do that for Frick's original "I have reason to (if I do p, do q)".
So, among person-affecting views and other views that reject absolute value, the asymmetry seems significantly more plausible than the reverse asymmetry, because reverse asymmetric views are harder to interpret as person-affecting or as rejecting absolute value than standard asymmetric views. Of course, this doesn't mean much if you're pretty sure of absolute value, as you seem to be.
I think the objection of regret-avoidance vs gladness-enticement may be missing how actualism is typically formulated. Here's "weak actualism" (not to be confused with only making weak asymmetry claims; the difference between weak and strong is who counts at all; strong actualism is much less plausible to me):
“They could say that what matters is whether the outcome of the action would have been better for those people who would have existed had it been performed. Thus:
Weak Actualism.—The moral status of any aj, actual or not, is determined by whether its outcome is better or worse for people in Saj than the outcomes of the other available actions", where aj is an action, and Saj is "the set of people who would exist (in the past, present, and future) if action aj were to be taken." (Hare, 2007)
"Self-Conditional Maximization. The permissible options are all and only the options that maximize self-conditional value." (Spencer, 2021)
So, under weak actualism, betterness/worseness is basically outcome-relative, because it's relative to who is actual. We're interested in whether the people in some outcome A are worse off than those same people (or nonexistence) in another outcome B, i.e. whether A is worse than B from the point of view of those in A. The unconditionally permissible options are exactly those that aren't worse than any other in this way, i.e. those that "maximize self-conditional value".
If B is just A with an additional miserable person and no other moral differences, and if B is actual, then B is worse than A for the people in B, because it's worse for the extra person, so it's not permitted under weak actualism. Still, A and B are equivalent for the people in A (or from the point of view of A).
If we try to reverse the actualist principles and get something like gladness-enticement and pronatalism, it's not totally clear how you should technically do this, and I think the result will seem spooky and lose its actualist/comparativist appeal.
On spookiness and actualist/comparativist appeal, suppose the outcome A is actual. If B is just A with an additional happy person and no other moral differences, then for A to be worse than B "from the point of view of A", which is actual, you need to refer to the interests of the non-actual extra person. That seems spooky and not at all in the spirit of actualism. If B is actual, then it would be better than A for the actual people (the people of B). However, that's already compatible with standard actualism, and doesn't rule out A unconditionally, because B isn't better for the people in A, and if you end up in A, there's no actual person for whom B is preferable.
On the technical difficulty of coming up with "actualist" principles to give gladness-enticement and pronatalism, I think you have to do one of the following:
1. An additional happy person in the actual outcome has to make that outcome unconditionally obligatory (rather than unconditional (im)permissibility as in standard actualism), which will mean multiple options are sometimes unconditionally obligatory, contrary to my understanding of unconditional obligation. For example, consider A and B, which differ just in the identity of one person who is very happy. We'd claim both A and B are unconditionally obligatory, but that can't be. That being said, I think there's a similar problem for standard actualism and miserable lives: if A and B differ only in the identity of a miserable person, then neither maximizes self-conditional value, and neither is permissible. One way to fix both views would be with something like saturating counterpart relations, or additional principles to break ties when there are multiple obligatory options or no permissible options. If I recall correctly, this is a common issue brought up and discussed around actualism, and Spencer (2021) in particular discusses Misery vs Moremisery and Misery vs Equalmisery. Also, multiple unconditionally obligatory options seems worse than no unconditionally permissible options.
2. From the actual world without an additional happy person, you need to decide which non-actual people to count and how to weigh them and their interests in non-actual worlds from the point of view of the actual world. There are usually multiple non-actual worlds to consider, so either you're left starting from pairwise comparisons and no counterpart to "maximize self-conditional value", or you sometimes need to weigh interests across 3 worlds simultaneously.
So, (something like) regret-avoidance falls out naturally from actualist intuitions, while (something like) gladness-enticement doesn't and faces more technical problems.
Hare, C. (2007). Voices from another world: Must we respect the interests of people who do not, and will never, exist?. Ethics, 117(3), 498-523.
Why the switch to "obligation"-talk? I take the pronatalist claim to just be that you have some *reason* to create each happy person (presumably in proportion to their welfare value). There's nothing paradoxical about that.
Working through the putative asymmetry in outcome-relative weak actualism, let's take world B as a base and compare both world A, which adds Joy, and world C, which adds Misery.
From B's perspective, the worlds are all value-equivalent. A says A is better. C says C is worse. Suppose B is actual. Asymmetrists want to say (i) we had good reason not to realize C, but (ii) no extra reason to realize A. How do they secure this? Appeal to self-conditional value gives us reasons against C, in that if we'd chosen C, it would've then been worse. But it's equally true that, if we'd chosen A, it would've then been better, suggesting that we should have self-conditional reasons for A. (B doesn't regard itself as any better or worse than either A or C, and so doesn't generate any competing self-conditional reasons when making these comparisons.) Where's the asymmetry between A and C?
I think an asymmetry may have been smuggled in to the account of permissibility as "exactly those [options] that aren't [self-conditionally] worse than any other". If we instead start from the more foundational normative question of what *reasons* there are in the situation (and insofar as we go for deontic concepts like 'permissibility' at all, take it to be determined in suitable fashion by the balance of those normative reasons), there doesn't appear to be any grounds for an asymmetry here after all.
(Though again, I do think the deeper issue here is that we should reject "the spirit of actualism". Even if B is actual, we should regard the extra suffering life in alternative C as a bad-making feature, resulting in world C being properly evaluated as *worse*. There's nothing "spooky" about this; it's entirely commonsensical that we should evaluate miserable lives as intrinsically bad.)
Mar 22, 2023·edited Mar 22, 2023Liked by Richard Y Chappell
"Where's the asymmetry between A and C?
I think an asymmetry may have been smuggled in to the account of permissibility as "exactly those [options] that aren't [self-conditionally] worse than any other". If we instead start from the more foundational normative question of what *reasons* there are in the situation (and insofar as we go for deontic concepts like 'permissibility' at all, take it to be determined in suitable fashion by the balance of those normative reasons), there doesn't appear to be any grounds for an asymmetry here after all."
How else would you weigh reasons so that you get symmetry (reasons to create good lives, and equally strong reasons against creating bad lives) in a way that's person-affecting and without requiring absolute value?
I think the asymmetry and maximizing self-conditional value can be justified on the basis of a stability/dominance/money pump argument. If C would happen, on Misery's account, A and B would have been better, and neither A nor B directs you back to C following self-conditional value. If you could, you'd pay to switch to A or B. Or, you would pay to shorten Misery's life to nothing and have the result look like B, but with an extra cost. However, if you would pay to go from C to A or B (or worlds close enough to them), choosing C first means getting money pumped in principle, and you should have just chosen A or B in the first place without the extra cost.
On the other hand, from A or B, self-conditional value doesn't direct you away, and there's no issue of dominance or money pumps on the basis of self-conditional value if you were to choose A or B. The value of A over B in A is other-conditional value from the perspective of B. Joy doesn't exist in B, so Joy has no account in B to consider or compel you from B towards A (or C). Self-conditional reasons in A only compel you to stick to A if you're already there.
Maybe we have other reasons to switch from B to A, but those don't come from B, so they don't seem like they could be actualist, person-affecting or merely comparative reasons.
There may also be stability/dominance/money pump arguments against actualism generally, so maybe that undermines this kind of argument, but I don't think it would completely defeat it.
In general, options that don't maximize self-conditional value seem unstable, in that they'll direct you away from them if you follow self-conditional value and they'll make you vulnerable to dominance or money pumps. On the other hand, any option that does maximize self-conditional value is stable, at least with respect to self-conditional value.
>> "Note that there is (as far as I’m aware) no philosophical basis for then denying the same reasoning as applied to non-comparative goods. A happy existence is clearly non-comparatively good for the person who enjoys it."
I don't know about _denying_ the reasoning for non-comparative goods, but I think there really is an asymmetry between harms and goods, which is that the badness of suffering is surely the most undeniable moral fact we know; there is, to my intuition, no comparably obvious moral fact about goodness. Even though I _have_ a happy existence, I am much less certain when I introspect on the matter that this constitutes some non-comparative good--I honestly cannot tell you that I feel benefitted by my reasonably happy existence. But I am as certain as I am about anything at all, that if I were to be tortured for 100 years, that would be bad, and that I would be harmed by such an existence. I just genuinely, truly, feel orders of magnitude more confident drawing conclusions based on the badness of suffering than on the goodness of well-being.
The above is compounded by the fact that there is _much_ more disagreement as to what even constitutes well-being than what constitutes suffering--it just seems to me undeniable that the role that suffering and well-being should play in our theorizing has to be asymmetrical, even if only for reasons of uncertainty over definition.
>> "denying that our kids are better than nothing, or that utopia is better than a barren rock"
>>"It’s deeply disrespectful to deny that another person (who is, themselves, happy to exist) is better than nothing."
FWIW, I don't feel disrespected by the denial that I am better than nothing, despite being currently happy with my existence. I think some of the claims you make along these lines in the article undermine the argument: they feel like emotive appeals intended to short-circuit arguments. Further, it seems to me that most of the disagreement comes down to intuitions about non-existence and how it should be valued relative to happy existence, so these emotive appeals simply end up reading as rather strident restatements of your intuitions on the matter (to me, at least, who doesn't share these intuitions). Finally, "disrespect" is an odd term of analysis to bring in here; I'm sure many deontologists object that the utilitarian denial of intrinsic rights is "disrespectful" to people, but I doubt that would bother you (nor should it, in my opinion).
>> "But the wrong view here could easily motivate anti-natalist “voluntary human extinction”, which would be literally the worst thing ever. It’s really worth not making mistakes that astronomically great."
I think the second sentence supplies a resolution to the first that takes a lot of the sting out of this argument as a criticism of the asymmetry: I presume much of what makes human extinction the worst thing ever is that it is _irrevocable_ in a way that almost no other decision can be--so under conditions of moral uncertainty, even theories that say that voluntary extinction is okay should be extremely unwilling to endorse it in practice.
Some very foundational disagreements about moral methodology and reasoning here!
> "it just seems to me undeniable that the role that suffering and well-being should play in our theorizing has to be asymmetrical, even if only for reasons of uncertainty over definition"
Uncertainty over definition doesn't entail uncertainty about whether the thing itself exists. And I'm here arguing against the "strong asymmetry" which claims that creating non-comparative goods counts for *nothing*. That's the view that lacks any basis. You could count goods for *less*, but that sort of "weak asymmetry" is a very different matter. That's compatible with my claim that we've every reason to recognize *some* positive reason to bring good lives into existence.
> "there is, to my intuition, no comparably obvious moral fact about goodness"
It's good to enjoy fun times spent with loved ones. That's at least as obvious as the badness of suffering. (Compare the common wisdom: "better to have loved and lost than never to have loved at all.")
> "emotive appeals intended to short-circuit arguments"
Um, no. They *constitute* (enthymematic) arguments. Maybe you reject the premise, as anyone can of any argument. But vivid counterexamples, and highlighting (or disputing) the repugnance of a view's implications, are the life-blood of moral philosophy, not a "short circuit". Getting these things right is absolutely central to possessing moral insight.
> Finally, "disrespect" is an odd term of analysis to bring in here; I'm sure many deontologists object that the utilitarian denial of intrinsic rights is "disrespectful" to people, but I doubt that would bother you
Again, I couldn't disagree more. It's really important to value people correctly! "Disrespect" is the standard term for not properly valuing people. And I take *very* seriously the charge that utilitarianism is "disrespectful" in this way -- I think it's a *much* more serious objection than superficial complaints about its extensional verdicts, because it gets much more to the heart of things. That's why I wrote 'Value Receptacles' - https://philpapers.org/rec/CHAVR - and also 'The Mere Means Objection': https://www.utilitarianism.net/objections-to-utilitarianism/mere-means/
One other worry for these views that say that creating happy lives isn't good is they inevitably imply in specific scenarios that it's not bad to gratuitously make future people's lives less good for trivial gains. https://benthams.substack.com/p/a-new-utterly-decisive-argument-against
I don't think they *inevitably* imply this. Some ("wide") views hold that that different-number cases are *incomparable* in value (or "on a par"), while applying standard utilitarian axiology to same-number cases. But I guess they would probably at least imply that it's not bad to create X mediocre new lives, for trivial gains, rather than 2X different awesome lives.
Suppose that one can create Alice with ten utility or John with 5 utility. It wouldn't be wrong to trade Alice with ten utility for John with 5 utility (by trade I mean go from a state where one will create Alice to John). But then they can trade John with 5 utility for Alice with 5 utility--and neither action is impermissible.
On "wide" views, it precisely *would* be wrong to create John rather than Alice. It's permissible to create no extra life at all. But (on these views) if you do create, you need to do the best you can with it.
Okay this more complicated series of trades. You trade John with 10 utility for creating no one for creating Alice with 5 utility for creating no one for creating John with 5 utility. None of the choices is itself impermissible.
I take it there are no other procreative options at each time. I.e. at t1 you can either keep John (@10) or switch to no-one. You switch. At t2 you can keep no-one or switch to Alice, etc.
Raises interesting issues about diachronic consistency. E.g. if you foresee where you'll end up, then the original choice is wrong (you aren't really switching to "no one", but to John@5, while John@10 is an available alternative). If you can't foresee the future options, it's maybe not so obviously impermissible to get tripped up in this way. I expect there are standard things that defenders of intransitivity / cyclic preferences say about these sorts of situations...?
But this runs afoul of the following obvious principle which says that the fact that some act gives you extra options doesn't count against it. So switching at t1, on this account, is wrong only because you'll get extra choices, but that's obviously crazy. You shouldn't think "oh no, this action gives me extra options! I won't take it then."
I think even your last point will depend on the particular view. I think most "wide" views, as I understand that term, would endorse the 2X awesome lives, because you get X times the difference between the average awesome life and the average mediocre life, which counts a lot, and the extra X awesome lives don't count against, assuming they're not worse than nothing (or if the awesome lives are bad, assuming they don't count much against, e.g. if the difference between an awesome life and a mediocre one is larger than the overall badness of an awesome life relative to nonexistence). I think Meacham's "Person-affecting views and saturating counterpart relations", Thomas, 2022's wide asymmetric views (both soft and hard), Frick's views and appropriately modified wide versions of actualism would endorse the 2X for these reasons, at least in a pairwise comparison. Even negative utilitarian views could endorse the 2X, assuming the 2X awesome lives have less bads in them than the X mediocre lives, but in principle awesome lives could have more bads than mediocre ones, even significantly more, as long as there are enough offsetting or outweighing goods.
I was assuming that the extra X lives would generate a lot of additional "incomparability" (as in Johan Gustafsson's discussions of "undistinguishedness"); but fair point that it may depend on the details of the view.
Gustafsson, J. (2020). Population axiology and the possibility of a fourth category of absolute value. Economics & Philosophy, 36: 81–110. https://doi.org/10.1017/S0266267119000087
>"there is (as far as I’m aware) no philosophical basis for then denying the same reasoning as applied to non-comparative goods."
Why is it your opponent who bears the burden of proof here? Even given the assumption that it is bad to create "miserable" individuals, applying the same reasoning to conclude that it is good to create "happy" individuals seems to require some version of the assumption that goodness and badness differ only in sign, rather than in dimension. This is a substantial assumption, and plausibly a false one (e.g. see Magnus Vinding's arguments here: centerforreducingsuffering.org/phenomenological-argument/). So do you think the assumption that value is basically one-dimensional, with good and bad differing only in sign, is well-supported? Or do you think your argument can work even without that assumption?
It's not clear to me what it means to ask whether goodness and badness are one dimension or two. But I do think there's a clear presumption of philosophical parity between the two, such that any putative asymmetries need to be explained rather than taken for granted. I'm happy to take that as self-evident.
Insofar as Vinding is claiming that there's nothing *desirable* (or good), opposing the way that suffering is *undesirable* (or bad), I just think that's insane. I read the link; I don't see anything "plausible" or worth engaging with there.
Really? That surprises me. Do you think the following is wrong?
"In my view, the strongest argument against the existence of a phenomenological counterpart to suffering is that introspection yields no sign of such a counterpart. When we introspectively examine the proposed candidates of positive experiences, such as those listed above, we do not find that they have any phenomenological properties that render them the dual opposites of suffering, or anti-suffering, as it were."
My own introspection yields pretty much the same result as Vinding. I take it that your own introspection yields something different (e.g. that a certain experience of pleasure could reasonably "offset" a given experience of pain to yield an overall zero-valenced experience).
What conclusion, if any, would you draw from the existence of people whose most basic intuitions about the value of their own experiences differ so much from yours?
I think it's wrong that that is any kind of argument. Positive and negative experiences aren't like matter and anti-matter, that literally cancel out (leaving "an overall zero-valenced experience") when combined. They compensate *normatively*, not *descriptively*.
To give an obvious example from common sense: “'Tis better to have loved and lost than never to have loved at all.” The suffering of grief is more than compensated for by the positive value of loving that person and getting to spend time with them while they were alive. This basic normative datum about compensation in no way depends upon combining the love and the grief to yield a weird combined experience with "zero valence".
I take your point about the compensation being normative rather than descriptive, but I don't think there is a "basic normative datum about [experiential] compensation," and I agree with Vinding that the impulse to posit such a datum comes from a mistaken account of experiential phenomenology combined with excessive systematizing zeal. There just isn't anything in our experiences (or at least in my experiences – perhaps there is in yours?) that could serve to ground a normative concept of "good experience" posessing philosophical parity with the concept of "bad experience" grounded by our experience of suffering.
Note, however, that I needed to add "experiential" to your proposed datum, because Vinding's criticism only targets assumptions about the normative value of subjective experiences. It has nothing to say about possible sources of value grounded in something other than the phenomenology of valenced experiences. It might undermine a "pro-existence" population ethics for a hedonic utilitarian, but it wouldn't necessarily undermine anyone else's.
(I don't know about you, but I find the notion that non-hedonic values can give us reasons for creating new lives much more intuitively plausible than the idea that hedonic values alone have the power to do so.)
I'm certainly inclined to think that the overwhelming bulk of what's valuable about our lives is non-hedonic. Raw pleasure isn't *that* big of a deal. But it's really just bonkers to deny that there's *any such thing* as positive experience.
I'm trying to imagine the response of ordinary people to your telling them, "On reflection, it doesn't seem to me that there's any such thing as happiness or good feelings. Don't you agree?" Pretty similar to if you suggested that grass was purple, I imagine.
This is a sleight of hand. Vinding never denied the existence of happiness, he denied the existence of "positive value" from mental states like happiness. I know you probably also think that's bonkers, but it's quite obviously not in the ballpark of obvious absurdity as denying the existence of happiness as a mental state.
If you're going around telling people that sort of thing, it helps if you wear special robes, or a silly hat, or something like that. They're much more likely to believe you that way.
(1) On the philosophical issue: Michael covers the matter well. But to justify the asymmetry in a more straightforward way - arguments for creating net-negative lives is self-defeating, while creating net-positive lives is question-begging.
If you ground normative value on conscious beings and their preferences & interests (i.e. my life and welfare matters because there is *this* perspective from which these things matter and are valuable), then if such beings/perspectives *don't exist and never will* unless we create them, there is no question-begging reason to create them (i.e. the argument goes - we should create such people because their lives matter; but their lives matter only because they want to live; but they would only want to live if we create such people and they exist in the first place).
In contrast, the reasons for not creating bad lives are self-defeating (i.e. if we create them, then their lives are bad and we shouldn't have created them). There is no circularity here - conditional on us creating them, we have reason to regret out decision, and so have reason not to create them in the first place.
(2) On the practical side - within the EA space, I do think totalist views on population don't necessarily lead to prioritizing existential risk. There are two reasons for this. (a) Firstly, saving lives, at whatever point, creates future lives (since people have kids who have kids who have kids etc), and to the extent that global fertility rates are dropping, it's much more valuable to save lives now then in the future (and in AMF et al's operating locations especially). (b) Secondly, unless you think there will be a population bounce-back post catastrophe (e.g. 90% of the world is wiped out), in the sense of the remaining population having more kids than they otherwise would have sans catastrophe (which is extremely unlikely, and with the converse remaining far more probable, in fact), such near-catastrophes are merely a levelling down of the future aggregate human population, and 100% dying in the catastrophe would indeed be 10% worse than 90% dying, not disproportionately worse.
Your (1) just assumes that we have structural reasons to avoid regrettable (self-conditionally negative) outcomes, but no structural reasons to pursue desirable (self-conditionally positive) ones. I don't see any reason to assume that.
But as I flagged to Michael, I think the deeper issue here is that resorting to merely structural reasons is itself undermotivated. We should just directly appreciate that miserable lives are bad, even in prospect (and, correspondingly, that awesome lives are good, even in prospect). I guess you can call this "question-begging" if you like; many obvious truths are.
This is a minor stylistic point, but I think it is better to avoid using this word in online writing, since it lends itself too easily to misunderstanding. You are using it here to describe views like "there is no positive-valued well-being," but it is also widely used to describe views like "there is no such thing as moral value at all," or even "there is no such thing as *objective* moral value (but subjective moral value can still be important)." These views are very different from each other, and since it would be tedious to specify which one you mean every time, it is probably better just to avoid using the word altogether. (My assumption here is that you aren't deliberately conflating different views as part of your rhetorical strategy.)
The tedium of explication is precisely why the short, familiar label is often worth using -- so long as the intended meaning is clear enough from context (or doesn't make a huge difference). In this case, my reference bracketing "outright nihilism" is to the view that there is no such thing as moral value at all. Just flagging that I'm not arguing here against radical skeptics or error theorists who think we've no basis for attributing value to anything. But for the rest of us, if we allow for value at all, we should surely acknowledge "good lives" as among the things of value.
(In other posts, I have described the "no positive value" view as "quasi-nihilistic", to distinguish it from "outright nihilism". I'm happy with this terminology.)
It's a tricky (and hotly debated) question whether metaethical views that deny *objective* value are aptly described as 'nihilistic' or not. Obviously they're distinguishable from views that also deny the importance of subjective value, but that hardly settles the question. The point is that 'nihilism' has certain bleak connotations, and the question in dispute is precisely whether the targeted view *warrants* those connotations. Critics say yes, defenders say no. Whoever is right, is right. I don't think there's any presumption that we should defer to the terminological preferences of the defenders of the view. "Assume I'm right and describe my view accordingly" is not, in general, a reasonable request to make of one's philosophical critics when they're trying to highlight what they see as a *problem* with the targeted view.
Mar 21, 2023·edited Mar 21, 2023Liked by Richard Y Chappell
I think "quasi-nihilistic" is less objectionable, since it will at least give readers pause and force them to consider whether or not they really understand what you are talking about.
As for the metaethical case, I don't think there is anything objectionable about presenting an argument that "moral antirealism amounts to a form of nihilism" (I think the argument would fail, but that's another matter). However, I would find it objectionable if someone were to assume that their belief they possess such an argument (whose success or failure remains contested) justifies them in referring to moral antirealists as "nihilists" in contexts other than the presentation of that argument. Apart from anything else, it would just muddy the waters of discourse and make it difficult to know what anybody is talking about, especially in an online context where we are always already subject to the temptation to play a little fast and loose with our terminology.
A silly analogy: Suppose I think I have an argument that non-naturalist realism entails theism, but most non-naturalists think my argument doesn't work. It would be bad discursive practice for me to subsequently refer to moral non-naturalism as "theism," because it would be a misrepresentation of their views. (I think this would still be true even if people who were not non-naturalists generally accepted my argument.)
Welfarism is the view that what's good (simpliciter) is just welfare, i.e. things being good for various individuals. (Note that this can aggregate: saving two lives is better than just saving one, but there's no larger entity that it is better *for*.)
By contrast, a nihilist might agree that something is "good for" an individual, but deny that this *matters* in any way. That is, they would deny that welfare is good (simpliciter).
Actualism and Frick's views imply the procreation asymmetry, but I wouldn't consider them absolute views or standard Existence Anticomparativism. Rather, I'd still consider them merely comparative views (although we could argue about definitions). They allow certain kinds of comparisons, but for principled reasons and without directly assuming asymmetry between apparent goods and apparent bads or asserting the existence of absolute harms while denying absolute benefits:
1. Actualism allows actual beings to compare their lives to nonexistence (or us to do so on their behalf), but not non-actual beings to compare their lives with anything (or us to do so on their behalf). That seems intuitive to me, because the actual and only the actual can have objects that exist in the universe we can point to in order to ground their interests (e.g. brains with interest-generating and identity-establishing structures*).
2. Frick's conditional reasons are of the form "I have reason to (if I do p, do q)" (or Bykvist and Campbell, 2021's contrastive version "I have reason to (if I do p, do q) rather than not (if I do p, do q)").
If all person-regarding reasons are actualist or conditional (as I think is assumed in actualism and Frick's views, respectively), then we get the procreation asymmetry. You also concede (or at least don't strongly object to) actualist reasons and so at least weak asymmetry in Rethinking the Asymmetry, as "partiality towards the actual". The existence of absolute reasons seems to require more than the existence of actualist reasons or Frick's conditional reasons, at least with respect to apparently good and bad lives, because actualist and conditional reasons (at least with respect to individual welfare) each have absolute counterparts that make stronger claims, and there are more absolute reasons. So, it seems that there's an extra burden of proof to establish absolute reasons.
I think we can also use Frick's conditional reasons (and only conditional reasons), along with transitivity and the independence of irrelevant alternatives, to establish absolute harms while denying absolute goods, basically antifrustrationism (e.g. Fehige, 1998).
*This is a bit subtle, because we want to consider the whole of the future, including future actual beings, not just presently actual beings. This depends on specific views of the nature of time, although I wouldn't be surprised if the differences turned out to be merely normative, anyway. It also seems like you're sympathetic to something similar, with "timeless difference" and 4-dimensionalism, as you discuss in https://www.philosophyetc.net/2009/08/must-harms-be-temporally-located.html. Another possibility is Bader's account.
Frick, J. (2020). Conditional reasons and the procreation asymmetry. Philosophical Perspectives, 34(1), 53-87. https://onlinelibrary.wiley.com/doi/full/10.1111/phpe.12139
Bykvist, K., & Campbell, T. (2021). Frick’s Defence of the Procreation Asymmetry2. CLIMATE ETHICS, 263. https://www.iffs.se/media/23375/climate_ethics_vol4_webb.pdf#page=264
Fehige, C. (1998). A Pareto principle for possible people. Preferences, 508-43. https://www.fehige.info/pdf/A_Pareto_Principle_for_Possible_People.pdf
Bader, R. M. (2021). The Asymmetry. Ethics and Existence: The Legacy of Derek Parfit, 15. https://homeweb.unifr.ch/BaderR/Pub/Asymmetry%20%28R.%20Bader%29.pdf
Thanks for the references!
(1) I take it that 'actualist reasons' = what I call 'person-directed' reasons in the OP. The quick argument for absolute (or undirected) reasons is that we need them to explain our reasons *not* to create Misery (a merely possibly person, who *would* be miserable were she to exist).
One could instead appeal to some kind of structural reasons of ratifiability: if we created Misery, we'd then have (actualist) reasons to regret it. But this just pushes the asymmetry back a step, for what is the justification for adding structural reasons of regret-avoidance but not structural reasons of gladness-enticement (we would have actualist reasons to be glad of Joy's existence, were we to create her)?
(2) I worry that Frick's conditional approach presupposes rather than justifies the asymmetry. After all, it seems a general feature of "conditional obligations" (like promises) that violating them is bad, whereas satisfying them is merely neutral rather than good. [Cf. https://www.philosophyetc.net/2021/10/puzzling-conditional-obligations.html ] But that's an abhorrent view to take of people: that our existence is possibly bad, and neutral at best!
This brings me to my central argument for absolute value/reasons, which is just that its denial seems to express an abhorrent/disrespectful attitude towards human life, as I explain in the OP. We should view each other more positively than that.
(You suggest that "there's an extra burden of proof to establish absolute reasons." I reject that picture of moral methodology. Occam's razor is no reason at all to ascribe less value to people, or to smaller numbers of people. Moral parsimony is not a virtue. We should simply endorse the moral views that are best-warranted on their first-order merits.)
"(2) I worry that Frick's conditional approach presupposes rather than justifies the asymmetry. After all, it seems a general feature of "conditional obligations" (like promises) that violating them is bad, whereas satisfying them is merely neutral rather than good. [Cf. https://www.philosophyetc.net/2021/10/puzzling-conditional-obligations.html ]"
I'm not sure either way if I would say it presupposes ethical asymmetry, but that falls out basically immediately, so I guess it's at least "close" to presupposing it. Like with actualism in my other reply, if you tried to reverse Frick's conditional reasons, what you get wouldn't capture any kind of view of existence being prior to welfare, so it's hard to interpret it in person-affecting or non-absolute terms: "I have reason to do p and do q". p and q can be switched in that statement without affecting its truth value, but you can't do that for Frick's original "I have reason to (if I do p, do q)".
So, among person-affecting views and other views that reject absolute value, the asymmetry seems significantly more plausible than the reverse asymmetry, because reverse asymmetric views are harder to interpret as person-affecting or as rejecting absolute value than standard asymmetric views. Of course, this doesn't mean much if you're pretty sure of absolute value, as you seem to be.
I think the objection of regret-avoidance vs gladness-enticement may be missing how actualism is typically formulated. Here's "weak actualism" (not to be confused with only making weak asymmetry claims; the difference between weak and strong is who counts at all; strong actualism is much less plausible to me):
“They could say that what matters is whether the outcome of the action would have been better for those people who would have existed had it been performed. Thus:
Weak Actualism.—The moral status of any aj, actual or not, is determined by whether its outcome is better or worse for people in Saj than the outcomes of the other available actions", where aj is an action, and Saj is "the set of people who would exist (in the past, present, and future) if action aj were to be taken." (Hare, 2007)
"Self-Conditional Maximization. The permissible options are all and only the options that maximize self-conditional value." (Spencer, 2021)
So, under weak actualism, betterness/worseness is basically outcome-relative, because it's relative to who is actual. We're interested in whether the people in some outcome A are worse off than those same people (or nonexistence) in another outcome B, i.e. whether A is worse than B from the point of view of those in A. The unconditionally permissible options are exactly those that aren't worse than any other in this way, i.e. those that "maximize self-conditional value".
If B is just A with an additional miserable person and no other moral differences, and if B is actual, then B is worse than A for the people in B, because it's worse for the extra person, so it's not permitted under weak actualism. Still, A and B are equivalent for the people in A (or from the point of view of A).
If we try to reverse the actualist principles and get something like gladness-enticement and pronatalism, it's not totally clear how you should technically do this, and I think the result will seem spooky and lose its actualist/comparativist appeal.
On spookiness and actualist/comparativist appeal, suppose the outcome A is actual. If B is just A with an additional happy person and no other moral differences, then for A to be worse than B "from the point of view of A", which is actual, you need to refer to the interests of the non-actual extra person. That seems spooky and not at all in the spirit of actualism. If B is actual, then it would be better than A for the actual people (the people of B). However, that's already compatible with standard actualism, and doesn't rule out A unconditionally, because B isn't better for the people in A, and if you end up in A, there's no actual person for whom B is preferable.
On the technical difficulty of coming up with "actualist" principles to give gladness-enticement and pronatalism, I think you have to do one of the following:
1. An additional happy person in the actual outcome has to make that outcome unconditionally obligatory (rather than unconditional (im)permissibility as in standard actualism), which will mean multiple options are sometimes unconditionally obligatory, contrary to my understanding of unconditional obligation. For example, consider A and B, which differ just in the identity of one person who is very happy. We'd claim both A and B are unconditionally obligatory, but that can't be. That being said, I think there's a similar problem for standard actualism and miserable lives: if A and B differ only in the identity of a miserable person, then neither maximizes self-conditional value, and neither is permissible. One way to fix both views would be with something like saturating counterpart relations, or additional principles to break ties when there are multiple obligatory options or no permissible options. If I recall correctly, this is a common issue brought up and discussed around actualism, and Spencer (2021) in particular discusses Misery vs Moremisery and Misery vs Equalmisery. Also, multiple unconditionally obligatory options seems worse than no unconditionally permissible options.
2. From the actual world without an additional happy person, you need to decide which non-actual people to count and how to weigh them and their interests in non-actual worlds from the point of view of the actual world. There are usually multiple non-actual worlds to consider, so either you're left starting from pairwise comparisons and no counterpart to "maximize self-conditional value", or you sometimes need to weigh interests across 3 worlds simultaneously.
So, (something like) regret-avoidance falls out naturally from actualist intuitions, while (something like) gladness-enticement doesn't and faces more technical problems.
Hare, C. (2007). Voices from another world: Must we respect the interests of people who do not, and will never, exist?. Ethics, 117(3), 498-523.
https://www.journals.uchicago.edu/doi/abs/10.1086/512172?journalCode=et
Spencer, J. (2021). The procreative asymmetry and the impossibility of elusive permission. Philosophical Studies, 178(11), 3819-3842.
https://link.springer.com/article/10.1007/s11098-021-01627-y
Why the switch to "obligation"-talk? I take the pronatalist claim to just be that you have some *reason* to create each happy person (presumably in proportion to their welfare value). There's nothing paradoxical about that.
Working through the putative asymmetry in outcome-relative weak actualism, let's take world B as a base and compare both world A, which adds Joy, and world C, which adds Misery.
From B's perspective, the worlds are all value-equivalent. A says A is better. C says C is worse. Suppose B is actual. Asymmetrists want to say (i) we had good reason not to realize C, but (ii) no extra reason to realize A. How do they secure this? Appeal to self-conditional value gives us reasons against C, in that if we'd chosen C, it would've then been worse. But it's equally true that, if we'd chosen A, it would've then been better, suggesting that we should have self-conditional reasons for A. (B doesn't regard itself as any better or worse than either A or C, and so doesn't generate any competing self-conditional reasons when making these comparisons.) Where's the asymmetry between A and C?
I think an asymmetry may have been smuggled in to the account of permissibility as "exactly those [options] that aren't [self-conditionally] worse than any other". If we instead start from the more foundational normative question of what *reasons* there are in the situation (and insofar as we go for deontic concepts like 'permissibility' at all, take it to be determined in suitable fashion by the balance of those normative reasons), there doesn't appear to be any grounds for an asymmetry here after all.
(Though again, I do think the deeper issue here is that we should reject "the spirit of actualism". Even if B is actual, we should regard the extra suffering life in alternative C as a bad-making feature, resulting in world C being properly evaluated as *worse*. There's nothing "spooky" about this; it's entirely commonsensical that we should evaluate miserable lives as intrinsically bad.)
"Where's the asymmetry between A and C?
I think an asymmetry may have been smuggled in to the account of permissibility as "exactly those [options] that aren't [self-conditionally] worse than any other". If we instead start from the more foundational normative question of what *reasons* there are in the situation (and insofar as we go for deontic concepts like 'permissibility' at all, take it to be determined in suitable fashion by the balance of those normative reasons), there doesn't appear to be any grounds for an asymmetry here after all."
How else would you weigh reasons so that you get symmetry (reasons to create good lives, and equally strong reasons against creating bad lives) in a way that's person-affecting and without requiring absolute value?
I think the asymmetry and maximizing self-conditional value can be justified on the basis of a stability/dominance/money pump argument. If C would happen, on Misery's account, A and B would have been better, and neither A nor B directs you back to C following self-conditional value. If you could, you'd pay to switch to A or B. Or, you would pay to shorten Misery's life to nothing and have the result look like B, but with an extra cost. However, if you would pay to go from C to A or B (or worlds close enough to them), choosing C first means getting money pumped in principle, and you should have just chosen A or B in the first place without the extra cost.
On the other hand, from A or B, self-conditional value doesn't direct you away, and there's no issue of dominance or money pumps on the basis of self-conditional value if you were to choose A or B. The value of A over B in A is other-conditional value from the perspective of B. Joy doesn't exist in B, so Joy has no account in B to consider or compel you from B towards A (or C). Self-conditional reasons in A only compel you to stick to A if you're already there.
Maybe we have other reasons to switch from B to A, but those don't come from B, so they don't seem like they could be actualist, person-affecting or merely comparative reasons.
There may also be stability/dominance/money pump arguments against actualism generally, so maybe that undermines this kind of argument, but I don't think it would completely defeat it.
In general, options that don't maximize self-conditional value seem unstable, in that they'll direct you away from them if you follow self-conditional value and they'll make you vulnerable to dominance or money pumps. On the other hand, any option that does maximize self-conditional value is stable, at least with respect to self-conditional value.
>> "Note that there is (as far as I’m aware) no philosophical basis for then denying the same reasoning as applied to non-comparative goods. A happy existence is clearly non-comparatively good for the person who enjoys it."
I don't know about _denying_ the reasoning for non-comparative goods, but I think there really is an asymmetry between harms and goods, which is that the badness of suffering is surely the most undeniable moral fact we know; there is, to my intuition, no comparably obvious moral fact about goodness. Even though I _have_ a happy existence, I am much less certain when I introspect on the matter that this constitutes some non-comparative good--I honestly cannot tell you that I feel benefitted by my reasonably happy existence. But I am as certain as I am about anything at all, that if I were to be tortured for 100 years, that would be bad, and that I would be harmed by such an existence. I just genuinely, truly, feel orders of magnitude more confident drawing conclusions based on the badness of suffering than on the goodness of well-being.
The above is compounded by the fact that there is _much_ more disagreement as to what even constitutes well-being than what constitutes suffering--it just seems to me undeniable that the role that suffering and well-being should play in our theorizing has to be asymmetrical, even if only for reasons of uncertainty over definition.
>> "denying that our kids are better than nothing, or that utopia is better than a barren rock"
>>"It’s deeply disrespectful to deny that another person (who is, themselves, happy to exist) is better than nothing."
FWIW, I don't feel disrespected by the denial that I am better than nothing, despite being currently happy with my existence. I think some of the claims you make along these lines in the article undermine the argument: they feel like emotive appeals intended to short-circuit arguments. Further, it seems to me that most of the disagreement comes down to intuitions about non-existence and how it should be valued relative to happy existence, so these emotive appeals simply end up reading as rather strident restatements of your intuitions on the matter (to me, at least, who doesn't share these intuitions). Finally, "disrespect" is an odd term of analysis to bring in here; I'm sure many deontologists object that the utilitarian denial of intrinsic rights is "disrespectful" to people, but I doubt that would bother you (nor should it, in my opinion).
>> "But the wrong view here could easily motivate anti-natalist “voluntary human extinction”, which would be literally the worst thing ever. It’s really worth not making mistakes that astronomically great."
I think the second sentence supplies a resolution to the first that takes a lot of the sting out of this argument as a criticism of the asymmetry: I presume much of what makes human extinction the worst thing ever is that it is _irrevocable_ in a way that almost no other decision can be--so under conditions of moral uncertainty, even theories that say that voluntary extinction is okay should be extremely unwilling to endorse it in practice.
Some very foundational disagreements about moral methodology and reasoning here!
> "it just seems to me undeniable that the role that suffering and well-being should play in our theorizing has to be asymmetrical, even if only for reasons of uncertainty over definition"
Uncertainty over definition doesn't entail uncertainty about whether the thing itself exists. And I'm here arguing against the "strong asymmetry" which claims that creating non-comparative goods counts for *nothing*. That's the view that lacks any basis. You could count goods for *less*, but that sort of "weak asymmetry" is a very different matter. That's compatible with my claim that we've every reason to recognize *some* positive reason to bring good lives into existence.
> "there is, to my intuition, no comparably obvious moral fact about goodness"
It's good to enjoy fun times spent with loved ones. That's at least as obvious as the badness of suffering. (Compare the common wisdom: "better to have loved and lost than never to have loved at all.")
> "emotive appeals intended to short-circuit arguments"
Um, no. They *constitute* (enthymematic) arguments. Maybe you reject the premise, as anyone can of any argument. But vivid counterexamples, and highlighting (or disputing) the repugnance of a view's implications, are the life-blood of moral philosophy, not a "short circuit". Getting these things right is absolutely central to possessing moral insight.
> Finally, "disrespect" is an odd term of analysis to bring in here; I'm sure many deontologists object that the utilitarian denial of intrinsic rights is "disrespectful" to people, but I doubt that would bother you
Again, I couldn't disagree more. It's really important to value people correctly! "Disrespect" is the standard term for not properly valuing people. And I take *very* seriously the charge that utilitarianism is "disrespectful" in this way -- I think it's a *much* more serious objection than superficial complaints about its extensional verdicts, because it gets much more to the heart of things. That's why I wrote 'Value Receptacles' - https://philpapers.org/rec/CHAVR - and also 'The Mere Means Objection': https://www.utilitarianism.net/objections-to-utilitarianism/mere-means/
One other worry for these views that say that creating happy lives isn't good is they inevitably imply in specific scenarios that it's not bad to gratuitously make future people's lives less good for trivial gains. https://benthams.substack.com/p/a-new-utterly-decisive-argument-against
I don't think they *inevitably* imply this. Some ("wide") views hold that that different-number cases are *incomparable* in value (or "on a par"), while applying standard utilitarian axiology to same-number cases. But I guess they would probably at least imply that it's not bad to create X mediocre new lives, for trivial gains, rather than 2X different awesome lives.
Suppose that one can create Alice with ten utility or John with 5 utility. It wouldn't be wrong to trade Alice with ten utility for John with 5 utility (by trade I mean go from a state where one will create Alice to John). But then they can trade John with 5 utility for Alice with 5 utility--and neither action is impermissible.
On "wide" views, it precisely *would* be wrong to create John rather than Alice. It's permissible to create no extra life at all. But (on these views) if you do create, you need to do the best you can with it.
Okay this more complicated series of trades. You trade John with 10 utility for creating no one for creating Alice with 5 utility for creating no one for creating John with 5 utility. None of the choices is itself impermissible.
I take it there are no other procreative options at each time. I.e. at t1 you can either keep John (@10) or switch to no-one. You switch. At t2 you can keep no-one or switch to Alice, etc.
Raises interesting issues about diachronic consistency. E.g. if you foresee where you'll end up, then the original choice is wrong (you aren't really switching to "no one", but to John@5, while John@10 is an available alternative). If you can't foresee the future options, it's maybe not so obviously impermissible to get tripped up in this way. I expect there are standard things that defenders of intransitivity / cyclic preferences say about these sorts of situations...?
But this runs afoul of the following obvious principle which says that the fact that some act gives you extra options doesn't count against it. So switching at t1, on this account, is wrong only because you'll get extra choices, but that's obviously crazy. You shouldn't think "oh no, this action gives me extra options! I won't take it then."
I think even your last point will depend on the particular view. I think most "wide" views, as I understand that term, would endorse the 2X awesome lives, because you get X times the difference between the average awesome life and the average mediocre life, which counts a lot, and the extra X awesome lives don't count against, assuming they're not worse than nothing (or if the awesome lives are bad, assuming they don't count much against, e.g. if the difference between an awesome life and a mediocre one is larger than the overall badness of an awesome life relative to nonexistence). I think Meacham's "Person-affecting views and saturating counterpart relations", Thomas, 2022's wide asymmetric views (both soft and hard), Frick's views and appropriately modified wide versions of actualism would endorse the 2X for these reasons, at least in a pairwise comparison. Even negative utilitarian views could endorse the 2X, assuming the 2X awesome lives have less bads in them than the X mediocre lives, but in principle awesome lives could have more bads than mediocre ones, even significantly more, as long as there are enough offsetting or outweighing goods.
Thomas, T. (2022). The Asymmetry, Uncertainty, and the Long Term. Philosophy and Phenomenological Research. https://onlinelibrary.wiley.com/doi/full/10.1111/phpr.12927
I was assuming that the extra X lives would generate a lot of additional "incomparability" (as in Johan Gustafsson's discussions of "undistinguishedness"); but fair point that it may depend on the details of the view.
Gustafsson, J. (2020). Population axiology and the possibility of a fourth category of absolute value. Economics & Philosophy, 36: 81–110. https://doi.org/10.1017/S0266267119000087
>"there is (as far as I’m aware) no philosophical basis for then denying the same reasoning as applied to non-comparative goods."
Why is it your opponent who bears the burden of proof here? Even given the assumption that it is bad to create "miserable" individuals, applying the same reasoning to conclude that it is good to create "happy" individuals seems to require some version of the assumption that goodness and badness differ only in sign, rather than in dimension. This is a substantial assumption, and plausibly a false one (e.g. see Magnus Vinding's arguments here: centerforreducingsuffering.org/phenomenological-argument/). So do you think the assumption that value is basically one-dimensional, with good and bad differing only in sign, is well-supported? Or do you think your argument can work even without that assumption?
It's not clear to me what it means to ask whether goodness and badness are one dimension or two. But I do think there's a clear presumption of philosophical parity between the two, such that any putative asymmetries need to be explained rather than taken for granted. I'm happy to take that as self-evident.
Insofar as Vinding is claiming that there's nothing *desirable* (or good), opposing the way that suffering is *undesirable* (or bad), I just think that's insane. I read the link; I don't see anything "plausible" or worth engaging with there.
Really? That surprises me. Do you think the following is wrong?
"In my view, the strongest argument against the existence of a phenomenological counterpart to suffering is that introspection yields no sign of such a counterpart. When we introspectively examine the proposed candidates of positive experiences, such as those listed above, we do not find that they have any phenomenological properties that render them the dual opposites of suffering, or anti-suffering, as it were."
My own introspection yields pretty much the same result as Vinding. I take it that your own introspection yields something different (e.g. that a certain experience of pleasure could reasonably "offset" a given experience of pain to yield an overall zero-valenced experience).
What conclusion, if any, would you draw from the existence of people whose most basic intuitions about the value of their own experiences differ so much from yours?
I think it's wrong that that is any kind of argument. Positive and negative experiences aren't like matter and anti-matter, that literally cancel out (leaving "an overall zero-valenced experience") when combined. They compensate *normatively*, not *descriptively*.
To give an obvious example from common sense: “'Tis better to have loved and lost than never to have loved at all.” The suffering of grief is more than compensated for by the positive value of loving that person and getting to spend time with them while they were alive. This basic normative datum about compensation in no way depends upon combining the love and the grief to yield a weird combined experience with "zero valence".
I take your point about the compensation being normative rather than descriptive, but I don't think there is a "basic normative datum about [experiential] compensation," and I agree with Vinding that the impulse to posit such a datum comes from a mistaken account of experiential phenomenology combined with excessive systematizing zeal. There just isn't anything in our experiences (or at least in my experiences – perhaps there is in yours?) that could serve to ground a normative concept of "good experience" posessing philosophical parity with the concept of "bad experience" grounded by our experience of suffering.
Note, however, that I needed to add "experiential" to your proposed datum, because Vinding's criticism only targets assumptions about the normative value of subjective experiences. It has nothing to say about possible sources of value grounded in something other than the phenomenology of valenced experiences. It might undermine a "pro-existence" population ethics for a hedonic utilitarian, but it wouldn't necessarily undermine anyone else's.
(I don't know about you, but I find the notion that non-hedonic values can give us reasons for creating new lives much more intuitively plausible than the idea that hedonic values alone have the power to do so.)
I'm certainly inclined to think that the overwhelming bulk of what's valuable about our lives is non-hedonic. Raw pleasure isn't *that* big of a deal. But it's really just bonkers to deny that there's *any such thing* as positive experience.
I'm trying to imagine the response of ordinary people to your telling them, "On reflection, it doesn't seem to me that there's any such thing as happiness or good feelings. Don't you agree?" Pretty similar to if you suggested that grass was purple, I imagine.
This is a sleight of hand. Vinding never denied the existence of happiness, he denied the existence of "positive value" from mental states like happiness. I know you probably also think that's bonkers, but it's quite obviously not in the ballpark of obvious absurdity as denying the existence of happiness as a mental state.
If you're going around telling people that sort of thing, it helps if you wear special robes, or a silly hat, or something like that. They're much more likely to believe you that way.
(1) On the philosophical issue: Michael covers the matter well. But to justify the asymmetry in a more straightforward way - arguments for creating net-negative lives is self-defeating, while creating net-positive lives is question-begging.
If you ground normative value on conscious beings and their preferences & interests (i.e. my life and welfare matters because there is *this* perspective from which these things matter and are valuable), then if such beings/perspectives *don't exist and never will* unless we create them, there is no question-begging reason to create them (i.e. the argument goes - we should create such people because their lives matter; but their lives matter only because they want to live; but they would only want to live if we create such people and they exist in the first place).
In contrast, the reasons for not creating bad lives are self-defeating (i.e. if we create them, then their lives are bad and we shouldn't have created them). There is no circularity here - conditional on us creating them, we have reason to regret out decision, and so have reason not to create them in the first place.
(2) On the practical side - within the EA space, I do think totalist views on population don't necessarily lead to prioritizing existential risk. There are two reasons for this. (a) Firstly, saving lives, at whatever point, creates future lives (since people have kids who have kids who have kids etc), and to the extent that global fertility rates are dropping, it's much more valuable to save lives now then in the future (and in AMF et al's operating locations especially). (b) Secondly, unless you think there will be a population bounce-back post catastrophe (e.g. 90% of the world is wiped out), in the sense of the remaining population having more kids than they otherwise would have sans catastrophe (which is extremely unlikely, and with the converse remaining far more probable, in fact), such near-catastrophes are merely a levelling down of the future aggregate human population, and 100% dying in the catastrophe would indeed be 10% worse than 90% dying, not disproportionately worse.
Your (1) just assumes that we have structural reasons to avoid regrettable (self-conditionally negative) outcomes, but no structural reasons to pursue desirable (self-conditionally positive) ones. I don't see any reason to assume that.
But as I flagged to Michael, I think the deeper issue here is that resorting to merely structural reasons is itself undermotivated. We should just directly appreciate that miserable lives are bad, even in prospect (and, correspondingly, that awesome lives are good, even in prospect). I guess you can call this "question-begging" if you like; many obvious truths are.
I didn’t know you had a child :).
>"nihilism"
This is a minor stylistic point, but I think it is better to avoid using this word in online writing, since it lends itself too easily to misunderstanding. You are using it here to describe views like "there is no positive-valued well-being," but it is also widely used to describe views like "there is no such thing as moral value at all," or even "there is no such thing as *objective* moral value (but subjective moral value can still be important)." These views are very different from each other, and since it would be tedious to specify which one you mean every time, it is probably better just to avoid using the word altogether. (My assumption here is that you aren't deliberately conflating different views as part of your rhetorical strategy.)
The tedium of explication is precisely why the short, familiar label is often worth using -- so long as the intended meaning is clear enough from context (or doesn't make a huge difference). In this case, my reference bracketing "outright nihilism" is to the view that there is no such thing as moral value at all. Just flagging that I'm not arguing here against radical skeptics or error theorists who think we've no basis for attributing value to anything. But for the rest of us, if we allow for value at all, we should surely acknowledge "good lives" as among the things of value.
(In other posts, I have described the "no positive value" view as "quasi-nihilistic", to distinguish it from "outright nihilism". I'm happy with this terminology.)
It's a tricky (and hotly debated) question whether metaethical views that deny *objective* value are aptly described as 'nihilistic' or not. Obviously they're distinguishable from views that also deny the importance of subjective value, but that hardly settles the question. The point is that 'nihilism' has certain bleak connotations, and the question in dispute is precisely whether the targeted view *warrants* those connotations. Critics say yes, defenders say no. Whoever is right, is right. I don't think there's any presumption that we should defer to the terminological preferences of the defenders of the view. "Assume I'm right and describe my view accordingly" is not, in general, a reasonable request to make of one's philosophical critics when they're trying to highlight what they see as a *problem* with the targeted view.
I think "quasi-nihilistic" is less objectionable, since it will at least give readers pause and force them to consider whether or not they really understand what you are talking about.
As for the metaethical case, I don't think there is anything objectionable about presenting an argument that "moral antirealism amounts to a form of nihilism" (I think the argument would fail, but that's another matter). However, I would find it objectionable if someone were to assume that their belief they possess such an argument (whose success or failure remains contested) justifies them in referring to moral antirealists as "nihilists" in contexts other than the presentation of that argument. Apart from anything else, it would just muddy the waters of discourse and make it difficult to know what anybody is talking about, especially in an online context where we are always already subject to the temptation to play a little fast and loose with our terminology.
A silly analogy: Suppose I think I have an argument that non-naturalist realism entails theism, but most non-naturalists think my argument doesn't work. It would be bad discursive practice for me to subsequently refer to moral non-naturalism as "theism," because it would be a misrepresentation of their views. (I think this would still be true even if people who were not non-naturalists generally accepted my argument.)
Welfarism is the view that what's good (simpliciter) is just welfare, i.e. things being good for various individuals. (Note that this can aggregate: saving two lives is better than just saving one, but there's no larger entity that it is better *for*.)
By contrast, a nihilist might agree that something is "good for" an individual, but deny that this *matters* in any way. That is, they would deny that welfare is good (simpliciter).