Andreas Mogensen’s paper “The weight of suffering” presents an interesting argument in favor of the axiological position that “there exists some depth of suffering that cannot be compensated for by any measure of well-being” — a position he calls “LTNU” (Mogensen, 2022, abstract). Mogensen then proceeds to explore how one might respond to that argument and thereby reject LTNU.
My aim in this post is to raise some critical points in response to this paper. As a preliminary note, I should say that I commend Mogensen for taking up this crucial issue regarding the weight of suffering, and for exploring it in an open-ended manner.
“The greatest cost of accepting LTNU”
Mogensen writes (p. 12):
the greatest cost of accepting LTNU is surely that it appears to support the desirability of human extinction or the extinction of all sentient life (Crisp, 2021).
This seems to be the main objection that Mogensen raises against LTNU. Yet two points are worth making in response to that objection.
First, LTNU need not imply the all-things-considered desirability of human extinction (and how extinction occurs can also be relevant, but more on this below).
Second, the objection rests on the assumption that extinction is bad, which one may reasonably disagree with (especially if extinction happens through, say, voluntary non-procreation).
I will elaborate on these two points in turn.
LTNU need not imply the desirability of extinction
Axiological reasons
As Mogensen observes, ”LTNU does not presuppose consequentialism, nor any other theory of normative ethics. It is a fragment of a population axiology” (p. 13). And this axiological principle may be combined with other axiological principles that would imply that extinction is bad, especially if it involves everyone getting killed.
After all, LTNU (as Mogensen defines it) only says something about the relative value of extremely bad lives and lives that (purportedly) have positive welfare levels. It does not say anything about the relative value of extremely bad lives and other potential bads, such as death, murder, rights violations, etc.1
In particular, one may hold that (purported) positive goods can never outweigh extremely bad lives, but that other bads, such as those mentioned above, are of comparable disvalue to extremely bad lives. (David Benatar appears to hold roughly such a view, in that he seems to consider death bad in itself, Benatar, 2006, pp. 211-221.)
Indeed, on some preference-based views, one need not even posit additional axiological principles beyond LTNU, since the preference frustration entailed by (involuntary) extinction may overall imply even worse welfare levels compared to non-extinction (depending on how one construes these views, as well as how bad the non-extinction scenario would be).
This also highlights the importance of specifying how the extinction in question occurs. After all, if every individual in a given world were to decide not to procreate, and thus voluntarily bring about extinction, then this need not involve potential bads such as murder, rights violations, or severe preference frustrations. In contrast, scenarios involving violent destruction would involve such potential bads. The former extinction scenario would likely be regarded as far less bad by most people, and perhaps even all-things-considered better than continued existence if the voluntary extinction prevents vast amounts of unbearable suffering.
Empirical reasons
Beyond the axiological reasons listed above, there are also empirical reasons why LTNU need not imply the desirability of human extinction. For instance, one may hold the empirical belief that humanity is likely to soon abolish the biology of suffering in all sentient life, in which case human extinction might be very bad by the lights of LTNU.
And even if one is less optimistic about such an abolitionist prospect, one may still believe that continued human existence would on the whole tend to reduce extreme suffering, such as by reducing the number of wild animals, or — more speculatively — by causing less suffering than would an alien civilization in humanity’s stead (Knutsson, 2021, sec. 4). (Note that I am not claiming that humanity is most likely to reduce suffering overall, but merely that it is unclear as an empirical matter what humanity’s net effect on suffering will be, and hence it is unclear whether LTNU alone would support human extinction.)
Of course, the points above do not apply to the extinction of all sentient life. But even here, there are important qualifications to be made. First, the extinction of all sentient life would not be optimal (by the lights on LTNU) if sentient life re-emerges later in worse ways (cf. Knutsson, 2021, sec. 4). Hence, we must be careful to distinguish a merely temporary extinction of sentient life from permanent extinction. After all, the potential impossibility of permanent extinction could imply the all-things-considered undesirability of temporary extinction. For example, one might hold an optimistic view of the future of humanity that implies that scenarios involving temporary extinction followed by a re-emergence of sentience (which might involve hundreds of millions of years of wild-animal suffering on many planets) would be worse than scenarios without temporary extinction, e.g. if one believes that humanity will soon abolish suffering for good throughout the accessible universe.
Moreover, even if we are talking about permanent extinction, one may still accept LTNU without necessarily believing that the extinction of all sentient life is on the whole desirable. In particular, if one combines LTNU with one or more of the axiological views outlined in the previous section — e.g. axiological views that assign significant disvalue to killings or to frustrated preferences — and if one holds an optimistic view of humanity’s ability to prevent suffering and other bads in the future, then one could maintain that non-extinction would overall be preferable to permanent extinction.
Would extinction be bad?
The intuition that extinction would be bad may be questioned in a variety of ways, and it seems worth separating that intuition from related yet distinct intuitions.
For instance, the badness of extinction might be conflated with the potential badness of death and murder, yet extinction need not involve murder, and it need not involve more death than non-extinction (in fact, earlier extinction would likely involve less death overall; the same point is made in Bergström, 2022, sec. VIII).
Additionally, there may be good reasons to doubt our intuitions about the badness of extinction (through voluntary non-procreation, say). One reason is that we might for evolutionary reasons have what Thomas Metzinger calls an existence bias, which strongly biases our intuitions to favor continued existence at almost any price.
Likewise, one may reasonably question our intuitions about the badness of an empty world. That is, we may intuitively feel like an empty world would be a horrifying prospect, but does such an intuition stand up to scrutiny? Some have argued that it does not, or at least that it might not (see e.g. Benatar, 2018; Crisp, 2021). In particular, some have argued that there is nothing bad or even suboptimal about an empty world (Ajantaival, 2022).
Miscellaneous comments
Suffering in Omelas: Far from the worst suffering
In his discussion of LTNU, Mogensen uses the miserable child in The Ones Who Walk Away from Omelas as an example of someone who endures intense suffering (i.e. a child who is locked up in isolation and who barely gets enough food to survive). Yet focusing on this example is arguably unfair to LTNU. That is, compared to how bad suffering can get, this is an exceedingly mild example (even as it is horrific in absolute terms). It would be more fair to focus on more extreme examples of suffering when discussing the plausibility of LTNU.
Indeed, a proponent of LTNU could place the misery threshold at a level of suffering that is much more severe than the misery experienced by the child in Omelas, in which case discussions of Omelas-level suffering have limited bearing on the plausibility of LTNU.
A thought experiment involving destruction
Mogensen writes the following in his discussion of Omelas-level suffering (p. 11):
If the child is imagined as inhabiting a faroff country, and if the boundless and generous contentment of Omelas is imagined as independent of her suffering, except in that it would have to be destroyed in the process of working to spare her from her misery, then I don’t find I have the same reaction as before [i.e. that the child should be spared from suffering].
Yet this talk about destruction is potentially misleading (when raised as an argument against LTNU). As noted above, one may endorse other axiological principles that render such destruction bad overall, without thinking that happy lives can outweigh extremely miserable lives. So this case does not clearly pit the value of happy lives against the disvalue of miserable lives, and hence it does not clearly serve to question the plausibility of LTNU.
Additionally, one can criticize this framing for potentially appealing to a status quo bias, which could be avoided by instead considering the issue from a neutral starting point where no beings exist, and where we are contemplating whether it would be better to create these beings or not. (I have made a similar critique of a similar framing here and here.)
“Disturbing implications”
Finally, I find it a bit unfortunate that the paper repeatedly refers to the “disturbing implications” of LTNU. First of all because it is not specified what these “disturbing implications” are exactly; and the one implication that is specified (that extinction would be desirable) is not one that strictly follows from LTNU, nor one that is shown to be disturbing (e.g. such a claim may plausibly be questioned in the case of voluntary extinction).
Moreover, I suspect that this framing can potentially distort the paper’s examination, since a key question of the paper is whether LTNU is all-things-considered more or less plausible (or “disturbing”) than its rejection. Yet the problematic implications of the rejection of LTNU are not characterized as disturbing, even though one could argue that such implications are even more disturbing.
References
Ajantaival, T. (2022). Peacefulness, nonviolence, and experientialist minimalism. Ungated
Benatar, D. (2006). Better Never to Have Been: The Harm of Coming into Existence. Oxford University Press.
Benatar, D. (2018). Is Extinction Bad? Ungated
Bergström, L. (1978/2022). The consequences of pessimism. Translated by Simon Knutsson. Ungated
Crisp, R. (2021). Would extinction be so bad? Ungated
Knutsson, S. (2021). The world destruction argument. Inquiry, 64(10), pp. 1004-1023. Ungated
Metzinger, T. (2017). Benevolent Artificial Anti-Natalism (BAAN). Ungated
Mogensen, A. (2022). The weight of suffering. Ungated
- The name “LTNU” — an abbreviation for “lexical threshold negative utilitarianism” — is thus quite unfortunate, since this axiological principle does not imply utilitarianism of any kind.[↩]