Reading notes / Notes de lecture

Better Never to Have Been

In this book, published in 2006, David Benatar will argue for the following claims :

He will then consider the implication of his views in topics such as procreative freedom, abortion, population, and extinction.

Chapter 1 - Introduction

Benatar starts by giving a very quick summary of why coming into existence is always a harm :

Although the good things in one’s life make it go better than it otherwise would have gone, one could not have been deprived by their absence if one had not existed. Those who never exist cannot be deprived. However, by coming into existence one does suffer quite serious harms that could not have befallen one had one not come into existence.

The focus of this book is on humans, but he makes it clear that his arguments apply to all sentient beings.

He then distinguish between the anti-natalist positions founded on a dislike of chilren or in the interests of potential parents in having more freedom, and his anti-natalism, founded on avoiding suffering. But no matter their origin, anti-natalist views runs counter to a strong pro-natal bias, rooted in the evolutionary origins of humans. This bias often manifest where people who did not procreate are seen as "selfish". Benanar objects : if one refrain from procreating in order to avoid inflicting the harm of coming into existence, they are being altruistic, not selfish.

Pro-natalism also manifests itself in the policies of nations, democratic or dictatorial, and in the moral realm :

Parents with dependents are some how thought to count for more. If, for example, there is some scarce resource—a donor kidney perhaps—and of the two potential recipients one is a parent of young children and one is not, the parent, all things being equal, will likely be favoured. To let a parent die is not only to thwart that person’s preference to be saved, but also the preferences of his or her children that their parent be saved. It is quite true, of course, that the death of the parent will harm more people, but there is nonetheless something to be said against favouring parents. Increasing one’s value by having children might be like increasing one’s value by taking hostages. We might find it unfair and decide not to reward it. That may make children’s lives worse, but must the cost of preventing that outcome be placed on the shoulders of those who do not have children?

Another distinction is made between his view, and the view that the world is too horible to bring someone into existence. He disagree with it in that there could be much less suffering, and yet procreation would still be unacceptable, since there is no net benefit in coming into existence.

Chapter 2 - Why coming into existence is always a harm

Can coming into existence ever be a harm?

Common sense suggest that a life can be so bad that coming into existence with such a life is most certainly a harm. But there is a non-identity problem (or paradox of future individuals) :

The problem arises in those cases where the only alternative to bringing a person into existence with a poor quality of life is not to bring that person into existence at all. In such circumstances it is impossible to bring the same person into existence without the condition that is thought to be harmful. This may occur, for instance, where prospective parents are carriers of a serious genetic disorder which, for one reason or another, they will pass on to their offspring. The choice is either to bring a defective child into existence or not to bring that child into existence at all. On other occasions the defective condition is not attributable to the person’s constitution, genetic or otherwise, but rather to his environment. This is the case with the fourteen-year old girl who has a baby but because of her own tender age is unable to provide it with adequate opportunities. If she conceives a child when she is older and better able to care for it, it will not be the same child (because it will have been formed from different gametes). So her alternative to bringing a socially compromised child into existence when she is fourteen years old is not to bring that child into existence at all, irrespective of whether she later has another child.

Lives worth living and lives not worth living

Some have made the strong claim that coming into existence cannot be a harm, even where impairments make a life not worth living. The following sort of argument is advanced for such claims :

  1. For something to harm somebody, it must make that person worse off.
  2. The ‘worse off’ relation is a relation between two states.
  3. Thus, for somebody to be worse off in some state (such as existence), the alternative state, with which it is compared, must be one in which he is less badly (or better) off.
  4. But non-existence is not a state in which anybody can be, and thus cannot be compared with existence.
  5. Thus coming into existence cannot be worse than never coming into existence.
  6. Therefore, coming into existence cannot be a harm.

One can object to the first premise :

For something to harm somebody, it might be sufficient that it be bad for that person on condition that the alternative would not have been bad. On this view of harm, coming into existence can be a harm.

Joel Feinberg objects differently, by disputing the assumption that to be worse off in a particular condition, it must be the case that one would have existed in the alternative condition with which it is compared. He offers an analogy with suicide :

When a person claims that his life is so bad that he would be better off dead, he need not mean literally that were he to die he would exist in some better state (although some people do believe this). Instead he may mean that he prefers not to be, rather than to continue living in his condition. He has determined that his life is not worth living—that it is not worth continuing to be. Just as life can be so bad that ceasing to exist is preferable, so life can be so bad that never coming into existence is preferable. Comparing somebody’s existence with his non-existence is not to compare two possible conditions of that person. Rather it is to compare his existence with an alternative state of affairs in which he does not exist.

"A life worth living" is ambiguous :

It has generally been thought that those cases where the impairment, although severe, is not so bad as to make life not worth living are more difficult than cases where the impairment is so great as to make life not worth living. It has been said that because the former, by definition, are cases of lives worth living, one cannot judge never existing to be preferable to existing with such a life. The force of this argument, however, rests on a crucial ambiguity in the expression ‘a life worth living’—an ambiguity I shall now probe.

Lives worth starting and lives worth continuing

The phrase "a life worth living" can mean either a life worth continuing (present-life sense) or a life worth starting (future-life sense). Making the distinction is important, because we usually require stronger justification for ending a life than for not starting one. For example, most people think that living life without a limb does not make life so bad that it is worth ending, but most people also think that it is better not to bring into existence somebody who will lack a limb. Benatar makes an analogy with cinema : a film might be bad enough that it would have been better not to have gone to see it, but not so bad that it is worth leaving before it finishes.

Because we got rid of the ambiguity, there is nothing paradoxical about the claim that it is preferable not to begin a life that would be worth continuing. But is the distinction between present-life cases and future-life cases morally relevant? Some arguments threaten to diminish the importance of this distinction. Benatar will reply to each.

First, one from Derek Parfit's :

He suggests that if I am benefited by having my life saved just after it started, (even if at the expense of acquiring some severe but non-catastrophic impairment), then it is not implausible to claim that I am benefited by having my life started (with such an impairment). This argument seeks to minimize the significance of the distinction between future-life and present-life cases. On this view it is not unreasonable to think that impairments that are inflicted in the course of saving a life are morally comparable to similar impairments that are inseparable from bringing a life into existence.

Problem : coming into existence in a morraly relevant sense in a long process rather than an event :

Although most people would agree that to save my life now at the cost of my leg would confer a net benefit on me, many fewer people would think that saving the life of a conceptus at the cost of its living a life without a leg constitutes a net benefit. That is why many more people support ‘therapeutic’ abortions even for non-catastrophic defects than condemn life-saving amputations on ordinary adults. Some people support even infanticide or at least passive euthanasia for neonates with severe but non-catastrophic disabilities even though they would not judge similar conduct to be in the interests of non-infant children and adults with such defects. Those who exist (in the morally relevant sense) have interests in existing. These interests, once fully developed, are typically very strong and thus, where there is a conflict, they override interests in not being impaired. However, where there are no (or very weak) interests in existing, causing impairments (by bringing people with defects into being) cannot be warranted by the protection of such interests. The scope of the class of beings without interests (or with very weak interests) in existing is a matter of dispute. (Does it include embryos, zygotes, infants?) In Chapter 5 I argue that at least zygotes, embryos, and fetuses until quite late in gestation have not begun existing in a morally relevant sense and that coming to exist in a morally relevant sense is a gradual process.

Can this gradualist view be used against Benatar? :

Now some might think that the gradualist view about coming into existence undermines my distinction between future-life cases and present-life cases. This, however, is not true. That the distinction between them is a gradual one does not render the distinction void. Nothing I have said excludes the possibility of a middle ground linking the two kinds of cases. Nor is the moral significance of the distinction compromised so long as one does not, as I do not, reject a moral sensitivity to the gradualism of the continuum that links clear future-life cases with clear present-life cases.

Second, one from Joel Feinberg. He starts by distinguishing between judgements by competent adults or mature older children that it would have been preferable if they had never come into existence, and similar judgements made by proxies on behalf of those who are so extremely impaired that they cannot make judgements themselves. For the latter, he thinks, it is insufficient that the judgement of the preferability of never existing be consistent with reason. It must be dictated (or required) by reason, and this requirement is met only when death would be preferable for the disabled person. For the former, he allows that the judgement be merely consistent with reason (that is, not irrational). Benatar :

The reason why this account conflicts with my distinction between present-life cases and future-life cases is that implicit in it is the requirement that we make judgements about future-life cases through the lens of present-life cases. Either life has to be so bad that it would not be worth continuing —Professor Feinberg’s standard for proxy decisions—or it has to be the case that already existing people with that disability would prefer never to have come into existence—his standard for those whose disabilities do not impair their competence to decide (retrospectively!) for themselves.

However, it is precisely because Professor Feinberg’s account requires us to adopt the perspective of already existing people that it is inadequate. In asking whether a life is worth starting, we should not have to consider whether it would not be worth continuing. Nor should we have to appeal to the preferences of already existing people about their own lives to make judgements about future lives. As I shall show in the second section of the next chapter, self-assessments of one’s life’s quality are unreliable.

Why coming into existence is always a harm

Common assumption : "one does no wrong by bringing into existence people whose lives will be good on balance", this is based on another : "being brought into existence (with decent life prospects) is a benefit (even though not being brought into existence is not a harm)"

Benatar claims that coming into existence is always a harm, but not that it is necessarily a harm. His argument do not apply to hypothetical cases in which a life contains only good and no bad. For those cases, he says that it is neither a harm nor a benefit, and we should be indifferent between such an existence and never existing. But no actual lives are like this.

For Benatar, everyone suffers at one point in life, and everyone dies (however, there are some views that consider death not to be a harm, more on that in chapter 7).

We infrequently contemplate the harms that await any new-born child—pain, disappointment, anxiety, grief, and death. For any given child we cannot predict what form these harms will take or how severe they will be, but we can be sure that at least some of them will occur. None of this befalls the non-existent. Only existers suffer harm.

Optimists will respond that life also contains pleasure, joy, satisfaction :

Not only bad things but also good things happen only to those who exist. Pleasure, joy, and satisfaction can only be had by existers. Thus, the cheerful will say, we must weigh up the pleasures of life against the evils. As long as the former outweigh the latter, the life is worth living. Coming into being with such a life is, on this view, a benefit.

The asymmetry of pleasure and pain

The problem with that optimist view is that there is an asymmetry between benefits and harms, for example between pleasure and pain. It is not controversial to say :

1) the presence of pain is bad.

2) the presence of pleasure is good.

But there is no symmetry regarding absence :

3) the absence of pain is good, even if that good is not enjoyed by anyone.

4) the absence of pleasure is not bad unless there is somebody for whom this absence is a deprivation.

Benatar goes on to defend claim (3) :

The judgement made in (3) is made with reference to the (potential) interests of a person who either does or does not exist. To this it might be objected that because (3) is part of the scenario under which this person never exists, (3) cannot say anything about an existing person. This objection would be mistaken because (3) can say something about a counterfactual case in which a person who does actually exist never did exist. Of the pain of an existing person, (3) says that the absence of this pain would have been good even if this could only have been achieved by the absence of the person who now suffers it. In other words, judged in terms of the interests of a person who now exists, the absence of the pain would have been good even though this person would then not have existed. Consider next what (3) says of the absent pain of one who never exists—of pain, the absence of which is ensured by not making a potential person actual. Claim (3) says that this absence is good when judged in terms of the interests of the person who would otherwise have existed. We may not know who that person would have been, but we can still say that whoever that person would have been, the avoidance of his or her pains is good when judged in terms of his or her potential interests. If there is any (obviously loose) sense in which the absent pain is good for the person who could have existed but does not exist, this is it. Clearly (3) does not entail the absurd literal claim that there is some actual person for whom the absent pain is good.

Then defend the asymmetry between (3) and (4) by showing it explanatory power. It can explain 4 other asymmetries :

First, the asymmetry between the duty to not bring suffering people into being, and the absence of duty to bring happy people into being. Possible objections :

It might be suggested that the reason why we have a duty to avoid bringing suffering people into being, but not a duty to bring happy people into existence, is that we have negative duties to avoid harm but no corresponding positive duties to bring about happiness. Judgements about our procreational duties are thus like judgements about all other duties. Now I agree that for those who deny that we have any positive duties, this would indeed be an alternative explanation to the one I have provided. However, even of those who do think that we have positive duties only a few also think that amongst these is a duty to bring happy people into existence.

It might now be suggested that there is also an alternative explanation why those who do accept positive duties do not usually think that these include a duty to bring happy people into existence. It is usually thought that our positive duties cannot include a duty to create lots of pleasure if that would require significant sacrifice on our part. Given that having children involves considerable sacrifice (at least to the pregnant woman), this, and not asymmetry, is the best explanation for why there is no duty to bring happy people into existence.

The problem with that view of positive duties taking sacrifice into account, is that we would have a duty to bring happy people into being if the sacrifice was low or null :

But this presupposes that the duty under discussion is an all-things-considered duty. However, the interests of potential people cannot ground even a defeasible duty to bring them into existence. Put another way, the asymmetry of procreative (all-things-considered) duties rests on another asymmetry—an asymmetry of procreative moral reasons. According to this asymmetry, although we have a strong moral reason, grounded in the interests of potential people, to avoid creating unhappy people, we have no strong moral reason (grounded in the interests of potential people) to create happy people. It follows that although the extent of the sacrifice may be relevant to other positive duties, this is moot in the case of a purported duty to bring happy people into existence.

Second, an asymetry of reasons to procreate based on the potential child's interests :

Whereas it is strange (if not incoherent) to give as a reason for having a child that the child one has will thereby be benefited, it is not strange to cite a potential child’s interests as a basis for avoiding bringing a child into existence. If having children were done for the purpose of thereby benefiting those children, then there would be greater moral reason for at least many people to have more children. In contrast to this, our concern for the welfare of potential children who would suffer is a sound basis for deciding not to have the child. If absent pleasures were bad irrespective of whether they were bad for anybody, then having children for their own sakes would not be odd. And if it were not the case that absent pains are good even where they are not good for anybody, then we could not say that it would be good to avoid bringing suffering children into existence.

Third, an asymmetry in our retrospective judgments about bringing people into existence or not :

Bringing people into existence as well as failing to bring people into existence can be regretted. However, only bringing people into existence can be regretted for the sake of the person whose existence was contingent on our decision. This is not because those who are not brought into existence are indeterminate. Instead it is because they never exist. We can regret, for the sake of an indeterminate but existent person that a benefit was not bestowed on him or her, but we cannot regret, for the sake of somebody who never exists and thus cannot thereby be deprived, a good that this never existent person never experiences. One might grieve about not having had children, but not because the children that one could have had have been deprived of existence. Remorse about not having children is remorse for ourselves—sorrow about having missed childbearing and child-rearing experiences. However, we do regret having brought into existence a child with an unhappy life, and we regret it for the child’s sake, even if also for our own sakes. The reason why we do not lament our failure to bring somebody into existence is because absent pleasures are not bad.

Finally, an asymmetry between distant suffering and unhabited portions of the Earth or universe :

Whereas, at least when we think of them, we rightly are sad for inhabitants of a foreign land whose lives are characterized by suffering, when we hear that some island is unpopulated, we are not similarly sad for the happy people who, had they existed, would have populated this island. Similarly, nobody really mourns for those who do not exist on Mars, feeling sorry for potential such beings that they cannot enjoy life. Yet, if we knew that there were sentient life on Mars but that Martians were suffering, we would regret this for them. The claim here need not (but could) be the strong one that we would regret their very existence. The fact that we would regret the suffering within their life is sufficient to support the asymmetry I am defending. The point is that we regret suffering but not the absent pleasures of those who could have existed.

An objection would be to say that while we do not regret the absent pleasure of the non-existent, we do not take joy in their absence of suffering (otherwise we would be overjoyed all the time). Benatar answer that the opposite of regret isn't joy :

Thus the important question is not whether we feel joy—the opposite of melancholy—about absent pains but whether the absent pain is the opposite of regrettable—what we might call ‘welcome’ or simply ‘good’. The answer, I have suggested, is affirmative. If we are asked whether the absent suffering is a good feature of never existing, we would have to say that it is.

Judgements supported by the asymmetry between 3 and 4 are not universally shared, taking the example of positive utilitarians :

For example, positive utilitarians—who are interested not only in minimizing pain but also in maximizing pleasure—would tend to lament the absence of additional possible pleasure even if there were nobody deprived of that pleasure. On their view, there is a duty to bring people into existence if that would increase happiness. This is not to say that all positive utilitarians must reject the view about the asymmetry of (3) and (4). Positive utilitarians who are sympathetic to the asymmetry could draw a distinction between (i) promoting the happiness of people (that exist, or will exist independently of one’s choices) and (ii) increasing happiness by making people. This is the now famous distinction between (i) making people happy and (ii) making happy people. Positive utilitarians who draw this distinction could then, consistent with positive utilitarianism, judge only (i) to be a requirement of morality. This is the preferable version of positive utilitarianism. Taking (ii) also to be a requirement of morality mistakenly assumes that the value of happiness is primary and the value of persons is derivative from this. However, it is not the case that people are valuable because they add extra happiness. Instead extra happiness is valuable because it is good for people—because it makes people’s lives go better. To think otherwise is to think that people are mere means to the production of happiness. Or, to use another famous image, it is to treat persons as mere vessels of happiness. But unlike a mere vessel, which is indifferent to how much of a valued substance it contains, a person cares about how much happiness he has.

Benatar shows that, given the asymetry, it is always a harm to come into existence, by comparing two scenarios, on in which a being called X exists, and one in which X never exists. This can be represented in Figure 2.1 :

Figure 2.1

Comparing existing with never existing

When we compare pain in both scenarios, non-existence wins (absence of pain > presence of pain). But for pleasure, the pleasure of the existent is not an advantage over its absence in non existence, since this absense is "not bad", instead of "bad". One could objects that "good" is still better than "not bad". However :

The mistake underlying this objection, however, is that it treats the absence of pleasure in Scenario B as though it were akin to the absence of pleasure in Scenario A—a possibility not reflected in my matrix, but which is implicit in (4) of my original description of asymmetry. There I said that the absence of pleasure is not bad unless there is somebody for whom this absence is a deprivation. The implication here is that where an absent pleasure is a deprivation it is bad. Now, obviously, when I say that it is bad, I do not mean that it is bad in the same way that the presence of pain is bad. What is meant is that the absent pleasure is relatively (rather than intrinsically) bad. In other words, it is worse than the presence of pleasure. But that is because X exists in Scenario A. It would have been better had X had the pleasure of which he is deprived. Instead of a pleasurable mental state, X has a neutral state. Absent pleasures in Scenario B, by contrast, are not neutral states of some person. They are no states of a person at all. Although the pleasures in A are better than the absent pleasures in A, the pleasures in A are not better than the absent pleasures in B.

An analogy :

Some people have difficulty understanding how (2) is not an advantage over (4). They should consider an analogy which, because it involves the comparison of two existent people is unlike the comparison between existence and non-existence in this way, but which nonetheless may be instructive. S (Sick) is prone to regular bouts of illness. Fortunately for him, he is also so constituted that he recovers quickly. H (Healthy) lacks the capacity for quick recovery, but he never gets sick. It is bad for S that he gets sick and it is good for him that he recovers quickly. It is good that H never gets sick, but it is not bad that he lacks the capacity to heal speedily. The capacity for quick recovery, although a good for S, is not a real advantage over H. This is because the absence of that capacity is not bad for H. This, in turn, is because the absence of that capacity is not a deprivation for H. H is not worse off than he would have been had he had the recuperative powers of S. S is not better off than H in any way, even though S is better off than he himself would have been had he lacked the capacity for rapid recovery.

Benatar attacks the "analysis of the cheerful", which weighs up (1) and (2). First by saying it ignores the right side of the diagram :

If we want to determine whether non-existence is preferable to existence, or vice versa, then we must compare the left- and the right-hand sides of the diagram, which represent the alternative scenarios in which X exists and in which X never exists. Comparing the upper and the lower quadrants on the left does not tell us whether Scenario A is better than Scenario B or vice versa. That is unless quadrants (3) and (4) are rendered irrelevant. One way in which that would be so is if they were both valued as ‘zero’. On this assumption A can be thought to be better than B if (2) is greater than (1), or to put it another way, if (2) minus (1) is greater than zero. But this poses a second problem. To value quadrants (3) and (4) at zero is to attach no positive value to (3) and this is incompatible with the asymmetry for which I have argued.

Second, by saying it conflates a life worth starting and a life worth continuing :

The cheerful tell us that existence is better than non-existence if (2) is greater than (1). But what is meant by ‘non-existence’ here? Does it mean ‘never existing’ or ‘ceasing to exist’? Those who look only at (1) and (2) do not seem to be distinguishing between never existing and ceasing to exist. For them, a life is worth living (that is, both starting and continuing) if (2) is greater than (1), otherwise it is not worth living (that is, neither worth starting nor continuing). The problem with this, I have already argued, is that there is good reason to distinguish between them. For a life to be not worth continuing, it must be worse than it need be for it not to be worth starting. Those who consider not only Scenario A but also Scenario B clearly are considering which lives are worth starting. To determine which lives are worth continuing, Scenario A would have to be compared with a third scenario, in which X ceases to exist.

Finally, the quality of a life is not determined simply by subtracting the bad from the good. As I shall show in the first section of the next chapter, assessing the quality of a life is much more complicated than this.

Another possible stance is to accept the asymmetry of Figure 2.1, agree that we need to compare Scenario A with Scenario B, but deny that this leads to the conclusion that B is always preferable to A. The argument is that we must assign positive, negative or neutral values to each of the quadrants. Figure 2.4 :

Figure 2.4

If (1) is at - n, (3) at + n, and (4) at 0, then coming into existence is preferable when (2) is greater than 2 * n. Benatar objects :

There are numerous problems with this. For instance, as I shall show in the first section of the next chapter, it is not only the ratio of pleasure to pain that determines the quality of a life, but also the sheer quantity of pain. Once a certain threshold of pain is passed, no amount of pleasure can compensate for it.

When we apply this quantitative logic to the Sick and Healthy analogy, we see how it is mistaken :

Figure 2.5

Following Figure 2.5, it would be better to be S than H if the value of (2) were more than twice the value of (1). (This presumably would be the case where the amount of suffering that (2) saves S is more than twice the amount S actually suffers.) But this cannot be right, for surely it is always better to be H (a person who never gets sick and is thus not disadvantaged by lacking the capacity for quick recovery). The whole point is that (2) is good for S but does not constitute an advantage over H. By assigning a positive charge to (2) and a ‘0’ to (4), Figure 2.5 suggests that (2) is an advantage over (4), but it quite clearly is not. The assignment of values in Figure 2.5, and hence also in Figure 2.4, must be mistaken.

Coming into existence will always be a harm, not matter how small :

There is a difference, I have indicated, between (a) saying that coming into existence is always a harm and (b) saying how great a harm it is. So far I have argued only for the first claim. The magnitude of the harm of existence varies from person to person, and in the next chapter I shall argue that the harm is very substantial for everybody. However, it must be stressed that one can endorse the view that coming into existence is always a harm and yet deny that the harm is great. Similarly, if one thinks that the harm of existence is not great, one cannot infer from that that existence is preferable to non-existence.

On this view, a life of bliss adulterated by a single pin-prick has no advantages over non-existence. In a world where all lives are like this, that tiny harm would be easily outweighed by the benefits to other, already existing humans. In our world however, no lives are like this.

Against not regretting one’s existence

Some people agree with Alfred Lord Tennyson that it is better to have loved and lost than never to have loved at all. They might think that they can apply similar reasoning to existence : it is better to have existed and lost (both by suffering within life and then by ceasing to exist) than never to have existed at all. But there is a huge difference between lovig and coming into existence. The person who never loves exists without loving and is thus deprived, which is "bad" according to Benatar. But one who never comes into existence is not deprived of anything, which is "not bad".

Most people do not regret their existence, but this fact does not make one’s existence better than non-existence, because :

[...] if one had not come into existence there would have been nobody to have missed the joy of leading that life and thus the absence of joy would not be bad. Notice, by contrast, that it makes sense to regret having come into existence if one does not enjoy one’s life. In this case, if one had not come into existence then no being would have suffered the life one leads. That is good, even though there would be nobody who would have enjoyed that good.

Now it may be objected that one cannot possibly be mistaken about whether one’s existence is preferable to non-existence. It might be said that just as one cannot be mistaken about whether one is in pain, one cannot be mistaken about whether one is glad to have been born. Thus if ‘I am glad to have been born’, a proposition to which many people would assent, is equivalent to ‘It is better that I came into existence’, then one cannot be mistaken about whether existence is better than non-existence. The problem with this line of reasoning is that these two propositions are not equivalent. Even if one cannot be mistaken about whether one currently is glad to have been born, it does not follow that one cannot be mistaken about whether it is better that one came into existence. We can imagine somebody being glad, at one stage in his life, that he came to be, and then (or earlier), perhaps in the midst of extreme agony, regretting his having come into existence. Now it cannot be the case that (all things considered) it is both better to have come into existence and better never to have come into existence. But that is exactly what we would have to say in such a case, if it were true that being glad or unhappy about having come into existence were equivalent to its actually being better or worse that one came into being. This is true even in those cases in which people do not change their minds about whether they are happy to have been born. Why so few people do change their minds is explained, at least in part, by the unduly rosy picture most people have about the quality of their own lives. In the coming chapter, I show that (with the exception of real pessimists, who may have an accurate view of how bad their lives are) people’s lives are much worse than they think.

Chapter 3 - How Bad Is Coming into Existence?

This chapter is a continuation of the previous one (the conclusion that coming into existence is always a harm tells us nothing about the magnitude of that harm). It also provides a basis independent of the asymmetry, for regretting one’s existence and for taking all actual cases of coming into existence to be harmful :

If people realized just how bad their lives were, they might grant that their coming into existence was a harm even if they deny that coming into existence would have been a harm had their lives contained but the smallest amount of bad.

Why life's quality is not the difference between the good and the bad

Substracting the bad from the good is not sufficient to assess how bad a life is. How the good and the bad are distributed plays a role too. Three distributional considerations are outlined :

Order :

For instance, a life in which all the good occurred in the first half, and uninterrupted bad characterized the second half, would be a lot worse than one in which the good and bad were more evenly distributed. This is true even if the total amount of good and bad were the same in each life. Similarly, a life of steadily inclining achievement and satisfaction is preferable to one that starts out bright in the very earliest years but gets progressively worse.

Intensity :

A life in which the pleasures were extraordinarily intense but correspondingly few, infrequent, and short-lived might be worse than a life with the same total amount of pleasure, but where the individual pleasures were less intense and more frequently distributed across the life. However, pleasures and other goods can also be distributed too widely within a life, thereby making them so mild as to be barely distinguishable from neutral states. A life so characterized might be worse than one in which there were a few more noticeable ‘highs’.

Length :

To be sure, the length of a life will interact dynamically with the quantity of good and bad. A long life with very little good would have to be characterized by significant quantities of bad, if only because the absence of sufficient good over such long periods would create tedium—a bad. Nevertheless, we can imagine lives of somewhat unequal length that share the same quantity of good and of bad. One life might have more neutral features, sufficiently evenly distributed over the life not to affect the quantity of good or bad. In such cases, one might plausibly judge the longer life to be better (if the life is of sufficient quality to be worth continuing) or worse (if it is not).

There is further (non-distributional) consideration, which is the threshold of badness :

Arguably, once a life reaches a certain threshold of badness (considering both the amount and the distribution of its badness), no quantity of good can outweigh it, because no amount of good could be worth that badness. It is just this assessment that Donald (‘Dax’) Cowart made of his own life—or at least of that part of his life following a gas explosion that burnt two-thirds of his body. He refused extremely painful, life- saving treatment, but the doctors ignored his wishes and treated him nonetheless. His life was saved, he achieved considerable success, and he reattained a satisfactory quality of life. Yet, he continued to maintain that these post-burn goods were not worth the costs of enduring the treatments to which he was subjected. No matter how much good followed his recovery, this could not outweigh, at least in his own assessment, the bad of the burns and treatment that he experienced.

Why self-assessments of one's life's quality are unreliable

Most people deny that their lives are bad, and they certainly deny that their lives are so bad as to make never existing preferable. Benatar argue that it is not the quality of their lives that explain this, but rather human psychological phenomenon. The first is the Pollyanna principle, a tendency towards optimism :

This manifests in many ways. First, there is an inclination to recall positive rather than negative experiences. For example, when asked to recall events from throughout their lives, subjects in a number of studies listed a much greater number of positive than negative experiences. This selective recall distorts our judgement of how well our lives have gone so far. It is not only assessments of our past that are biased, but also our projections or expectations about the future. We tend to have an exaggerated view of how good things will be. The Pollyannaism typical of recall and projection is also characteristic of subjective judgements about current and overall well-being. Many studies have consistently shown that self-assessments of well-being are markedly skewed toward the positive end of the spectrum. For instance, very few people describe themselves as ‘not too happy’. Instead, the overwhelming majority claims to be either ‘pretty happy’ or ‘very happy’. Indeed, most people believe that they are better off than most others or than the average person.

Most of the factors that plausibly improve the quality of a person’s life do not commensurately influence self-assessments of that quality (where they influence them at all). For example, although there is a correlation between people’s own rankings of their health and their subjective assessments of well-being, objective assessments of people’s health, judging by physical symptoms, are not as good a predictor of peoples’ subjective evaluations of their well-being. Even among those whose dissatisfaction with their health does lead to lower self-reported well-being, most report levels of satisfaction toward the positive end of the spectrum. Within any given country, the poor are nearly (but not quite) as happy as the rich are. Nor do education and occupation make much (even though they do make some) difference. Although there is some disagreement about how much each of the above and other factors affect subjective assessments of well-being, it is clear that even the sorts of events that one would have thought would make people ‘very unhappy’ have this effect on only a very small proportion of people.

There is also some kind of adaptaion to bad events, that makes our assessments of well-being unreliable and that explains some of the Pollyannianism just mentionned :

When a person’s objective well-being takes a turn for the worse, there is, at first, a significant subjective dissatisfaction. However, there is a tendency then to adapt to the new situation and to adjust one’s expectations accordingly. Although there is some dispute about how much adaptation occurs and how the extent of the adaptation varies in different domains of life, there is agreement that adaptation does occur. [...] Because the subjective sense of well-being tracks recent change in the level of well-being better than it tracks a person’s actual level of well-being, it is an unreliable indicator of the latter.

The third factor is comparing ourselves with others :

It is not so much how well one’s life goes as how well it goes in comparison with others that determines one’s judgement about how well one’s life is going. Thus self-assessments are a better indicator of the comparative rather than actual quality of one’s life. One effect of this is that those negative features of life that are shared by everybody are inert in people’s judgements about their own well-being. Since these features are very relevant, overlooking them leads to unreliable judgements.

Those three factors reinforce each other :

Of these three psychological phenomena, it is only Pollyannaism that inclines people unequivocally towards more positive assessments of how well their life is going. We adapt not only to negative situations but also to positive ones, and we compare ourselves not only with those who are worse off but also with those who are better off than we are. However, given the force of Pollyannaism, both adaptation and comparison operate both from an optimistic baseline and under the influence of optimistic biases. For example, people are more prone to comparing them- selves with those who are worse off than with those who are better off. Thus, in the best cases, adaptation and comparison reinforce Pollyannaism. In the worst cases, they mitigate it but do not negate it entirely. When we do adapt to the good or compare ourselves with those who are better off than ourselves, our self-assessments are less positive than they otherwise would be, but they do not usually cause them to become negative.

Their existence is not surprising, given how they operate in an evolutionnary perspective.

Three views about the quality of life, and why life goes badly on all of them

Three kinds of theories about the quality of a life :

Hedonistic theories

Included in negative mental states are discomfort, pain, suffering, distress, guilt, shame, irritation, boredom, anxiety, frustration, stress, fear, grief, sadness, and loneliness.

Positive mental states are of two kinds :

Some pleasures have both components : eating a tasty meal while hungry brings both relief from hunger and the intrinsic pleasure of fine-tasting food.

Neutral ones are neither negative, nor positive.

For Benatar, daily life is filled with negative mental states that we tend to overlook because of the aformentionned psychological reasons :

These include hunger, thirst, bowel and bladder distension (as these organs become filled), tiredness, stress, thermal discomfort (that is, feeling either too hot or too cold), and itch. For billions of people, at least some of these discomforts are chronic. These people cannot relieve their hunger, escape the cold, or avoid the stress. However, even those who can find some relief do not do so immediately or perfectly, and thus experience them to some extent every day. In fact, if we think about it, significant periods of each day are marked by some or other of these states. For example, unless one is eating and drinking so regularly as to prevent hunger and thirst or countering them as they arise, one is likely hungry and thirsty for a few hours a day.

This is characteristic of a healthy daily life, but some experience more badness :

Chronic ailments and advancing age make matters worse. Aches, pains, lethargy, and sometimes frustration from disability become an experiential backdrop for everything else. Now add those discomforts, pains, and sufferings that are experienced either less frequently or only by some (though nonetheless very many) people. These include allergies, headaches, frustration, irritation, colds, menstrual pains, hot flushes, nausea, hypoglycaemia, seizures, guilt, shame, boredom, sadness, depression, loneliness, body-image dissatisfaction, the ravages of AIDS, of cancer, and of other such life-threatening diseases, and grief and bereavement.

Of course there are also neutral states, relief pleasures and intrinsic pleasures, but :

[...] there would be something absurd about living for neutral states or relief pleasures, or about starting a life in order to create more neutral conscious states or to produce more relief pleasure. Neutral states and relief pleasures can be valuable only in so far as they displace negative states. The argument that it is better never to come into existence explains why it is also absurd to start a life for the intrinsic pleasures that that life will contain. The reason for this is that even the intrinsic pleasures of existing do not constitute a net benefit over never existing. Once alive, it is good to have them, but they are purchased at the cost of life’s misfortune—a cost that is quite considerable.

Desire-fulfilment theories

What has just been said is relevant to desire-fulfilment theories, since we desire positive mental state and the absence of negatives ones. However, there is a clear difference, because there can be positive mental states in the absence of fulfilled desires, when :

And there can be fulfilled desires in the absence of positive mental states, when :

One cannot be mistaken on whether or not one experience a positive, neutral or negative mental state right now. For desires the scope of error is larger : one can be mistaken right now about whether one’s desires have been fulfilled (unless those desires are for pleasures). Given Pollyannianism, the error will tend towards the positive in self-assessments of well-being.

Benatar notes that our lives contain very few satisfied desires, and a lot of unsatisfied ones. First, he considers the desires for what we lack :

Such a desire must be present before it can be satisfied and thus we endure a period of frustration before the desire is fulfilled. It is logically possible for desires to be fulfilled very soon after they arise, but given the way the world is, this does not usually happen. Instead, we usually persist in a state of desire for a period of time. This time may vary—from minutes to decades. As I said before, one usually waits at least a couple of hours until hunger is satiated (unless one is on a ‘hunger-prevention’ or a ‘nip-hunger-in-the-bud’ diet). One waits still longer to get rest when one is tired. Children wait years to gain independence. Adolescents and adults can wait years to fulfil desires for personal satisfaction or professional success. Where one’s desires are fulfilled, this fulfilment is often ephemeral. One desires public office and is elected but not re-elected. One’s desire to be married is eventually fulfilled, but then one gets divorced. One wants a holiday but it ends (too soon). Often one’s desires are never fulfilled. One yearns to be free, but dies incarcerated or oppressed. One seeks wisdom but never attains it. One hankers after being beautiful but is congenitally and irreversibly ugly. One aspires to great wealth and influence, but remains poor and impotent all one’s life. One has a desire not to believe falsehoods, but unknowingly clings to such beliefs all one’s life. Very few people ever attain the kind of control over their lives and circumstances that they would like.

Second, the desires to not lose what we already have. These desires are already fulfilled by definition, but the fulfilment doesn't last. One for example desire to not lose health and youth, but this loss happens all too quickly. As if that's not bad enough, fulfilling a desire often makes another one arise. Professor Maslow noted that :

[...] need gratifications lead only to temporary happiness which in turn tends to be succeeded by another and (hopefully) higher discontent. It looks as if the human hope for eternal happiness can never be fulfilled. Certainly happiness does come and is obtainable and is real. But it looks as if we must accept its intrinsic transience, especially if we focus on its more intense forms.

Schopenauer also noted that life is constantly striving and willing (a state of dissatisfaction). Once a desire is satisfied, another one arise. He also claims that there are no intrinsic pleasures, only reliefs from suffering. But one does not need to go as far to agree with Benatar that "dissatisfaction does and must pervade life".

An optimist objection would be to say that there is something valueable in the striving itself :

There are two ways of making sense of this objection on a desire-fulfilment view. The first is to say that in addition to desiring whatever we desire, we desire striving to fulfil that desire. Thus the striving fulfils a desire on the way to fulfilling another desire. As a result our desires are not as unsatisfied as I have suggested they are. The second way of making sense of the objection is to say that whether or not we desire the period of deprivation or the striving to fulfil our (other) desires, such a precursor to desire fulfilment makes the eventual fulfilment that much sweeter.

On the first interpretation, he responds by saying that some do enjoys the journey to satisfaction but not everyone : some writers enjoy having created their poem/book, but hate the process. Some gardeners hate gardening but do so to be able to eat. People desire to get cured from cancer, but nobody would desire the process, treatments and side-effects.

On the second, he also agree that some desires are like that (food taste sweeter when hungry, winning the race feels better if one trained), but again it can't be true for all desires. And even among those of which it is true, for at least some of them it would be better if no striving was necessary :

Freedom may be valued more if it were long desired or the result of a protracted struggle, but it would still be better not to have been deprived of freedom all that time. Long incarceration followed by freedom simply is not better than lifelong freedom. In other words, we must not mistakenly think that the redeeming features of deprivation and striving are actually advantages over more rapid desire fulfilment.

So far Benatar only showed how much unfulfilment there is. He also claim that people overestimate the significance of the fufilment that do exist :

On desire-fulfilment views, our lives go well to the extent that our desires are satisfied. However, the state of having one’s desires fulfilled can be attained in one of two ways :

(a) having fulfilled whatever desires one has, or

(b) having only those desires that will be fulfilled.

A crude desire-fulfilment view will not distinguish between these two ways of having one’s desires fulfilled. The problem with such a crude view is that a terrible life could be transformed into a splendid one by expunging desires or by altering what one desires. If, for instance, one came to desire the various features of one’s doleful existence, one’s life would thereby transmute from the miserable to the magnificent. This is hard to swallow. It might seem (or feel) as though one’s life had so improved, but it surely would not actually have been so transformed (even though it would actually have improved in a more limited way by feeling better). The question is whether a more plausible version of a desire-fulfilment view can be constructed—one that judges a life to go better when desire-fulfilment is attained via (a) rather than (b). As this is a problem internal to desire-fulfilment theory, I shall not pursue this question. Suffice it to say that if such a version of the desire-fulfilment view cannot be formulated, then so much the worse for the desire-fulfilment view. But what if such a version could be constructed? We would then need to notice that because of the psychological phenomena I have outlined, (b) would account for much of our desire-fulfilment. Our desires are formulated and shaped in response to the limits of our situation. Therefore, our lives are much worse than they would be if our desire-fulfilment were exclusively (or even primarily) attributable to (a).

Objective list theories

Here too there is some overlap with the previous theories. Pleasurable mental states, the absence of painful ones, and the fulfilment of some desires must be on the list of objective goods.

The problem is that, just like our desires adapt to circumstances, what we would put on an objective list will be determined by the limits of what we can expect. The lists are not truly from an objective perspective (sub specie aeternitatis), but from a human one (sub specie humanitatis) :

For instance, since none of us lives until age 240, people tend not to think that failing to reach that age makes one’s life go less well. However, most people regard it as tragic when somebody dies at forty (at least if that person’s quality of life was comparatively good). But why should a death at ninety not be tragic if a death at forty is? The only answer can be that our judgement is constrained by our circumstances. We do not take that which is beyond our reach as something that would be a crucial good. But why must it be that the good life is within our reach? Perhaps the good life is something that is impossible to attain. It certainly sounds as though a life that is devoid of any discomfort, pain, suffering, distress, stress, anxiety, frustration, and boredom, that lasts for much longer than ninety years, and that is filled with much more of what is good would be better than the sort of life the luckiest humans have. Why then do we not judge our lives in terms of that (impossible) standard?

Another good candidate for the objective list is the meaning of life. Most people would say life has no meaning, looking at it from the cosmic perspective. Other might object and say that life has meaning, looking at it from the human perspective :

A life devoted to the service of humanity, for example, can be meaningful, sub specie humanitatis, even if, from the perspective of the universe, it is not. Other lives, though, such as that of the man who devotes his life to counting the number of blades of grass on different lawns, would lack meaning sub specie humanitatis. The grass-counter’s life could be meaningful, however, from his own subjective perspective, if he derived satisfaction from his unusual life’s project. That his life could be meaningful from this perspective leads many people to think that the subjective perspective is unsatisfactory. But why should we think that the perspective of humanity is any more reliable than the perspective of the individual? From the perspective of the universe, the lives of both the philanthropist and the grass-counter are meaningless (which is not to say that philanthropy is not better than grass counting).

A world of suffering

Even if optimists reject the pessimistic view of ordinary healthy life, they are on weak grounds when we consider all the suffering the world contains (again Benatar focus on humans, but it's even larger if you consider non-human suffering). He cites numbers of deaths from natural disasters, hunger (and those who suffers from hunger/malnutrition without dying), diseases, accidents... Then add the intentional killings and suffering, including obviously wars, but also violence from private citizens : rape, murder, children abuse etc... Sometimes the suffering is so great that people take their own lives. Such numbers makes procreation a kind of gambling with someone else welfare :

A charmed life is so rare that for every one such life there are millions of wretched lives. Some know that their baby will be among the unfortunate. Nobody knows, however, that their baby will be one of the allegedly lucky few. Great suffering could await any person that is brought into existence. Even the most privileged people could give birth to a child that will suffer unbearably, be raped, assaulted, or be murdered brutally. The optimist surely bears the burden of justifying this procreational Russian roulette. Given that there are no real advantages over never existing for those who are brought into existence, it is hard to see how the significant risk of serious harm could be justified. If we count not only the unusually severe harms that anybody could endure, but also the quite routine ones of ordinary human life, then we find that matters are still worse for cheery procreators. It shows that they play Russian roulette with a fully loaded gun—aimed, of course, not at their own heads, but at those of their future offspring.

Chapter 4 - Having Children: The Anti-Natal View


No duty to proceate

Some believe that there is a duty to procreate, wich have various scope and justification :

(1) Scope: A duty to procreate can be understood as

(a) a duty to have some children

(b) a duty to have as many children as one can

(2) Justification: Purported duties to procreate could be based on

(a) the interests of those brought into existence

(b) other considerations, such as the interests of others, utility, divine commands, and so on

Benatar's arguments are a fatal blow to the 2.a category. It also challenges the 2.b category, but do not logically prevent them, because one could agree that coming into existence is a harm, and yet say that the considerations that ground the duty outweigh the harm. But if the harm is as severe as Benatar suggested, then it's highly implausible that the considerations outweight it, even more so in the 1.b cases.

Is there a duty not to procreate?

Some people think there are sometimes a duty not to procreate (when the life of the child will be unusually bad), but Benatar's question is about all possible beings. He considers the cost of non-procreation by making a distinction between three kinds of interests the potential parents have :

Obviously non-procreation comes at the cost of frustrating procreative interests. Coital interests can be fullfiled without procreation via contraceptive methods, which is a small effort to do in order to prevent the harm of coming into existence. Parenting interests can be fulfiled by adoption, as long as there are unwanted children.

Children cannot be brought into the world for their own sake's, even if procreators think so and use it as a reason. Benatar suspects that most people who decide to have a child do it to serve their own procreative and related interests (and sometimes the interests of other, such as the nation or the famillies of the procreators). Serving one’s own interests is not always bad, but where doing so inflicts significant harm on others, it is usually not justified.

A way to defend having children even if accepting that coming into existence is a harm would be to deny that the harm is great. And even if one agree that the harm is great, one could point to the fact that most poeple do not regret being born :

In defence of the permissibility of having children, it might be suggested that it is morally significant that most people whose lives go relatively well do not see their having come into existence as a harm. They do not regret having come into existence. My arguments suggest that these views may be less than rational, but that, it might be suggested, does not rob them of all their moral significance. Because most people who live (relatively) comfortable lives are happy to have come into existence, prospective parents of such people are justified in assuming that if they have children they too will feel this way. Given that it is not possible to obtain consent from people prior to their existence to bring them into existence, this presumption might play a key role in a justification for having children. Where we can presume that those whom we bring into existence will not mind that we do, we are entitled, the argument might go, to give expression to our procreational and other interests. Where these interests can be met by having either a child with a relatively good life or a child with a relatively bad life it would be wrong if the parents brought the latter into existence, even where that child would also not regret its existence. This is because if the prospective parents are to satisfy their procreational interests they must do so with as little cost as possible. The less bad the life they bring into being, the less the cost. Some costs (such as where the offspring would lead a sub-minimally decent life) are so great that they would always override the parents’ interests.

Those cases in which the offspring turn out to regret their existence are exceedingly tragic, but where parents cannot reasonably foresee this, we cannot say, the argument would suggest, that they do wrong to follow their important interests in having children. Things would be quite different, according to this argument, if the majority or even a sizeable minority of people regretted coming into existence. Under such circumstances the above justification for having children certainly would be doomed. However, given that most people do not regret their having come into existence, does the argument work?

The argument is problematic :

Its form has been widely criticized in other contexts, because of its inability to rule out those harmful interferences in people’s lives (such as indoctrination) that effect a subsequent endorsement of the interference. Coming to endorse the views one is indoctrinated to hold is one form of adaptive preference—where an interference comes to be endorsed. However, there are other kinds of adaptive preference of which we are also suspicious. Desired goods that prove unattainable can cease to be desired (‘sour grapes’). The reverse is also true. It is not uncommon for people to find themselves in unfortunate circumstances (being forced to feed on lemons) and adapt their preferences to suit their predicament (‘sweet lemons’). If coming into existence is as great a harm as I have suggested, and if that is a heavy psychological burden to bear, then it is quite possible that we could be engaged in a mass self-deception about how wonderful things are for us. If that is so, then it might not matter, contrary to what is claimed by the procreative argument just sketched, that most people do not regret their having come into existence. Armed with a strong argument for the harmfulness of slavery, we would not take slaves’ endorsement of their enslavement as a justification for their enslavement, particularly if we could point to some rationally questionable psychological phenomenon that explained the slaves’ contentment. If that is so, and if coming into existence is as great a harm as I have argued it is, then we should not take the widespread contentment with having come into existence as a justification for having children.

It might be objected that a duty not to procreate is too demanding a requirement. Earlier we saw that many people would agree that there is sometimes a duty not to procreate, namely when the offspring would suffer terribly :

In such cases many people are prepared to admit that it would be wrong to have a child. But notice that the burden for those who must desist from producing progeny in such cases is no lighter than the burden any other potential human breeder faces in forgoing having children. If the burden is not too great for the former, then it ought not to be too great for the latter. Where the difference is thought to lie is in the quality of life of the offspring. In other words, it is thought that non-procreation is required of those whose children’s quality of life would be unacceptably low, but we cannot make such a demand of those whose offspring would be ‘normal’. Notice, however, that this is not an argument about the magnitude of the burden of non-procreation, but rather one about when that burden may be imposed. I could accept that non-procreation should only be required when the children produced would lead very poor quality lives. This is because I have argued that all lives fall into this category. Those who think that a few lives do not fall into this category are not (much) better equipped to defend the objection that non-procreation is too demanding. This is because they must surely be moved by the fact that we cannot tell, when we deliberate about whether to bring somebody into existence, whether that life will be one of the few lives that is not of a very poor quality. It seems, then, that those who accept that coming into existence is a great harm and that there is a duty not to procreate in cases where the offspring would suffer great harm by being brought into existence must accept that a duty not to procreate is not too demanding.

If I am mistaken about this, however, and it is not immoral to have children, my arguments in Chapters 2 and 3 show, at the very least, that it is preferable not to have children. Although our potential offspring may not regret coming into existence, they certainly would not regret not coming into existence. Since it is actually very much not in their interests to come into being, the morally desirable course of action is to ensure that they do not.

Procreative freedom

Understanding the purported right

Benatar understand the right to procreative freedom as a negative right to choose whether or not to produce children. It's clear that this right, in its moral form, is incompatible with the duty not to procreate. However, there could be a legal right to do so, as legal rights give people the liberty to do things that may be, or regarded to be wrong.

Grounding the right on autonomy

The legal right can be grounded on the basis of autonomy, but a legal right to have children is not an absolute entitlement, it is a very strong presumption in favour of having children. Presumptions can be defeated, and as John Robertson said : "those who would limit procreative choice have the burden of showing that the reproductive actions at issue would create such substantial harm that they could justifiably be limited". If we accept that coming into existence is always a harm, then the presumption is always defeated, and a right that is always defeated is not really a right.

Grounding the right on futility

Denying the right would have to be linked to state enforcment, and this would be unpractical and too costly. The state would have to be highly intrusive, and be able to distinguish between people who procreate wittingly or negligently, and those who do so accidentally. In both cases, the state will have to require abortion. All of this will drive pregnancy underground, making it unsafe. Those moral costs are too much to pay.

Grounding the right on disagreement

We can imagine a society in which non-procreation is ensured without the invasions of privacy and bodily intrusions mentionned above. For example the state could release a contraceptive subtance in the air or the drinking water. In such a society, one could still defend the right to procreate from the fact that Benatar's views are highly contested :

The view that coming into existence is always a harm is highly contested. Even if this view is nonetheless correct, the mere fact that it enjoys so little support shows that ordinary people can disagree about this. And where such disagreement exists about whether some action is (unjustifiably) harmful, the state should grant people the right to choose whether or not to engage in such actions. This argument suggests a qualification of the famous harm principle. According to this principle, in its unqualified form, states may prohibit activities only when they harm non-consenting parties. The qualification states that cases in which there is disagreement between ordinary people about whether some action is harmful do not fall within the scope of the principle.

Benatar replies by making an analogy with slavery :

Consider the case of slavery in a slave-owning society—or at least that sort of slave-owning society where slavery is defended by a view that slaves are naturally suited to being slaves. In such a society we find large numbers of people who believe that slavery does not harm slaves. Indeed they might even believe that slavery benefits slaves. They might listen to the arguments of a few abolitionists that slavery is harmful to slaves, and reply that their claims are highly contested and thus exempted from the harm principle. Although that conclusion would be accepted readily by slavery’s defenders, it is quite clear that neither the abolitionists in that society nor we who have the benefit of some temporal or geographic distance from slavery would be impressed by this argument. Surely there ought not to be a legal right to own slaves even when slavery’s harmfulness is contested? This shows that the mere fact that an activity’s status as harmful is contested does not show that the harm principle is inapplicable or that people should have a legal right to engage in it.

Grounding the right on reasonable disagreement

We can draw a further distinction between reasonable and unreasonable disagrement :

For a disagreement to be reasonable the reasons for one view on a matter must not be sufficiently stronger than reasons for conflicting views that accepting one of the conflicting views is actually (rather than merely perceived to be) unreasonable. Now the problem with this standard is that of differentiating between disagreements that are actually unreasonable and those that are only perceived to be so.

Benatar thinks that there can be no reasonable disagrements about the harmfulness of chattel slavery. But what about a more controversial case like meat eating? His conclusion :

There is then, for now, a strong case for recognizing a legal right to reproductive freedom. I can envisage circumstances in which this case would crumble, but on account of the strong biases toward rejecting the view that coming into existence is a harm, the case would be in utter ruins before a rights-respecting state withdrew a legal right to reproduce. If matters were that clear, the loss of such a right would not be regrettable. Indeed, the loss of such a right in liberal societies will cease to be regrettable long before it ever is lost—if it is ever lost in a society that remains liberal. In the interim, we can defend a legal right to produce children even while thinking that there is a moral duty not to bring new people into existence.

Assisted and artificial reproduction

Benatar summarize the interactions between reproductive ethics and sexual ethics :

Reproductive ethics

Sexual ethics

He obviously espouse the anti-reproductive view in both cases. Those views blame not only the reproducing, but also those who assist them : it is wrong to help somebody inflict the harm of bringing somebody into existence. However, even if it's morally wrong, there is still a legal right to do so, derived from the negative legal right of reproducing freedom the patients have. But this does not imply, for the patients, a positive legal right to medical assistance for reproduction.

Treating future people as mere means

Some cases of reproduction are opposed because of the Kantian requirement not to treat people merely as a means. For example, having a second child in order to save the first child (bone marrow transplant, in the case of leukemia), or cloning. But Benatar points out that this apply to all ordinary cases of reproduction :

Children are brought into existence not in acts of great altruism, designed to bring the benefit of life to some pitiful non-being suspended in the metaphysical void and thereby denied the joys of life.

It might be suggested that cloning is worse, because it is an act of narcissism (wanting an exact replica of yourself). But parents who choose adoption could say the same to parents who reproduce :

They could argue that it is narcissistic for a couple to want to create a child in their combined image, from a mixture of their genes. The point is that both cloning and usual methods of reproduction may be narcissistic, but neither is it the case that each kind of reproduction must necessarily be characterized in this way.

Cloning, therefore, is no more or no less problematic than regular reproduction in this regard. But having a child in order to save a child is less problematic than ordinary cases of reproduction. In ordinary reproduction people produce children (a) to satisfy their procreative or parenting interests; (b) to provide siblings to existing children; (c) to propagate the species, nation, tribe, or family; or (d) for no reason at all. These are all clearly weaker reasons for producing a child than is the goal of saving the life of an existing person.

Chapter 5 - Abortion: The ‘Pro-Death’ View

Most people think that a reason must be provided for having or procuring an abortion. Some pro-choice think it's a regrettable act even when justified. The burden must be reversed if :

(1) coming into existence is a harm, and (2) somebody has not yet come into existence at the particular stage of gestation when the abortion is to be performed. If both of these conditions are met, then the burden of proof is shifted to those who would not abort (at the specified stage of gestation). The failure to abort is what must be defended. The greater the harm of existence, the harder it will be to defend that failure. If a third condition is met—(3) coming into existence (in ordinary cases) is as great a harm as I have suggested it is—then the failure to abort (at the specified gestational age) may never, or almost never, be justified.

Benatar distinguish between coming into existence in a biological sense and in a morally relevant sense. The former mean the beginning of a new organism, and the latter mean the beginning of an entity’s morally relevant interests. We didn't come into existence in a biological sense prior to conception, since both a sperm and an ovum are necessary, and two cannot equal one. In the case of twins, it might be later than conception. But this question can be bypassed if one is interested in the morally relevant sense and if one comes into existence in the morally relevant sense after even the latest reasonable estimate of when one comes into existence in the biological sense, which is the view of Benatar. To understand when we come into existence in the morally relevant sense, he examines different kinds of interests.

Four kinds of interests

1. Functional interests: The first sort of interest is that which an artefact, such as a car or a computer, is sometimes said to have. Because artefacts have functions, some things can promote and others impede those functions. Those things that facilitate an artefact’s functioning are said to be good for the artefact, or to be in its interests, and those things that compromise its functioning are said to be bad for it, or against its interests. Thus, rust is bad for a car and having wheels is good.

2. Biotic interests: Plants have a different kind of interest. Like artefacts, they function, and their functioning can be fostered or impaired. However, unlike artefacts, plants are alive. Their functions and associated interests are biotic.

3. Conscious interest: Conscious animals also function, and as with plants, their functions are biological. But there is something that it feels like to be a conscious being. The associated interests I shall call conscious interests. But this term requires clarification. By conscious interest I do not mean an interest that one consciously has—that one is explicitly aware of having—but rather an interest that can only be had by those who are conscious. One may, for instance, have an interest in avoiding pain, without being aware that one has such an interest.

4. Reflective interests: Some animals—typically most humans— are not only conscious, but are also characterized by various higher-order cognitive capacities, including self-awareness, language, symbolization, and abstract reasoning. Such anim- als are not only conscious, but are also ‘reflective’. They have interests in the reflective sense that they can be explicitly interested in their interests.

They are incremental (machines have 1, plants have 1 and 2, animals have 1, 2 and 3, humans have all four). Having interests is necessary for having moral standing, but not sufficent. The question is which interests are relevant.

Which interests are morally considerable ?

To Benatar, it's hard to make a compeling argument for a particular kind of interest. Such arguments will rely heavily on intuitions. This is impacting his strategy :

[...] my argument strategy will be to point out the implications of regarding biotic interests as morally relevant and to show that most if not all pro-lifers do not embrace them.

As I shall show later, fetuses only become conscious quite late in the gestational period. Thus, if conscious interests are the most basic morally relevant interests, fetuses will acquire morally relevant interests only very late. One way of grounding a pro-life argument would be to claim that biotic interests are also morally relevant. However, if biotic interests count morally, a principle of equality would require that equal biotic interests count equally. Thus, it cannot be only human biotic interests that are relevant. The interests of plants, bacteria, viruses, and so forth must count as much as the biotic interests of human embryos and preconscious fetuses. But those are implications that very few (if any) pro-lifers will embrace. Consistency requires, then, that they do not ground their view on a claim to the moral relevance of biotic interests. (Of course, this does not mean that there is no other way to support a pro-life position, and I shall consider some other arguments later.)

Those who take biotic interests to be morally relevant do not deny that conscious interests are also relevant. They object only to setting the threshold of relevance above biotic interests. There is another challenge to those who take conscious interests to be the minimum morally relevant interests. This challenge comes from those who would set the threshold above conscious interests—at the level of reflective interests. The implications of this view are also implausible. If only reflective interests count morally, then there can be nothing (directly) wrong with torturing beings that are conscious but not self-conscious—most animals and all human neonates. We can reject the view that only reflective interests count.

When does consciousness begin?

For Benatar, the scientific litterature about fetal pain indicate the following conclusion :

[...] there is non-negligible evidence to think that from around twenty-eight to thirty weeks gestational age, fetuses are conscious, at least in some minimal sense. Given the evidence and the gradual nature of the developmental process, it is highly unlikely that the earliest manifestation of consciousness is fully formed. It is much more likely that the level of consciousness evolves. Indeed, in humans, consciousness also gradually develops into self-consciousness. Thus, conscious interests do not suddenly arise. Instead they emerge gradually, even if not at a constant pace.

Interests in continued existence

If we come into existence around that time, and that is better never to have been, then it is better to be aborted before that time. But it does not follow that abortion after that is morally wrong. One could think it's wrong to inflict pain on a consious entity, but not wrong to kill it painlessly. Michael Tooley think so, and has an argument :

  1. The statement ‘A has a right to continue to exist as a subject of experiences and other mental states’ is roughly synonymous with the statement ‘A is a subject of experiences and other mental states, A is capable of desiring to continue to exist as a subject of experiences and other mental states, and if A does desire to continue to exist as such an entity, then others are under a prima facie obligation not to prevent him from doing so.’
  2. To have a desire is to want a certain proposition to be true.
  3. To want a proposition to be true one must understand that proposition.
  4. One cannot understand a given proposition unless one has the concepts involved in it.
  5. Therefore, the desires one can have are limited by the concepts one possesses.
  6. Neither a fetus (at any stage of its development) nor a young infant can have concepts of itself as a subject of experiences and other mental states.
  7. Therefore, neither a fetus nor an infant can have a right to continue to exist.

Benatar ponders on the argument and reject it, but think there is a kernel of truth to it :

To say, as I have suggested, that a minimally conscious entity can have an interest in continued existence is not to say that that interest is anything like as strong as that of a self-conscious entity. Where the interest in continued living is derivative from quite rudimentary interests in further pleasurable experiences, it is much weaker than it becomes when self-consciousness and projects and goals emerge. Then the being is much more invested in its own life and stands to lose much more by dying or being killed. That the earliest interests are weak ones, however, does not mean that they are not interests at all.

He espouse a gradual view of moral intersts, which makes sense because the properties that are relevant for interests (such as consciousness and self-consciousness) arise gradually. Two famous arguments threaten this gradual view :

The Golden Rule

Richard Hare employed the Golden Rule against abortion. In its positive form the rule states : "we should do unto others as we would have them do to us". Hare extends this to "we should do to others as we are glad was done to us". Since we are glad that nobody terminated the pregnancy that resulted in our birth, we should do the same for the birth of a person having a life like ours.

For Benatar, it is not true that everybody is glad not to have been aborted. Hare responds that they must wish that had they been glad to have been born, then nobody should have aborted them. But the problem is that it takes the preference to have been born as the standard/moral touchstone :

Had he taken the preference not to have been born as the standard, then it could be said that those who are glad to have been born must wish that had they not been glad, then somebody should have aborted them. It is obvious that had either kind of person had the opposite preference to the one he does have, the Golden Rule argument would produce the opposite conclusion to the one it does produce when his preference is the way it actually is. Thus Professor’s Hare’s response to the case of those who are not glad to have been born will not do.

Which preference should prevail? Some argue that most fetus grow into people with the preference to have been born, so this preference is more statistically reliable. Benatar gives two reasons why it should not prevail.

The first one is the principle of caution. If one is mistaken in the assumed preference to have been born, the person may not actually have this preference, and will suffer a lifetime. If one is mistaken in the opposite preference and abort the featus, nobody will suffer. A reply can be made that the fœtus does suffer from it but 1) Hare does not think that they suffer when aborted and 2) :

The entire point of an argument from potentiality is to show that abortion can be wrong even if the fetus does not, as a fetus, have properties that are reasons for not killing it.

The second one is that coming into existence is always a serious harm. So the preference to have been born is uninformed, and so it's odd to use the Golden Rule with it.

It also has benn noted that the first premiss (the extention of the Golden Rule) is false :

There is a difference between being glad that somebody did something for one and thinking that that person was obligated to have done as he did. Not everything that we might wish to be done (or are glad was done) is something that we think should be done (or should have been done). We can wish to be treated in ways that we recognize others are not duty-bound to treat us (or we them).

A "future like ours"

Don Marquis's argument starts with the assumption that killing adults is wrong :

The best explanation of why this is wrong, he says, is that the loss of one’s life deprives one of the value of one’s future. When one is killed one is deprived of all future pleasures, and of the ability to pursue one’s present and future goals and projects. But most fetuses have a valuable future like ours. Thus, concludes Professor Marquis, it must also be wrong to kill these fetuses.

He notes the virtues of the argument :

  1. It avoid speciesm : If there are non-human animals who also have valuable futures then it would be wrong to kill them too. And it would not necessarily be wrong to kill those humans, including fetuses, whose future quality of life promises to be so poor that they do not have valuable futures.
  2. It avoids problems arising from the personhood criterion. On this view, killing children and infants is obviously wrong, just like killing adults.
  3. It does not say that abortion is wrong because it involves the killing of potential persons. Such arguments are unable to explain why potential persons are entitled to the same treatment as actual persons. The future-like-ours argument is based on an actual property of the fetus—that it has a future like ours—rather than some potential property.

He then consider two alternative views, that he rejects. On the desire account, killing is wrong because it thwarts the important desire people have to continue living. But it can't explain why it's wrong to kill depressives or people in a coma. One the discontinuation account, killing us is wrong because it involves the discontinuation of the valuable experiences, activities, and projects of living. Fetuses have none of these, so abortion before the development of consciousness would not be wrong on the discontinuation account.

Don Marquis says that it cannot be the mere discontinuation of experiences that is wrong. If the future experiences will be ones of unmitigated suffering, discontinuation may actually be preferable. Thus, the discontinuation account cannot work unless it refers to the value of the experiences that may be discontinued. Moreover, he says, the nature of the immediately past experiences of a person are not relevant. It makes no difference, he says, whether a person has been in intolerable pain, has been in a coma, or has been enjoying a life of value. He concludes that it is only the value of the person’s future that matters. If that is so, he says, then the discontinuation account must collapse into a future-like-ours account.

But for Benatar this is too quick :

A discontinuation account may say that, although the value of the future is necessary for explaining the wrong of killing (those with a valuable future), it is not sufficient. Such an account may hold that only a being with morally relevant interests can have a morally relevant interest in its valuable future. Thus it is the discontinuation of a life of a being that already has morally relevant interests that is wrong. In other words, for killing to be wrong, the future must be a valuable one, but it must also be the future of a being that already counts morally.

Now, Don Marquis might respond that all entities with a future like ours do, in virtue of having such a future, have a morally considerable interest—the interest in enjoying that future.

But why does he think that a human fœtus now has such an interest? His answer seems to be that we can by then uniquely identify the entity that will later enjoy that future. We can't choose non-arbitrarly a subject deprived of the future before that stage, which is why the argument doesn't rule out contraception.

Benatar talks about an odd implications of this view : if the genetic material was entirerly in the sperm and that the ovum was just feeding it, then contraception would be ruled out :

However, it is hard to see how that can make a difference to whether contraception is morally akin to murder. How, in other words, can mere genetic individuation make all the difference between whether or not it is wrong to prevent a future like ours? If it really is a future like ours that counts, then why should it be the future of genetically complete organisms? My alternative proposal, which avoids the odd implication of this, is that what counts is the valuable future of those entities with morally relevant interests. Discontinuing the valuable life of a being with morally relevant interests in that life is (prima facie) wrong.

Another advantage of Benatar's account over the future-like-ours account is that if the value of the future is all that counts, then killing a fœtus would be worse than to kill a 30 years old, since the fœtus has a longer future :

The greater deprivation makes sense when we are comparing the death of a thirty-year old with that of a nonagenarian, where most people take the former to be worse. However, it makes much less sense when comparing the deaths of the fetus and the thirty-year old, where many of us take the latter to be much worse. The best explanation for this is that a fetus has not yet acquired the interest in its own existence that the thirty-year old has. The case of the thirty-year old and the nonagenarian can be explained in one of two ways. It could be that both have equal interests in continued life but the nonagenarian has less life left. Alternatively, in some cases only, it could be that the nonagenarian’s interest in living has already begun to decline, perhaps on account of life’s becoming worse with advancing age and decrepitude.


Benatar adopts what he calls the "pro-death view" :

Combining the view that fetuses lack moral standing in the earlier stages of pregnancy with the view that it is always a harm to come into existence turns the prevailing presumptions about abortion on their head. Instead of a presumption in favour of continuing pregnancy, we should adopt a presumption, at least in the earlier stages of pregnancy, against carrying a fetus to term.

Such a view might be of interest to legal pro-lifers :

One of its valuable features is that it offers a unique challenge to those pro-lifers who reject a legal right to abortion. Whereas a legal pro-choice position does not require a pro-lifer to have an abortion—it allows a choice—a legal pro-life position does prevent a pro-choicer from having an abortion. Those who think that the law should embody the pro-life position might want to ask themselves what they would say about a lobby group that, contrary to my arguments in Chapter 4 but in accordance with pro-lifers’ commitment to the restriction of procreative freedom, recommended that the law become pro-death. A legal pro-death policy would require even pro-lifers to have abortions. Faced with this idea, legal pro-lifers might have a newfound interest in the value of choice.

Chapter 6 - Population and extinction

Central questions to these two topics are :

  1. "How many people should there be?" Benatar will, unsurprisingly, defend the case for "zero".
  2. "Is the prospect of human extinction something to be regretted?" :

    I shall answer that although the process of extinction may be regrettable, and although the prospect of human extinction may, in some ways, be bad for us, it would be better, all things considered, if there were no more people (and indeed no more conscious life).

  3. "Given the fact of extinction, should it come earlier or later?" :

    Here I shall argue that although very imminent extinction would be worse for us, earlier extinction nonetheless would be better than later extinction. This is because earlier extinction guarantees against the significant harm of future lives that would otherwise be started. I shall show, however, that on some views, the creation of a limited number of new people may be justified. If that is so, then although extinction need not come as early as it could it should still come earlier rather than later.


There is a distinction between two questions : how many people should there be at a given time/period, and how many people should there be in total (the cumulative question). Since for Benatar, coming into existence is a serious harm, a cumulative population of only one person is already overpopulation.

The cumulative population is more than 106 billion people, according to a 2004 statistic.

Although it would have been better had none of the more than 106 billion people come into existence, these people (among whom you and I are included) can no longer be prevented. For this reason, many might wish to focus on the question how many more people there may be—in the cumulative sense, rather than at some specified time in the future. The ideal answer here is again ‘zero’, although that ideal is being violated about every second.

Solving problems in moral theory about population

Professor Parfit’s population problems

In Reasons and Persons Derek Parfit discuss the non-identity problem. How do we explain the common judgment that starting a life of poor quality is wrong? It can't be explained with what Parfit calls "person-affecting" views of morality. Such views imply that an action is bad if people are affected for the worse.

Such a view, he says, cannot solve the non-identity problem because in non-identity cases those who are brought into existence cannot be worse off than they would otherwise have been, because they would not otherwise have been.

Parfit goes on the hunt for Theory X, that must solve the non-identity problem but without generating other problems. He considers impersonal views of morality :

Whereas person-affecting views maintain that something can only be bad if it is worse for somebody, impersonal views are not concerned about the effects on particular people. Instead they examine outcomes more impersonally. If people’s lives go better in one possible outcome than the other, then the better outcome is to be preferred even though nobody is better off in that scenario.

It solves the non-identity problem, but generates other problems. Benatar distinguish between two kinds of impersonal views :

On the impersonal total view, a small population with a high quality of life is worse than a larger population with a smaller quality of life, so long as there are enough people in the larger population to outweigh the lower quality of life. This is because only the total matters, not the average.

Figure 6.1

Parfit suggests that the conclusion is repugnant. The impersonal average view avoid this conclusion, but it cannot be Theory X, since it comes with its own problems :

To show why this is the case, Derek Parfit asks us to imagine another two worlds. In the first world everybody had a very high quality of life. In the second world, in addition to all these people with their same high quality of life there are additional people who, although not quite as well off, nonetheless have lives that are well worth living. These sorts of cases, Derek Parfit calls ‘Mere Addition’. More specifically, mere addition occurs ‘when, in one of two outcomes, there exist extra people (1) who have lives worth living, (2) who affect no one else, and (3) whose existence does not involve social injustice’.

The second world is made worse by the addition of people, Parfit takes this to be implausible.

Why anti-natalism is compatible with Theory X

In chapter 2 Benatar argued that the non-identity problem can be solved. One of the way to do it is to say that even if coming into existence is not worse, it may still be bad for the person who comes into existence. Since the alternative is not bad we can say that the person is thereby harmed. This doesn't work with the person-affecting view, understood as the view that something ‘is bad if people are affected for the worse’. But that first formulation is more restrictive than needed. Later in his book, Derek Parfit distinguish between the narrow and the wide versions of the person-affecting view :

The Narrow Person-Affecting View : An outcome is worse for people (in the narrow sense) if the occurrence of X rather than Y would be either worse, or bad, for the X-people.

He believes this formulation cannot solve the non-identity problem because he thinks it is not bad for people to come into existence so long as they have a life worth living. However, we saw that the phrase ‘a life worth living’ is ambiguous. It might mean either ‘a life worth starting’ or ‘a life worth continuing’. Considering the argument that coming into existence is always a harm, it follows that no lives are worth starting (even if some lives are worth continuing). Thus coming into existence is always bad for a person even if one thinks that it is not worse for that person.

Benatar turns to explain how his anti-natalist arguments affects the impersonal views. First, the impersonal total view :

Others might suggest that if the impersonal total view takes account of my argument it can avoid the repugnant conclusion. My argument may be seen to explain that the repugnant conclusion arises because of the mistaken assumption that it is good to have extra lives that are worth continuing. The impersonal total view can be revised to avoid both this mistake and the resultant conclusion. One way in which this can be done is to restrict the scope of the impersonal total view in such a way that it applies only to people who do exist or will anyway exist and not to questions about how many people should exist. In other words, it can be seen as a principle to maximize the happiness of the existent, but not to affect the number of existers. This revision, however, comes at an obvious cost. The revised view ceases to provide guidance on how many people there should be.

Then, the impersonal average view :

[...] the impersonal average view says that we should not add extra lives if they lower the average well-being of all humans who have ever lived. The implication that additional lives ‘worth living’—read ‘worth continuing’—should not be added is taken to be implausible. My arguments, however, show that it is not. If no lives are worth starting, it is not a defect in a theory that it precludes adding new lives that are not worth starting, even if those lives would be worth continuing. It would indeed have been better if no people had been added to the Edenic lives of Adam and Eve.

But on this view, we would be obliged to start new lives if it would raise the average quality of life of all people who have every lived. This is at odd with Benatar's conclusions. Both impersonal views make the mistake of valuing people only to the extent that they increase (total or average) happiness. They mistakenly assume that the value of happiness is primary and the value of persons is derivative from this. However, it is not the case that people are valuable because they add extra happiness. Instead extra happiness is valuable because it is good for people.

To avoid that mistake, impersonal views can be modified to not seek the greatest total or average happiness but rather the smallest total or average unhappiness :

This way of revising the impersonal views has two advantages. First, it preserves the impersonal view’s ability to provide guidance on how many people there should be. Secondly, it generates the conclusion for which I have argued—namely that the ideal population size is ‘zero’. The way to minimize unhappiness is for there to be no people (or other conscious beings). The lowest total unhappiness and the lowest average unhappiness are both zero unhappiness, and zero unhappiness, at least in the real world, is achieved by having zero people.

Those who want to resurrect the repugnant conclusion and mere addition problems by imagining a world in which no lives contain any bad but differ only in how much good they contain face a number of problems :

First problem :

[...] it is not clear that we can even make sense of such a world, given the interaction between the good and bad in a life. As I showed in Chapter 3, a life that contained very little good would have to contain some bad—namely the tedium of long stretches of absent goods. The only way this could be avoided would be if the life’s duration were shortened, but the shortening of life is another bad.

The second one, about the repugnant conclusion :

What is repugnant about the repugnant conclusion is the suggestion (entailed by the impersonal total view) that a world filled with lives barely worth living is better than a world containing many fewer lives of much greater quality. But how could lives be barely worth living (read ‘worth continuing’) if they contained no bad—and the absence of more good were not bad? In other words, how can a life containing only good and no bad be barely worth continuing? If the lives in Z are actually quite well worth living then preferring Z over A is no longer repugnant (even if one thinks that it is still mistaken).

The last one, about mere addition :

It is true that if future possible lives were known to contain no bad, even the unhappiness-minimizing version of the impersonal average view could not rule out mere addition. However, the question is whether this would be problematic. Much of the reason why mere addition is seen as a problem is that the average impersonal view’s rejection of mere addition runs counter to an implicit assumption that it is good to have extra lives that are worth living. The average impersonal view says that it can be bad to have extra lives that are worth living (if these extra lives lower average well-being). Where lives contain some bad I have shown that my argument says that the impersonal average view is right to reject mere addition. Although my argument does not show that the impersonal average view is right to reject mere addition in cases where the extra lives contain no bad, it helps nevertheless to overcome the problem. Remember that, following my argument in Chapter 2, a (hypothetical) life that contains some good but no bad is not worse than never existing—but neither is it any better than never existing. Following my argument, there is no way to choose between (a) never existing and (b) coming into existence with a life that contains no bad whatsoever. This makes the impersonal average view’s judgement of mere addition less implausible. If it is better to have extra lives worth living and the impersonal average view suggests that it is worse, then there is a serious problem. However, if on one criterion, there is no way of choosing in favour of or against mere addition, and the impersonal average view suggests we choose against, then there need not be any contradiction. The impersonal average view can be seen as layering a further (impersonal) condition upon a judgement that mere addition is neither better nor worse for those who are added.

In conclusion, Benatar's view that coming into existence is always a harm thus does much of what Theory X needs to do. It solves the non-identity problem, avoids the repugnant conclusion and mere addition problem, and explains one of Derek Parfit's asymetry :

While it would be wrong to have a child that would have a life not worth continuing, there is no moral reason to have a child that would have a life well worth continuing.

But Benatar do not say that his view is Theory X. His view is an argument about whether there should be more people, while Theory X is a general theory about morality that can also deal satisfactorily with population questions. The fact that his argument, unlike so many others, appears compatible with Theory X in all these ways provides some further grounds for taking his argument seriously.


Derek Parfit thinks that contractarianism cannot provide guidance on how many people there should be :

On the ideal contractarian view, principles of justice are chosen in what John Rawls calls the ‘original position’—a hypothetical position in which impartiality is ensured by denying parties in the position knowledge of particular facts about themselves. The problem, however, says Derek Parfit, is that parties in the original position must know that they exist. But to assume, when choosing principles that affect future people, that we shall certainly exist, he says, ‘is like assuming, when choosing a principle that would disadvantage women, that we shall certainly be men’. This is a problem because it is essential to ideal contractarianism ‘that we do not know whether we would bear the brunt of some chosen principle’.

The problem with this objection is that the analogy does not hold, since non-existent people can't ‘bear the brunt’ of any principle :

A principle that results in some possible people never becoming actual does not impose any costs on those people. Nobody is disadvantaged by not coming into existence. Rivka Weinberg makes the same point in a different way. She says that ‘existence is not a distributable benefit’ and thus neither ‘people in general nor individuals in particular will be disadvantaged by the assumption of an existent perspective’.

Derek Parfit thinks we can't alter the original position of contractarianism so that parties do not know whether they will exist, but Benatar thinks we can. In that case, the parties would maximize the position of the worst off (what wome call "maximin"), and the ideal population will again be zero.

For Michael Baykes, maximin produces this conclusion only if it is uttilities that are being distributed. If it is primary goods (goods that are needed to secure all other goods), then the opposite conclusion is reached :

[...] worst off are the nonexistent, for they do not receive any primary goods. The next worse off class consists of those who may or may not exist, and if they exist they will receive some primary goods. Consequently one should bring as many people as possible into existence.

However, as we saw in Chapter 2, absent goods are not bad if there is nobody who is deprived by their absence, thus the non-existent are not the worst off.

Some rejects maximin on the grounds that it produces the zero conclusion and think that parties should reason probabilistically. It makes no difference whether they may or not :

This is because it is always very bad to come into existence. Thus the probability of a bad outcome is one hundred per cent. How bad the outcome is—very, very, very bad or just very bad—is a matter of probability. However, that does not matter in the current context, given that one already knows that any outcome in which one exists holds no advantages for oneself over an outcome in which one does not. Thus even those who think that (a) probabilities should be taken into account, (b) the interests of parents and children should be balanced, and (c) procreation should be permitted ‘only when it would not be irrational’ will be led to the same conclusion as those who choose to maximin. If my argument is right, it is always irrational to prefer to come into existence. Rational impartial parties would choose not to exist and the upshot of this is zero population.

Phased extinction

When decreasing population decreases quality of life

In some circumstances, a decrease in population causes a decrease in quality of life. One way it can happen is when a population shrink too rapidly at the result of lower birth rates. A larger part of the population is now unproductive because of age, so younger people can't produce enough to sustain the quality of life. Another way is when the absolute size of the population falls beneath some threshold.

Thus on Benatar's view, the only way that creating new people is acceptable is if we are aiming at phasing out people :

Unless humanity ends suddenly, the final people whether they exist sooner or later, will likely suffer much. There is some sense in making sure that fewer people suffer this fate. This can be done by steadily reducing the number of people.

Benatar imagine two scenarios (the width represent population size, the negative axis is the quality of life) :

Figure 6.2

A is a population that would exist if procreation continued, but at about seventy-five per cent of replacement rate. B is the population that would exist if we ceased procreation with immediate effect.

The absence of the new generation makes life in B worse for a lot of people. Benatar face the following questions :

  1. May we ever create new lives in order to improve the quality of existent lives?
  2. If so, under what conditions may we do so?

Reducing population to zero

The narrow person-affecting view, it will be recalled, is the view that :

An outcome is worse for people (in the narrow sense) if the occurrence of X rather than Y would be either worse, or bad, for the X-people.

This view can show why world B in Figure 6.2 is worse for the 1-people and why world A is worse for the 2-people. But it cannot provides guidance about the two above questions. We have to use the wide person-affecting view :

An outcome is worse for people (in the wide sense) if the occurrence of X would be less good for the X-people than the occurrence of Y would be for the Y-people.

Here too there are average and total versions. Benatar express them negatively :

Negative Wide Total Person-Affecting View :

An outcome is worse for people if the total net harm to the X-people by the occurrence of X would be greater than the total net harm to the Y-people by the occurrence of Y.

Negative Wide Average Person-Affecting View :

An outcome X is worse for people if the average net harm per person to the X-people by the occurrence of X would be greater than the average net harm per person to the Y-people by the occurrence of Y.

On both views, B is worse than A, so we would be justified in giving birth in order to reduce the harm of existing people. Benatar take the average view to be the less plausible one :

Adding extra lives does not increase the average harm per person if the quality of life of the new people is the same or better than those who preceded them. Thus twelve billion poor quality lives is no worse, on the average view, than six billion lives of the same quality. But it surely must be worse to inflict that same harm on double the number of people.

The total view can avoid this problem, and thus can provide an answer to the questions : we may create new lives when we thereby minimize the total harm that people suffer. There are possible objections :

It might be the case that these views are impersonal after all, and only person-affecting in appearance :

An impersonal view does not become a person-affecting one simply because it is stated in a way that sounds person-affecting. Impersonal principles are not concerned with the impact an action has on particular people but are rather concerned with the impact an action has on people in general. It is unsurprising, therefore, that such views will not be able to account adequately for concerns about the harm done to particular people by being brought into existence.

Pehaps there is a version of the wide person-affecting view that avoid this problem. But now Benatar will focus on views that take into account the possible objections, for example, the rights/deontological view. Attributing rights to potential people is problematic, so we can think of this in terms of duties to not bring new people into existence.

If it's an absolute duty, it doesn't matter how much harm to already existent people it causes. If it isn't, it can be thwarted if bringing new people causes a significant ammount of harm reduction (more than equivalent). The stronger the non-absolute right, the greater the harm reduction to others must be.

Benatar summarize the views and their implications :

1) Person-Affecting View : Cannot answer the question.

2) Negative Average View : Incompatible with the anti-natalist argument

3) Negative Total View : We may create new people where the total amount of harm in doing so is equivalent to, or less than, the harm that would be suffered by existing people if the new people were not created.

4) Rights/Deontological View : Creating new people cannot be justified by mere reduction in total harm

4a) More stringent rights or duty view : Creating new people can never be justified by any reduction in total harm, no matter how great that reduction may be.

4b) Less stringent rights or duty view : Creating new people may be justified by substantial (but not mere) reduction in total harm.

Whether or not the conditions of the total view or the less stringent rights or duty view are met, ordinary procreators or potential procreators cannot currently appeal to them to justify their reproducing. This is because the population-related quality of life problems currently faced are those resulting from increasing not decreasing population. And even if the population growth started to taper off or transform into gradual population decline, that would still not be enough. It is only in situations of very rapid population reduction or of reduction back to levels that humans exceeded millennia ago, that questions about creating people to reduce harm could even arise. We are nowhere near there.


Human extinction will certainly occur, some hope as late as possible. The fact that it will certainly occur makes Benatar's argument an optimistic one (things will be some day the way they should be : without any humans). There are however concerns about extinction :

  1. Where extinction is brought about by killing, it might be thought to be bad because it cuts lives short. This is not a problem for Benatar, since he doesn't suggest extinction by killing, but rather by not procreating.
  2. Whichever way extinction is brought about, it might be thought to be bad for those who precede it. The last generation would bear the burden of "no future" : its hopes and desires for the future will be thwarted. It will also suffer the burden of society breaking down. But since it will happen anyway, it is better sooner than later, since if it is later there is the cost of the suffering of the intervening new generations.
  3. The state of extinction might be thought to be bad, in itself. On this view, a world in which there are no people (or other conscious beings) is regrettable in its own right, irrespective of the significance of this state of affairs for earlier beings. This concern, although widely spread, can't be made sense of if one accepts (as most people do) the asymmetry of pain and pleasure. On this asymmetry, a world without humans is better for those humans who would otherwise have existed. It could be objected that such a world would be worse in other ways, for example, it would lack moral agents and rational deliberators, and it would be somewhat less diverse. Benatar replies :

    First, what is so special about a world that contains moral agents and rational deliberators? That humans value a world that contains beings such as themselves says more about their inappropriate sense of self-importance than it does about the world. (Is the world intrinsically better for having six-legged animals? And if so, why? Would it be better still if there were also seven-legged animals?) Although humans may value moral agency and rational deliberation, it is far from clear that these features of our world have value sub specie aeternitatis. Thus if there were no more humans there would also be nobody to regret that state of affairs. Nor is it clear why a less diverse world is worse if there is nobody deprived of that diversity. Finally, even if we think that such factors as moral agency, rationality, and diversity enhance the world, it is highly implausible that their value outweighs the vast amount of suffering that comes with human life.

Chapter 7 - Conclusion

Countering the counter-intuitiveness objection

It is obviously not reasonable to reject a view just because it is counter-intuitive (for example, abolishing slavery was very counter-intuitive in the past). Benatar's conclusion follow from views that are accepted by most people and are quite reasonable, such as the asymetry described in Chapter 2. One could take it as a reductio ad absurdum of the asymetry and reject it, but this pose several problems.

First, we must consider what rejecting it commit us to :

Of course, there are various ways of rejecting asymmetry, but the least implausible way would be by denying that absent pleasures are ‘not bad’ and claiming instead that they are ‘bad’. This would commit us to saying that we do have a (strong?) moral reason and thus a presumptive duty, based on the interests of future possible happy people, to create those people. It would also commit us to saying that we can create a child for that child’s sake and that we should regret, for the sake of those happy people whom we could have created but did not create, that we did not create them. Finally, it would commit us not only to regretting that parts of the earth and all the rest of the universe are uninhabited, but also to regretting this out of concern for those who could otherwise have come into existence in these places.

Matters become still worse if we attempt to abandon asymmetry in another way—by claming that absent pains in Scenario B are merely ‘not bad’. That would commit us to saying that we have no moral reason, grounded in the interests of a possible future suffering person, to avoid creating that person. We could no longer regret, based on the interests of a suffering child, that we created that child. Nor could we regret, for the sake of miserable people suffering in some part of the world, that they were ever created.

A second problem is that the dominant intitions in these matters are untrustworthy :

First, why should we think that it is acceptable to cause great harm to somebody—which the arguments in Chapter 3 show we do whenever we create a child—when we could avoid doing so without depriving that person of anything? In other words, how reliable can an intuition be if, even absent the interests of others, it allows the infliction of great harm that could have been avoided without any cost to the person who is harmed? Such an intuition would not be worthy of respect in any other context. Why should it be thought to have such force only in procreative contexts?

Secondly, we have excellent reason for thinking that pro-natal intuitions are the product of (at least non-rational, but possibly irrational) psychological forces. As I showed in Chapter 3, there are pervasive and powerful features of human psychology that lead people to think that their lives are better than they really are. Thus their judgements are unreliable. Moreover, there is a good evolutionary explanation for the deep-seated belief that people do not harm their children seriously by bringing them into existence. Those who do not have this belief are less likely to reproduce. Those with reproduction-enhancing beliefs are more likely to breed and pass on whatever attributes incline one to such beliefs.

There is a third problem for those who want to treat Benatar's argument as a reductio ad absurdum of the asymmetry :

their argument could also be used by a species doomed to lives much worse than our own. Although we might see their lives as great harms, if they were subject to the kinds of optimistic psychological forces characteristic of humans they too would argue that it is counter-intuitive to claim that they were harmed by being brought into existence. That which would not be counter-intuitive from our perspective would be counter-intuitive from theirs. Yet we can see, with the benefit of some distance from their lives, that little store should be placed on their intuitions about this matter. Something similar can be said about the common human intuition that creating (most) humans is not a harm.

Responding to the optimist

Optimists and pessimists might disagree about facts (what is or will be the case) or about evaluation (what is or will be the case is good/bad). The view that coming into existence is always a serious harm is pessimistic in both a factual sense (there is much more negative things in human lives than people realize) and evaluative sense (life's goods do not make life worth starting). There is the optimist twist of Chapter 6, that eventually sentient life will end, which is a good thing on Benatar's view (and a twist on the twist : a pessimist can worry that it will end in a very far future).

Benatar rightfully note that an optimist view can't be right merely because it is cheerful (and a pessimist one merely because it is grim). Evidence is what matter. The "smug macho tone" of people who says that the pessimist must "grind and bear it" is no good either :

This view is defective for the same reason that macho views about other kinds of suffering are defective. It is an indifference to or inappropriate denial of suffering, whether one’s own or that of others. The injunction to ‘look on the bright side’ should be greeted with a large dose of both scepticism and cynicism. To insist that the bright side is always the right side is to put ideology before the evidence.

Death and suicide

Benatar's view that coming into existence is always a harm does not imply that death is better than continuing to exist. Future-life and present-life cases are different :

the existent can have interests in continuing to exist, and thus harms that make life not worth continuing must be sufficiently severe to defeat those interests. By contrast, the non-existent have no interest in coming into existence. Therefore, the avoidance of even lesser harms—or, on my view, any harm—will be decisive.

If death is a harm, then it is one of the reasons why coming into existence is a harm (everyone will have to die at some point). There has been objections to the idea that death can be a harm :

Epicurus famously argued that death is not bad for the one who dies because so long as one exists, one is not dead, and once death arrives one no longer exists. Thus, my being dead (in contrast to my dying) is not something that I can experience. Nor is it a condition in which I can be. Instead it is a condition in which I am not. Accordingly my death is not something that can be bad for me. Lucretius, a disciple of Epicurus’ and thus also an Epicurean, advanced a further argument against death’s being a harm. He argued that since we do not regret the period of non-existence before we came into being, we should not regret the non-existence that follows our lives.

Such a view is at odds with :

Benatar compares counter-intuitiveness in the case of his view and in the case of the Epicurean view :

First, the Epicurean conclusion is more radically counter-intuitive than my conclusion. I suspect that more people think, and feel more strongly, that murder harms the victim than who think that coming into existence is not a harm. Indeed there are very many people who believe that coming into existence is often a harm and there are still more people who believe that it is never a benefit even if they think that it is not also a harm. Yet there are very few people who truly believe that murder does not harm the victim. Even where the victim’s life was of a poor quality, it is widely thought that killing that person without his consent (where consent could have been obtained) is to wrong him. Secondly, a precautionary principle applies asymmetrically to the two views. If the Epicurean is wrong, then people’s acting on the Epicurean argument (by killing others or themselves) would seriously harm those who were killed. By contrast, if my view is mistaken, people’s acting on my view (by having failed to procreate) would not harm those who were not brought into existence.

But since counter-intuitiveness isn't enough to dismiss a view, Benatar responds. First to Lucretius : there is an difference between pre-vital non-existence and post-mortem non-existence :

Whereas any one of us could live longer, none of us could have come into existence much earlier. This argument becomes very powerful when we recognize the kind of existence that we value. It is not some ‘metaphysical essence’, but rather a thicker, richer conception of the self, that embodies one’s particular memories, beliefs, commitments, desires, aspirations, and so on. One’s identity, in this thicker sense, is constructed from one’s particular history. But even if one’s metaphysical essence could have come into existence earlier, the history of that being would have been so different that it would not be the same person as one is. Yet, things are quite different at the other end of life. Personal histories—biographies—can be lengthened by not dying sooner. Once one is, one can continue to be for longer. But an earlier coming into existence would have been the coming into existence of a different person—one with whom one might have very little in common.

Then he responds to Epicurus, using the depravation account : death is bad for the person who dies because it deprives that person of future life and positive features. So in this view death is sometimes bad, but not always (it depends on the future of the life).

One objection is that advocates of the deprivation account cannot say when the harm of death occurs. If it's after death, no one can be harmed, if it's before, it entails backwards causation. A possible response would be that the harm occurs "always" or "eternally" :

George Pitcher offers a helpful analogy. He says that if ‘the world should be blasted to smithereens during the next presidency. . .this would make it true (be responsible for the fact) that even now, during [the current president’s] term, he is the penultimate president of the United States’. Similarly, one’s later death makes it true that even now one is doomed not to live longer than one will. Just as there is no backward causation in the case of the penultimate president, so there is no backward causation in a death that harms one all along.

Another objection to the deprivation account is to simply deny that those who have ceased to exist can be deprived of anything :

David Suits, for example, argues that although the ante-mortem person may indeed be worse off than he would otherwise have been had he lived longer, being worse off in this ‘purely relational’ way is not thought to be sufficient to show that he is harmed. He argues further that even if it were, there cannot be real deprivation if there is nobody left to be deprived. One can only be deprived if one exists.

We have come to an impasse : it's a matter of thinking if death is a special case of depravation, or if it must be treated like any other cases.

Those who reject the Epicurian view can hold one of the following : a) Death is always a harm. b) Death is always a benefit. c) Death is sometimes a harm and sometimes a benefit.

Benatar adopts c), which is the common sense view. But his version allows for death to be a benefit more often than the usual view. That is because it may be an irrational love for life that prevent some suicide in lives that are actually very bad.

The fact that suicide can have a very negative impact on those close to us is part of the tragedy of coming into existence :

We find ourselves in a kind of trap. We have already come into existence. To end our existence causes immense pain to those we love and for whom we care. Potential procreators would do well to consider this trap they lay when they produce offspring.

Misanthropy and philanthropy

This book might sounds like misanthropy, but since Benatar argue against harming people, it is a kind of philanthropic view (ableit an odd one). It is also a zoophilic view (not in a sexual sense) in the case of other sentient animals. There is however a misanthropic argument against having children and for human extinction, based on the harm that the person will cause if brought into existence. It doesn't go as far, since it concern only the human species, but it is perfectly compatible with the philanthropic arguments.

Final words :

It is unlikely that many people will take to heart the conclusion that coming into existence is always a harm. It is even less likely that many people will stop having children. By contrast, it is quite likely that my views either will be ignored or will be dismissed. As this response will account for a great deal of suffering between now and the demise of humanity, it cannot plausibly be thought of as philanthropic. That is not to say that it is motivated by any malice towards humans, but it does result from a self-deceptive indifference to the harm of coming into existence.