SciELO - Scientific Electronic Library Online

 
vol.27 número2Some curious cases against cognitive human enhancementO caráter híbrido do conceito de referência em Ricœur índice de autoresíndice de assuntospesquisa de artigos
Home Pagelista alfabética de periódicos  

Serviços Personalizados

Journal

Artigo

Indicadores

Links relacionados

  • Não possue artigos similaresSimilares em SciELO

Compartilhar


Revista Diacrítica

versão impressa ISSN 0807-8967

Diacrítica vol.27 no.2 Braga  2013

 

Good to die

Rainer Ebert*

*Rice University, School of Humanities, Philosophy Department, Texas 77251-1892, USA.

rainerebert@gmail.com

 

ABSTRACT

Among those who reject the Epicurean claim that death is not bad for the one who dies, it is popularly held that death is bad for the one who dies, when it is bad for the one who dies, because it deprives the one who dies of the good things that otherwise would have fallen into her life. This view is known as the deprivation account of the value of death, and Fred Feldman is one of its most prominent defenders. In this paper, I explain why I believe that Feldman’s argument for the occasional badness of death fails. While staying within an Epicurean framework, I offer an alternative that adequately accounts for a significant range of widely held intuitions about prudential value. My account implies that death is almost always good for the one who dies, yet often less good than not dying. Finally, I discuss some puzzles that remain for my account and hint at possible ways to address them.

Keywords: Hedonism; Epicurus; death; value theory; Feldman.

 

RESUMO

Entre aqueles que rejeitam a defesa de Epicuro de que a morte não é nociva para quem morre, é popularmente considerado que a morte é nociva para quem morre, quando nociva para aquele que morre, porque o priva das coisas boas que poderiam ocorrer na sua vida. Esta via é conhecida como a apreciação de privação do valor da morte, da qual Fred Feldman é um dos defensores mais proeminentes. Neste artigo explico porque creio que o argumento de Feldman para a maldade ocasional da morte falha. No seio dum enquadramento epicurista, proponho uma alternativa que responde de forma adequada a um conjunto significativo de intuições amplamente difundidas sobre o valor prudencial. A minha apreciação implica que a morte é quase sempre boa para aquele que morre, mas muitas vezes menos boa do que não morrer. Por último, discuto alguns enigmas que permanecem na minha análise e possíveis formas de resolvê-los.

Palavras chave: Hedonismo; Epicuro; morte; teoria do valor; Feldman.

 

Epicureans claim that, contrary to common belief, the event of death is not bad for the one who dies.[1] The arguments they present in support of this claim are diverse. One such argument asserts that death is neither good nor bad for the one who dies, because it results in an eternal experiential blank and hence does not lead to any experiences, good or bad. And only (certain kinds of) experiences are prudentially good or bad in themselves. Epicureans get their name and inspiration from the ancient Greek philosopher Epicurus who wrote in a letter to Menoeceus that

death, the most terrifying of ills, is nothing to us, since as long as we exist death is not with us; but when death comes, then we do not exist. It does not then concern either the living or the dead, since for the former it is not, and the latter are no more. (…) [T]he wise man neither seeks to escape life nor fears the cessation of life, for neither does life offend him nor does the absence of life seem to be any evil (…) (Epicurus, 1940: 31).

In his book Confrontations with the Reaper (Feldman, 1992), Fred Feldman challenges the Epicurean position on the value of death and argues that, even if the Epicureans are correct in assuming that (i) death marks the point in time at which a being ceases to exist, (ii) the non-existent do not experience anything at all, and (iii) “all good and evil consist in sensation,” (Epicurus, 1940: 31) there are nevertheless good reasons to believe that death is bad for the one who dies. Feldman is a proponent of the so-called deprivation approach.[2] According to this approach, death is bad for the person who dies, when it is bad for the person who dies, because it deprives that person from goods that would have fallen into her life if she had continued to live.[3]

In this paper, I will briefly summarize Feldman’s argument and explain why I believe it fails. I will argue that death is almost always good for the one who dies, yet often less good than not dying – a position consistent with, but stronger than, the Epicurean claim that death is not bad for the one who dies. In doing so, I shall follow Feldman’s example and work within an Epicurean framework. I shall assume, without providing any justification, that we exist neither before birth or conception (or whenever life begins), nor after death. I shall further adopt a hedonistic theory of what is ultimately good for people. I will end by discussing some puzzles that remain for my account.

Feldman on the evil of death

The short paragraph from Epicurus’ letter to Menoeceus that I quoted above drew a good deal of attention in the philosophical literature, both critical and favorable.[4] Different philosophers extracted different arguments with different conclusions from that paragraph. Even though Epicurus exegesis is interesting and important in its own right, I shall not get into that business here. We will instead be concerned with the particular Epicurean argument against the prudential badness of death that Feldman presents in his book. This argument may or may not be an argument Epicurus actually meant to produce, but it is worth considering – partly because of the far-reaching implications it could have, if sound, say for questions about the morality of killing animals, both human and non-human, or the rationality of suicide.

Before we can state the Epicurean argument, we need to introduce some terminology. Following a long tradition in philosophy, Feldman distinguishes two kinds of badness-for-a-person: Badness-for-a-person of a thing either lies entirely in that thing, or depends on its relation to some other thing. The former kind of badness is called intrinsic badness, the latter extrinsic badness. According to hedonism, “painful experiences are the only things that are intrinsically bad for a person” (Feldman, 1992: 133), and enjoyable (or pleasurable) experiences are the only things that are intrinsically good for a person. Cars, of course, are usually also good for people, but only extrinsically so. We use cars to get to and from university, work, amusement parks and our favorite holiday destinations. Their value for us derives from the contingent fact that they play a role in bringing about experiences we find enjoyable. The same applies to money, loving parents and other things we commonly consider good for us to have. In contrast, “illness, poverty, injustice, and ignorance (…) [are extrinsically bad for us] because they happen to be connected to pain” (Ibidem).[5]

As death as such is not a painful experience, it is not intrinsically bad for us.[6] This is a trivial result, given our hedonistic framework. Hence, to say something interesting, the Epicurean argument against the prudential badness of death must conclude that death is not extrinsically bad for us. Feldman formulates this argument as follows:

1. Each person stops existing at the moment of death.

2. If (1), then no one feels any pain while dead.

3. If no one feels any pain while dead, then death does not lead to anything intrinsically bad for the one who dies.

4. If death does not lead to anything bad for the one who is dead, then death is not extrinsically bad for the one who is dead.

5. Therefore, death is not extrinsically bad for the one who is dead. (Idem, 136)

With Feldman, we shall grant (1), (2), and (3).[7] (4) is based on what Feldman calls the causal hypothesis: “CP: If something is extrinsically bad for a person, then it is bad for him or her because it leads to later intrinsic bads for him or her” (Idem, 135).

Earlier, I preliminarily described extrinsic badness as consisting in its being somehow connected with intrinsic badness. CP is one way to specify this connection. Feldman believes it is the wrong way and offers two (what he takes to be) counterexamples to CP. Both counterexamples are designed to show that CP represents an “overly narrow view” (Idem, 137) with regard to the relation between intrinsic and extrinsic badness. Since they are relevantly similar, it will be sufficient to reproduce only one of them here.

Consider John, a young man who is accepted by two colleges, College A and College B. Unlike College B, College A does not offer any philosophy courses. John “decides to attend College A (…) [and] spends four happy years [there]. Suppose he never learns anything about philosophy” (Ibidem), leads a reasonably happy life and dies. However, John “has outstanding aptitude for philosophy and (…) would have enjoyed it enormously if he had been given the opportunity” (Ibidem). If he had chosen to go to College B, he “would have become a philosophy major, and his life would have been much happier” (Ibidem).

John’s choice to go to College A did not lead to any later intrinsic prudential bads. Neither was his time at College A painful, nor was going there the cause for any pain later on in his life. CP would have us believe that his choice to go to College A was not bad for John. Feldman disagrees and suggests that John’s choice “was a misfortune for this young man. It’s a pity; too bad for him” (Ibidem). Going to College A deprived him of all the additional goods that would have fallen into his life had he chosen to go to College B. Feldman takes this example to show that “[s]ome things are bad for us even though they are not themselves painful experiences, and they do not lead to any painful experiences” (Idem, 138).[8] In order to accommodate his intuition with regard to John’s choice, he offers the following alternative account of extrinsic value for a person (S) of a state of affairs (P): “D: The extrinsic value for S of P = the difference between the intrinsic value for S of the life S would lead if P is true and the intrinsic value for S of the life S would lead if P is false” (Idem, 150), where P is considered true, if, and only if, P obtains.[9]

Note that, according to D, “a state of affairs can be extrinsically bad [or good] for a person whether it occurs before he exists, while he exists, or after he exists” (Idem, 152). Examples for these respective cases are my parent’s winning the lottery before I am conceived, my being hit by a bus now, and my not travelling to Tuvalu one hundred years from now.[10]

D implies principles for both the extrinsic badness and the extrinsic goodness of things, or states of affairs, for a person: “EI: Something is extrinsically bad for a person if and only if he or she would have been intrinsically better off if it had not taken place” (Idem, 138). EI*: Something is extrinsically good for a person, if, and only if, he or she would have been intrinsically worse off if it had not taken place.

EI not only allows us to say that John’s choice to go to College A was extrinsically bad for John because he would have been intrinsically better off if he had chosen College B, but also explains why death is bad for us (when it is bad for us):

Suppose Jones, a healthy man of 30 years leading a reasonably happy life, gets shot in the head from behind on the streets. He dies instantly. Neither did he take any note of his approaching killer, nor did the bullet cause him any pain. If getting shot was bad for Jones, CP is false and we can ask why it was bad for Jones to get shot. According to EI, what explains why death was bad for Jones is the fact that “his life is on the whole intrinsically less valuable for him than it would have been if he had not died when he in fact died” (Idem, 139). Jones’ death was bad for him because it deprived him of net prudential good that would have been part of his life otherwise. In contrast, if the life Jones had ahead of him at the moment he got shot had contained more pain than enjoyment, death would not have been bad, but good, for him. Accordingly, Feldman holds that death is sometimes bad for the one who dies, but not always.

Before we conclude this section, a brief ontological note is in order. By defining D exclusively in terms of states of affairs and then applying that principle to the event of death, Feldman implicitly commits himself to the controversial view that events are states of affairs. Feldman defines the event of a person’s death as the event that, necessarily, occurs precisely when that person dies, i.e., when that person “makes the transition from being alive to being dead” (Idem, 109). The event of a person’s death, occurring at t, might well be identical to a state of affairs, say the state of affairs of that person’s dying at t. However, while I accept Feldman’s definition of the event of a person’s death, I will not follow him in making such an ontological commitment. Instead, when formulating my account of extrinsic value, I will more generally talk about states of affairs or events, and then apply that account to the event of death.

Rejecting the deprivation account

While EI might seem plausible initially, I believe it confuses things that are merely less good for us with things that are prudentially bad.

Suppose Claudio is at a club with his friends, sipping a drink at the bar. At the other end of the bar, there are two attractive women who catch his attention, Rebeca and Joana. Claudio exchanges some smiles with them, and eventually plucks up the courage to start a conversation with one of them. His choice falls on Rebeca. From that night on, Claudio and Rebeca are inseparable. They become best friends; fall in love with each other. Eventually, they marry, and happily spend the rest of their lives together. Looking back at that fateful night shortly before his death, Claudio thinks of his decision to take heart and talk to Rebeca as the best decision of his life. What he does not know, and never will know, is that, if he had chosen to start a conversation with Joana instead, he would have spent his life with her. And he would have been even happier with Joana than he was with Rebeca. His life would have been slightly higher in prudential value. Hence, EI would have us believe that Claudio’s choice to start a conversation with Rebeca was bad for him, because he would have been intrinsically better off if he had not started a conversation with Rebeca. In the closest possible world in which he does not start a conversation with Rebeca at the bar, he starts a conversation with Joana and is better off than in the actual world. While we might agree that choosing Rebeca was worse for Claudio than choosing Joana would have been, it seems absurd to say that his choice to start a conversation with Rebeca was bad for him. After all, that choice brought him much happiness and enriched his life more than he ever anticipated.[11]

Things could be different if Claudio somehow learned later on in his life that he would have been even happier with Joana. In that case, knowledge about what he missed out on might cause intense and persistent feelings of regret that outweigh all the good that being with Rebeca brings to his life. If so, it seems plausible to say that Claudio has been deprived of the additional goods starting a conversation with Joana would have brought to his life. This suggests that deprivation, properly understood, generally requires that the one who is deprived feels deprived and suffers as a result of the deprivation. David B. Suits, too, suggests an experience requirement for deprivation proper and says about death that to think of it “as a deprivation is to misuse the usual sense of deprivation as having one’s hopes, aspirations and expectations thwarted or frustrated” (Suits, 2001: 77).

If you find the case of Claudio unconvincing, consider the following example by Aaron Smuts:

Buridan, a chocolate-chip cookie lover, walks into a bakery to buy a snack. In the display case sit two apparently equally scrumptious cookies. Although he was at ?rst unable to decide between the two, Buridan chooses the one on the right. He takes tremendous pleasure in eating the decadent cookie. But, unbeknownst to Buridan, the cookie on the left had an extra chocolate chip. He would have enjoyed that cookie slightly more. (Smuts, 2012: 208)

According to Feldman’s deprivation account, Buridan’s choice to buy and eat the cookie on the right was bad for him. If he had not chosen that cookie but the other one instead, he would have been intrinsically better off. Yet, Buridan’s choice to buy the cookie on the right did not lead to any bad for Buridan. On the contrary, he takes great pleasure in eating the cookie he actually chose. His actual choice might have been worse for him than the alternative choice would have been, but it was nevertheless a good choice to buy the cookie on the right, notwithstanding EI’s counterintuitive implication to the contrary.

EI also implies that I have suffered a great misfortune today in that NASA did not call me and offer me a job as an astronaut.[12] It is not like I expected them to call me. But, if they had called me, my life would have been so much better for me. I would have enjoyed receiving astronaut training at Johnson Space Center tremendously. And then going to the International Space Station… But unless I had reason to believe that NASA might call and am hence disappointed that they in fact did not, it is hardly intuitive to say that harm has befallen me. After all, while many goods that could have been part of my life are not, nothing bad happened to me, just like nothing bad happened to me even though an ounce of gold did not materialize in front of me just now. My day was just fine, and I go on leading a reasonably happy life.[13]

I believe that something similar should be said about the case of John. It is true that it would have been better for John to go to College B, but that does not make his choice to go to College A bad. As we have seen in the examples above, less good is not always bad. John had a good life and did not suffer from being “deprived” of the opportunity to engage in philosophy. He can look back on a successful and satisfying career, and he will rightly consider his choice of college a good choice.

A hybrid account of extrinsic prudential value

In the previous section, I argued that Feldman inflates the notion of what is bad for us beyond what is intuitive. EI holds that a thing is extrinsically bad for a person if that person would have been intrinsically better off without it. But, often, that is simply not true. The fact that a thing is worse for a person than the alternative would have been makes that thing less good, but not necessarily bad. Claudio’s choice to start a conversation with Rebeca, Buridan’s choice to buy the cookie on the right, and John’s choice to go to College A all are more plausibly described as choices that were good for them than as choices that were bad for them.

In light of EI’s overreach, we might be tempted to go back to a causal account of the connection between intrinsic and extrinsic value for a person, and say that

(A) a state of affairs or event is extrinsically good for a person (S), if, and only if, it leads to net intrinsic good for S, while

(B) a state of affairs or event is extrinsically bad for a person (S), if, and only if, it leads to net intrinsic bad for S.[14]

Neither Claudio’s decision to start a conversation with Rebeca, nor my not being called by NASA, led to more good than bad for either one of us. Claudio and Rebeca had an exceptionally good marriage, and, even though I would have enjoyed visiting the ISS, I did not suffer because I did not get the chance to do so. B hence respects our intuitions with regard to both cases, and the others. However, A and B have their own problems. That becomes immediately apparent when we apply these principles to the case of death. Since death marks the end of our existence, it cannot possibly cause us any pain or pleasure. Therefore, if A and B were true, death would always be neutral. That is, we could assign neither a positive nor a negative prudential value to any death.

Consider Feldman’s example of a “very old and unhappy person. Suppose that further life for this person will inevitably contain more pain than pleasure” (Feldman, 1992: 140). Feldman holds, and I think plausibly so, that a painless death would be extrinsically good for that person.[15] His explanation, of course, consists in an appeal to EI*, which seems to provide the right kind of explanation in the case at hand. From the perspective of this paper, that suggests a curious asymmetry between pleasure and pain. While we concluded in the previous section that less pleasure is bad for people only if it results in frustration or other unpleasant experiences, less pain – in contrast – seem to be always good. This is an important point that deserves elaboration: Imagine you have a splitting headache and decide to take a pain killer. While your pain will reduce, taking a pain killer will not bring about any intrinsically good experiences. Nevertheless, it seems natural to say that taking the pain killer was good for you. Similarly, being anesthetized prior to surgery is usually good for you insofar as it spares you much of the excruciating pain you would experience otherwise. What makes taking a pain killer and being anesthetized prudentially good in these situations is the fact that your life would contain more pain if you had not taken the pain killer, and if you had not been anesthetized, respectively. Reduction and absence of pain are good for a person, even if that person does not enjoy these things. This peculiar kind of goodness-for-a-person defies A, but it is captured by EI*.

We need to look for an account of positive prudential value that accommodates our intuitions with regard to pain killers and anesthesia as well as the case of Claudio and the other examples discussed in the previous section. The preliminary causal account that is the conjunction of A and B and Feldman’s counterfactual account both are only partly successful, yet each account seems to catch an important aspect of the account of goodness-for-a-person we are looking for. While the causal account correctly disregards deprivations that nobody suffers, the counterfactual account recognizes that we can benefit people not only by doing them good but also by taking them out of their misery.[16] This suggests a hybrid account that incorporates both aspects and holds that

(A’) a state of affairs/event (P) is extrinsically good for a person (S), if, and only if,

i) P leads to net intrinsic good for S, or

ii) the life of S contains more pain in the closest possible world in which P does not obtain/occur.

In the closest possible world in which you do not take a pain killer, your splitting headache continues until it finally goes away. From then onward your life will be about the same in that possible world and the actual world. Therefore, your life in the closest possible world in which you do not take a pain killer overall contains more pain than your life in the actual world. A’ hence yields the intuitive result that taking a pain killer was good for you. Similarly, A’ offers an explanation why being anesthetized is usually good for those about to undergo surgery.

A’ and B taken together, however, open up the possibility that something is both good for a person (S) and bad for S insofar as a state of affairs/event (P) can lead to net intrinsic bad for S while, at the same time, the life of S would contain more pain in the closest possible world in which P does not obtain/occur. This is not a satisfying result. While things might be good for us in one regard and, at the same time, bad for us in another regard, we do not want to say that something can be good for us, all things considered, and, at the same time, bad for us, all things considered.[17] We want to know which one it is – good, or bad, or neutral.[18]

In order to solve this problem, let us have a look at an example that satisfies the right-hand sides of both A’ and B: The active ingredient of popular pain killers such as Advil is ibuprofen. Common adverse effects of ibuprofen include nausea and other effects we typically find unpleasant. Suppose, once again, you have a splitting headache and decide to take an Advil tablet. Your headache goes away but you end up with mild nausea caused by the medication. Taking the Advil tablet caused you some pain and no pleasure. However, at the same time, your life would otherwise contain even more pain. We would hence want to say that taking Advil was good for you. While you are now feeling a little sick, you got rid of a splitting headache and feel much better overall. Similarly, exercising, even though you might take pain in it, is not bad for you insofar as it leads to a more healthy life and hence less unpleasant experiences associated with illness. Things that cause more pain than pleasure are extrinsically bad for you only if your life would have been less painful otherwise. Hence, I propose that

(B’) a state of affairs/event (P) is extrinsically bad for a person (S), if, and only if,

i) P leads to net intrinsic bad for S, and

ii) the life of S contains less pain in the closest possible world in which P does not obtain/occur.

In contrast to Feldman’s account of prudential value, B’ correctly distinguishes between things that both have comparative disvalue for us and are bad for us, and things that have comparative disvalue for us but fail to qualify as genuine evils.[19] Contracting malaria is an example for the former kind of thing. An ounce of gold not materializing in front of me just now was one of the many examples for the latter kind of thing that I presented in the previous section. Discussing Feldman’s account of extrinsic prudential value, Kai Draper gives a very similar example. He notes that D, or more specifically EI,

would imply that I have suffered a terrible misfortune today in that I did not find Aladdin’s lamp and hence have not been granted three wishes by an omnipotent genie. For the intrinsic value for me of the life I would lead should I find the lamp would be far greater than the intrinsic value for me of the life I would lead should I not find it. Hence, given D, the disvalue for me of not finding the lamp is enormous. (Draper, 1999: 389)

It is not hard to imagine that the comparative disvalue for me of not finding Aladdin’s lamp is greater than the comparative disvalue for you of being punched in the face by an angry drunk. Yet, it would be a peculiar use of language, if I said: “You think you just suffered a terrible harm? What that drunk did to you is nothing in comparison to what just happened to me: I didn’t find Aladdin’s lamp!” Even more, failing to find Aladdin’s lamp does not seem to be an evil at all in any important sense. There are things that have comparative disvalue for a person, but are not bad for that person. As correctly recognized by A’, some of these things – not including failing to find Aladdin’s lamp, but including, for example, Claudio’s decision to start a conversation with Rebeca – are in fact prudential goods. And that is why Feldman’s comparativist account of extrinsic prudential value fails.[20]

The alternative account presented in this section, the conjunction of A’ and B’, does not go as far as defenders of CP who hold that extrinsic value is never purely comparative. It retains counterfactual conditions for both extrinsic goodness-for-a-person and extrinsic badness-for-a-person, yet adds further conditions to generate what I would consider plausible results for all examples considered so far. But what does my hybrid account say about death?

The prudential value of death

We are assuming that a person’s death marks the end of that person’s existence and awareness. As we are further assuming hedonism, the event of death hence cannot possibly lead to either net intrinsic prudential good, nor net intrinsic prudential bad. Therefore, death neither meets A’ i), nor B’ i), and the question about the prudential value of death turns on whether death meets A’ ii). Let us recall the case of Jones, the healthy businessman leading a reasonably happy life who gets shot in the head and dies instantly. Suppose in the closest possible world in which Jones does not get killed he continues to live a good life that, from the point in time at which he actually does get killed onwards, contains more pleasure than pain. In this possible world, too, he will eventually have to go through a process of dying. And dying usually involves at least some pain. And even if he is lucky enough to die painlessly, he will surely experience some pain between the point in time at which he actually gets shot and the point in time at which he dies in the closest possible world. People often experience anxiety, depression, boredom, loneliness, etc., and hardly ever is anybody completely free of all such unpleasant emotions for more than a few hours at a stretch. In general, it is safe to assume that, for almost all people, their lives would contain less pain overall if they ceased to exist. With regard to their extrinsic prudential value, almost all deaths are relevantly like the death of Jones. A’ implies that death is almost always good for people because their lives would almost always contain more pain otherwise. On the other hand, B’ implies that death is never bad for people because neither could their lives have possibly contained less pain if they had continued to live, nor can death possibly result in later pains.[21] Note that I am not saying that death is not worse than continuing to live. In fact, it would have been better for Jones not to get killed. But just like starting a conversation with Rebeca rather than with Joana was merely less good, yet not bad for Claudio, being deprived of the net good that could have fallen into his life is not bad for Jones insofar as death makes it impossible for Jones to suffer this deprivation. Death is a deprivation no one suffers. At the same time, death is good for the one who dies because it spares her pain that would occur in the future if she continued to live. This is a strange result, but maybe that should not surprise us. Death is a strange thing![22]

The asymmetry between pleasure and pain on which our conclusion about the prudential value of death rests has significant explanatory power. It accommodates our intuitions – or mine, at any rate – with regard to all cases considered in this paper so far and, to just give one further example, could be part of an explanation why it makes little sense to be sad for those possible Martians who, had they existed, would have had an enjoyable life containing no pain whatsoever, yet rational to consider not coming into existence good for those other possible Martians who, had they existed, would spend their lives in excruciating pain.[23] Even though my approach implies a view of the prudential value of death that is not widely held, it provides a coherent rational basis for fairly common intuitions concerning a wide range of cases, and hence should not be flippantly rejected.

Some puzzles

My account of the extrinsic prudential value of states of affairs and events, while avoiding some, inherits a number of problems from Feldman’s account on which it is partly built. In what follows, I will briefly discuss some of these problems.

A’ and B’ both essentially refer to the closest possible world in which a certain state of affairs does not obtain or in which a certain event does not occur. Now there might be more than one closest possible world. Feldman considers such a case in a footnote (Feldman, 1992: 236 f. n. 7). A structurally similar example will serve to reveal a weakness of my account. Suppose, once again, that you have a splitting headache and decide to take an Advil tablet. Your headache goes away but you end up with mild nausea. We want to know whether taking Advil was good for you, or bad for you. Suppose there are two possible worlds, equally close and closest to the actual world, in which you do not take Advil. In one of these worlds your headache persists for a considerable amount of time and, accordingly, your life as a whole contains more pain than in the actual world. In the other possible world, you remember that you have an aspirin tablet in the pocket of your shirt, and you decide to take it. Your headache goes away and there are no adverse effects. Your life as a whole contains less pain in this possible world than in the actual world. Had the former possible world been the only closest possible world, then A’ would have us say that taking Advil was good for you. In contrast, had the latter possible world been the only closest possible world, then B’ would have us say that taking Advil was bad for you. However, as it is, A’ and B’ render us clueless about what to say. It is implausible that there should be no way in principle to decide whether Advil was good or bad for you, all things considered, even if we stipulate that we know all facts about your life in each of the three worlds we discussed. In the actual world, the Advil you took made away with your headache and caused a lesser ill: mild nausea. You never discover the aspirin tablet in your shirt pocket; no unpleasant “Damn it, if only I had remembered!” Looking back at your decision to take Advil, you would probably say it was a good choice. And who would not agree?

Suppose you get into a bar fight. Somebody throws a bottle at you. The bottle hits your nose and causes it to fracture. The result is pain, an unpleasant visit to the doctor and a few weeks with an irritating bandage on your face. My guess is that none of us would find it easy to think of the bar incident as something that was good for you. However, that is what A’ commits us to, if we further assume that you would have lost an eye otherwise. Perhaps we can explain away our sense that breaking your nose was bad for you. We could point out that most people would likely have not found it odd if, when in your situation, the doctor had told them how lucky they were. “That was close. If the bottle had hit you just a single inch toward either one of your ears, you would have lost one of your eyes forever. You are one lucky woman. It’s a good thing you only broke your nose in the fight.” If we do not want to bite this bullet, we can slightly revise the hybrid account of extrinsic prudential value. We could say that for a state of affairs/event (P) to be extrinsically good for a person (S) in a case where P does not lead to net intrinsic good for S, P must lead to the reduction of actual pain that temporarily precedes P. If we revise A’ accordingly, we thereby remove every reference to possible worlds from my account. Such a revised account would not only account for the intuition that breaking your nose was bad for you, but also lead to a plausible result in the previous example insofar as taking Advil led to the reduction of actual pain that was already there when you took the tablet. Note, however, that in those cases in which a state of affairs or an event (P) does not lead to net intrinsic good, the reduction of actual pain that temporarily precedes P is merely a necessary and not a sufficient condition for P being prudentially good. If P also causes more pain than would have occurred otherwise, P is not good for the person who suffers that pain. We are back to “would have” considerations, and hence to possible worlds. I will stop this line of thought here with the humble satisfaction of having found necessary conditions for something having extrinsic prudential value, leaving a loose end for others to deal with. I nevertheless hope that I said enough to make this approach seem worthy of further development.

Before we conclude this paper, a warning might be in place for those who are inclined to hold on to an essentially counterfactual account of extrinsic prudential value. In one of his numerous examples, Feldman asks whether it would be bad for him to die en route to Europe on an airplane trip. “D directs us to consider the life [he] (…) would lead if [he does] (…) die en route to Europe on this trip, and to consider the value for [him] (…) of this life” (Idem, 151). He then proceeds to calculate the value-for-him of the life he actually leads and the value-for-him of a life that would end in an airplane crash. His calculations are based on an assumption Feldman takes to be innocent, namely that “past pleasures and pains would be unaffected” (Ibidem) if he died in an airplane crash. This assumption is problematic. Just like we can imagine that a small bird hitting the engine of an airplane can cause death and suffering, and hence have a tremendous effect on the lives of the passengers, a possible world with no bird hitting the engine may require a significantly different past. We might call this reverse butterfly effect. It is very difficult, if not practically impossible, to know how the distant past would have had to be for a bird now not to hit the airplane engine. “Seldom, if ever, can we find a clearly true counterfactual about how the past would be different if the present were somehow different. Such a counterfactual, unless clearly false, normally is not clear one way or the other” (Lewis, 1979: 455). While this observation raises no conceptual problems for either Feldman’s or my account, it is worth pointing out that both accounts might be of little use in practice with regard to many states of affairs and events we would wish to evaluate.[24]

Conclusion

In this paper, I argued that the event of death is almost always good for the one who dies. This conclusion is conditional on the assumptions that a person stops existing at the moment of death, and that welfare hedonism is true. In adopting an Epicurean framework, I followed Feldman’s example, and I argued against his deprivation account of death’s supposed occasional badness. The alternative I developed in this paper adequately accounts for a significant range of widely held intuitions about prudential value. My hybrid account of extrinsic prudential value implies that death, while being almost always good for the one who dies, is often less good than not dying. I concluded with a discussion of some puzzles that remain for my account, and I hinted at possible ways to deal with these puzzles.

 

Acknowledgements

I am grateful to Jennifer Blumenthal-Barby, Christopher D. Belshaw, Susanne Burri and Fred Feldman for their insightful comments which helped me to make significant improvements to this paper, and for their encouragement. I am also grateful to an anonymous reviewer for constructive comments and suggestions. I presented earlier versions of this paper at the Departments of Philosophy at Jahangirnagar University in Savar and Rice University in Houston, and at the IV Meetings on Ethics and Political Philosophy at the University of Minho in Braga. I would like to thank the students and faculty members present on those occasions for many valuable questions and comments. Finally, I am indebted to the Department of Philosophy at Rice University for its generous financial support for the research that led to this paper.

 

References

Benatar, David (2006), Better Never to Have Been, Oxford, Oxford University Press.         [ Links ]

Bradley, Ben (2007), “How Bad Is Death?”, Canadian Journal of Philosophy 37, pp. 111-127.         [ Links ]

— (2009), Well-Being and Death, Oxford, Oxford University Press.         [ Links ]

Draper, Kai (1999), “Disappointment, Sadness, and Death”, The Philosophical Review 108, pp. 387-414.         [ Links ]

— (2004), “Epicurean Equanimity Towards Death”, Philosophy and Phenomenological Research 69, pp. 92-114.         [ Links ]

Epicurus (1940), “Letter to Menoeceus”, in Whitney J. Oates (ed.), The Stoic and Epicurean Philosophers, New York, The Modern Library.         [ Links ]

Feldman, Fred (1992), Confrontations with the Reaper, Oxford, Oxford University Press.         [ Links ]

— (2002), “The Good Life: A Defense of Attitudinal Hedonism”, Philosophy and Phenomenological Research 65, pp. 604-628.         [ Links ]

Lewis, David (1979), “Counterfactual Dependence and Time’s Arrow”, Noûs 13, pp. 455-476.         [ Links ]

Luper, Steven (2004), “Posthumous Harm”, American Philosophical Quarterly 41, pp. 63-72.         [ Links ]

— (2009), Death, available at http://plato.stanford.edu/entries/death/, accessed on 08/09/2013.         [ Links ]

McMahan, Jeff (1988), “Death and the Value of Life”, Ethics 99, pp. 32-61.         [ Links ]

Nagel, Thomas (1979), Mortal Questions, Cambridge, Cambridge University Press.         [ Links ]

Nozick, Robert (1974), Anarchy, State, and Utopia, New York, Basic Books.         [ Links ]

Smuts, Aaron (2012), “Less good but not bad: In defense of Epicureanism about death”, Pacific Philosophical Quarterly 93, pp. 197-227.         [ Links ]

Suits, David B. (2001), “Why death is not bad for the one who died”, American Philosophical Quarterly 38, pp. 69-84.         [ Links ]

 

[Submetido em 15 de junho de 2013, re-submetido em 31 de agosto de 2013 e aceite para publicação em 10 de setembro de 2013]

 

Notes

[1] Whenever I write “death” in this paper, I shall always mean the event of death, i.e., the event that occurs at the end of the process of dying. The event of death must be carefully distinguished from both dying and the state of being dead. Dying, of course, can be, and often is, painful and hence bad for the one who is dying. Another distinction that must be made is that between badness-for-somebody and other kinds of badness. Even if it is true that death is not bad for the one who dies, death might still be bad in other ways. For example, your death might be bad for those who care for you, or for those who depend on you. And the death of a poet or a great scientist might be bad for the world, even if there is nobody who grieves her death.

[2] Feldman is, by far, not the only author who advocates a deprivation approach to the value of death. Other advocates of this approach include Thomas Nagel (1979), Jeff McMahan (1988), and Ben Bradley (2009).

[3] Even though I will conveniently talk about persons and people, most of what I say in this paper equally applies to all beings who are capable of experiencing pleasure and displeasure, including human beings who are not persons, all non-human mammals, birds, amphibians, reptiles and fish, and maybe some invertebrates.

[4] Relevant literature can be found in the bibliography of (Feldman, 1992). In particular, cf. p. 129 n. 4 and p. 144 n. 2.

[5] Pain here must not be understood as sensory pain (e.g., the feeling usually associated with bodily injury). While sensory pain is a feeling we typically dislike, it is not always unpleasant. Sometimes, sensory pain can even be enjoyable (e.g., erotic spanking). Also, not all unpleasant experiences are painful (in the narrow sense): think of anxiety, fears and other ills. In his paper, “The Good Life: A Defense of Attitudinal Hedonism” (Feldman,2002), Feldman argues that what is ultimately bad for us is taking displeasure in things. We will pretend that, in the current context, he loosely speaks of pain, instead of using the more precise term, attitudinal pain, for the sake of convenience only (although he sometimes slips and talks about “feeling pain”).

[6] Remember that we use the term “death” to refer to the instantaneous event of death.

[7] While (2) is uncontroversial, (1) and (3) are not. (3), for example, rests on the controversial assumption that only pain is intrinsically bad for us. Robert Nozick’s experience machine (Nozick, 1974: 42-45) and Nagel’s deceived businessman (Nagel, 1979) both have been taken to be counterexamples to this assumption. However, the question whether these alleged counterexamples are successful is beyond the scope of this paper.

[8] Given that Epicureans believe that not even death, which prevents us from having any further experiences ever, is bad for us, it is unlikely that they would share Feldman’s intuition that merely getting less of the good stuff than one could have gotten, while still getting some, is bad for us. Hence, Feldman’s College example would probably not convince Epicureans that their causal hypothesis is false.

[9] Bradley defends a similar principle: “CMV: The overall value, for a person x, of an actually occurring or obtaining event or state e = the value of x’s actual life minus what the value of x’s life would have been had e not occurred or obtained” (Bradley, 2007: 115). The counter-examples to D that I will present in this paper also apply to CMV.

[10] I will not be travelling to Tuvalu one hundred years from now because there will be no thing that is me then. I will die in less than hundred years. However, if I were to travel to Tuvalu one hundred years from now, and if I had a lot of fun there, then the prudential value of my life would be higher, all else being equal. Hence, D implies that not travelling to Tuvalu one hundred years from now is extrinsically bad for me.

[11] Aaron Smuts discusses structurally similar counterexamples to the deprivation account in (Smuts, 2012: 205 ff.).

[12] Even though distinctions between the terms “evil,” “bad,” “harm,” and “misfortune” can be made, I am using them interchangeably throughout this paper to refer to those things, and only those things, that are (intrinsically or extrinsically) bad for people.

[13] A defender of Feldman’s account could reply that there are no negative facts and, hence, not being called is not a state of affairs in the relevant sense. I do not find this objection convincing. After all, not being hit by a car seems to be good for the one who would have been hit. In any case, the previous two examples still stand and we can disregard the puzzles arising from negative facts for the purpose of this paper.

[14] When I say that something leads to net intrinsic good for S, I mean that it leads to more pleasant experiences that are S’s than painful experiences that are S’s. Similarly, something leads to net intrinsic bad for S, if, and only if, it leads to more painful experiences that are S’s than pleasant experiences that are S’s.

[15] As will become clear later, my judgment diverges from Feldman’s in cases where further life will contain more pleasure than pain.

[16] Kai Draper calls benefits of the latter kind negative benefits, cf. (Draper, 2004: 102).

[17] Given a particular state of affairs or event (P), and given a particular person (S), P is either good, bad, or neutral for S, all things considered.

[18] And, even if we cannot know which one it is, we surely want things to be either good, or bad, or neutral.

[19] The comparative value/disvalue for a person (S) of a state of affairs or event (P) is the difference between the intrinsic value for S of the life S would lead if P is true and the intrinsic value for S of the life S would lead if P is false.

[20] For a good discussion of comparativism more generally, see (Luper, 2009).

[21] This might actually not be true. See my discussion of so-called backtrackers below.

[22] Cf. the first part of (Feldman, 1992), “The Nature of Death”.

[23] This example is loosely based on a similar example David Benatar provides in support of an asymmetry between pleasure and pain that is related, yet not identical, to the one I am defending; cf. (Benatar, 2006: 35). Of course, the example raises a number of difficult questions that I cannot address in this paper. E.g., does it even make sense to talk about what is good or bad for possible beings that will never exist?

[24] In some cases of death, what Lewis calls back-tracking counterfactuals might lead to perplexing results. Suppose I die from cancer. If I had not died from cancer, I would likely not have had terminal cancer in the first place. The closest possible world in which I do not die from cancer would hence not only contain additional pains I will experience in future, but it would also not contain cancer pains I experienced in the past in the actual world. If the latter pains outweigh the former pains, my cancer death was not good for me. Note that it would still not be bad insofar as it did not cause any pain. We would have to say that death had no extrinsic value for me. It seems odd to say that death is less good if it is preceded by a terrible diseases but this just shows the profound difficulties of evaluating counterfactuals.