RESPONDING TO THE TIMING ARGUMENT

: AccordingtotheTiming Argument, death isnotbadforthe individual who dies, because there is no time at which it could be bad for her. Defenders of thebadnessof death haveobjectedtothis influential argument, typically by arguing that there are times at which death is bad for its victim. In this paper, I argue that a number of these writers have been concerned with quite different formulations of the Timing Argument. Further, and more importantly, I show that their objections to the Timing Argument fail as attempts to refute the argument in its most challenging form.


Introduction
In an attempt to convince Menoeceus that 'death is nothing to us,' Epicurus famously pointed out that 'as long as we exist death is not with us; but when death comes, then we do not exist.' 1 On a popular interpretation of these words, Epicurus argues that because death can be bad for its victim neither before nor after it has occurred, it cannot be bad for her at all. 2 This is the rough idea behind the Timing Argument, which is commonly reconstructed in the following way 3 : 1 Anything that is bad for a person is bad for her at some time. 2 There is no time at which death is bad for the person who dies. 3 Therefore, death is not bad for the person who dies.
In order to resist the Epicurean conclusion that death is not bad for the one who dies, anti-Epicureans typically argue that, contrary to premise (2), there are times at which death is bad for its victim.In this paper, I argue that, on closer inspection, a number of these anti-Epicurean critics have been suggests, must be the same as asking for the time at which this condition holds.In other words, Feldman understands (TQ) in the following way: (TQ1) At which times is a person's life (as a whole) intrinsically worse than the life (as a whole) that the person would have had if her death had not occurred?While it is not evident what the correct answer to this question is, it seems fairly clear that it is either 'at all times' or 'at no time' (i.e., atemporally). 5or, plausibly, the total amount of intrinsic value a life contains cannot vary over time; that one life is intrinsically worse than another is as invariable as, for example, the fact that Emily Brontë's life was shorter than Franz Kafka's.
As far as I know, no one has accused Feldman of giving the wrong answer to what he takes to be the appropriate interpretation of the question of when death is bad for its victim.However, many writers have found his interpretation misplaced with respect to the Timing Argument: (TQ1) fails to capture what makes (TQ) a challenging question. 6I believe this diagnosis is correct.In particular, what makes (TQ1) a less than ideal interpretation of (TQ)at least in the context of discussing the Timing Argumentis that even if we could not identify a time of the kind with which (TQ1) is concerned, we would not thereby have any reason to doubt that death can be bad for its victim.To see why, note that if there is no such time in the case of deathi.e., if the relevant condition holds atemporally, rather than eternallythen, plausibly, there is no such time in the case of any other intuitively bad event either.For instance, if the relevant condition holds atemporally, then there is no time at which a person's actual life (as a whole) is worse than the life (as a whole) that she would have had if she had not, for example, broken her arm.But surely, this is not a good reason to deny the badness of arm fractures.That would be like denying that Brontë's life was shorter than Kafka's on the grounds that there is no time at which Brontë's life was shorter than Kafka's.
It has sometimes been suggested that, rather than providing the times at which a person's death is bad for her, Feldman provides the times at which it is true that the person's death is bad for her. 7In other words, the question to which Feldman has given a (perhaps correct) answer, according to some writers, is 'At which times is it true that the person's death is bad for her?' But the interpretation of (TQ) that he ought to answer in order to refute the Timing Argument, they suggest, is rather the following: (TQ2) At which times t is it true that a person's death is bad for her at t? Two comments are in order.First, it should be noted that, as Jens Johansson points out, it does not seem to be some idea of true propositions' being eternally true that drives Feldman to his eternalist stance.Feldman's proposal seems to appeal only to the idea that if one life in its entirety is intrinsically worse (better) than another life in its entirety, then the first life is always intrinsically worse (better) than the second life.Accordingly, even a person who believes that true propositions are timelessly true could accept that one possible life is intrinsically worse than another at all times. 8Second, and more importantly, (TQ2) as such sheds no light on the part of (TQ) that needs to be clarified, namely, what it is for an event, such as death, to be bad for a person at a time.Hence, unless we already have an idea of what it is for something to be bad for a person at a time, (TQ2) is not of much help when trying to find a suitable interpretation of (TQ).Indeed, while (TQ1) is a suboptimal candidate for other reasons, it at least offers a way of understanding a sentence of the form 'event E is bad for person S at time t.' (TQ2), on the other hand, does not spell out the notion of an event being bad for someone at a time at all.Thus, (TQ2) is not an improvement over (TQ1).
The main lesson to draw from Feldman's response to Epicurus is that a suitable interpretation of (TQ) should provide a sense of an event's being bad for a person at a time, such that the times at which breaking one's arm and other bad non-death events are bad for their subjects are, at least in principle, easily located, whereas the times at which the event of death is bad for the deceased are not.On Feldman's interpretation, there is nothing about death that makes it particularly difficult to locate the times at which it is bad for its victim; if there are times at which the badness of any event is located, then the badness of death, too, is located at precisely those times.This takes us to what, from an Epicurean point of view, seems to be the most promising interpretation of (TQ): (TQ3) At which times t is a person's well-being level at t lower than it would have been if her death had not occurred? 9like (TQ1)and, for that matter, (TQ2) -(TQ3) points to what seems to be a noteworthy difference between death and other intuitively bad events.To see how this difference emerges and why it poses a threat to the badness of death, let us formulate a new version of the Timing Argument in accordance with (TQ3): (1*) An event E is (overall) bad for a person S only if there is a time t such that S's well-being level at t is lower than it would have been if E had not occurred.
(2*) There is no time t such that S's well-being level at t is lower than it would have been if the event of S's death had not occurred.(3*) Therefore, the event of S's death is not (overall) bad for S.
In what follows, I will refer to this formulation of the Timing Argument as 'the Timing Argument*.'To see why it poses a serious challenge to defenders of the badness of death, consider first what seems to be a straightforward reason to accept premise (1*).
With respect to events that we normally consider to be bad for us, such as breaking one's arm, getting fired, having a headache, and going into a depression, it seems clear that there are times at which the subjects of these events are worse off than they would have been if the events had not occurred.If I break my arm, for instance, then the times after the accident at which I am in pain are presumably such that my well-being level at those times is lower than it would have been had I not broken my arm.Other bad events make their subjects occupy a lower well-being level at times without causing them any painful experiences (or whatever might be thought to have negative intrinsic value for a person).To borrow an example from Bradley, suppose that, unbeknownst to me, someone has left baseball tickets in my mailbox, but before I find out about the tickets, someone steals them from my mailbox. 10Plausibly, assuming that I would have enjoyed the game, someone's stealing the tickets is bad for me, even if it does not cause me anything intrinsically bad.But although there is no time at which any intrinsic bad befalls me as a result of the theft, I nonetheless occupy a lower well-being level at certain times after the thefte.g., during the game that I missthan I would otherwise have done.(Similarly, the times at which I am in pain because of my broken arm need not be the only times at which my well-being level is lower than it would otherwise have been; after the pain has vanished, the arm fracture might still prevent me at certain times from enjoying things that I would otherwise have enjoyed.)Moreover, if it were to turn out that neither my arm fracture nor the baseball ticket theft actually made me occupy a lower well-being level at any time than I would otherwise have done, then, intuitively, this is a reason to doubt that either of these events was in fact bad for me.This suggests that the fact that events that we typically consider to be bad for us fulfill the necessary condition in premise (1*) is not a merely contingent fact: in order for a particular event to be bad for me, the event must be such that I am worse off at some time than I would have been if the event had not occurred.If this is correct, then so is premise (1*).
Next, consider premise (2*).Could there be a time at which a person would have been better off if her death had not occurred?It seems not.When a person dies, she goes out of existence, and so does her well-being level.(At least, this seems to be the intuitively correct thing to say.I return to this issue in Section 4.) Thus, no matter how high her well-being level would have been at times after her actual death if her actual death had not occurred, there is no time after death at which she occupies a lower well-being level than she would otherwise have doneafter all, occupying a lower well-being level at a given time requires occupying a well-being level at that time.It is not much easier to see how a person's well-being level could be lower at times before death than it would have been if her death had not occurred.Plausibly, just as my well-being level on my fifth birthday cannot be affected by what happens today, my well-being today cannot be affected by my eventual death (but see note 31).Those who wish to reject premise (2*), then, can do so only at the cost of giving up certain very plausible claims.Now, it might be objected that since premise (2*) is hard to deny, and it is obvious that death is, at least in some cases, bad for the person who dies, the event of death is a good counterexample to premise (1*).The worry, however, is that this gives rise to an unappetizing asymmetry with regard to death and other bad events: while nearly all bad events are such that their subjects are worse off at some time than they would have been if the event had not occurred, the event of death is not.Several writers have shown reluctance to make death a special case in this way. 11Their reluctance is understandable; ideally, anti-Epicureans should be able to defend their view by showing that the badness of death is no more mysterious than the badness of other, more ordinary, evils.In all axiologically relevant respects, death is supposed to be similar to, for example, being robbed of baseball tickets without ever finding out about it.But by treating death differently, anti-Epicureans 'only give comfort to the Epicureans, who will view their doubts about the evil of death as having been justified by our need to make distinctions between ordinary evil and the evil of death.' 12 Note that I am not suggesting that the Timing Argument* is the most natural way of formulating the Timing Argument, or that it correctly captures what we actually mean when we talk about events being bad for people at times.What I am suggesting is that the Timing Argument* is an interpretation of the Timing Argument that, unlike other interpretations, makes the argument seriously challenging.This is important, for suppose that the most plausible way to understand a sentence of the form 'E is bad for S at t' does not correspond to how the premises of the Timing Argument* are formulated (or to how (TQ3) is formulated).Would this automatically make the Timing Argument* an unsuitable formulation of the Timing Argument?No, not if we are primarily interested, as we should be, in investigating whether or not death can be bad for the one who dies.If that is our primary goal, then we ought to focus on the formulation that makes the argument as challenging as possible, and the Timing Argument* is at least a good candidate for being such a formulationeven if it deviates from what might be considered the most natural interpretation of the Timing Argument. 13As we shall see in the next section, however, several critics of the Timing Argument have been guided by interpretations of what it is for an event to be bad for a person at a time, which, while natural, are only relevant to formulations of the Timing Argument that are less challenging than the Timing Argument*.

LAMONT AND WOLF
Julian Lamont and Aaron Wolf argue that the time at which death is bad for the person who dies is the time at which the person dies. 14Given how they understand the question of the timing of death's badness, what they suggest is quite natural.Consider first Lamont's proposal.Concerning the question of when death is bad for the one who dies, he writes: As should be obvious, in many cases it is straightforward for us to answer these important 'when' questions.When did Peter harm Paul? -8 pm, last Wednesday, when he punched him in the nose; 10.15 am on December 15, when he used Paul's card to withdraw all the money out of his bank account, etc.The problem has been that many have viewed the parallel question in the case of death as more puzzling and not so easy to answer.15 According to Lamont, 'the parallel question in the case of death' is no more difficult to answer than the question of when Peter harms Paul: just as Peter harms Paul at the time of Peter's harmful actions, death harms its victim at the time of its occurrence.The question of when a harmful death harms its victim, then, is simply a question of when the harmful eventdeathoccurs.As an answer to this question, Lamont's 'at the time of death' seems indisputably correct.
Wolf argues that death is bad for us when it makes us worse off.He suggests that reflecting on cases of non-lethal deprivations 'lends support for the view that deprivations make us worse off at two times: when the event that deprives us occurs, and when the events of which we are being deprived would have occurred.' 16However, because things cannot go well, badly, or neutrally for a person when she does not exist, Wolf argues, we must make an exception in the case of death and say that death makes its victim worse off only at the time of the event of death.While I do not share Wolf's intuition that deprivations normally make their victim's worse off at two times, there is clearly a sense in which it is true that death makes its victim worse off at the time of death (in cases in which death constitutes a deprivation).If we take the question 'When does death make its victim worse off?' to mean 'When does death cause its victim to be worse off?' then 'at the time of death' appears to be the most appropriate answer.Plausibly, if death causes its victim to be worse off, then the time at which it does that is the time of death itself.
But even if Lamont and Wolf's suggestion that death is bad for its victim at the time of death is plausible on some interpretations of an event's being bad for a person at a time, it is unsatisfactory as an answer to (TQ3).For note that, even if death makes its victim worse off, on the whole, than she would have been if her death had not occurred, it need not be the case that the person is worse off at the time of death than she would otherwise have been. 17For instance, we can imagine a person who, at the time of death, would have been in great pain were it not for her death, but whose future as a whole would nonetheless have been on balance intrinsically good for her.Anti-Epicureans should find such a death bad for her.In such a case, however, the time of death is not a time at which the victim is worse off than she would otherwise have been.Thus, if we want an answer to (TQ3) that holds for more than some bad deaths, then the answer 'at the time of death' is not what we are looking for.

FISCHER
Several philosophers have argued that death is bad for its victim at times after death, but few of them have been concerned with (TQ3) and the Timing Argument*. 18Consider first John Martin Fischer, who finds the timing of death's badness one of the most puzzling issues about death. 19However, he argues, acknowledging that events can be bad for us without being able to causally affect us should help us see that there is a simple answer to the question of the time of death's badness.To illustrate that an event can be bad for a person even though it occurs at a spatial distance too big for the event to be able to causally affect its victim, Fischer describes a case in which you are betrayed by two people at a colony orbiting Mars while you are on earth.
The Mars example drives this point home: one can be harmed by something that is spatially remote and from which one is causally isolated (barring violations of the laws of physics).But if this is so, then it is plausible to say that one can be harmed by temporally distant events (from which one is causally isolated).If you are harmed by the betrayal in the Mars case, why not also say you can be harmed by your death, even though the death occurs after you cease to exist?These reflections, then, suggest simple answers to some of the most perplexing puzzles pertaining to the badness of death.[…] The time of the misfortune is the time during which you are dead.20   Fischer takes the difficulty of finding a time for death's badness to be connected to what might be considered a more general problem, namely, how something can be bad for a person even though it cannot causally affect her.But once we, by reflecting on cases like the Mars example, realize that spatially distant things can be bad for us without being able to causally affect us, it should be clear that temporally distant things, such as our own eventual deaths, can be bad for us without being able to causally affect us.Therefore, Fischer concludes, just as a misfortune can be spatially located in the vicinity of Mars, a misfortune can be temporally located at times when you no longer exist.And if the misfortune is death, then the misfortune must be located at 'the time during which you are dead.'Fischer's approach to determining the time of death's badness, then, is importantly similar to Lamont's.While they end up locating the badness of death at different times, both of them seem to offer an answer to the question of where the bad thing, or the 'misfortune,' is temporally located.The reason that their answers differ nonetheless seems to be that they have different misfortunes in mind.In Lamont's case, the relevant misfortune is the event of death.In Fischer's case, it is rather the status of being dead.At least, that is what the above quote suggests; arguably, the claim that 'death occurs after you cease to exist' as well as the claim that death is something from which you are 'causally isolated' make sense only if death is understood not as the event of death, but as the status of being dead.If this is the misfortune Fischer has in mind, then it seems correct that the misfortune lasts from the time of the event of death to the end of time.
Clearly, though, Fischer is not concerned here with the more challenging task of locating the times that (TQ3) asks for.Nor do the times at which Fischer locates the badness of death correspond to the times at which the deceased's well-being level is lower than it would have been if her death had not occurred.First, in order for any time after death to be such that the dead person is worse off at that time than she would otherwise have been, it must be the case that the person has a well-being level when she is dead.As noted in Section 2, however, the most natural view is that dead people lack well-being levels, and Fischer does not provide any reason to think otherwise.Second, Fischer does not seem to locate the relevant times even if we assume (controversially) that a dead person does have a well-being level.For note that if I were to die tomorrow, then it would not be the case that I am worse off a thousand years from now, say, than I would have been if my death had not occurred.Obviously, a thousand years from now I will be dead anyway, and no matter whether I will have a well-being level after death or not, I will not have a negative well-being level at that time.But if death is bad for me at all times when I am dead, as Fischer suggests, then this includes times a thousand years from now.
Recently, Fischer has offered a different and more detailed (partial) defense of the view that, in cases where death is bad for its victim, it is bad for her after death. 21The primary aim of the defense is to show how advocates of this view can successfully address the so-called Problem of Predication: if, as seems reasonable, a person goes out of existence when she dies, how can she have the property of 'being harmed' after her death?This problem, Fischer argues, can be solved by a 'truthmaking account' of the time of death's badness. 22A crucial component of this account is the idea that a particular death is bad for the person who dies in virtue of a comparative proposition of the following kind being true: the person's actual life is on-balance worse than the life she would have had if her death had not occurred (cf. the Deprivation Account).To determine the time at which a particular death is bad, Fischer suggests, we should focus on the truthmaker of this comparative proposition.More specifically, he suggests that '[t]he time of death's badness for the deceased individual is the time of the occurrence of the events included in the truthmaker of the comparative proposition.' 23In Fischer's view, these truthmaking events are all located during the period of time that begins right after the person's death in the actual world and ends at the person's death in the hypothetical world, that is, the closest possible world in which her actual death does not occur.After all, it is during this period of time that all the events that make the hypothetical life better than the actual life take place.Therefore, according to Fischer, in cases where death is bad for the deceased, it is bad for her during this entire period of time.And since this truthmaking approach allows us to locate the badness of death after death without positing that the victim of deathor, for that matter, anyone elsehas intrinsic properties at a time when she does not exist, Fischer concludes that the Problem of Predication can be avoided.
Whether or not Fischer's truthmaking account of the time of death's badness manages to avoid the Problem of Predication, it does not seem to address the Timing Argument*.Locating the time of the occurrence of the events included in the relevant truthmaker is not the same as locating the time at which a person's well-being level is lower than it would have been if her death had not occurred, and only the latter is relevant to answering the challenging timing question, (TQ3).Moreover, as with Feldman's suggestion, the question to which Fischer does provide an answeri.e., the question about when the relevant truthmaker obtainsmay be a difficult metaphysical question as such, but it does not seem to be more difficult than analogous questions about bad non-death events.By contrast, (TQ3) does seem more difficult than analogous questions about such events.
Could the times at which Fischer locates the badness of death nonetheless be the times at which the deceased is worse off than she would otherwise have been?Note that unlike Fischer's earlier proposal, his more recent proposal does not locate the badness of death at all times after death, but only at those times after death that are prior to the person's death in the closest world in which her actual death does not occur.Hence, it does not imply that a person's death is bad for her a thousand years after it has occurred.It does imply, however, that a person's death is bad for her at all times between her actual death and her hypothetical deathindeed, this is something that Fischer emphasizes. 24But surely, even if we assume (again, controversially) that a dead person has a well-being level of zero, far from all cases of bad deaths are such that the deceased has a lower well-being level at all those times than she would have had if her death had not occurred.Presumably, there are many times at which the deceased would otherwise have occupied a neutral or negative well-being level if her death had not occurred, and so would not have had a higher well-being level than a dead person at those times.

PURVES
Purves, too, locates the badness at times after death, and his reasons for doing so are related to Fischer's more recent proposal.In Purves's view, (TQ3) renders the question of when death is bad for its victim 'needlessly difficult to answer.' 25 If we want to refute the Timing Argument, he suggests, we should not try to find a time such that the deceased's well-being level is lower at that time than it would have been if her death had not occurred, but instead locate the times that 'ground the fact that someone's life as a whole contained less value than it would have contained had her actual death not occurred.' 26And these times, he claims, are more easily located.They are the times at which, if the person had not died, the states of affairs that would have obtained at those times would have contributed more to the person's total receipt of intrinsic value than the states of affairs that actually obtain at that time.Since the states of affairs that obtain after a person's death do not contribute to her receipt of intrinsic value at all, Purves holds that death is bad for its victim at all times after death at which the states of affairs that would have obtained if her death had not occurred would have made an on balance positive contribution to her receipt of intrinsic goods.Thus, he claims to be able to locate the badness of death after death without making the controversial assumption that a person occupies a well-being level after death.
Purves may well be right that the times that ground the fact that a deceased person's actual life contains less value than it would have done if the person's death had not occurred are located after death.But unless these times are also such that the person is worse off at these times than she would otherwise have been, these are not the times with which (TQ3) and the Timing Argument* are concerned.Remember that the reason for focusing on this particular formulation of the Timing Argument is that it seems to make the argument as strong as possible.As Purves himself points out, however, the times that he thinks ought to be located in order to temporally locate the badness of death are not as hard to locate as the times that (TQ3) asks for.But if this is so, then this is a reason not to formulate the Timing Argument in accordance with Purves's suggestion.In other words, the fact that Purves's interpretation of (TQ) makes it easier to answer than if interpreted as ( TQ3) is precisely why we should not interpret it in that way. 27nlike Fischer, Purves does not locate the badness of death at all times after death at which the deceased would have been alive if her death had not occurred.Rather, on Purves's view, the badness fluctuates over this period of time depending on whether or not the state of affairs that would otherwise have obtained at each moment would have made a positive contribution to the person's total receipt of intrinsic value.As a response to (TQ3), therefore, the times that Purves suggests are more plausible than those provided by Fischer.But, of course, the main problem of locating the (TQ3)-relevant times after death remains: it is doubtful that a person occupies any well-being level at all when she is dead, and hence doubtful whether she can be worse off (or better off) than anyone or anything after death.

ROBSON
In a discussion of how the growing block view of time can help solve certain puzzles concerning the badness of death, Jon Robson argues that death is bad for the person who dies at the time of death and continues to be bad for her at all subsequent times. 28Robson is thus yet another advocate of the view that death is bad for its victim after death.However, as with the suggestions considered in the previous sections, Robson's suggestion does not provide an answer to the relevant timing question, (TQ3), and thus fails to address the Timing Argument*.
Although Robson does not offer any detailed explanation of why it is bad to die, he suggests, roughly, that death is bad for its victim because death is the cessation of a person's conscious life and 'it is better (all else being equal) to have consciousness than to lack it.' 29As for the times at which death is bad for the deceased, he thinks it plausible that it starts being bad for her at the time of her death, at which she ceases to be a conscious being, and remains bad for her for as long as she stays in this non-conscious condition, that is, forevermore.This is not because he thinks that the deceased is worse off at these times than she would have been if her death had not occurred -Robson finds counterfactual approaches to determining the badness of events 'misguided across the board.' 30 Rather, his view seems to be that, intuitively, being deprived of some particular good, such as consciousness, is bad for as long as one lacks this good.
Perhaps Robson is right that this is the intuitively right thing to say about deprivations.But even if it is, the times at which he locates the badness of death are plausibly not relevant to (TQ3).For even if we assume, contrary to most standard theories of well-being, that, other things being equal, it is better for a person to have consciousness than to lack it, a dead and hence non-conscious person cannot be worse off at times after death than she would otherwise have been unless she has a well-being level at those times.Showing that the dead have well-being levels, however, is a challenge that Robson does not address.Moreover, as noted in my discussion of Fischer's first suggestion, a person who dies today is clearly not worse off a thousand years from now than she would otherwise have beeneven if we grant that she has a well-being level after death.For this reason too, then, the times at which Robson locates the badness of death are not plausibly the times that (TQ3) asks for.

Who occupies a well-being level?
The only times that the anti-Epicurean should look for when attempting to reject the Timing Argument*in particular, premise (2*)are the times at which a person is worse off than she would have been if her death had not occurred.As noted in Section 2, the main obstacle to locating these times is that people do not seem to have well-being levels after deathan assumption that none of the proposals considered in Section 3 provided any reason to doubt. 31In this final section, I want to return to this assumption and say a bit more about the costs of rejecting it.Before closing, I suggest that even if considerations concerning the difficulty of occupying a well-being level speak against rejecting premise (2*), they may well speak in favor of rejecting premise (1*) instead.Some find it obvious that a person does not occupy a well-being level when she is dead. 32I think this is too quick.After all, given that my well-being level at a time is plausibly determined by my receipt of intrinsic goods over intrinsic bads at that time, it is natural to think that I have a well-being level of zero at times when I receive no intrinsic goods or bads.Indeed, with respect to any typical human adult, it seems clear that if she receives nothing intrinsically good or bad at a given time, then she occupies a well-being level of zero at that time.Considerations of this sort have led anti-Epicureans such as Bradley and Feit to suggest that since a person does not receive anything of intrinsic value at any time at which she is dead, she has a well-being level of zero, rather than no well-being level, at all times after her death. 33owever, even if the view that a person has a well-being level of zero after death is worth taking seriously, it has certain counterintuitive implications that cannot be ignored.Suppose, for instance, that I currently occupy a negative well-being level.If Bradley and Feit are right, then every dead person is now better off than I am.This seems odd; no matter how bad a day I am having today, it is hard to see how, for example, Epicurus or Elizabeth Taylor could have a better one.Moreover, it seems wrong to say that someone who occupies a negative well-being level during the last moments of life, perhaps because of a painful dying process, gets a well-being boost immediately after death.A more natural view is that, in cases in which death comes as a merciful release, the person who dies goes from faring badly to not faring at all.
A further problem for Bradley's and Feit's view is that if simply not receiving any intrinsic goods or bads at a time is sufficient for occupying a well-being level of zero at that time, then dead people are not the only dubious members of the well-being club; shoes, rocks, and other inanimate objects, too, fulfill the relevant condition. 34But, intuitively, just as a shoe, for instance, does not occupy a zero level of health because it is neither healthy nor unhealthy, it does not occupy a zero level of well-being because it receives no intrinsic goods or bads.
If it makes sense to ascribe well-being levels to dead people and shoes, then well-being levels must be quite different from, for example, health levels.Bradley thinks that they are.In defending the view that dead people have well-being levels, he suggests that 'we stop thinking of well-being as being very much like health, and rather think of it as essentially connected to prudential deliberation.' 35Because it is prudentially correct for a person to prefer a future in which she fares well to a future in which she no longer exists, Bradley argues, we should conclude that she is better off in the first future than she is in the second, and so that she has a well-being level at times when she does not exist.On this way of thinking about well-being levels, however, occupying a level on the well-being scale is radically different from occupying a level on any other scale that might be thought of as comparable to the well-being scale.As David Hershenov notes, it is also prudentially reasonable for a person to be indifferent between two possible worlds in which she never exists. 36Thus, if Bradley is right about the essential connection between well-being and prudence, it seems that we all now have well-being levels (of zero) in possible worlds in which we never come into existence.While this is an interesting view that may come in handy for solving certain puzzles about, for example, the harm of creation, it requires that we give up some very plausible assumptions about well-being in particular, and properties and relations more generally.This is a concession that the anti-Epicurean should avoid if possible.
The prospects of a successful defense of the view that people have well-being levels when they are dead may not be hopelessly dim, but neither are they very bright.Let us suppose, then, that this view is false, and hence that occupying a well-being level at a time is harder than Bradley and Feit think it is.One might then wonder how hard it is for an individual to occupy a well-being level at a time.More specifically, one might wonder what accounts for the fact, if it is a fact, that human adults typically have well-being levels whereas dead people do not.
I believe reflection on this question suggests an alternative way for the anti-Epicurean to object to the Timing Argument*: to criticize premise (1*). 37To begin to see how, note that having a well-being level at a given time plausibly requires more than merely existing at that time.For example, the fact that my left shoe exists right now does not seem to imply that it currently has a well-being level.Nor does it seem enough, for something to have a well-being level, that it is alive in the biological sense.Plants are living things, but arguably lack well-being levelsat least in the relevant, nonderivative, sense of 'well-being level.'What is the difference, then, between typical human adults, on the one hand, and plants and shoes, on the other, that makes it correct to attribute well-being levels to the former, but not to the latter?A promising suggestion is that the relevant difference has to do with the fact that, unlike shoes and plants, typical human adults have psychological features.This way of drawing the line between those who have well-being levels and those who do not is intuitively plausible; the reason that I cannot be better or worse off than the oak tree outside my window, for instance, seems to be that the oak tree has no psychology whatsoever, and so cannot be faring well or badly at all.Moreover, since there is no psychological difference between a shoe and a dead person, the present suggestion plausibly implies that a person loses her well-being level when she dies.However, it arguably also implies that some individuals lack well-being levels at times when they are still alive.For example, a person who falls into a permanent comai.e., a coma in which she spends the rest of her lifedoes not seem to have the psychology required for having a well-being level when she is in the coma; with respect to psychological features, a comatose person who remains comatose until her death appears to be on a par with a dead person.But if such a permanently comatose person does not have a well-being level when she is comatose, then it follows from premise (1*) that no matter how much better her life would have been had she not fallen into the coma, falling into the coma was not bad for her.This is hard to swallow.Hence, there is reason to be skeptical about premise (1*). 38aturally, it might be objected that the event of falling into a permanent coma is too similar to the event of death to be dialectically interesting. 39et us therefore note that cases of temporary coma, too, cause trouble for premise (1*).For if we accept that a person in a permanent coma lacks the psychological features required for having a well-being level, then it is difficult to see why we should not also accept that individuals who are only temporarily comatose lack these features.Even if a temporarily comatose person eventually wakes up from the coma, she typically does not have any notable psychological features while comatose.And if having a well-being level at a given time requires having such features at that time, it does not matter whether she has those features at a later time, when she is no longer in a coma.Thus, if premise (1*) is true, then it is not bad for a person to fall into a temporary coma unless there is some time after the coma at which her well-being level is lower than it would otherwise have been.But this just seems wrong.Suppose someone falls into a five-week-long coma that deprives her of five weeks of happiness but that does not have any negative (or positive) effects on her well-being level afterwards.Falling into such a coma is obviously bad for her, and so premise (1*) yields an incorrect verdict.
Moreover, if premise (1*) is true, then we seem to end up with some arbitrary results concerning different cases of coma.Suppose someone falls into a five-week-long coma after which she regains her psychological functions, and hence her well-being level, but never actually receives any intrinsic goods or bads before she dies only a day later.Suppose also that she would have had a positive, instead of a neutral, well-being level during the last day if she had not fallen into the coma.Since there are times after the coma at which her well-being level would have been higher had she not fallen into the coma, there is no reason for a proponent of premise (1*) to deny that the coma is bad for her.However, if she had stayed comatose until her death, a proponent of premise (1*) would have to deny that the coma is bad for her.Intuitively, though, whether the person stays comatose one more day or just fails to actually receive any intrinsic goods or bads that day should not make any difference to whether the coma is bad for her or not.
In defense of premise (1*), the Epicurean could argue that, contrary to what I have suggested, a permanently comatose individual does occupy a well-being level while comatose, and so (1*) does not imply that falling into a permanent coma cannot be bad.But by doing this, the Epicurean ends up with another arbitrary verdict: while it is never bad to die, it is sometimes bad to fall into a permanent coma.But, intuitively, if one of these events is bad for its victim, then so is the other.
40 I am grateful to Krister Bykvist, Victor Moberger, two anonymous referees for Pacific Philosophical Quarterly, participants at the Joint Stockholm/Uppsala Seminar in Practical Philosophy, and, especially, Jens Johansson for very helpful comments.Work on this paper was supported by grant 2019-00511 from the Swedish Research Council.