Some of you, this past week, have been carrying a reply. Ever since Epicurus was allowed to say that death is nothing to the one who dies, you have wanted to answer him with a single sentence: but she lost her future. Tonight we state that reply properly, put it in its strongest form — and then we meet a second argument that takes it and points it in a direction no one thinks to defend.
This is Philosophy for Us — philosophy for everyone, no degree required.
This is the fifth of eight episodes on grief and dying. Last time we built Epicurus's argument to its full strength — death is nothing to the one who dies, because once she is dead there is no longer anyone for it to be bad for — and we could not refute it. Tonight we build the natural reply, that death robbed her of her future; and then we meet a second argument, older than the reply itself, that takes that very reply and aims it backward, at the nothingness before she was born.
The reply is the one nearly everyone reaches for. Epicurus keeps asking us to find the dead person having a bad time, and of course we cannot, because she is having no time at all. But that was never the claim. The claim is that death is bad not because it feels bad to her but because it took something from her — the years that were coming, and everything those years would have held. It names the person who was harmed: the one who lost all of it.
So we do two things tonight. First we give that reply and build it to its strongest form, just as we built Epicurus — and it is a serious reply, defended by careful philosophers, and it will feel like the moment we finally answer him. Then we turn it around, because a second argument has a question waiting, and it is a simple one. If death is bad because it robbed her of all the time ahead of her, what about all the time behind her — the endless stretch before she was ever born, when she also did not exist? That nothing took just as much. She could have come into the world a century earlier and had those years, and she did not; and no one grieves those missing centuries, and no one ever dreaded them on her behalf. So why is the one nothing a horror and the other nothing simply nothing at all?
That is tonight. Build the reply to full strength, then meet the argument that uses it against us. Two hard arguments, and we will be holding both by the end. Let us begin with the reply.
So we give the reply — the one many of you have been holding since last time.
Last episode we built Epicurus to full strength and could not refute him. But we wanted to, and the objection we wanted to raise was not a wild one. It was specific, and it is the one nearly every thoughtful person reaches for on first meeting the argument. It runs like this. You keep asking me to find the dead person having a bad time, Epicurus, and of course I cannot, because she is having no time at all. But that was never my claim. I never said death is bad because it feels bad to her. I said it is bad because it took something from her. It took her future — every year she had coming, and everything those years would have held. The harm is not something she feels. It is something she loses.
That reply is worth examining slowly, because the whole episode turns on it. Consider what it actually does. Epicurus built his case on a single hidden assumption: that for something to be bad for you, it has to reach you — register on you, be felt by you. Pain is bad because you are in it. Remove all experience and you remove the only place badness could live. That was the first premise of his argument, and the deepest, and we granted it its full strength last time. The reply does not evade that premise. It denies it outright: a thing can be bad for you that you never feel at all.
And the striking part is that we already believe this everywhere else in life. Consider a man whose closest friends have quietly mocked him for years, behind his back, to everyone he knows — and he never finds out. He dies believing himself well-loved. Was he harmed? Most of us say yes without hesitation. He was betrayed; he was made a fool of; something was taken from him — his standing among the people he trusted — and it was taken whether or not he ever became aware of it. Or consider a promise made at a deathbed and broken the day after the funeral, the will ignored, the estate plundered. The dead woman feels none of it, and still we want to say she was wronged. We make judgments of this kind constantly. We hold that a person can be harmed by what happens outside their experience, even after their experience has ended altogether. Epicurus needs us to believe that badness exists only where feeling reaches. On reflection, we do not believe that, and never did.
The reply, then, in its sharpened form. Death is bad for the one who dies — not because of anything it is like, since it is not like anything; there is no one for it to be like anything to — but because of what it subtracts. It cancels a future. It removes the rest of a life: the ordinary mornings, the work left unfinished, the people not yet met, the years that were coming and now are not. The badness of death is not a dark experience hidden somewhere we cannot observe. It is a gap — the difference between the life she would have had if she had lived and the life she actually got, which stopped. That gap is real, it has a size, and it is bad for the person whose gap it is.
This reply has a name and an author, and both are worth stating, because we are about to rest a great deal of weight on it. The author is Thomas Nagel — an American philosopher, still living, one of the most careful of the last century — who set it out in a short, spare essay with the plainest possible title. He called it "Death." Nagel's step is the one the objection reached for. The evil of death, he argues, lies not in any positive bad thing it contains, but in what it deprives you of. To be deprived of something is to be without a good you would otherwise have had. The prisoner is deprived of freedom; the starving man, of food; and the dead, Nagel says, are deprived of the rest of their lives, and of all the good those lives would have held. Death is bad in exact proportion to the goods it cancels. That is why a death at twenty strikes us as worse than a death at ninety, without our having to do any sums: more was taken.
See what this does to the question that defeated us last time. Epicurus said: point to the one death is bad for, and you will find the place empty. And we pointed, and it was empty — because we were looking for a person standing at the grave, undergoing the bad experience of being dead, and there is no such person and no such experience. The deprivation reply says we were looking in the wrong place. Do not look for someone at the harm, feeling it. Look for the one the future was taken from. And there she is — not a ghost at the graveside, but the living person she was, with a whole life ahead of her that was cancelled. The subject was not missing after all. The subject is the one whose future this was.
That is the reply, and it feels as though it wants to be the answer — the moment we finally defeat Epicurus. It may be. But "it feels like the answer" is precisely the feeling we have learned to distrust, three episodes running. So before we accept it, we do what we have done throughout: we build it to its strongest form, and we make it survive Epicurus's hardest reply, not his weakest. Because he has a reply too, and it is sharper than it first appears.
Start where Epicurus is strongest, because a reply built against his weakest form will look like a victory that is not one. His strong form makes two demands, and the deprivation account has to meet both. Take them one at a time.
The first is the one we have already begun to dismantle: that a harm has to be felt. Epicurus stakes everything on it — no experience, no harm, and the dead experience nothing. So the account's first task is to break that link cleanly rather than halfway. And it breaks it with cases — ordinary ones, in which we already judge in the way the account requires, so that the work is done by our own settled sense and not by any philosopher's device.
Nagel offers two, and both are worth keeping. The first we met a moment ago: the man ridiculed behind his back his whole life, warm to his face, who never learns of it and dies believing himself beloved. We have already agreed he was harmed. What matters in returning to him is the direction of the judgment, which is easy to get backwards. It is not that the betrayal is bad because it would hurt him if he found out. It is the other way round: finding out would hurt because the thing is already bad. The badness is present in the situation, fully formed, before any experience arrives — and in this case no experience ever arrives, and the badness does not go anywhere. So "bad for you" cannot mean "felt by you." We have just judged a clear case in which the two come apart.
The second case is harder, and reaches deeper. Imagine an intelligent, capable adult, at the height of her powers, who suffers a brain injury that leaves her with the mind of a contented infant. She is happy, placid, wanting for nothing she can conceive of, and she will live out her days that way, untroubled. Has a misfortune befallen her? Every instinct says yes — a terrible one. But the contented infant she now is is not suffering. If we look for the victim in the way Epicurus instructed — someone present, feeling the badness — we will not find her, because the person in front of us is perfectly content. The one the misfortune happened to is the adult: the person with all those cancelled capacities, the life of the mind that has been switched off. The harm is the gap between what she was, and could have gone on being, and what she has been reduced to. And that harm is entirely real even though no one anywhere is feeling it. That is the deprivation account in a single case — and it is not death yet; it is only a life cut down to a fraction of itself. Death is simply the limit: the fraction taken all the way to zero.
So the first premise holds, and holds firmly. A harm is a deprivation — a being-without of goods one would otherwise have had — and deprivations are bad whether or not anyone is present to feel them. Epicurus's deepest assumption has been broken on cases we settled for ourselves.
Now the second demand, and this is the one people forget to build against, which is why they think the reply easier than it is. Epicurus, or his sharper successors, will grant everything so far and then ask a single quiet question that has defeated a good many confident replies. Grant all of it, they say: no experience is needed. But a misfortune still has to befall someone at some time. So tell me — when is death bad for her? Name the moment. And the reply immediately runs into difficulty. It is not bad for her now, while she is alive; she is fine, she is reading, she is having lunch. It is not bad for her after she is dead; there is no longer any her for anything to be bad for — which was the whole difficulty we began with. Before, she exists but the harm has not happened; after, the harm has happened but she does not exist. We are back at the empty place, approached from a new angle. So, Epicurus says: you have told me what the harm is — a deprivation — but you cannot tell me when it occurs, and a harm that can never be assigned a time begins to look like no harm at all.
Here is the answer, and it is the most important step in the whole reply, so it is worth taking slowly. The demand to name the exact moment is not as innocent as it sounds. It smuggles back Epicurus's own picture. Why should a harm have to occur at a precise instant? Because pains do. A stubbed toe is bad at one moment and better a few minutes later; experiences are dated, they sit at points on a clock. But the deprivation account has just spent its first half showing that this harm is not an experience — and once that is granted, the right to demand that it behave like one is given up as well. Not every misfortune carries a timestamp. When was the deceived man harmed by the mockery he never knew of? There is no answer, and the absence of one does not trouble us in the least, because the harm was never the sort of thing that happens at a moment. It is a standing fact about his life — that it contained this falseness, that it fell short of what he took himself to have. In the same way, the deprivation of death attaches to the person, across the whole comparison between the life she would have led and the life she got. It is not an event within her biography, waiting to be dated. It is a true thing about her: that her life was cut short of the goods it would otherwise have held. Ask when it is bad for her, and the honest reply is that the question is malformed. It demands of a deprivation an alibi that only an experience could supply.
And that, built squarely, is the point at which the reply holds. Consider how much ground we have recovered, because it is a great deal, and it is genuine. Last episode the challenge — name the one death is bad for — left us pointing at an empty place we could not fill, and careful people cannot fill it either. Tonight we have filled it. The subject is her: not a phantom at the grave, but the person whose future was cancelled, identified not by standing somewhere feeling the badness but by being the one to whom this whole shortened life belongs. The harm is the cancelled future. It requires no experience and no timestamp, and the demand for either turns out to be Epicurus quietly asking the deprivation to be a pain, so that he can win again. A hard-headed philosopher — not a sentimental one — follows this chain and concludes: yes, that answers the no-subject argument. Death is bad for the one who dies, and I can say who, and I can say in virtue of what. That is not a feeling about the argument. That is the argument standing on its feet.
So here the discipline we have kept all along cuts in a direction it never has before. In every previous episode, the rule that building an argument is not the same as endorsing it restrained our recoil — it kept us from dismissing a view we disliked. Tonight it restrains our satisfaction. For we do not dislike this reply. We have wanted to give it, we have now built it to full strength, and it feels like vindication — as though we had returned to Epicurus and defeated him on his own ground. Perhaps we have. But "it feels like winning" is the impression we have learned to keep at arm's length, and there is one loose end to show before any celebration. The deprivation account works by a rule: to be robbed of a future one would otherwise have had is a real harm, felt or not. It is a clean rule, and a powerful one. And a rule that clean does not get to run only in the direction we find comfortable. Hold the result where it is, fully earned — and then turn with me, because the same rule can be aimed at a place it never occurred to us to defend.
The argument belongs to Lucretius — a Roman poet, writing perhaps two hundred and fifty years after Epicurus, and a convinced follower. He set out the entire Epicurean cure for the fear of death in one long poem, On the Nature of Things. He is Epicurus's heir, pursuing the same end, trying to remove the same fear at the same root. And he brings the sharpest form of the argument anyone has produced. He does not need the no-subject argument we struggled with last week; he has one that works more quickly.
Here is his passage, in plain words. Look back, Lucretius says — look back over the whole stretch of time before you were born. All those ages. Was any of it bad for you? Did it distress you, torment you, cost you a single uneasy hour? Of course not. You were not there. It was nothing to you. And then the line that gives this episode its title: nature holds that up to you as a mirror. The endless nothing before your birth is the exact reflection of the endless nothing after your death. Look at the one already behind you — it was peaceful, it was painless, it was nothing — and you are looking at the other.
Last week, that might have sounded like one more Epicurean consolation, the kind of claim we would build up and then test. But we are no longer standing where we stood last week. We are standing in front of the deprivation account we have just built, the one that answered Epicurus — and here is what the symmetry does to that account. Lucretius is not going to dispute our rule. He is going to use it.
We said death is bad because it robs her of time she would otherwise have had — the years ahead, and all the goods contained in them. Grant that, says Lucretius; hold to it. Now consider the years behind her — the entire eternity before she was born. She did not have those either. And she could have. She could have come into the world a generation earlier, or a century, and had all of it: more decades, more mornings, more of exactly the goods you are grieving on the other end. That time was just as real, just as full of possible life, and she received none of it. By your own rule — to be without a life one could otherwise have had is a harm — being born when she was born deprived her of an enormous stretch of life, every bit as much as dying when she died. The two losses have the same shape. They are reflections of each other, and the rule cannot tell them apart.
And here is the force of it, because we do not believe the two losses are the same, and we know we do not. No one grieves the time before. No one stands at a graveside distressed that the woman was not born in 1850 and granted her hundred extra years in advance. No one lies awake dreading, on a dying friend's behalf, the centuries he "missed" before he existed. Grief runs in one direction only: forward. The entire weight of our mourning and our dread is directed at the nothing ahead, and we feel nothing whatever about the nothing behind — and it is the same nothing. So Lucretius puts the matter plainly: one of three things must be true. Either the rule is wrong, and we are back with Epicurus — death deprives no one of anything that matters. Or the rule is right, and you have failed your entire life to grieve a loss exactly as large as the one now breaking you — a loss you cannot even bring yourself to attempt to mourn. Or there is some real difference between the nothing ahead and the nothing behind, a difference that makes one a theft and the other simply nothing — and if there is, you have been living by it without ever once accounting for it. Name it.
Weigh how strong this is, because the temptation is to treat it as a trick, and it is not a trick. It is the same clean rule we built a few minutes ago — that deprivation is a real harm, felt or not — simply refusing to run in only the one direction we find comfortable. Lucretius is not playing on words. He has identified an asymmetry — dread of what lies ahead, complete ease about what lies behind — that our own best account of why death is bad cannot support. We reach for the difference and cannot say what it is. Try it now, before any help: why is the nothing ahead a horror and the nothing behind simply nothing? Whatever came to mind first, hold on to it, because the serious replies are all versions of it, and each one costs more than it first appears to.
Three philosophers have tried to pay that cost, and it is worth taking them in order, because each one fails at a different point.
The first attempt is Nagel's own; he saw the objection coming and tried to answer it. The two nothings, he says, only look symmetric. Death can take more of this life from her: she could have lived another forty years and still been herself, the same person, with more behind her. But the time before her birth could not have held more of this life — because a person born a century earlier, into different circumstances, at a wholly different moment, would not be her; she would be someone else altogether. So death deprives her of more of her own life, whereas an earlier birth could not have; it could only have substituted a different person. The asymmetry is real after all. That is a genuine answer — but notice where it is weak. We allowed ourselves to imagine her living longer and remaining herself without any difficulty. Why is imagining her born a little earlier, and remaining herself, suddenly impossible? If a year added at the end leaves her the same person, why does a year added at the front turn her into a stranger? Perhaps Nagel is right that there is a deep difference between the two ends of a life. But he has not shown it; he has asserted the very thing in question. The objection is weakened, perhaps, but not removed.
The second attempt is Derek Parfit's, and it goes the other way. Perhaps there is no deep difference out in the world at all; perhaps it lies in us. We are simply built to care about the future and not the past. Parfit gives a thought experiment that makes the bias hard to deny. Imagine you wake in a hospital, still foggy, and are told that one of two things is the case. Either you came through a long and agonising operation yesterday — ten hours of it, now over, wiped from your memory — or you face a short one this afternoon, one hour, under anaesthetic that will wear off. Which do you hope is true? Everyone hopes the long agony is the one already behind them. We will gladly take ten hours of past pain over one hour of pain still to come. We are heavily biased toward the future: past suffering we shrug off, future suffering we dread. And the same holds for goods — we long for pleasures still ahead and barely miss those long gone. So perhaps this is the whole answer to Lucretius: we mourn the nothing ahead and not the nothing behind because we are future-biased about everything, and death is simply future loss, birth simply past loss. But here is the cost, and Parfit himself is the one who supplies it. He was not sure the bias is rational. He raised it as much to question it as to rely on it. And if the bias is only a quirk of our constitution, with no justification beneath it, then it does not answer Lucretius; it merely restates him. "We care more about the future" is a description of the very asymmetry he asked us to defend. It names the thing. It does not earn it.
The third attempt tries to earn it. Two philosophers, Anthony Brueckner and John Martin Fischer, take Parfit's bias and argue that it is not a mere quirk but fitting. Anticipation, they say, is the appropriate attitude to take toward future goods, and is simply not available toward past ones: it makes sense to look forward to a pleasure that is coming, and no sense to mourn one you never had, long behind you. So we are rationally moved by the loss of our future and rationally unmoved by the "loss" of our deeper past, and the asymmetry rests on reason rather than mere reflex. This is the most worked-out of the three attempts. And still, the whole weight rests on the word "fitting." Is forward-looking anticipation genuinely the rational attitude, or merely the one we happen to have, given a justification after the fact? That is what Lucretius called into doubt, and it cannot be settled by repeating that the attitude feels right, because whether it is right is precisely what is in question. It is a serious reply. It is not a closing one.
Three attempts to defend the asymmetry, and each leaves a gap. Which means the symmetry argument still stands, exactly where Lucretius set it. And notice what that does to the result we reached earlier. It does not destroy it: the deprivation account still answers Epicurus, still names the subject, still holds that death robbed her. But it attaches to that result a debt we cannot pay. Keep the deprivation account, and you owe an explanation of why the loss going forward is a horror while the identical loss going backward is nothing — and no one has produced one that holds. The two arguments are now locked together: the reply that names the harm, and the symmetry that uses that very naming to ask why we grieve in one direction only.
One further point, and then the question is yours. I have been applying the symmetry to her — to the one you lost, to her two nothings, the one before her birth and the one after her death — because this whole arc has concerned grief for someone else. But you may already have felt it turn toward you. Lucretius aimed it at you first: your birth, your death, your two nothings. That is a colder question, and we will go fully into it in a later series — your own death, and the stranger question of whether you would even want to escape it. Not tonight. Tonight we keep it where the grief actually is: on her. And here is what it leaves us holding.
Take stock of what has moved tonight, because something real has, and it is worth being exact about the gain.
This week you were carrying a reply, ready to give: she lost her future. And you half-believed it was the obvious thing Epicurus had been avoiding — that you would return, state it, and he would give way. What changed tonight is not that the reply failed. It did not. We built it to its strongest form, and it did what we hoped: it broke Epicurus's deepest premise — that a harm has to be felt — and it filled the empty place. We can now say who death is bad for, and in virtue of what: the one whose future was cancelled. That is a genuine answer to the argument that defeated us last week, and it was earned.
But it did not come free. It came attached to an obligation we cannot meet. The very rule that named the harm — that to be robbed of a life one would otherwise have had is bad for you — turned, in Lucretius's hands, to point at the whole eternity before she was born, when she also was not alive and also could have been, and asked why we do not grieve that by exactly the same measure. And we do not, and we cannot make ourselves, and we cannot cleanly say why not. Three of the most careful attempts in the field each left a gap. So we did not come away with a clean answer. We came away with an answer that leaves an obligation unmet.
And here is where tonight is decided. There are two cheap ways to make the discomfort stop, and both are forbidden. One is to reject Lucretius's argument outright. The other is to accept it as consolation.
Rejecting it sounds like this: the deprivation account wins, and the symmetry is a mere trick. Obviously the future matters more than the past — everyone knows you cannot relive what is behind you — so death is a real theft and birth is not, and Lucretius was simply being clever. That is what relief wants to say. Do not. We watched three serious philosophers try to turn that "obviously" into an argument, and we watched all three come up short. "The future just matters more" is not the answer to Lucretius; it is a restatement of the exact thing he asked us to defend. To reject his argument now would be to do to Lucretius precisely what we refused to do to Epicurus last week — to dismiss a strong argument because its conclusion frightens us. We built it. We do not get to un-build it because we dislike where it points.
Accepting it as consolation is the opposite mistake, and just as cheap. A calmer temperament settles into it and feels wise: the symmetry wins; the nothing after she died is just the peaceful nothing before she was born; I do not mourn the one, so I can set down the other; her death took nothing real; be at peace. Look hard at that, because it is last week's Epicurean comfort in different clothing. It declares the symmetry the winner and walks off soothed — and to do that it has to pretend the deprivation account was answered. It was not. If the bias toward the future is rational — and careful people argue, firmly, that it is — then the symmetry fails and death really did rob her of something the time before her birth never could. To leave at peace, on the grounds that it is all symmetric, is to let one half of the question stand for the whole simply because that half brings calm. That is not rigour. It is the oldest consolation there is, under a new name.
So we do not get to reject the argument and we do not get to dissolve into it. We leave both arguments standing, and we keep looking at them. We cannot stop believing death robbed her — the deprivation account is ours now; we built it; it named her; and everything in us that grieves is grieving a theft. And we cannot say why the nothing ahead of her is a horror while the identical nothing behind her is simply nothing — we reached for the difference, the whole field reached for it, and came up empty. Both of those are true at the same time tonight. They do not fit together, and I am going to leave them not fitting — forcing a fit would be a verdict no one can support. That is where the two arguments leave us: both upright, holding up two nothings we cannot tell apart, and we can neither dismiss the symmetry nor turn away from it. We did not settle whether death is bad. We removed the right to think the answer obvious — and we removed it in both directions, which is the only honest way to remove it.
And here is where it goes next. For two episodes we have stood with the dead — with the question of whether death is bad, which is a question about her, about metaphysics, about the one who is gone. Next time we return to the living. Back to your grief itself — not the death, the grief, the feeling you are actually carrying. And we ask the question we have been careful not to raise: is grief even the kind of thing that answers to reasons? And if it is, can those reasons change, or run out? Can a grief be wrong to keep, or wrong to let go of? That is where we turn — not whether the dead were wronged, but whether the grief is answerable to anything at all.
Before then, one piece of work to do alone. Here is a line you have said, or had said to you, and it sounds like plain truth:
"The saddest deaths are the young ones. She had her whole life ahead of her."
Build it — and notice that its strongest form is pure deprivation, the exact account we built tonight: a young death is the worst kind precisely because it cancels the most future, the most years, the most unlived goods. That is a real point, and it is why we rank a child's death above an old man's without doing any sums. Make the case so well that anyone who holds it would nod. Then — and this is the new part, the part last week could not have asked of you — turn the symmetry on it yourself. By that same rule, did the young person not also "lose" a whole life behind them, the long stretch before they ever existed? We do not grieve that at all. So what licenses counting only the future they were robbed of and not the past they were never given? And then keep the discipline, which is the hard part: do not land on "so young deaths are not really worse, be comforted" — that is the consolation — and do not land on "so the symmetry is obviously a trick, of course young deaths are worse" — that is the dismissal. Build it, test it against the symmetry, and come away with the real question still open, exactly as wide as you found it. That one you carry out on your own.
You can now name who death was bad for — the one whose future was cancelled — and that is a real gain; do not lose it. But naming the harm is as far as it goes. You cannot say why you grieve the future that was taken from them and feel nothing about the identical past they were never given. You can name the harm, and you cannot justify the shape of your own grief — both at once. That is not a failure of thinking. It is the place honest people have refused to leave since the argument was first made: standing in front of two nothings they can neither tell apart nor turn away from, and holding them both.
Thanks for listening. I'll see you next time.