Introduction: Bernard Williams's Makropulos case

In his ‘The Makropulos case: reflections on the tedium of immortality’ (Reference Williams1973), Bernard Williams presents a famously striking dilemma for immortality. If immortality were actual, we would either eventually become hopelessly bored of the very activities and projects which give our current lives meaning or we would have to take on new fundamental desires and projects to rejuvenate our lives. I will call these ‘Lemma One’ and ‘Lemma Two’ respectively. The static boredom of Lemma One is clearly not desirable. But Lemma Two, argues Williams, is also problematic because it constitutes a fundamental loss of character. In other words, immortality either entails the sacrifice of a meaningful, non-tedious life, or it entails a fundamental change in one's character. How does Williams support this claim?

Williams builds his argument around the example of Elina Makropulos – a fictional character from Karel Capek's 1922 play The Makropulos Case. Makropulos is given a potion which grants her immortality and everlasting health (albeit conditional upon taking the potion regularly). As Williams retells it, Makropulos becomes disenchanted with the world after merely a few centuries:

The lack of life which Makropulos experiences is not merely a contingent state of affairs, argues Williams. Rather, her eventual boredom is in some sense unavoidable – a necessary consequence of immortality.

But why should Williams think that boredom is unavoidable for an immortal? For it seems at first glance that Makropulos could have taken up some new project or set of desires. There are, after all, a wide range of activities and desires which are good candidates for providing one's life with a sense of meaning. Makropulos could choose from among these to stave off boredom. Indeed, some have argued that it is precisely the capacity to develop an interest in a range of projects which makes human life meaningful. Charles Taliaferro defends this view, claiming that personhood is a ‘non-time enclosed good’ largely because we have the capacity to experience ‘a rich, perhaps endless variety of time enclosed goods’ (Taliaferro (Reference Taliaferro1990), 369). That is, the good living forever is dependent upon the possibility of taking part in a perhaps indefinite range of projects and goods. Why does Williams reject such a possibility?

The short answer is that Williams thinks shifting through a wide range of fundamentally different projects or desires would constitute a loss of character – a loss of one's very self. Why is this the case? Williams thinks of ‘categorical desires’ as desires which make one's life worth living (Williams (Reference Williams1973), 86). Such desires importantly do not depend upon the assumption that one is alive. For example: while the contingent desire to be free from intense pain depends upon the assumption that I'm alive, my categorical desire to publish a book of poetry does not depend upon the presupposition that I'm alive. Instead, it is the sort of thing that I want to stay alive for. My desire to publish a book of poetry is itself one of the desires which – if left unsatisfied – would make my death premature and therefore a significant loss to me. Categorical desires are therefore desires which ‘drive’ a person forward into choosing to live longer.

Building on this distinction, Williams suggests that the maintenance of a particular character (that is, a core set of categorical desires) is a necessary condition for the kind of identity across time that matters to us.Footnote ¹ Williams then invites us to imagine the relation of such a character to an immortal life. Eventually, even the most exciting of a person's categorical desires will (say, after 1,000 years) fail to enliven and motivate her. She will then have to develop new categorical desires to stave off boredom. But her character will consequently be lost – she will no longer be the same person in a vital sense. For it was the original set of categorical desires which gave her reasons to go on living in the first place. Furthermore, she will be required to change her categorical desires again in another 1,000 years. At some point during immortality, such a person's character becomes unrelated in any meaningful way to one's character in one's ‘normal’ life.

In the remainder of this article, I will outline reasons for doubting the force of Williams's argument. I take an approach which is similar in spirit to Timothy Chappell's (Reference Chappell2009) response to Williams, in that I argue that Lemma One and Lemma Two each fail to admit of realistic alternatives. However, I extend the literature on Williams's argument in the following ways. First, I present a response to Lemma One which is more detailed than either Chappell or Bortolotti & Nagasawa (Reference Bortolotti and Nagasawa2009).Footnote ² Second, Taliaferro (Reference Taliaferro1990) and Bruckner (Reference Bruckner2012) have each challenged Lemma Two, detailing the ways in which a particular character can persist even when a person cycles through a very wide range of categorical desires. However, I engage with philosophers who have defended Williams's argument (or presented alternative versions of it) since Taliaferro's, Chappell's, and Bruckner's papers were published. For instance, I address in some detail the arguments of Matheson (Reference Matheson, Matheson and Nagasawa2017) and Scheffler (Reference Scheffler2013) – arguments which have not received considerable critical response in the literature. Building on the distinction between first-order and second-order desires in moral philosophy, I develop a unique response to Lemma Two. Finally, I respond to Lemma Two on the grounds that one can experience temporal scarcity (and therefore character-forming choices) even in immortality.

Lemma One: can boredom be undone?

Lemma One has received some commentary and critique in the literature on immortality.Footnote ³ A. W. Moore (Reference Moore2006) considers the possibility that the vivacity of one's memories might decay in immortality – making boredom impossible. The picture is something like this: I as a persisting human being go into a deep sleep after a full human life. I (that is, the particular human organism I am) then awake with no memories of my past life – but with an opportunity to experience a full and varied life without irremediable tedium. This process might be repeated indefinitely. But, as Williams notes, it is difficult to see how this option will overcome the objection, first, that it is not me who wakes up. For, whether or not we accept the full thrust of Locke's famous account of personhood,Footnote ⁴ memory seems to play some vital role in the kind of persistence that matters to us. And, second, it feels as though the human being who lives these serial lives is somehow trapped in a great, unrealized tragedy of amnesia.

Perhaps more promising than the above scenario is one in which we only gradually lose our memories as we gain new ones.Footnote ⁵ Here, we can imagine an equally eerie alternative to Christopher Nolan's philosophical film Memento – one in which a temporally distant portion of one's long-term memory is lost each day. Under this scenario, the immortal person each day undergoes the total decay and loss of a particular set of memories from her experiences of, say, eighty years past.Footnote ⁶ The upside is that such a person can experience as novel and exciting all those projects which she only took part in eighty years or more ago. Although this view presents a certain balance between achieving diachronic identity on the one hand and avoiding boredom on the other, it is not without its problems. Though the person may not be aware that she wakes up each day with a set of memories decayed and forever lost, there remains a certain sense of tragedy in the fact that a person undergoes a form of amnesia each and every day for eternity. This is to say nothing of the worry that such a person might eventually gather non-memory-based evidence of her past lives – leaving her disturbed by the photos and traces left by her unrememberable actions from more than eighty years ago.

So both the series of psychologically disjointed lives and the constantly decaying memory versions of immortality seem undesirable. Because of these and other worries, I will not focus upon the notion of memory decay to answer Lemma One. Rather, I want to focus on the possibility that one's phenomenal awareness of the world might be deeply altered in the afterlife – making irremediable boredom an impossibility. In addition to providing an answer to Williams, this line of thought has the benefit of drawing from the tradition of pre-Williamsian conceptions of immortality. It is therefore in no way an ad hoc response to the Williams worry about immortality.

Wordsworth on tedium, wonder and immortality

Thus begins William Wordsworth's haunting masterpiece, ‘Ode: intimations of immortality from recollections of early childhood’. Remarkably, we see in Wordsworth's poem precisely the same sort of experience which stands behind Williams's Makropulos case. They each express in their own way the reality that youth possesses a vigour and wonder which cannot be regained at later points in life (for Makropulos, much later in life). Our propensity to boredom seems a learned – and yet perhaps an unavoidable – disposition. Notwithstanding their agreement about this fact, Wordsworth and Williams draw precisely the opposite conclusions from their experience. The latter claims that immortality would just present more of the ever-increasing boredom which eventually characterizes our mortal life, while the former is deeply moved by an impression of the soul's essential immortality. Wordsworth uses the fact of our increasing dullness and lack of wonder to suggest that the soul in some sense has eternal origins. The wonder and profundity of childhood experiences show that our soul has its origin in God.Footnote ⁸ Wordsworth writes:

However, this glorious youth and wonder of childhood does not last. We find of the world's glory that ‘At length the Man perceives it die away, | And fade into the common light of day.’ Nonetheless, the driving theme of Wordsworth's poem is that our tendency towards boredom and disengagement from the world in later life is a profound loss and a sign that something has gone wrong with our perceptual capacities. Or, at the very least, something has gone wrong with our emotional and spiritual engagement with what we perceive in the world. Because the soul is deeply attuned to wonder and spiritual engagement in our earliest years, intimates Wordsworth, it must originate in God. And anything which originates with God must have an immortal existence of some kind. I will of course not attempt to explicate or defend any argument for the soul's immortality on the basis of Wordsworth's verses. Rather, I want to point out that Wordsworth's speculation about the change in our dispositions from wonder to disenchantment between childhood and adulthood brings to the fore an important insight: the increased human tendency to boredom with more exposure to the world seems a contingent psychological fact about human persons. And while it may indeed be true that those who have progressed further into life often tend towards a disposition of boredom because of their over-exposure to the things of life, it is not unthinkable that the human mind could be altered in such a way that it retains its sense of wonder through immortal ages. Wordsworth subtly suggests that just such an eternal youthfulness is a God-like disposition to the world.

Jonathan Edwards on the afterlife

Wordsworth, however, was not the first thinker to note the possibility that the human mind in immortality might enjoy a state of ineluctable wonder. The American philosopher and theologian Jonathan Edwards anticipates this line of reasoning in his own conjectures about the mind's make-up in the afterlife. Indeed, Edwards cites three distinct contingencies which God controls and which therefore may be transformed in the afterlife. Our psychological disposition towards the world, the perceptible beauty and harmony in the created order, and our perceptual faculties themselves may all be enlivened and transformed in unimaginable ways in the hereafter.

First, Edwards writes that in the afterlife, ‘’tis the directly reverse of what ’tis on earth; for there, by length of time things become more and more youthful, that is, more vigorous, active, tender and beautiful’ (Edwards (Reference Edwards and Schaefer1996), 341). Like Wordsworth after him, Edwards believed that the disposition of the human mind to grow dull over time could actually be reversed in the afterlife. Furthermore, God may also change the created order itself, so as to bring about radically new perceptual and experiential possibilities. The beauties in the current created order are already magnificent. But there's no reason to doubt that ‘God can contrive matter so that there shall be other sort of proportions, that may be quite of a different kind’ (ibid., 328). This would create the possibility of perceptual experiences which are currently ‘inconceivable’ but which are ‘vastly more ravishing and exquisite’ than any we have experienced (ibid., 328). In other words, whole genres of perception and of beauty could be awaiting us in the afterlife – since the world in which we will exist may be in some sense discontinuous with the world in which we currently live.

Finally, our capacities to perceive such exquisite beauties may be deepened significantly. Edwards argues that our current capacities only capture the beauty of some things – visually, aurally, and perhaps scientifically. But in the afterlife we may learn to perceive a plenitude of musical and visual ratios at once: ‘perhaps we shall be able fully and easily to apprehend the beauty, where respect is to be had to thousands of different ratios at once to make up the harmony. Such kind of beauties, when fully perceived, are far the sweetest’ (ibid., 329).

These claims may seem ungrounded conjectures to some readers. However, what's at stake is not whether or not these things are in themselves probable. Rather, Edwards's claims represent a defence of the possibility that immortality might avoid Williams's worries. For Edwards is arguing that, given theism, human beings’ cognitive capacities could be altered in the afterlife. It is also worth stressing that these significant dispositional alterations need not undermine the numerical identity of the persons involved. At the very least, such questions are distinct and therefore deserve treatment by those who seek to give an account of the possibility of embodied, personal identity across death.Footnote ¹⁰ This is also compatible with memories being carried over from this life to the afterlife, for God may preserve those memories while changing an individual's disposition to the acquisition of new experiences.

However, an important objection remains. It is plausible that a person who maintained his character by engaging eternally in a determinate, finite set of projects – even with an unearthly freedom from boredom – would somehow still embody a tragic existence. We can recall in this regard the Greek myth of Sisyphus – in which an individual is burdened with a single, meaningless task which he attempts over and over for all eternity. The deplorableness of such a life would not, we tend to think, be erased by a change in Sisyphus's psychological disposition to boredom. So, perhaps even Wordsworth's and Edwards's conjectures don't get immortality off the hook. However, we can imagine a different case – one in which, for instance, Vincent van Gogh is revived, made immortal, and given the task of painting forever. Assuming van Gogh both lacks the psychological propensity to irremediable boredom and is given a universe of landscapes to paint, is it tragic for him to carry on for ever? This case is not obviously tragic in the same way as the Sisyphus case, with the clear disanalogy between the act of senselessly rolling a stone up an incline and the much more meaningful act of creating works of art. But perhaps answers to this question will diverge. So I shall move on to address Lemma Two.

Lemma Two: the character question

Lemma Two represents Williams's claim that an engaging immortality (full of a wide range of diverse projects and desires) would entail a loss of character in some deeply important sense. Williams is unfortunately unclear as to what precisely this loss of character consists in and why it would be unacceptable. His criterion for a meaningful immortality is that ‘the state in which I survive should be one which, to me looking forward, will be adequately related, in the life it presents, to those aims which I now have in wanting to survive at all’ (Williams (Reference Williams1973), 91). What does this mean? There seem to be three (mutually compatible) possibilities. The first is that, if I were immortal, I would eventually lose my narrative sense of self. That is, if I should engage in a very long sequence of interesting and varied projects into eternity, I will lose the capacity to conceive of my life as a narrative whole. The second possible reading is that one's character is constituted by a core set of stable, categorical desires. An engaging immortality would require me to change these desires and therefore lose my character. The final reading is that Williams thinks character is bound up with mortality because it is bound up with choosing between alternative courses of action – something which could not exist in immortality. I will discuss these in turn.

The narrative identity reading of Lemma Two

First, is Williams worried that an immortal person cannot maintain a sense of the narrative of her life through eternity? That is, Williams may be worried that if I experience an indefinite range of categorical desires across time, with an indefinite number of relationships, I will lose any sense of my life as a unique narrative unity. In short, there will come a time when I will be unable to narrate my life meaningfully – for the simple fact that there are billions of projects and years of memories to choose from in constructing a narrative. This is indeed a worrying thought.Footnote ¹¹ However, should it be decisive? It seems not, for the reason that our current lives also include a staggeringly large number of experiences which would be left out of any brief, narratival reckoning. However, when asked to summarize our lives, we simply narrate the most significant experiences in order to provide an order to the whole. I find it difficult to believe that an analogous exercise should be impossible in immortality. It is indeed true that one will not be able to recall the entire sense of one's billion-year-long life at any moment, but this is true (to a much lesser extent) of our mortal lives as well.Footnote ¹² While there is more to be said about this worry, the crux of Williams's Lemma Two lies elsewhere.

The categorical desires reading of Lemma Two: Matheson's extension of the Makropulos case

The second possible reading of Lemma Two is that Williams intends to argue that character is bound up with a certain constellation of categorical desires. Another way of saying this is that immortality will eventually involve a loss of my will to live. I have a finite possible set of categorical desires, and when these desires are exhausted, I will have no desire to live my life – I will want to live some other life than my own (current) life. It seems highly plausible that this is an accurate reading of Williams. Indeed, it is compatible with the third and final reading I discuss below. I believe Williams presents these as two related but distinct arguments in support of Lemma Two.

I will contest Williams's claim. At first pass, my argument is that one could take up an indefinite number of projects and goals and desires through time – without ever losing oneself. Charles Taliaferro and Donald Bruckner each argue for just such a response to Williams.Footnote ¹³ But some defenders of Williams's Lemma Two provide a riposte at this juncture. If a certain person were instantly or gradually to change all her categorical desires (but maintain strict numerical identity), then she would cease to exist in any way that matters to her. That is, she would lose what some philosophers have called her ‘practical identity’ (Matheson (Reference Matheson, Matheson and Nagasawa2017), 391ff.). And just such a change would be required to stave off boredom. So, since Taliaferro and Bruckner's responses depend upon this kind of radical alteration of categorical desires, the identity of the individual cannot really be maintained in immortality. I will now consider Matheson's defence of Williams in detail because doing so will help us judge the plausibility of Lemma Two.

There are two claims involved in Matheson's defence. First, Matheson accepts Williams's suggestion that an engaging immortality would eventually entail a total shift in one's set of categorical desires. Second, Matheson accepts the claim that a total shift in one's categorical desires would constitute an unacceptable loss of one's character. Each of these aspects of Lemma Two can be challenged. Matheson employs a thought experiment in defence of the first claim – inviting us to imagine an alternative to the Makropulos elixir. Matheson's elixir, in addition to granting an individual immortality, also causes ‘a slow . . . change of character’ (Matheson (Reference Matheson, Matheson and Nagasawa2017), 398). This transformation of desires is both gradual (guaranteeing continuity of numerical identity) and total (changing all one's categorical desires to their opposites). Matheson claims that no one would choose the elixir because we ‘would rather cease to exist than become something we (currently) despise’ (ibid., 398). This is plausible, as such an elixir would cause one to despise all those activities which one currently enjoys, to possess an aversion to one's current loved ones, etc. On the assumption that no one would take such an elixir, Matheson draws the general conclusion that ‘we care more about maintaining our practical identities than we do about preserving our existence’.Footnote ¹⁴ In other words, character in Williams's sense matters more than the survival of our numerical selves. But such a radical and unacceptable change of character would be the only way to avoid utter boredom in immortality. What can be said of Matheson's version of the tedium objection?

The regretful reprobate

In order to judge Matheson's claim, we can employ a version of the distinction between second-order desires and first-order desires.Footnote ¹⁵ Second-order desires are desires about first-order desires. For our purposes, we can say that second-order desires can be in some cases essential to one's character – they contribute to one's character by ruling out the adoption of certain first-order desires and helping one to choose particular first-order desires.Footnote ¹⁶ So my desire that I should always desire to uphold my moral obligations is a second-order desire. This governing desire prevents me from adopting the desire to live the life of a pure aesthete. On the other hand, first-order desires might be desires for goods or states of affairs – such as the desire to write a poem or the desire to see the Sahara Desert. First-order desires can serve as categorical desires – they can be the kinds of desires which constitute a reason for me to go on living. But second-order desires seem plausible candidates for the status of categorical desires as well. For instance, I might have the abiding desire to see the day in which I desire only to act in such a way that my actions will not harm others. It is possible that, for me, the unmet desire to achieve freedom from desires which cause harm can constitute a profound reason for me to go on living. It is not simply a desire which is conditional upon the assumption that I'm alive.

This distinction presents a possible reply to Matheson's thought experiment. We can imagine a person with a disparity between his first- and second-order desires who would gladly choose Matheson's elixir. We can imagine a person with a keen awareness that he has overwhelmingly bad first-order desires. Historical and literary examples serve us well here. The character Kichijiro from Shusaku Endo's novel Silence (Endo (Reference Endo1980) and the Martin Scorsese film by the same name) presents us with a self-consciously cowardly character – a character who desperately desires to become someone more courageous. In the most extreme case, such a person wishes that all his desires should be transformed into their opposites – because this would make him into a virtuous person. Let's call this person the regretful reprobate. Furthermore, the normal ways of changing one's own character – such as the practice of consistently acting in ways which eventually cause one to desire to carry out the right action – are unavailable to the regretful reprobate. This is because he cannot (even on one occasion) stop doing the selfish or cowardly things he deeply desires to do. It seems that such a person would choose Matheson's elixir as a welcome reprieve from his own character; Matheson's elixir simply has the handy side effect that it guarantees immortality! And if one can reasonably hope to change one's character in the most radical of ways (as in the case of Matheson's elixir), then even changes of desire in immortality perhaps do not entail an unacceptable loss of self.

Furthermore, this case might help us to think about practical versus numerical identity. In such a case as the regretful reprobate, it is precisely numerical identity which he aims to preserve. Such a person cares more about the survival of numerically the same self than about the maintenance of his current character/practical identity. Indeed, it seems to me that the notion of radical ethical improvement presupposes numerical identity and prioritizes it to practical identity. This consideration perhaps tells against the claim that there is something called practical identity which is in every case the bearer of concern for ourselves more than numerical identity.Footnote ¹⁷

But do we have an answer to the tedium objection? It seems that Matheson can respond here that the regretful reprobate – in order to be motivated to take the elixir at all – must at least have some second-order desire to have virtuous desires. But then the elixir will have the unfortunate side effect of changing the regretful reprobate's original, second-order desire to have only virtuous desires into a desire to not have virtuous desires. So perhaps Matheson is correct that even in the most extreme cases we would not choose his elixir.

Must character change in immortality?

However, perhaps we have moved too quickly here. For there are vital disanalogies between Matheson's elixir case and the immortality case. The first disanalogy is that the immortality case may allow one to maintain some second-order desires, while Matheson's elixir case does not. So, in order to mirror the case of immortality, Matheson's elixir thought experiment would need to be changed to cause first-order desires to change and yet maintain some governing, second-order desires. If the regretful reprobate finds himself with first-order desires to attain greedy and self-serving pleasures (and yet the second-order desire to desire the selfless thing), then he would welcome this latest version of the elixir. However, if Matheson's elixir example is to do the work of supporting the tedium objection to immortality, then the elixir must transform all of one's desires. For if it is reasonable to totally transform one's first-order categorical desires radically, then it also seems reasonable to transform one's first-order categorical desires on several occasions. It is perhaps even reasonable to change them indefinitely. For, should one have some general second-order desires which are categorical (and which are compatible with changes in first-order desires), then one could maintain a stable character while cycling through many first-order desires. One could thereby avoid boredom in immortality while maintaining a particular character in the relevant sense.

So I might have a second-order categorical desire to desire things which bring more good into the cosmos. Provided there is always more good to be done, then I will always have the opportunity to change my first-order desires to achieve the promotion of a new kind of good. And while it makes sense to think of one getting bored of a particular set of first-order categorical desires (such as the desire to paint magnificent landscapes on earth), it is much less plausible that one could become bored with a general and second-order desire such as ‘desiring to desire to bring good into the world’. For the distressing thing about the Matheson elixir was that it should cause my categorical desires to become diametrically opposed to those desires which I currently possess (as in the case that I come to desire the downfall of my loved ones). Immortality need not entail that – even if we cycle through a huge range of desires.

Furthermore, the person whose first-order categorical desires change (but whose second-order desires remain stable) satisfies the Williams/Matheson necessary condition for identity across time. Namely, there is a set of identity-conferring categorical desires (such as the desire to desire that the world should be filled with increasing amounts of aesthetic or moral goods) which can be sustained through eternity. There may, after all, be an indefinite number of worlds which require cultivating in an indefinite number of ways. Therefore, Matheson's (and Williams's) general conclusion that one's identity is deeply bound up with the maintenance of a certain set of categorical desires turns out to be compatible with immortality. As long as we maintain a core set of identity-conferring, second-order desires, we may range over many first-order desires which accord with those second-order ones. What should we conclude about Williams's original argument then? In short: because Williams does not distinguish between first- and second-order categorical desires, he mistakenly assumes that one's character cannot possibly survive immortality. But even if we accept that maintenance of some set of categorical desires is a necessary condition for immortality, we may still hope for that to obtain in an engaging afterlife. This is possible through the maintenance of second-order desires which govern a (possibly) indefinite range of first-order desires. So, Matheson's example – which seemed to nicely illustrate Williams's argument – has actually demonstrated a weakness in his charge that immortality must be tedious.

At this juncture, Williams or Matheson might respond that we care most deeply about the maintenance of our first-order categorical desires. So the claim that we can maintain a core set of second-order, categorical desires in immortality is simply a verbal victory for the defender of immortality. We must maintain the set of first-order desires which we now possess in order to maintain our practical identity. This response, however, presents us with too strong a criterion for identity. For what Williams's view then amounts to is the claim that the maintenance of a small, core set of first-order desires is a necessary condition for the reasonableness of hoping for one's survival and well-being in the future. But many of our ordinary (this-worldly) hopes for the future involve hoping for the flourishing of ourselves after we've changed our core desires and projects.

So, under this version of Williams's argument, a person who has the projects and goals of an art student at Time1 (henceforth T1) but who believes she will have the projects and goals of a restaurateur and mother at Time2 (henceforth T2) cannot reasonably hope for the good of the restaurateur at T2. This would be the case even if the person at T1 had perfect access to the facts which will make it the case that the restaurateur at T2 possesses numerical identity with the student at T1. So, if we take Williams's argument as a general claim about the necessity of maintaining one's current first-order desires, then we cannot reasonably hope to survive significant life and career changes. But we clearly can reasonably hope for the future flourishing of our numerically identical selves in the mundane, this-worldly cases. This is true even in those cases when we reasonably expect we will have deeply altered first-order desires. It seems I only give up my hope for such a future self when I am convinced that the future version of myself will embrace the opposites of some of those second-order (governing) desires which I find essential to my character. For if all my current first-order desires are necessary for my identity, then the reasonableness of hope for normal life will also be put in jeopardy.

The ‘temporal scarcity’ version of Lemma Two

We now come to the final version of Lemma Two. In perhaps the clearest insight into Williams's views of the relationship between character and immortality, Williams considers the mythical character Teiresias. This mythical figure lived an immortal life with a variety of desires and projects which were cumulative in his memory. Critiquing this kind of conception of immortality, Williams writes: ‘One thing the fantasy has to ignore is the connexion, both as cause and consequence, between having one range of experiences rather than another, wishing to engage in one sort of thing rather than another, and having a character’ (Williams (Reference Williams1973), 94). Williams does not elaborate on what he means here. But the most pressing consideration seems to be this: choosing between alternative (and incompatible) courses of action is a necessary condition for the existence and expression of character. But immortality entails that one need never choose between alternative courses of action. This is because one always has an infinite amount of future time to do what one might wish to do. So, immortality is incompatible with the existence of a particular character.

We see once again that recent philosophers have echoed Williams's original insight in more detail. Samuel Scheffler, in his Reference Scheffler2013 book Death and the Afterlife, argues along similar lines. In Scheffler's reckoning, the ‘temporal scarcity’ which characterizes our mortal lives is the very precondition of the meaningfulness of decisions – and indeed of our lives as a whole (Scheffler (Reference Scheffler2013), 99). This argument is reiterated in Brian Ribeiro's ‘The problem of heaven’, where he claims that, ‘All the tension of choice disappears in the literal boundlessness of my options and the infinite time in which I choose’ (Ribeiro (Reference Ribeiro2011), 50). A very closely related criticism is presented in the riveting short story, ‘The immortal’, by Jose Louise Borges (Reference Borges1964). Though Scheffler, Ribeiro, and Williams express themselves in their own ways, I believe the worries here are very similar. What can be said about them?

The temporal scarcity version of Lemma Two can be challenged on the following grounds: there are in fact many meaningful choices whose meaningfulness or value (despite depending upon some kind of temporal scarcity) does not depend upon my life being finite in length. Since character (according to Williams) depends upon meaningfully choosing one thing rather than another, character can continue in immortality. For we can still have the capacity to choose between alternative courses of action. Why think this is the case? The most direct way to understand what goes wrong with the Scheffler/Ribeiro version of Lemma Two is to note a distinction. That is, there is a distinction between the scarcity of an undifferentiated moment of time in relation to the notion of a bounded life (which will cease if I am made immortal) and the scarcity of particular moments of time in relation to every other moment of time and in relation to external states of affairs (which need not cease even if I am made immortal). The vital insight of this distinction is that valuing and choosing one thing rather than another may be dependent upon temporal scarcity of some kind, without thereby being dependent upon the temporal scarcity of my time considered as an undifferentiated series of moments. So my own immortality does not preclude the possibility of the existence of some (or even a great many) particularly ‘scarce’ moments which shall never come around again. But it is the presumed total removal of temporal scarcity which was supposed to make character and valuing impossible in immortality. But, it seems that immortality will remove the scarcity of undifferentiated time without necessarily threatening the reality of the temporal scarcity of particular times. So instances in which we must choose one thing rather than another can continue in immortality.

This claim is best illustrated through a thought experiment. We can imagine some future in which one group of human beings on earth is granted immortality (or achieve it themselves through scientific and medical research). However, there are also two nations of non-immortal human beings who are on the verge of war with one another. The Immortals happen to be in a position to stop the impending catastrophe. Under the assumption of their immortality, would the immortal humans simply relax, claiming, ‘We have an eternity ahead of us – why rush to intervene?’ I see little reason to think they would necessarily respond in that way. What this example shows is that we do not only consider decisions momentous in virtue of the shortness of our life. Second, it is perfectly possible that states of affairs in a world in which we are immortal may be such that they require our action and decision just as immediately and urgently as similar situations in this world. This fact is not dependent upon whether we will or will not live forever. And if there are times at which we must choose between some time-sensitive courses of action, then character in Williams's sense will be affected by that decision. This implies that, since character would continue to bear some relationship to our choices in immortality, Lemma Two does not stand.

Now, it may be objected here that we are allowing a world in which some persons do not share immortality, which is problematic for some reason. I see no reason to admit this as incompatible with the prospect of some people's immortality. However, even granting such an objection does not lessen the significance of the distinction between undifferentiated and specified temporal scarcity. For we can also imagine a world in which all human beings are immortal but in which non-sentient species exist with the possibility of extinction. When these species are in danger, would our immortal selves not be faced with momentous decisions – ones which we would not disregard on the basis of the surplus of our own indefinite store of (generic, undifferentiated) time?

Furthermore, it does not seem to be the case that immortality itself rules out even the most ambivalent human emotions and experiences such as tragedy and homesickness. Even an immortal being could be placed in situations in which she had to choose between time-bound and incommensurable goods, with the consequence that a tragic decision is unavoidable. But this shows that immortality is compatible with urgency and (by implication) the practices of valuing and prioritizing goods. These practices of valuing lead to dynamic developments in one's character – preserving each person's particularity. Even an emotion like homesickness would be possible in immortality because one's confinement to a particular place at each particular time means that we may sometimes yearn to be at another place. For instance, I may be required during immortal life to run some humanitarian mission on a distant planet – all the while desiring that I should be at home on earth bringing in the autumn harvest with my wife and children. By choosing one rather than the other at a particular time, I develop into one kind of person rather than another. These are only a few of the possible cases in which both valuing and possessing a character are shown to be in principle compatible with immortality.