I think nothing matters, and I’m quite upset about it. JL Mackie also famously saw that error theory can be a bit of a downer, that ‘denial of objective values can carry with it an extreme emotional reaction, a feeling that nothing matters at all, that life has lost its purpose’. He thought this reaction symptomatic of endowing our subjective ‘concerns and purposes’ with a ‘fictitious external authority’ (Mackie 1977, 34).

Bernard Williams was not an error theorist. But he also rejected strongly objective values. His positive view about ethics is not easy to summarise, but includes a relativistic stance-dependence where we have only our ethical concepts and point of view.Footnote 1 Williams acknowledged that the positions he thereby rejected—Kantianism and the ‘morality system’ in particular—had consoling aspects, not least in their insulation from luck and unfairness.

No stranger to pessimism, we might therefore expect Williams to sympathise with upset at the loss of the consolation that luck-immune objective justification would offer. In a posthumously-published but widely-cited paper, he wrote that a metaethically irrealist view ‘can make people feel that human activities are absurd, because we invest them with an importance they do not really possess’ (Williams 2006, 137). But Williams was scathing about such feelings, calling them a ‘muddle’. His polemical powers are on full display, not so much responding to as mocking his target, Bertrand Russell.

I’ll argue that there need be no muddle in mourning the loss of absolute importance and objective justification. I’ll distinguish several different ways we might lose those things and why we might care. So Williams was wrong, and I’ll later argue that Williams-style internalism about reasons is in some respects even more depressing than stark global nihilism.

1 Normative Nihilism

Williams might dismiss the following taxonomy as tedious or plodding. It’s a mark of how powerful his polemics were that I—who never met him—am almost afraid of what he might have to say about it. But the subtle differences between metaethical views do matter (and perhaps I’m being unfair: he was certainly not afraid of detail).

I accept, or at least am tempted by

Global Normative Nihilism

There are no normative reasons for either belief or action.

Normativity—the crucial ingredient in normative or justifying reasons—is not a feature of the world. We can truly describe whether someone’s beliefs, desires, and actions are coherent, truthful, and systematic, but what we say lacks normative content. She has no reason to be coherent, truthful, or systematic, and there is no justification to be found.

Williams had a far less—but still somewhat!—skeptical view about reasons. His ‘Humeanism’ or ‘internalism’ accepts that we have genuinely normative reasons, but ties them to our motivations:

Austere Internalism

An agent has a reason to A iff Aing would promote one of her desires, perhaps subject to procedural purification.

Note the ‘iff’. Some versions of internalism replace this with an ‘only if’, in which case some of the arguments to follow would have to be weakened. I’m also using ‘desire’ as a shorthand for all the contents of what Williams (1981a) calls our ‘subjective motivational set’.

Austere internalism says that our reasons track our desires, without any particular restrictions on their content: we only have reason against committing murder insofar as that’s the better option with respect to our desires. Our desires can make murder what we have most reason to do, because there are no categorical (desire-independent) reasons, so none against murder. Williams is the most famous defender of austere internalism—see also Williams (1995)—but he is far from alone; Sobel (2016) also defends a broadly similar account.

I use ‘austere’ to distinguish this view from ‘moderate’ internalism—see especially Schroeder (2007) and Markovits (2014)—which accepts that all reasons are tied to desires but nevertheless exploits the ‘procedural purification’ proviso to revive categorical or agent-neutral reasons, normally corresponding to moral reasons. From now on I’ll drop the qualification, because the austere is my focus.

Neither Nihilism nor Austere Internalism says anything about morality without a linking principle such as

Moral Rationalism

Moral obligations constitute or entail reasons for action.Footnote 2

Together with Moral Rationalism, either of Internalism and Nihilism almost forces us into

Moral Error Theory

All positive moral claims are false.

Error Theory denies that torturing puppies for fun is wrong. Moral obligations would entail categorical reasons for action, but Internalism and Nihilism deny that such reasons exist. And similarly for moral theories that don’t trade in obligations. An appeal to Moral Rationalism is a natural interpretation of Mackie’s queerness argument, for example: if torturing puppies were wrong then we’d have categorical reason not to torture puppies, but there are no such reasons, which would be queer.

Internalism only almost forces us to Error Theory, because we might escape by appeal to (Harman 1975)-style moral relativism. On that view, morality provides reasons, but they are not categorical because the content of morality depends on what we desire, construed very broadly. But as we’ll see, in dropping an absolute or objective moral perspective such relativism loses much of what is consoling about morality.

There is a cluster of views resiling from the traditional objective picture of morality—of an absolute moral stance issuing categorical reasons—and despite their differences, views in this cluster agree that some kind of objectivity, whether absolute moral obligations or categorical reasons, is lost.

We might also worry that we lack absolute importance, that our lives are meaningless, and that there is no purpose or point to our existence. The connections between these theses run deep. For example, if we have categorical reason to promote and protect the well-being of our fellow humans, then that well-being matters and is absolutely important, because it provides reasons for any agent capable of having them. Williams links the evaluative with the important, seeing the judgement that something is ‘just better’ as an instance of the notion of ‘absolute importance, that last relic of the still enchanted world’ (Williams 2006, 141, emphasis in original).

But what about meaning—isn’t it a stretch to say that normative reasons could provide a ‘purpose or point’ to our lives? If ‘purpose’ is interpreted literally as involving intentionality—as it is by (Benatar 2017, 46)—then indeed categorical reasons might not bring meaning. But I think if we have normative reasons for action then our lives aren’t really pointless. There are certain things we ought to do, and especially if our reasons are categorical then the to-be-doneness of those things is somehow built into them. They call out to us. It might not be the purpose or point we were hoping for, but it is some kind of purpose. The teleology is there: if we have reasons and don’t act in accord with them, then we’re not functioning correctly.

These several skeptical metaethical theses agree that some absolute or objective standard is missing. But does it make sense to get upset about that?

2 The Coherence of Nihilistic Despair

I find the views just surveyed depressing. Insofar as I accept one of them, I am saddened or depressed by what I take to be true. I’ll call such an attitude ‘nihilistic despair’, a slightly overblown name that encompasses weaker negative emotions such as discomfort or angst.

I’m not alone: many find moral error theory and its cousins a little unsettling. In conversation I’ve often been told that error theory would be a sad truth, perhaps with the implication that error theorists show a disreputable character or cast of mind. Error theorists may respond that they are simply facing the unpalatable truth. No matter who is right here, an emotional reaction to denial of objectivity is common, acknowledged (though not endorsed) even by Mackie.

Susan Wolf writes that the question of a ‘purpose or point to human existence’ has a depressing answer: there is such a purpose or point only if there is a God to provide one, and the latter is unlikely (Wolf 2007, 1). Sharon Street is a metaethical anti-realist, and claims that things matter, but dependent entirely on us caring about them. So we should think of nihilism ‘not as a philosophical position that might, to our dismay, turn out to be correct, independently of what we think or hope—but rather as a state of mind we might fall into—false as long as we don’t fall into it, but true as soon as we do’ (Street 2017, 148). The prophylaxis is causal, including a good eating and sleeping regime.

Jonas Olson (a nihilist in my terms) writes that for many people ‘error theory is emotionally difficult to accept […] it may feel sickening to accept that none of [the past century’s atrocities] were in fact morally wrong’ (Olson 2014, 143). He suggests an explanation rooted in the affective origin of belief in the moral wrongness of such atrocities.

I don’t think this can quite be the whole story: Street’s vertigo also accompanies more generally nominalist metaphysical views where, at risk of overstatement, chairs exist only insofar as we think they do. And my aesthetic judgements also have an affective origin. Outside ethical antirealism there is less fear or sadness or dismay, though Kieran Setiya has recently argued that they aren’t entirely absent in the metaphysical or epistemic case (Setiya 2021, especially pp. 51–52). Revisionary metaphysical pictures are unsettling, but there seems to be something particularly upsetting about global normative nihilism and other disavowals of metaethical objectivity.

Like other emotions, despair has cognitive and conative components: we feel despair about something. A non-idiosyncratic account of nihilistic despair will identify aspects of nihilism that are recognisably, if contingently, sad. Contingently sad because if nihilism is true there’s no reason to find those things sad, as there’s no reason for anything.

If nihilism is false then the nihilist has (perhaps blamelessly—metaethics is difficult) gone wrong somewhere and believes a false view. Nevertheless, our mistaken nihilist may have reasons of coherence or rational requirements to be upset about nihilism’s putative truth. Our question is what nihilism looks like from the inside.

Despair is cognitively distinguished from other broadly negative emotions by a belief that its target is immutable, a lost cause. Despair at the impending death of a friend would normally be a mistake if we could easily save her. Nihilistic despair certainly qualifies here: whatever the correct metaethical theory is, it is a necessary truth beyond our power to change. (Even on Street’s view where we have some power, we are powerless to change the truth of Street’s view.)

On the conative side, despair is often marked by infectious hopelessness and the sapping of motivation. I can’t bring my friend back to life, so I can’t bring myself to get up and go to work. Grief is also a reaction to immutable loss. Grief can lead to resignation, where we also recognise our powerlessness, but in a way that mitigates our sadness. We learn to live with it: our emotional reaction is eventually muted, and (the rest of) life goes on. If we don’t learn to live with it, then grief may become despair. A lapsed moral realist may experience nihilistic grief, if she’s come to think the objective goodness she built her life around was a myth or fiction all along. The sunk costs of religion make converting to atheism fraught.

So despair and grief constitutively involve regarding a putatively immutable fact as somehow bad, sad, or depressing—attaching a negative valence to it. This can make nihilistic despair look incoherent: isn’t judging it bad that nothing is bad clearly self-undermining? Indeed, it’s surprisingly common to claim that nihilistic despair is nonsensical or muddled. I’ll interpret ‘muddled’ to mean either self-contradictory or manifesting a weaker form of incoherence: muddled attitudes don’t ultimately hang together.

Bertrand Russell famously argued that the modern scientific worldview suggests that we are insignificant, and offers us only a ‘firm foundation of unyielding despair’ (Russell 1985). In his wide-ranging The Human Prejudice, Williams claims that the titular attitude (which allegedly underpins meat-eating, for example) is not really a prejudice, but a consequence of the fact that we humans are the only beings who can provide and understand reasons. It’s therefore no surprise that human being is either a basic ethical concept, or very close to one through an ethical notion of species membership and loyalty.

For my purposes the important strand of Williams’s argument is metaethical. He claims that there’s no wholly objective moral point of view and it’s thus ‘a total illusion to think that this [human] enterprise can be licensed in some respects and condemned in others by credentials that come from another [non-human] source’ (Williams 2006, 147). If we try to occupy an objective moral point of view, we fall into that illusion. The ‘sources’ he considers are absolute importance in the cosmic scheme of things and utilitarianism.

Rejecting absolute importance as incompatible with the modern disenchanted worldview, Williams is surprisingly sympathetic to the thought that the size of the universe has a metaethical upshot. I defend this thought elsewhere, but most contemporary philosophers think it a somewhat silly mistake, so Williams is a welcome ally on that point.Footnote 3 He argues that utilitarianism has untenable first-order consequences, not least—and memorably—in our dealings with disgusting or conquering aliens.

(Williams 2006, 137) distinguishes two claims about absolute importance (what I also call ‘cosmic significance’). First, the metaethical claim that absolute importance—of any kind—is a myth, ‘a relic of a world not yet thoroughly disenchanted’. If we accept a connection between importance and reasons, as I’m claiming we should, then this corresponds to either Internalism (there is no absolute importance) or Nihilism (there is no importance of any kind, beyond our psychological state of regarding things as important).

The second claim is first-order, that though there is such importance to be had, humanity lacks it—‘that humans do have a definite measure of importance in the scheme of things, but that it is very low’.

Williams appeals to this distinction in rejecting nihilistic despair as a muddle. Mocking Russell’s ‘self–pitying and at the same time self-glorifying rhetoric’—and it is indeed a little high-flown—Williams claims the muddle is not even worth engaging with philosophically, because ‘the feelings probably come from some place which that comment will not reach’ (Williams 2006, 137). Directed at Bertrand Russell, such dismissive rhetoric had better have solid philosophical backing. It doesn’t:

if the idea of absolute importance in the scheme of things is an illusion, a relic of a world not yet thoroughly disenchanted, then there is no other point of view except ours in which our activities can have or lack a significance. (Williams 2006, 137)

Williams argues that it doesn’t make sense to be upset that humanity is cosmically unimportant because the very idea of the latter is an illusion. Nothing is or could be cosmically important, so there’s no disappointing first-order cosmic judgement to be disappointed by. He claims that Russell must be conflating there is no absolute cosmic scale of importance with the cosmic scale of importance doesn’t think much of us, and so the feelings involve a muddle.

But Williams is mistaken. If there’s no cosmic point of view, then we are not significant from a cosmic point of view. There’s nothing muddled in getting upset about that. The very fact that there is no test of cosmic significance is depressing to some of us, from our own point of view. Why must ‘I wish there were a standard of cosmic significance to measure my activities against’ be a muddle? Similarly, there need be no muddle in an atheist regretting that she is not blessed by God, if that is part of regretting that there’s no God to dispense or withhold blessings.

Perhaps the idea is that absolute importance is outright incoherent and so can’t sensibly be the target of a (frustrated, in this case) wish? Some of Williams’s language does suggest this—a few pages later, he writes that ‘we can’t get a hold of that idea [absolute betterness] at all’ (Williams 2006, 141)—but no argument is given that metaethical realism is not only false but outright incoherent, and even Mackie doesn’t think moral properties akin to round squares. In fact I think there may be something to claims of incoherence, because the idea of a reason is hard to spell out. But Williams himself outlines one (seemingly) coherent way to vindicate cosmic importance: what God cares about. We’ve learnt that God doesn’t exist, not that the very idea of Him is incoherent.

Even if absolute importance is an incoherent idea in the context of our modern disenchanted conceptual scheme, it doesn’t follow that it’s a muddle to be upset that we lack it. Williams himself argues elsewhere that the impossibility of improvement is a good defence (or theodicy) when demanding a justification from (or of) God for the world’s evils. But he also accepts that ‘we have many mere wishes that go against possibility’, and defends a kind of unhappiness that ‘perhaps does presuppose the defeated expectation of something better, but not as a focus of complaint: what has failed is not justification but hope’ (Williams 2009, 54 and 55). And a failure of hope is one way of characterising nihilistic despair. So Williams’s argument is a misfire, and an unfortunately uncharitable one.

More recently, Guy Kahane has also argued that nihilistic despair is muddled, that if we get upset at the unchangeable truth of evaluative nihilism—which says that nothing is bad—then our attitudes ‘make no sense’ (Kahane 2017, 331). If we regard it as bad that nothing is bad, then two components of this state contradict each other and the state is a muddle. For Kahane the muddle lies not in the nonexistence of what is missed, but in contradicting oneself by missing it.

As Kahane notes, a similar argument was made by Hare (1972) in a broadly emotivist framework: if to think something matters is to care about it, then to think nothing matters is to care about nothing. Since we do care about things, things matter to us. Moreover, it would be contradictory to think nothing matters and to be upset about it, insofar as upset is a kind of caring.

Kahane’s argument is intriguing, but I think it also goes wrong. If he’s right, then given belief in nihilism any emotional reaction to anything, or at least any reaction with an evaluative component, is also senseless. Sadness at the death of a dog makes no sense, because that death is not bad. Responses to nihilism itself simply bring out the latent contradiction, and Kahane may endorse this implication, given his characterisation of the nihilist. Something similar is defended by Keller (2017), who argues that for the error theorist love is a mistake, because love posits objective reasons for action.

But that the negative valence in nihilistic despair is the kind of evaluative judgement that nihilism rejects is a substantive claim, and not one I accept. Nihilistic despair involves an unhappy judgement about the (supposed) fact that nothing is objectively bad. Such despair is indeed structurally muddled if the unhappy judgement is or implies an attribution of objective badness. This is really a question for theorists of the emotions, but I’m not inclined to—and I don’t think other nihilists should—accept such an understanding of the attitude constituting despair. Instead, I claim, despair can simply rest on the thought that the truth is at odds with what one most deeply desires (or hopes). I want to be categorically important, and I’ll argue that many of us do.

Indulging in nihilistic despair might not be sensible: it takes time and effort that could be spent promoting other desires. But, I’ve claimed, such despair does make sense, and need not be muddled or incoherent. To show that it’s a muddle we’d need to show that there’s a conflict between the attitude of despairing at nihilism and that of accepting nihilism. The most promising route to this would be if the emotion of despair constitutively involved an evaluative judgement, so that despairing at some X constitutively involved regarding X as objectively bad. Accepting nihilism means accepting that nothing is objectively bad, so in that case regarding this (supposed) truth with despair would be muddled because we’d be judging it bad that nothing is bad. But it’s a substantive claim, and one I reject (but which to be fair is accepted by many), that despair—especially despair felt by the nihilist—implicates such a judgement of objective badness.

3 What We Miss About Morality

Another substantive question—what upsets us about nihilism?—often goes unanswered. We might ask what’s objectionable about an apparent loss of external authority or why we should take measures to avoid nothing mattering. Whether as philosophical position or state of mind, why might Nihilism provoke dismay?

I have sometimes been told that nihilistic despair is a foolish overreaction to the falsity of robust metaethical realism: what about more moderate, naturalistic metaethical realist views? But that misses the point, because the question is conditional: what follows if we think nihilism true? By seeing what’s depressing about nihilism we can also understand whether more moderate views are less glum. We might also ask whether realist metaethical views are also depressing, in other ways of course. If they are, then we face a kind of philosophical pessimism: there are only two options, both bad. (I’ll continue to use evaluative language, in the hope that it doesn’t cause confusion. Here ‘bad’ means ‘unwelcome’ or ‘undesired’.) But that would take us too far afield.

There are several things we might lose when we lose morality or reasons. Morality—or belief in it—could be a powerful causal route to things we value. If there is a casual connection between your believing moral error theory and your acting in ways I’d prefer you didn’t, then if I expect error theory to spread then I may lose some hope for the future.

Morality may also be a source of intrinsic value. For example, if we think that moral virtue is necessarily prudentially valuable—Hooker (1998) thinks it isn’t, but I think it would be—then when we lose morality we lose a source of prudential value. The nihilist may think that compliance with our reasons would have been prudentially valuable. If this would have been intrinsic value, then it’s lost and there’s no substitute.

Somewhere in the middle, we may think morality constitutively part of something we want. Joshua Blanchard claims that morality can supply meaning to life in the face of poor treatment: ‘unjustified harms and oppression cause one’s life to go poorly, but one’s life is better if one has available an authoritative protest […] moral standing provides the strong with a reason not to harm the weak’.Footnote 4 For Blanchard, having such a protest to hand—even when ignored—can provide an irreplaceable sort of meaning. And insofar as meaning is good or desired, its loss is something to regret.

In a similar vein, Williams noted the Kantian hope to convict the immoralist of irrationality on his own terms, which can give us a kind of satisfaction.Footnote 5 This foreshadows the view I’ll defend below, that morality consoles us that wrong actions are somehow also bad for the immoral agent herself, even if she doesn’t see it.

Even those who deny that anything is depressing about nihilism per se might still concede that it is often packaged with depressing non-metaethical theses. It’s no coincidence that—like Mackie—many people are attracted both to error theory and to atheism, as part of the disenchanted modern scientific world view. That view is dismaying to many as Russell showed us (a little self-pityingly—Williams was right about that) and Tolstoy puts it especially clearly via his puppet Levin:

In infinite time, in the infinity of matter, in infinite space, a bubble-organism separates itself, and that bubble holds out for a little while and then bursts, and that bubble is – me.

Levin dismisses this as a falsity, ‘the cruel mockery of some evil power’ (Tolstoy, Pevear, and Volokhnosky 2000, 788–89). Given such connections, perhaps upset at nihilism is really upset at what it implies about mortality, for example? For many people there is clearly something to this.

But there also seems to be something distinctively sad about the loss of absolute importance. Many of us—including perhaps a plurality of contemporary analytic metaethicists—accept atheism, evolution and the rest but hold onto moral realism. For example, even in his defence of pessimism, Benatar (2017) appeals to moral considerations. Adding error theory or nihilism to this package would make it even more pessimistic, but why?

4 Significance and Luck

Morality and reasons are constitutive means to things we value, and nihilism strips us of those things. We have twin desires for significance and insulation from luck that realist views (claim to) satisfy, and nihilism doesn’t. The discussion to follow clearly owes much to Williams, in particular to his Moral Luck and Persons, Character, and Morality. Much as in the latter he criticises ‘recent work in moral philosophy […] of basically Kantian inspiration’, I’ll stipulate a somewhat hazy realist foil:Footnote 6

Plausible Moral Realism

The well-being of other people (or some other feature of them) provides categorical reason to act in certain ways—in particular, not to harm them.

Plausible Moral Realism conjoins a metaethical thesis—that we have categorical moral reasons—with a first-order thesis about what those reasons are. Almost nobody thinks that there are categorical reasons but that they don’t include at least some resembling those of Plausible Moral Realism.

The Kantian Formula of Humanity, for example, tells us to act so that you use humanity, as much in your own person as in the person of every other, always at the same time as end and never merely as means (Wood 2008 [Ak 4:429]). Assuming it’s not too anachronistic to cast his views in terms of reasons, the Formula tells us that we always have reason not to act in certain ways—namely, not to treat other people or ourselves as mere means. But Plausible Moral Realism is not itself a Kantian view, and is entailed for example by the views of Hooker (2000) and Crisp (2006).

Plausible Moral Realism says that we matter. Suppose that I am walking down a Berlin street and trip on a paving slab as a tram passes by. I am dependent on the stranger next to me—call her Sarah—not to push me under the tram’s wheels. My trip would give her plausible deniability: if someone sees her push me, she might claim she was trying to help me and just panicked under pressure. In normal circumstances, Plausible Moral Realism says that Sarah has very strong moral reasons not to push me, and probably to help me. I matter in that I provide categorical reasons. In that sense, I am absolutely important.

That doesn’t mean Sarah will treat me well. But when paired with a non-skeptical moral epistemology, Plausible Moral Realism plausibly implies that the connection between the moral truth and her helping me (or at least not pushing me) is non-accidental, since she is capable of grasping moral reasons and acting on them. That itself is reassuring. If we can grasp moral reasons and we have at least a weak tendency to comply with them, then we should be more optimistic about how our fellow humans will act on the whole. But this causal-statistical claim isn’t the most important consoling aspect of Plausible Moral Realism.

It claims that we have certain reasons, but the concept of a normative reason is hard to analyse. Perhaps incomprehensible, but Williams offers an internalist account of reasons, and so seems not to think so. Reasons guide us in our choices; even if we don’t follow them, we should. Here I’ll make an assumption: if you have a reason, then it’s good for you (in some sense) to act in accord with it.

A strong version of the assumption is that pace Hooker (1998), moral virtue and reasons-compliance more broadly is intrinsically prudentially valuable. A weaker version is that our reasons are necessarily tied to what benefits us. We cannot have a reason to φ if φing doesn’t benefit us in any way whatsoever. This is a substantive assumption but a Williams-friendly one. For example, internalism about reasons together with a desire-satisfaction theory of well-being means that on the whole, when we act as we have reason to, we tend to increase our well-being. In a less loaded way, it could simply be that when we follow our reasons, we avoid acting in a way that is incoherent by our own lights, which would in some sense be bad for us.

This admittedly mysterious connection between actions and our own good is central to the concept of genuine normativity. The good of complying with one’s reasons is the good of reasonableness, and it’s a component of our welfare, either intrinsically or as an instrument to, for example, desire-satisfaction. Reasonableness is a prudential good, or it would be if there were any normative reasons. The connection between morality, reasons, and welfare partially explains the power of Susan Wolf’s ‘moral sainthood’ challenge to contemporary moral theories: they give us reasons which we don’t think it would be at all good for us to follow (Wolf 1982, 419).

An even weaker version of the assumption is that we unreflectively believe that acting on our reasons is good for us, even if that belief cannot ultimately be borne out.

We have seen that Plausible Moral Realism offers us a kind of significance as patients: we are important and others have reason to treat us well. Together with the reason-welfare connection, it says that Sarah’s helping me dodge the tram benefits her: she is more reasonable. Anyone who unjustifiably harms or kills me is made worse off, because they lose reasonableness. Maybe not worse off in every respect, and maybe not worse off than me. But even though I lose my life, my enemy takes a debit to (at least one aspect of) her welfare. I have at least partial revenge because her reasonableness bears a necessary connection with treating me well.

To hark back to Williams’s distinction, we have at least some importance in the scheme of things, and hence (granting my assumption) some impact on the welfare of others. We are categorically significant in that necessarily, others’ welfare depends on their treating us well. This is the truth in Blanchard’s claim that (even as a patient, subject to the will of another) an authoritative protest makes our lives go better (Blanchard 2020, 124). Perhaps this reflects badly on me, but the benefit of a categorical reason not to harm me is not meaning but revenge: if I’m going down (under a tram), then she’s coming down (morally) with me.

Fear often brings out our need for such revenge. I used to live three miles away from my local rail station along a fast and unlit rural road. When cycling home from the station at night, I would get particularly nervous when I heard behind me a large vehicle. As the road in front of me was lit up by their headlights, I took some comfort that since I was wearing reflective clothing and several lights, if they were to hit and kill me it would be their fault and they’d have no excuse. Part of this comfort was of course legal and social, rooted in the the punishment and ostracism the driver would (hopefully!) face, along with having to ‘live with’ having killed me. But not all: if a driver kills me through culpable negligence, that is worse for her too.

Granting some connection between reasons and prudence, Plausible Moral Realism also offers us a luck-insulated source of well-being as agents: if we act as we have moral reason to do, we gain at least some reasonableness. We may sacrifice other elements of well-being, because we might for example have most reason to face intense physical pain and rescue the victim of a trolley problem. But the moral reason and associated welfare remains, even if outweighed.

Special features of morality and reasons insulate them, as Williams highlighted, from luck.Footnote 7 If ‘ought’ implies ‘can’, there must be a way we can act in accordance with our reasons, and those reasons will partially vindicate that action no matter how things causally turn out. Even those versions of realism that allow for moral luck do not typically say that we are always hostage to it. Moral reasons are a route to increased welfare at least partly and usually insulated from luck.

But if Global Normative Nihilism is true, then there are no normative reasons and so no good of reasonableness. It robs us of luck-insulated welfare as agents, because there is simply no normative sense in which our actions are correct or incorrect. Sartre calls such a phenomenon abandonment: without a God or ‘luminous realm of values … [without] any values or commands that could legitimise our behaviour … we are left alone, without excuse’.

He argues that such abandonment, together with a commitment to each person as ‘a legislator deciding for the whole of mankind’, brings a ‘sense of complete and profound responsibility’ (Sartre 1948, 34 and 30). Here I part company with Sartre, because part of nihilism’s point is that we do not legislate for anyone else or even our future selves. There is no legislature. So I adopt the evocative ‘abandonment’ terminology with the caveat that I do not have quite the existentialist sense in mind. We are abandoned in the sense that there is no guidance for our actions.

This is a double-whammy for us agents: we may simply want guidance for its own sake, and in particular we may want luck-immune guidance. In Sartre’s most famous example, a student must choose between two weighty courses of action—caring for his mother and fighting for France—and cannot do both. At least absent a genuine moral dilemma, Plausible Moral Realism tells the student that if he can identify the permissible thing to do, then he can follow that course and be justified. The damage to France or to his mother is a high but acceptable price. Neither course is cost-free, but insofar as he chooses correctly he is justified and his life goes well to that extent. (If we are attached to a fair and luck-immune picture of morality, the fact that moral dilemmas would upset such a picture is some motivation to reject their possibility and to argue that ‘ought’ implies ‘can’. It’s also no coincidence that Williams objected to an insistence on tragedy-free rankings of outcomes.)

As patients, Nihilism robs us of categorical significance.

For the Nihilist, there is no sense in which the particles that make up my body necessarily matter more than any other similarly-sized clump of matter. None provide categorical reasons for action, so we are on a par with the other constituents of the universe, including nonhuman animals and distant, inanimate matter. We have no special reason-giving status and so indeed lack absolute importance in the scheme of things.

If you’ve accepted my assumptions, you’ll agree that Nihilism is grim. We thought there was a necessary connection between my welfare and that of anyone who acts, but there isn’t. There need be no sense in which Sarah is worse off if she pushes me in front of the tram, and so there may be no vehicle for my revenge. Not only is it somewhat more likely that people will harm me (as I argued above), those who do so need be taking no debit to their welfare. If the driver’s conscience doesn’t trouble him for killing me, he need be making no mistake.

To recap, Plausible Moral Realism can be comforting in two ways. Even if in the long run we are all dead, for now we are not abandoned and our actions may be what they ought, or not. They are assessable against an absolute, luck-immune, and inescapable standard, compliance with which is good for us. Second, anyone who treats us in certain ways—violating the absolute standard—is making a mistake, and taking a pro tanto hit to their own welfare. Nihilism denies us these consolations, and so frustrates deep desires.

If Nihilism is true, then it is in no sense objectively sad. This is the truth in Kahane’s argument. Whether each of us will find it sad depends on whether we have the desires in question. Clearly, many do.

Desire for significance does seem common—fiction is full of stories in which human action turns the tide of the cosmos. Perhaps the desire for significance has an evolutionary basis. Quentin Smith suggests that we live in ‘an illusion of meaningfulness’, which allows us to survive and reproduce (Smith 2003, 53). Kahane, however, is as dismissive of an evolutionary basis as Smith takes it to be obvious, claiming that ‘it’s not as if evolution had made us react with fear and despair to nihilism as we instinctively fear snakes…’ (Kahane 2017, 6). But neither offers (or claims to offer) arguments for these evolutionary claims, and so I put this matter aside. Kahane claims that coming to believe nihilism leads to a near-complete decay of subjective concern and action (I disagree). If he is right, then non-nihilists may be more evolutionarily successful, and so perhaps fearing nihilism would be evolutionarily adaptive.

But this is entirely speculative. If such attitudes are nearly universal, we might have a secondary quality account of nihilistic despair: nihilism is such as to provoke sadness or despair in normal observers in normal conditions.

5 Internalism about Reasons is Even Sadder (in one respect)

I’ve argued that nihilism is depressing. Williams-style internalism says that there are normative reasons, but all are (on typical versions) connected to our desires. It might be thought that because such a view is closer to moral realism—it accepts some normative truths—it must be less depressing than nihilism.

I’ll argue the opposite: Internalism is sadder in at least one respect. Of course Nihilism may remain sadder in others, and I don’t attempt to assess the overall balance. The respect I have in mind is that Internalism re-exposes us to luck as agents (through what reasons we have) and as patients it says that others might do better if they harm us.

Consider Williams’s ‘hard case’, who has no motivation to be nicer to his wife and therefore has no reason to be nicer to her, even though we may truly call him ‘ungrateful, inconsiderate, hard, sexist, nasty, selfish, brutal, and many other disadvantageous things’ (Williams 1995, 39). From our point of view, of course. What the man is exactly guilty of is not clear. Some of these insults would be appropriately directed at an abusive husband, but he is also described as lacking a reason to be ‘nicer to his wife’, which suggests lesser wrongdoing.

Internalism says that we all lack necessary significance. Our well-being may indeed provide reasons for others, but contingently on the desires of those others. There is genuine normativity in the world, but it does not assign us any distinctive or categorical status. Whether any person or group matters to an agent depends on that agent’s motivational set. We do fail a test of being categorically important in the scheme of things, though so does everything else.

As he is standardly understood, the hard case has no desires that would be promoted by being nicer to his wife, or they are outweighed by his patriarchal desires. He makes no more mistake in treating her badly and ignoring her welfare than I—as someone who does not care about cricket—do when I ignore cricket. This comparison may be offensive, but that is part of the point: according to austere internalism, my and his actions are on a par in terms of desire satisfaction, and those are the only terms. The view knocks us off the pedestal that Plausible Moral Realism erects for us, the pedestal for those whose needs nobody may ignore.

If acting in accord with one’s reasons contributes to welfare, then insofar as he gains things he values (such as domination of the household), the hard case gains reasonableness and hence well-being. Hence Internalism adds an extra way in which it’s better for him to treat his wife badly: not only does he get what he wants, getting what he wants brings him the good of reasonableness. Internalism perverts the Kantian ambition I mentioned above: he is irrational on his own terms insofar as he does not treat others badly.

Whether this is a mere conceptual possibility or something more threatening depends on how often we have strong internal reasons to treat others poorly (in particular, to treat them in ways that would have been forbidden by Plausible Moral Realism). That’s an empirical question about our desires, but the Hard Case is just one of indefinitely many plausible examples we can think of. When we lose the categorical, we lose the claim that there must always be some way in which someone who wants to push me under a tram is irrational, or going against her strongest reasons.Footnote 8

This can be particularly depressing if we assume that inter-agent comparisons can be made, and not just counterfactual comparisons with the same agent acting differently. Given that assumption, Internalism says that the hard case may be normatively better off than both his wife and us. Plausible Moral Realism offered me revenge on Sarah. Even if I lose a physical confrontation with her or even my life, I can be better off than her in one respect. I can be reasonable, and she is not. She may be rich, but I am rich in spirit. Nihilism dismissed reasonableness as a myth. It was depressing to think there’s no spiritual wealth, but Internalism rubs salt in the wound: there is robust welfare, but it bears no necessary connection to treating me well, and in pushing me in front of a tram she might be gaining reasonableness. Perhaps the eudaimonic difference between us is increased by reasons: she gains reasonableness, and I lose my life. Being rich in coins makes her richer in spirit.

Plausible Moral Realism offered a luck-insulated component of welfare as agents, and this is substantially weakened by internalism. There is some luck-insulated reasonableness on offer: your desires provide reasons, and you can be more or less reasonable insofar as you comply with them. The reasons we have affect our welfare, and are exposed to luck and contingency in that some desires are easier to satisfy, and so on. We do not all share categorical reasons, which would have been a kind of level playing field.

Turning from luck to abandonment, internalism offers us only the most limited kind of guidance. It endorses the goals we already have, perhaps subject to procedural purification and the correction of factual mistakes, and the like. Truly external vindication of our actions and preferences is not on offer.

To my knowledge, only Manne (2014) has recently argued that internalism about reasons is sad or depressing. Her internalism is somewhat different to that sketched here: for Manne, an agent has a reason to do something only if they can be interpersonally reasoned into doing it. Our reasons depend on what we can be talked into, which may not be precisely what would satisfy our desires. But the core of the view is shared with Austere Internalism.

Much of Manne’s discussion focuses on Williams’s abusive husband. It may be that we cannot reason him into treating his wife better, she claims, because ‘our tongues are tied by his motivational deficits. This is admittedly sad, but it may be true nonetheless’ (Manne 2014, 111). So he doesn’t have a reason to treat his wife better. So far, I agree with Manne’s diagnosis: her version of internalism implies that people, including his wife, have no categorical significance, in that they don’t necessarily provide reasons.

But by what independent standard are his motivations deficient? Plausible Moral Realism had an answer, but that’s not available here. I think what’s truly sad is that he need not be deficient: he could be functioning perfectly well, albeit in a way we’d condemn.

(Manne 2014, 115) thinks it sad that ‘there are some people who are most plausibly interpreted as being beyond the reach of being reasoned with about what they are doing’. But there’s nothing particular to internalism about that: wrongdoers will always be with us, whether or not they have categorical reason to be nicer. She describes individuals like the hard case as ‘unguided missiles’, but I think this exactly the wrong way around. Internalism says he may be a guided missile, may be acting perfectly reasonably. Sadly, he is guided to act in ways that harm others. And—to strain the metaphor—insofar as it is good for a missile when its guidance system functions well, he does better insofar as he harms others. (According to Plausible Moral Realism his guidance system must be malfunctioning in a way that’s bad for him; Nihilism makes him truly unguided.)

Many of us feel that internalism about reasons is akin to nihilism. I’ve argued that it is more depressing than outright nihilism, in one way at least. Rather than being a fair and equalising force, normativity can exacerbate luck and inequality.

6 Conclusion

I’ve defended nihilistic despair: we need not be muddled in being upset at the loss of what morality promised us. If there is no objective moral point of view—in particular no objective point of view that provides reasons for action and contributes to our well-being—then we lack categorical significance and are more exposed to luck. We can find that prospect depressing without incoherence or particularly odd desires.

If nihilism is true, then we must set aside the notion of an objectively appropriate or fitting response to its truth. But I’ve argued that nihilistic despair need be no muddle and is natural given desires for significance and guidance. I also sketched an argument that those desires are common.

But perhaps not universal—you might not have abandonment issues, and may even experience nihilistic joy if you prefer the rejection of objective values, having desires better satisfied by the absence of such guidance. Imagine you wish to be a terrible tyrant and have the strength to make this a realistic possibility. If there are categorical reasons not to have your rivals killed then you must either give up some welfare from moral reasons or from tyranny, because you can’t have both. If Nihilism is true, then you face no such choice; if Internalism is true, then you have reasons to be a tyrant.

Closer to home, someone we might call the moralist seeks objective grounding or support for her morally-laden actions. She seeks objective support for the claim that she is more reasonable than those who push others in front of trams. But she’s content with arbitrariness or contingency in non-moral matters. Merely internal support will do for her choice of tea or place to live. Many of us are moralist in this sense.

At the other extreme, some people can’t bear rational arbitrariness. Such a person desires objective grounding or support for her most trivial preferences. If she likes red wine and her friend doesn’t, she wants assurance that one of them is wrong, absent some vindication for the divergence (such as neurological difference). She wants reassurance that everything she does is objectively best—or at least objectively acceptable, in cases of incommensurability and the like.Footnote 9

This person will find Nihilism and Austere Internalism quite disturbing, but perhaps also many versions of Plausible Moral Realism, which allow for desire-based variation in many normative reasons. This most rationally insecure person will experience Street’s vertigo about aesthetic pursuits and the reasons to pursue her hobbies. There is no best hobby that ought to be pursued.

As we might expect, nihilistic despair focuses on the peculiar features of absolute justification or moral realism. Those things are causally inert, but they (in standard forms) make us absolutely important and offer an inescapable and important standard for the assessment of our actions, and the actions of others towards us. Upset at the loss of such importance and assessment may not be heroic, but it is understandable. Exactly how upset you are—and exactly what you are upset about—will depend on which desires you have, and thus which desires are unsatisfied by the metaethical truth.