2. The conditionality of theories of intergenerational justiceFootnote ¹

The effects of climate change will extend far into the future, long after currently living people have stopped existing. A very prominent way of understanding the ethical implications of this is to use the theoretical resources provided by theories of intergenerational justice. Shue, for instance, writes that ‘rapid climate change places current and future generations in precisely the kind of general circumstances that call for the construction of rights-protecting institutions’ (Shue Reference Shue and Arnold2011), and there are many other examples in the literature of the basic insight that underlies this approach, namely the idea that we should conceptualise all or a significant part of the potential harm caused by climate change in terms of potential violations to rights (see e.g. Bell Reference Bell2011; Caney Reference Caney2008, Reference Caney2009, Reference Caney, Gardiner, Caney, Jamieson and Shue2010; Hayward Reference Hayward2007; Page Reference Page2006; Vanderheiden Reference Vanderheiden2008).

There are well-known problems with extending theories of justice to cover relations with future non-contemporaries (see Meyer [Reference Meyer and Zalta2015] for a description of these. See also Beckerman and Pasek [Reference Beckerman and Pasek2001]; Herstein [Reference Herstein2008] for scepticism about extending justice to cover intergenerational relations) and in this paper I am concerned with none of them. Rather, I am going to simply assume in this paper that theories of justice can be extended to cover intergenerational relations, because what I am interested in here is some of the assumptions that must necessarily be made for this extension to work.

The basic and fundamental assumption can be summed up succinctly: in order to care about the future it is necessary that there be a future. More specifically, theories of intergenerational justice need to make assumptions about the existence, number and identity of future agents in order to be able to say something both meaningful and useful about the obligations that currently living people have towards them. We can see this clearly using climate change as an illustrative example.

First, in order to even begin the task of identifying which future consequences of climate change will likely cause rights violations, it is necessary to assume that there will be agents who exist in the future at all. If there are no such agents, then nothing currently living agents do will or even could lead to rights violations of future generations.

Second, it is necessary to assume that future agents will be of a certain sort; that is to say, we need to assume what kinds of interests those agents will have. Without this assumption, we do not know which consequences will actually be agent-affecting in the future, and which consequently could potentially cause or constitute rights violations. If, for example, we cannot say of future agents that they will require what we currently think of as food to survive or flourish, then we cannot claim that actions of currently living agents that affect the ability of future agents to obtain (what we think of as) food are actions with agent-affecting future consequences.

Third, a theory of intergenerational justice needs to make assumptions about how many future people there will be. Consider, for example, the possibility that the actions of currently living people will lead to drought in the future. ‘Drought’, is not an agent-independent feature of the world; rather, it depends crucially on how much water there is relative to how many agents there are – what would be far too much beer for one philosopher could legitimately be described as a drought for a rugby team. So even after we claim that currently living agents have an obligation to ensure that the rights of future agents to a satisfactory amount of acceptable drinking water are not violated, on its own this claim is insufficient to translate that right into an account of the duties of currently living agents. That translation requires making an assumption about the number of future agents there will be. The same point can be made about many other potential future consequences of the actions of currently living agents. A world with 20 billion agents in the future will require much more of currently living agents than a world with a million.

This also shows that not only do theories of intergenerational justice rely on certain assumptions about the future, but they are also highly sensitive to the precise assumptions that are used. Indeed, the very same theory will say very different things depending on what it assumes about existence, identity and number. For example, in the extreme case, a theory which assumes that there will be no future agents will tell us that there are no restrictions on the actions of currently living agents qua the future consequences of those actions, while exactly the same theory will impose duties on currently living agents if it uses the assumption that there will be a future population of 15 billion people.

In other words, theories of intergenerational justice take a conditional form. The consequent is whatever the theory says the duties of (for e.g.) intergenerational justice are; the antecedent is, if there are future agents of a certain type and given number. If the world were to end tomorrow, from the perspective of justice there would be no point in even asking what to do about the future consequences of climate change, let alone actually doing anything about them.

3. Two features of the intergenerational justice conditional

I now want to pick out two features of the intergenerational justice conditional. The first feature can be seen by contrasting it with a different and familiar ethical conditional, namely: if there turns out to be a baby drowning in a pond, you ought to save it. This conditional tells people what they ought to do when the antecedent comes true. But the intergenerational justice conditional is more like: if there turns out to be a baby drowning in a pond, you ought to take swimming lessons now. This conditional tells agents to act now on the basis that the antecedent will at some point in the future come true.

Unlike many other conditionals, therefore, which only require sacrifice at the point when the antecedent comes true, theories of intergenerational justice which claim that currently living agents ought, for instance, to reduce their total emissions, are theories which require sacrifice from (at least some of) those agents now. That is to say, from the temporal perspective of a currently existing agent, these theories do not, in the end, make conditional claims. They make the categorical one: Act now on the basis that the metaphysical assumptions will come true.Footnote ²

The second feature of the intergenerational justice conditional that I want to pick out is that the antecedent is something that is in principle entirely within human control: it is in principle possible for humans to collectively decide to let the human race peter out. Further, even at a more practical level, at the very least it is true that the number of future agents can, within certain limits, be a matter of human choice. This choice, we have seen, makes an enormous difference to what any given theory will understand as the agent-affecting consequences of climate change, and consequently what it will see as the problem, and consequently what it will tell us we ought to do about that problem.

4. The ethics of metaphysics

The name ‘the ethics of metaphysics’ is my term of art to cover a number of questions, all of which are specific variants of one basic question (see Jonas [Reference Jonas1984] and Heyd [Reference Heyd1994] for what in my view is the best work on both identifying and discussing many of the issues in the ethics of metaphysics). While absolutely nothing of substance turns on my use of this term of art, it would be well to briefly explain why I chose it.Footnote ³ It is not straightforward to give a non-trivial and uncontroversial description of what the subject matter of metaphysics is (see van Inwagen and Sullivan Reference van Inwagen, Sullivan and Zalta2016). This means that not everyone will agree on what counts as a metaphysical question. It also means, however, that it is defensible to hold that one of the questions of metaphysics is: why is there something rather than nothing? The ethics of metaphysics introduces an ethical element, because it asks: why should there be something rather than nothing? And, if we decide there should be something rather than nothing, what should there be? In the context this paper is concerned with, i.e. future generations, this means the following three questions:Footnote ⁴

1. (a) Should there be agents in the future?

2. (b) How many agents should there be?

3. (c) What kinds of agents should there be?

These three questions are generated simply by turning the assumptions about the existence, number and identity of future people into objects of ethical inquiry. I now turn to arguing that it is necessary to do the ethics of metaphysics if we are to develop a properly comprehensive understanding of our ethical relation to the future, and to future generations.

5. Why we need to do the ethics of metaphysics

In this section, I give two important reasons for why it is necessary to ethically investigate and justify any particular assumptions that theories of intergenerational justice make about the existence, identity and number of future people; why, in other words, we must do the ethics of metaphysics.

First, as we have seen, theories of intergenerational justice must make assumptions about future population in order to say something meaningful and reasonably determinate about our obligations to future generations. These assumptions are not given facts about the future that must simply be accepted. Rather, their truth is in principle under entirely human control. It is the choices we make, through collective population policies and individual procreative decisions, which determine whether there will be people in the future, and how many of them there will be.

As we have seen above, theories of intergenerational justice demand sacrifice of currently living agents now on the basis of those assumptions, for e.g. the claim that currently living people have a duty to reduce emissions, despite the associated decline in welfare, because of the likely effects of unchecked emitting on future generations. But given that whether the assumptions come true is under human control, it is entirely legitimate for an agent to ask: ‘Why should I accept the ethical necessity of making these sacrifices? The sacrifices you require me to make are only necessary if there will be future agents. And not only am I am not going to make any future agents, I think there shouldn’t be any future agents at all’.Footnote ⁵ The point is not that this response cannot be addressed, the point is that it needs to be addressed: if you are going to ask an agent to sacrifice something now, for the sake of a future she not only does not want to create but thinks ought not to be created, you have to somehow justify to her that it is worth bringing that future about, and it is worth enough for her to be required to make sacrifices for its sake

The second reason for the necessity of the ethics of metaphysics stems from the transformed condition of human action, and in particular its enormously increased power. As Jonas (Reference Jonas1984) points out, ‘with certain developments of our powers the nature of human action has changed’ (1), and one of the changes he has in mind is the ability human beings now have to decide (in the extreme case) whether ‘there will be a world for the generations of man to come’ (10). Heyd reiterates the truth and ethical importance of this claim, when he states that

The argument here is very simple: Power brings responsibility in its wake, and our powers over the existence and fate of the world come with some responsibility. At the very least, this is the responsibility of being able to justify an entirely ‘irresponsible’ use of it. And it is ultimately impossible to do that, or to determine what would be responsible use, without asking the fundamental questions asked by the ethics of metaphysics. Even when someone makes the assumption that there will be no future agents, for example, and concludes therefore that we may use our powers however we like, it is necessary to ask: is that right? Or should we make it the case that humanity will continue to exist?

To my mind, each of these two reasons taken on their own constitute a strong independent case for the necessity of the ethics of metaphysics. In case some find the second more convincing than the first, it may be worth explicitly noting here that the second argument makes no reference to sacrifices imposed on currently living people, and can therefore be accepted even if the first is rejected. I have tried to suggest, however, that both should be accepted.

A selective summary is in order before proceeding. The two main claims at this point in the paper are that (i) theories of intergenerational justice rely on assumptions about the existence, number and identity of future generations; and (ii) these assumptions must be made the object of direct ethical inquiry, i.e. we must do the ethics of metaphysics. I now turn to arguing that theories of intergenerational justice lack the theoretical resources to properly ethically investigate the assumptions they rely on.

6. Justice and the ethics of metaphysics: Why theories of intergenerational justice cannot deal adequately with questions in the ethics of metaphysics

The first question in the ethics of metaphysics, as it is outlined here, concerns the continued existence of humanity. It concerns whether that continued existence is valuable, and if it is, how valuable it is in relation to other values. My claim in the rest of this paper is that theories of intergenerational justice lack the theoretical resources to allow us to properly answer this question. This theoretical poverty of justice can be illustrated by considering what it can tell us about whether we have a duty to ensure the continued existence of humanity.

On the understanding that justice is a person-affecting value, justice applies to relations between people (a) once they exist or (b) on the assumption that they will exist. This means that justice cannot ground a duty of currently living people to future agents to bring them into existence. It is clear, I think, that justice cannot ground such a duty. It is clear because ‘the nonexistent makes no demands and can therefore not suffer violation of its rights. It may have rights when it exists, but it does not have them by virtue of the mere possibility that it will one day exist. Above all, it has no right to exist at all before it in fact exists. The claim to existence begins only with existence’ (Jonas Reference Jonas1984, 39. My emphasis). Indeed, it is misleading (but unavoidable) to say things like ‘the nonexistent has no right to exist’, because the grammar of the sentence suggests that the nonexistent is a subject of some kind. It is not. ‘It’ is nothing. Attributing anything, let alone a right, to nonexistence is not just wrong, but incoherent.

The key point about nothing is that it is not a thing – it is the absence of things. As Kolakowski (Reference Kolakowski and Kolakowska2001) says, ‘“Nothingness” is no more than an illicit substantive formed from the particle of negation – the simple and useful word No.’ (16). It’s language that’s deceiving us here – because ‘nothing’ can be the subject and object of sentences, we can slide into thinking it is something. And of course, that is precisely what nothing is not.

This is not the sceptical claim that ‘future generations – of unborn people – cannot be said to have any rights,’ (Beckerman and Pasek Reference Beckerman and Pasek2001, 14) nor, specifically, scepticism about future agents having at least some of those rights vis-à-vis currently living agents (based, for e.g. on the view that the non-identity problem cannot be solved). The claim is only that ‘possible future people do not have a right to existence’ (Meyer Reference Meyer, Shelp and Fotion1997, 139). This is entirely compatible with believing that future people do have rights vis-à-vis currently existing agents, as long as one accepts those rights are contingent (See for example Feinberg Reference Feinberg and Blackstone1974).

If this is correct, a theory of justice has two options left in attempting to ground a duty to ensure the continued existence of humanity. It can claim there is a duty of justice to ensure that there will be agents in the future either by arguing (a) that justice is impersonal in some way or (b) it is a duty of justice to currently existing agents. I begin with the impersonalist attempt.

7. Justice as an impersonal value?

There are many different ways to understand what it means for value, or a value, to be person-affecting. To begin with the most famous, Parfit describes ‘The Person-Affecting View, or V: It will be worse if people are affected for the worse’ (Parfit Reference Parfit1984, 370). The idea behind this is that ‘what is bad must be bad for someone’ (363). There are a good number of sometimes competing, sometimes complementary, interpretations of this basic idea, its relevance, and its implications. I do not want to enter that debate. Luckily, it isn’t necessary, I think, because what I want to do here is argue that even if one holds that justice is an impersonal value, this doesn’t help in allowing justice to ground duties towards making some particular ontological assumptions about the existence, identity and a number of future agents come true. For this, it is only necessary to say what it would be for justice to be impersonal, and all I will say is this: whatever else the view contains, it must contain the claim that justice has value over and above any value it may have for and/or to persons. The basic idea can be expressed like this: in a world where there were no persons, and would never be persons, justice would still be a value. This might appear to be a way for theories of justice to ground duties to make certain metaphysical assumptions come true. By claiming that the value of justice has nothing to do with people, it seems to open up the theoretical space for duties of justice which have nothing to do with people. Impersonalist theories of justice may, in other words, be able to ground a duty that there be future agents.

Before examining (and rejecting) this possibility, I want to stress again just how unusual the impersonalist view is when it comes to justice. So for instance, in a discussion of libertarianism in the context of intergenerational justice, where the authors say ‘we do not attempt to defend libertarianism. Instead, we work out the most plausible version thereof and identify its implications for intergenerational justice,’ (Steiner and Vallentyne Reference Steiner, Vallentyne, Gosseries and Meyer2012, 50) they pick out a particular sense of justice that they will focus on. This sense is justice as ‘what we owe each other’, and as is clear simply from that, they point out that ‘justice as what we owe each other is not concerned with impersonal duties (duties owed to no one, i.e. that do not correspond to anyone’s rights). If there are impersonal duties, then something can be just but nonetheless morally impermissible’ (50–51).

This is stipulative fiat, of course, but it is still instructive, because it tells us something about how theorists of intergenerational justice understand the value of justice. Or take another example. A different pair of authors begin their paper with this question: ‘What do we owe to future people as a matter of social justice?’ (Meyer and Roser Reference Meyer, Roser, Gosseries and Meyer2012, 219). They say that they will ‘follow a broad understanding of intergenerational justice: justice considerations are relevant to decisions that are likely to affect the existence, number and identity of future people if – with respect to these decisions – future generations can be viewed as holding legitimate claims or rights against present generations, who in turn stand under correlative duties to future generations’ (219). Again, on this view, justice is relevant to intergenerational relations if future people have rights against current people, and these generate correlative duties for the latter towards the former. That is to say, on this understanding too the duties of justice are duties owed by people to people.

Indeed, it is remarkably difficult to actually find an account in the literature that claims that there are impersonal duties of justice – in our context, an account which claims that there is a duty of justice that there be future people which is not a duty owed either to future or currently living people, or both. I confess, in fact, that I don’t know of even one such view.

The closest I could find to such an idea is the suggestion that ‘a complete account of intergenerational justice must appeal, at least in part, to ‘identity-independent’ principles. Identity-independent principles evaluate acts and policies according to the extent that they promote the quality of life independently of how particular individuals fare in alternative scenarios’ (Page Reference Page2006, 166). According to the author, such principles, ‘when focused solely on human life … are referred to as ‘impersonal’ principles’ (166).

Now, these may certainly be called impersonal principles if one feels like it, but they are not impersonal in the sense relevant to us. If these principles are what the impersonalist strategy uses, then it cannot generate a duty of justice that there be future people. Why not? Because these principles, while not caring about particular individuals, ‘evaluate acts and policies according to the extent that they promote the quality of life’. For these principles to tell us anything about how we should act towards the future, it has to be assumed that there will be agents in the future whose quality of life can be promoted. If this is what is meant by impersonalist principles of justice, then impersonalist principles do not allow us to ground a duty of justice to make any particular set of ontological assumptions come true.

In a word, the relevant sense of impersonal is not identity-independent, but existence-independent. We are looking for a way in which principles of justice could tell us that we have a duty to make the assumption of existence (and the others) come true, not for a way in which they can tell us what we owe people assumed to exist even if their existence (and identity) is contingent on our actions. This is not done by the only candidate I know of for impersonal principles of intergenerational justice.

Here is one alternative possibility.Footnote ⁶ The view is this. First, we deny the claim that the duties of justice are purely duties owed by persons to persons. Second, we say that there is a second type of duty of justice. Third, we say that this is the duty that there be justice. This duty is not owed to any agents, but simply (I imagine) comes from some background duty towards the value of justice.

This background duty can be understood in terms of the natural duty of justice.Footnote ⁷ According to Rawls, ‘the natural duty of justice … requires us to support and to comply with just institutions that exist and apply to us. It also constrains us to further just arrangements not yet established, at least when this can be done without too much cost to ourselves’ (Rawls Reference Rawls1999, 99). Now the first part of that duty can’t help us, because it deals with existing institutions as they apply to existing people. The second is where the action must be. That second part, to further just arrangements not yet established, of course makes the most sense when one assumes there will be agents in the future to whom those arrangements will apply. On this understanding, the duty to further the creation of just arrangements is really a duty to those future agents. But for our reinterpretation here, this will not do. That second part must be understood as a duty to further just arrangements for the sake of justice itself.

Now, if there is such a duty, and if it is true that for the value of justice to be instantiated it is necessary that there be people, then we can derive a duty of justice that there be people. Generalising, the duty to ensure that there will be people would be understood as part of the duty to create the preconditions necessary for the value of justice to be instantiated, with the duty to create those preconditions itself derived from the duty that the value of justice be instantiated.

There are a huge number of problems with understanding this view, of course, and an even greater number with understanding why one should believe it. But my concern here is not with whether one should have this view but whether, if one has it, one can avoid the necessity of the ethics of metaphysics. I do not think one can.

The first issue is as follows. On the view outlined, the duty to bring about the preconditions of justice (including but not restricted to bringing about the existence of agents in the future) is somehow connected to the value of justice. The first question to ask is: how exactly should we understand this connection? Put another way, what does fulfilling this duty to create the circumstances of justice achieve in terms of the value of justice?

It is clear that simply securing the preconditions of justice does not guarantee that there will be justice – securing the preconditions of justice does not secure the existence of justice, it only secures the possibility of its existence. Further, securing the possibility of the existence of justice also secures the possibility of injustice, because the preconditions of both are the same. This I think is clear – the preconditions which secure the possibility of justice are preconditions which allow us to speak of a situation in terms of justice. And one of ways to speak of these situations in such terms is to say that the situation is unjust. That is to say, discharging the duty to bring about the preconditions of justice means creating not only a world where there could be justice, but simultaneously and necessarily also creating a world where there could be injustice.

Very well. One option now is to hold that there is a duty to bring about the preconditions of justice even though this at the same time brings about the possibility of injustice. But if this is the interpretive option taken, then the value being appealed to is not, ultimately, the value of justice. Why not? Because on this option the duty is interpreted to hold even though fulfilling it could lead to injustice. The ground of the duty must therefore lie in it creating the possibility, rather than the reality, of justice. Consequently, what is being appealed to is the value of there being a world in which there could be justice, not a world in which there is justice. And this appeal is a view within the ethics of metaphysics rather than theories of justice. It is a claim about what there should be which is grounded in a claim of the value of there being the possibility of value, rather than being grounded purely in a claim about the value of there being justice.

So this view, to avoid the necessity of the ethics of metaphysics, must interpret the duty in some other way. What might other interpretations be? One, I suppose, could be that the duty is to be interpreted as being a duty to bring about the preconditions of justice when doing so is associated with a higher likelihood of bringing about the existence of justice rather than injustice. But this too is a claim in the ethics of metaphysics; or at least, it presupposes such claims. Why? Because such an interpretation expresses a view on the relative weight of the existence of the value of justice versus the value (if any) of the possibility of there being value; and, as I have suggested, the latter claim is a claim in the ethics of metaphysics.

We can generalise the point thus. Any interpretation of the duty to bring about the preconditions of justice must recognise the fact that bringing about the preconditions of justice necessarily and simultaneously brings about the preconditions of injustice. It must recognise this because it is just a conceptual truth. But once it recognises this, it is forced to take a view on the question of whether there is a value to bringing about the possibility of value (in this case justice) – and whatever view one takes on this question will be a view within the ethics of metaphysics rather than theories of justice.

So much for impersonal theories of justice. There is one final strategy justice can try to use to ground a duty to ensure the continued existence of humanity, namely to argue that there are of duties of justice to create future agents which are owed to currently existing agents. I now turn to examining this, and to arguing that even on this strategy it is not possible to evade with the necessity of the ethics of metaphysics.

8. Duties of justice to currently living agents to ensure the continued existence of humanity

The attempt is to justify a duty of justice to currently living agents that there be a future, and of a certain sort. Clearly, any such justification will require, as a necessary but insufficient element, an account of why it matters so much to currently living agents there is this certain sort of future. This is insufficient, incidentally, because as has been pointed out,

That is to say, in addition to arguing that it is important to currently living agents that there be a future, it is necessary to argue that it is important enough to justify duties of justice to those agents to bring it about.

Now, there are different specific ways to argue for this view, and I think all of them share some version of a simple and powerful idea, and the specific arguments are attempts to somehow express it. That idea, in its strongest form, it is something like this: for the present to matter, it is necessary that there is a future. In some way, the fact that there is a future confers meaning on the present (See for e.g. Heyd Reference Heyd1994; Meyer Reference Meyer, Shelp and Fotion1997; Scheffler Reference Scheffler2013).

So for instance, when Thompson (Reference Thompson, Gosseries and Meyer2012) writes of lifetime-transcending interests and attempts to ‘establish that they play an important role in the lives of individuals and the formation of their identities,’ (35–36) her argument ‘concentrates on two ideas: that lifetime-transcending interests are essential for a meaningful life and that they are a pre-requisite for making a rational plan of life’ (36). Following Partridge, she claims that ‘“self-transcendence”, as he calls it, gives meaning to our lives and projects, especially when we face the fact of our own mortality,’ (36) and she attempts to strengthen Partridge’s point by arguing that even those individuals who do not have lifetime- transcending interests are ‘riding piggyback on the lifetime-transcending interests of others’ (36). As an example of this, according to her, is the ‘millionaire ((who)) would not get satisfaction from owning great works of art and denying them to the public gaze if he did not believe in the value of these works as important contributions to an artistic heritage’ (36). The idea is that even the selfish millionaire, gloating privately over his art collection, can himself only value the art collection if he sees it as part of an activity or practice whose value and existence transcends the limits of his temporal self.

I do not want to criticise this argument as an argument; I am interested in what one must believe to believe it. And I think that one has to believe some sort of claim like the one I gave above, about the existence of the future somehow conferring meaning on the present. A different way of putting it might be to say that the underlying idea, at its most basic, is something like: life is not enough. Or, perhaps more precisely, life – understood as our temporal span – is enough only if it involves something that goes beyond its temporal span.Footnote ⁸

Funding

This work was supported by an Austrian Science Fund [grant number P12345-B51].