1
Can we act unjustly toward the future?

There is a strong sense in which the ability of humankind to deeply change the world and the future has increased over the last decades, which led to calling our times the Anthropocene. It seems to follow that our ability to harm the future has risen accordingly and that this entails a rising responsibility. Yet, like a grain of sand in the gears, the "non-identity" problem, unveiled by Parfit in the late 1970s, has kept the philosophical community busy ever since.1 For this problem questions the possibility of meaningfully characterizing our actions as potentially harmful to the future. In a sense, work by philosophers on the "non-identity" problem is akin to that of the biomedical community facing a new pathogen. The fact that a pathogen has been newly discovered does not mean that it cannot be dangerous. Instead of denying its existence or importance, we should try to precisely assess its nature, implications, and possible remedies.

I will present and defend three main strategies to address the non-identity problem. They are partly complementary. Each of them relies on different assumptions about the justice-harm nexus. Each seeks to show that while the non-identity problem is serious, it does not render the idea of significant obligations to the future meaningless. The chapter is structured as follows. I first present the non-identity problem. Next, I discuss the link between harm and justice. I then move to the three "rescue" strategies. I explore whether relying on a new grammar - namely, alternative concepts of harm - can offer a solution. I then present the containment option. It sticks to an ordinary concept of harm, while calling for rethinking the nature of our intergenerational obligations. Finally, I explore the severance alternative. Here, the idea is to free concerns of justice to the future from heavily relying on the concept of harm.

Let me add two words of caution. First, having gained an initial grasp of the non-identity problem, unfamiliar readers might feel cheated and react with disbelief or even irritation. They might think: "I can't believe this! Philosophers have made it up! How can they possibly spend time at all on such a far-fetched problem while some of Earth's fundamentals are collapsing?" I hope to show that, while surprising, the problem is serious. This holds regardless of whether one thinks that too much intellectual energy has been spent on it and too little on other philosophical and non-philosophical problems. Because acting meaningfully matters, we cannot just dismiss it. Second, some may argue that the fact that the non-identity problem threatens the meaningfulness of duties to the future is rather good news. It should make us feel relieved and light-hearted, whatever we do. It should be seen as a cure rather than a pathogen. Anyone with a lasting sense of strong duties to the future would resist such a conclusion, though.

The non-identity problem

What does the non-identity problem consist in? By way of illustration, consider the idea of optimal age of reproduction.2 It stresses the timing dimension of the non-identity problem. In many societies, social pressure or even legal restrictions - such as sexual majority ages or lower/upper age limits for medically assisted reproduction - push people not to reproduce outside an age window. This may rest on several justifications. We may want to guarantee proper sexual consent or limit health-care costs. But we may also be imposing such age restrictions out of concern for the interest of the to-be-born child. Aren't children better off with parents equipped with sufficient judgment abilities and with sufficient additional longevity to be able to accompany them into adulthood? Such justifications actually face the non-identity problem. Had the person decided to conceive a child at another age, it would have been another child, not this one. Hence, for this very child, this is the best possible life. In what sense could this person then meaningfully claim to have been harmed, or even wronged? This is the challenge.

The non-identity problem has to do with a trivial fact, one of puzzling and potentially radical consequences. Let me begin by dissipating a few possible confusions. The expression "non-identity" does not refer to the mere fact that future people are likely to qualitatively differ from us, whether in being more short-sighted, fatter, or less red haired. Neither does it single out the likely fact that the preferences of future people might differ from ours, whether in being more passionate about rap music, more sensitive to gender bias, or less keen on natural environments. Nor does the expression merely point to the fact that the goods - natural and cultural resources - at the disposal of future generations are likely to differ from ours. The focus is not merely on whether future people's lungs will be more resistant to pollution, whether they will form weaker preferences for clean air, whether their air will be more polluted, or whether they will benefit from social arrangements well adjusted to episodes of air pollution.

Instead, when we talk about "non-identity," we are in fact referring to the interaction between someone's internal (e.g. one's legs) and external (e.g. one's bike) resources and a person's numerical identity. What does this mean? When we consider "identical twins," they are identical in a qualitative, "type" sense, at least genetically. And yet, since they are two separate persons, they are non-identical in another, numerical, "token" sense. Consider now the interaction between a person's numerical identity and her qualitative identity, the latter referring here to the state of her internal resources only (such as the state of her lungs). In an ordinary setting, one can modify a person's physical features (e.g. turn her black hair red) or the content of her preferences (e.g. convince her to prefer red hair) while claiming that she still remains the same person, in the numerical sense. In contrast, in a non-identity setting, you cannot merely modify certain features of a person's condition, because the only way of doing so requires bringing another person into existence instead. In such a context, the only way in which we can ensure a person having red hair rather than black hair is to bring another person (with black hair) into existence instead. The type change requires a token change. Improvement can only be achieved through selection. In a non-identity context, qualitative change cannot be achieved without numerical change. In such a specific context, changing a person necessarily implies changing person, i.e. bringing about a different person.

What is true of the person's internal resources is also true of her external circumstances. In a normal setting, if Pablo lives in a polluted environment, it is possible to improve Pablo's environment without the need for Pablo to be substituted with Marta. In a non-identity setting, adopting a policy to clean up the environment will also entail that it is Marta rather than Pablo who will be born. Changing a person's environment necessarily implies changing person.

At this stage, the reader may remain skeptical. Do any real-world situations actually fit such descriptions? What is the underlying story? As a matter of fact, most of our policies have an effect on the timing of people's actions. For instance, energy, employment, education, and transportation policies influence when people get up, the time they spend fetching food, when they are back home, and . when they make love. The last fact explains the unintended and indirect impact of non-reproductive policies on who will be born. A lot of non-reproduction-related policies affect the timing of our reproductive acts. As a result, they also affect the identity of who will come to existence. This is the compelling starting point of the non-identity problem. We still need to understand how that can lead to a problem. For some authors, it threatens no less than the very possibility of obligations of justice to the future.3 We will show that it is not necessarily the case.

In order to grasp things further, consider three other types of directly reproductive choices - besides our "optimal age" example. First case: human cloning. It involves a direct connection between a policy choice and the identity of who will be born. It also helps us to clearly separate qualitative from numerical identity issues. People can be concerned about cloning humans for various reasons.4 One of them rests with the concern that cloning involves something wrong toward the new person that results from the cloning exercise - hereinafter "the clone." Yet, in the absence of cloning, the clone would not have existed. How can we claim that this life is worse than an alternative life for this person if this is the only possible one for her? How could an action possibly be regarded as harmful to someone if that action is necessary for that person to come to existence? Cloning confronts us with two competing claims, each relying on a different concept of identity. Those worried about cloning may be concerned with the fact that two individuals are being created...