The Weight of The Past

This chapter asks whether it is ever required or even permissible to give some weight to the values of your past selves. It argues that sometimes, when a past self has made a sacrifice that benefits the current self, it is at least permissible and sometimes required to give weight to the permissible values of that past self. To do this, the chapter introduces a principle descended from H. L. A. Hart’s (1955, ‘Are there any natural rights?’ Philosophical Review, 64) Principle of Mutual Restriction and John Rawls’s (1975, A Theory of Justice) Principle of Fairness.

The Problem of Weighting Changing Selves

This chapter summarizes the conclusions of the first part of the book and introduces the work of the second part. The first part proposed the Aggregate Utility Solution to the problem of choosing for changing selves; it argued for a particular version of this solution that aggregates particular utilities using weighted averages; and it defended that version against the objection that the utilities of different selves are incomparable and the objection that we do not know enough about our future utilities to make our decisions in these ways. This second part asks how to set the weights to which this version of the Aggregate Utility Solution appeals.

Can We Compare the Utilities of Different Selves?

This chapter asks whether we can compare the values of different selves of the same individual in such a way that we can measure those values on the same scale using utilities. It surveys a number of different proposals adapted from the literature on interpersonal utility comparisons, and concludes that none of the existing proposals will work. It presents an alternative and argues that it does work. This answers an objection raised by R. A. Briggs (2015),. ‘Transformative Experience and Interpersonal Utility Comparisons’ in Res Philosophica, 92(2).

The Road Ahead

This final chapter of the book summarizes the conclusions of the preceding chapters and looks forward to how the Aggregate Utility Solution might be generalized so that it applies not only to the expected utility theory framework, but also to other frameworks for rational decision-making. In particular, it explains how we might extend the Aggregate Utility Solution to the framework of imprecise credences and utilities, and to the framework of risk-sensitive decision theories.

I’ll be Glad I Did It—So, I’ll Do It

This chapter considers a form of practical reasoning explored by Elizabeth Harman (2009, ‘I’ll be glad I did it’ reasoning and the significance of future desires, in Philosophical Perspectives, 23(1)), which she dubs ‘I’ll be glad I did it’ reasoning, and which she rejects in general. In such reasoning, when faced with a choice between two options, if I would later think that one option is better than the other were I to choose that option, that gives me reason to choose that option. This chapter distinguishes between a number of varieties of such reasoning and applies the Aggregate Utility Solution to diagnose the problem with it. The resulting diagnosis turns out to be quite different from Harman’s.

Do We Know Enough to Make Decisions This Way?

This chapter considers two related arguments that, if successful, tell against the Aggregate Utility Solution to the problem of choosing for changing selves. The first is due to L. A. Paul (2014), ‘Voluntary Benefits from Wrongdoing’, Journal of Applied Philosophy, 31(4); the second is due to Sarah Moss (2018), Probabilistic Knowledge. Both claim that we cannot know enough to make decisions in the way proposed by the Aggregate Utility Solution. Paul appeals to considerations of authenticity; Moss appeals to knowledge norms for decision-making. The chapter responds to those arguments.

Why Aggregate Using Weighted Averages?

This chapter defends the claim that we should aggregate an individual’s different utility functions at different times by taking a weighted average of them. It introduces the Principle of Minimal Mutilation, which says that an aggregate of a collection of attitudes should lie as close as possible to the attitudes it aggregates. By giving a mathematically precise account of the distance between different utility functions, the chapter appeals to a mathematical result to show that only weighted averages do this. It then defends the practice of taking weighted averages in the case of credences against objections due to Elkin & Wheeler (2016) ‘Resolving Peer Disagreements Through Imprecise Probabilities’, Noûs, May, among others.

The Aggregate Utility Solution II: The Solution Itself

The previous chapter concluded that we should aggregate credences and utilities first. This chapter considers how we should aggregate them. It claims that we should use the individual’s current credence function together with a weighted average of the utility functions of the individual’s past, present, and future selves. It then responds to an objection based on the partition sensitivity of the proposed solution. It proposes that an individual should aggregate the utilities assigned to those states of the world that specify everything that any of the individual’s various selves care about.

The Problem of Choosing for Changing Selves

This chapter introduces the problem that the rest of the book seeks to solve, namely, the problem of choosing for changing selves. The chapter presents a series of examples. In each, an individual’s values change over the course of their life, and they must make a decision at a particular time in their life. These examples motivate our problem, and will be drawn on time and again in what follows. The chapter gives an overview of the structure of the book along with a brief précis of each chapter.

The Nearer The Dearer

This chapter asks whether it is rationally permissible to assign less weight to the values of my other selves in line with the proximity of the values of those other selves to my current values. It argues that it is. It then introduces the Stoicism objection to the Aggregate Utility Solution. This says that you should never be rationally required to change your values in order to increase your utility, and it claims that this is required by the Aggregate Utility Solution in certain situations. The chapter then explains how the conclusions of the earlier part of the chapter give us the tools to answer that objection.