Future selves, paternalism and our rational powers

ABSTRACT

This paper challenges the two aims of Michael Cholbi’s Rational Will View (RWV) which are to (1) offer an account of why paternalism is presumptively or pro tanto wrong and (2) relate the relative wrongness of paternalistic interventions to the rational powers that such interventions target (Sections 1 and 2). Some of a paternalizee’s choices harm their future selves in ways that would be wrong if they were done to others. I claim this challenges Cholbi’s second aim (2) because the cases his account deems particularly wrong turn out to be not to be as wrongful as expected (Section 3). When this second aim is challenged, it has knock-on effects on the capacity of the RWV to discern which cases of paternalism are generally more wrongful than others, which undermines Cholbi’s first aim (1). I consider responses on behalf of Cholbi’s view but conclude that the account is insufficient on its own to vindicate its two aims (Section 4). Finally, I draw on recent work that adopts ideas from the practical reasoning literature to help determine paternalism’s wrongness (Section 5). I argue this helps Cholbi’s view withstand my objections, but we must remain skeptical of why interceding with rational powers is particularly wrong.

KEYWORDS:

• Paternalism
• reasons
• personal identity
• Parfit
• Cholbi

1. Introduction

Many contemporary philosophers working on paternalism – roughly, when X interferes with Y in order to benefit Y – tend to defend the view that paternalism is presumptively or pro tanto wrong (Birks 2018; Cholbi 2017; Coons and Weber 2013; Enoch 2016; Groll 2012; van Oosterum 2024; Parry, MS). This is often understood as the claim that paternalism is generally, but not always, all-things-considered wrong. One of the appealing features of holding such a view is it manages to capture a powerful intuition: we should generally allow individuals to make certain choices so long as they do not cause harm to others (Mill, [1861]2010). In other words, it is generally wrong to interfere with one’s choices if one harms only oneself. The view that Michael Cholbi (2017) develops in ‘Paternalism and our Rational Powers’ is one intended to fit within the contours of this contemporary philosophical view of paternalism’s wrongness. In Section 2, I introduce his view before raising a problem for his two aims which are to (1) offer an account of why paternalism is presumptively or pro tanto wrong and (2) relate the relative wrongness of paternalistic interventions to the rational powers that such interventions target. I draw on Parfit’s (1986, Section 107) work on personal identity to leverage certain insights that undermine the second aim which has negative knock-on effects that undermine the first aim of his paper (Section 3). The insights offered by Parfit here have very recently been picked up by Dominic Wilkinson (2023). Wilkinson’s claim is that we can have reason to protect a medical patient’s future self from the bad decisions of their present self, which raises interesting questions for paternalism in the context of public health ethics and medical ethics. My aim here is more general which is to bring Parfit’s insights in contact with recent debates about what makes paternalism wrong and precisifies it with concepts from the practical reasoning literature. Finally, I consider some potential responses on behalf of Rational Will View adherents (Section 4) but argue that the view requires amending via a kind of sophisticated casuistry (Section 5).

2. The rational will view

Cholbi’s Rational Will View draws on but develops in detail similar views that other philosophers have defended (Quong 2011; Shiffrin 2000). Cholbi’s project and that of other philosophers with adjacent views is to identify the distinctive explanation for why paternalism is wrong. Some philosophers might be wondering why such an explanation is required. The skepticism behind the value of such a project might go like this: there seems to be nothing distinctively wrong with paternalism. After all, how can there be something wrong with benefitting someone?

Cholbi’s (2017, 133) ‘rational will view’ provides an answer:

According to the rational will view, the wrong-making feature of paternalism is that paternalism amounts to denying equal moral status to another because it intrudes into her rational agency [own emphasis] in ways that indicate judgments of inferiority.

The distinctive feature of Cholbi’s account is its claim that we can identify the wrongness of paternalistic intervention by investigating how it affects the ‘rational powers’ that constitute our rational agency. The rational powers are the capacities necessary for individuals to ‘choose, intend and act on reasons’ and they are important for us to be practically rational agents, that is, to be able to devise and follow through our own plans of life. Cholbi argues that there are three rational powers. They are what he calls ‘recognition, discrimination and satisfaction’. Recognition is a capacity agents have to perceive ends as choiceworthy and which ‘give us reasons to adopt attitudes toward certain objects’ (2017, 134). Discrimination is a capacity to rationally select between certain choices or ends one recognizes as minimally choiceworthy. Satisfaction is the capacity to pursue those choices and the means towards our ends. The important point is that one’s paternalistic interference towards a paternalizee can target an agent’s recognition, discrimination, or satisfaction separately.

Take Cholbi’s case of the recovered alcoholic, C, and her adult child, D. C wants to figure out how to prevent D from suffering with alcohol addiction. C can remove alcohol from the list of ends that D deems choiceworthy, which interferes with D’s rational power of recognition, perhaps by an elaborate campaign of censorship. Alternatively, C can make drinking alcohol seem less choiceworthy which interferes with D’s rational power to discriminate between certain choices or ends. Finally, C can interfere with D’s rational power of satisfaction by making D unable to act on choices related to drinking alcohol. (Sometimes a paternalistic interference might have consequences for all three of an agent’s rational powers, even if it only targets one particular power. This is a point Cholbi (2017, n. 10) recognizes, but I will grant the possible separation of these powers for what follows).

The final part of Cholbi’s account is perhaps the most interesting. He notes that these rational powers are largely constitutive of our practical identity. He suspects that what makes paternalism seem so objectionable is the way in which paternalizers ‘colonize our identities as rational wills’ (Cholbi 2017, 136). He argues that his account of the three rational powers is able to help determine the wrongfulness of paternalistic acts. His claim is that the reasons to justify paternalistic interventions affecting the power of recognition are stronger than those that justify interfering with discrimination or those that interfere with satisfaction (Cholbi 2017, 139). This is thus a claim that hierarchically orders the wrongs of paternalism based on which power they affect, with the interventions that affect the power of recognition being the hardest to justify or the most wrong. To figure out how wrong paternalism is, we must look to which of the rational powers a paternalizer intercedes in. This is how we can start to answer the question of what makes paternalism generally, but not always all-things-considered wrong.

Therefore, a crucial component of Cholbi’s rational will view is the claim that the most wrongful paternalistic interferences are those that interfere with our power of recognition. While Cholbi provides a lengthier defense of this, we can provide a succinct one here: such interferences are the ones that most deeply affect our practical identities and, if that’s true, then how we exercise the power should ultimately be left up to us. Interferences with the other rational powers are less wrong since they become significant only after the inputs given by the power of recognition. I will argue against Cholbi’s view by showing how there are cases of interference with the power of recognition that do not seem wrong. As such, we require more explanation for how to avoid the worrisome implication for his argument that we cannot rely on which of the powers we interfere with as a guide to how wrong paternalism is. To bolster the argument, I also consider cases where it is unclear interferences with recognition are really more wrongful than interferences with discrimination or satisfaction. As such, the distinctive appeal of the rational will view is diminished.

3. A (future selves) problem for the rational will view

In this section, I highlight a problem for Cholbi’s rational will view. Let’s revisit the claim that there are certain choices that we can make that deeply affect our practical identities. It is not implausible to argue that the choices we make can affect our future selves in particular ways. Without embroiling ourselves too much in the complexity of personal identity questions, we can say that D (the adult child of Cholbi’s recovered alcoholic) will be psychologically continuous with themselves, perhaps to varying degrees, 50 years into the future. Such psychological continuity with oneself over time can be the important relation we have with our ‘successive selves’ (Parfit 1986). Of course, things can happen to D that might make us doubt how psychologically continuous they are with themselves in the future. Perhaps D at T1 might suffer devastating injuries from a car accident and remain in a vegetative state for the rest of their lives at T2. D in a vegetative state might be very psychologically discontinuous with D at T1, but perhaps there is some minor degree of continuity that enables us to say they are identical with one another.

At this point, it is worth carefully unpacking the senses of ‘identity’ that are at play in Cholbi’s work and in that of Parfit.¹ Cholbi’s use of the term ‘practical identity’ outlines a description of the commitments, values and concerns unique to a person which is important to who they take themselves to be (Cholbi 2022, Chapter 1; Korsgaard 1989). In this practical conception of identity, the question of a person’s persistence over time has less to do with metaphysics and more to do with what practical reasons they have to act in line with those commitments, values, and concerns over the course of their lives. By contrast, it is thought that Parfit’s work on identity is purely about answering the metaphysical question of whether the person that existed at T1 is the same as the person that exists right now at T2.² This view has practical implications too, of course. Famously, Parfit (1986, Section 95) described feeling less angst about his own death on the basis of his own psychological continuity account of identity. But, according to Korsgaard (1989, 121), Parfit’s view of identity leaves out the crucial or essential aspect of identity which is that our identity is practical or ‘something that we do [emphasis original]’.³

Consider a non-paternalistic example to illustrate the difference provided by Cholbi’s philosophical work on grief. Someone’s practical identity may significantly change when we experience bereavement because that person whom they had a relationship with shaped their self-conception and what they valued. But the person is likely still psychologically continuous with themselves immediately after the bereavement so, metaphysically, there is no change in their identity (in other words, they are numerically identical with themselves). In what way are these two senses of identity related to one another? One of Parfit’s (1986, Chapter 12) core claims is that identity is not what matters prudentially, that is, the basis of a person’s concern about what happens to them over time is not whether they are numerically identical but, roughly, the degrees of psychological connectedness and continuity they exhibit with themselves over time.⁴ These can also be referred to as the ‘prudential unity relations’, the relations that ground a current self’s greater or lesser concern about what happens to their future self (McMahan, 2002; Wilkinson, 2023).⁵ Here then is where the two senses relate: the choices we make (e.g. to become an alcoholic or grieve) can have profound effects not just on our practical identity but on the prudential unity relations that connect our current self with our future self. The weaker those relations become, the less clearly it seems one’s future self will be related to their previous self who made the choice. As a result, it may seem more plausible to interpret the choices the previous self makes as similar enough to choices that would affect other people.⁶

For Cholbi, the power of recognition is a crucial part of how we constitute our practical identities over time and, thus, how we use it or fail to use it, can have far-reaching implications for our future selves. This may turn out to be very relevant for the wrongness of paternalism. Why? If we concur with Parfit’s views as glossed above, it could be wrong to let people do things to their future selves that it would be wrong for them to do to other people (Parfit, 1986, Section 107; Wilkinson, 2023, 397).

Consider if Cholbi’s recovered alcoholic did not interfere with D. Given certain assumptions by Cholbi about D’s genetic predisposition towards alcoholism, it is likely that D, 50 years from now, might suffer the worst consequences of being an alcoholic, such as liver failure, depression, cognition problems etc. Let’s call D before the alcoholism sets in, ‘D-present’ and D who might suffer from alcoholism, ‘D-future’. Why might D’s life go like this? Without interference by C, D perceives as choiceworthy drinking alcohol, sometimes to excess. This will have downstream consequences for their ability to discriminate between alcohol and nonalcohol-related choices and for them to satisfy their nonalcohol-related ends. Let’s assume the key catalyst to D’s slide into alcoholism comes from exercising their power of recognition completely unencumbered by others. D-future has had their practical identity profoundly shaped by D-present’s exercise of their power of recognition. Despite some degrees of psychological continuity between the two, D-future looks like a very different person to D-present and has had their life go considerably worse. What ought we to think about this case?

I believe it reveals that it would not be wrong for C to have interceded in D’s power of recognition. To my mind, it resembles an alternative case where interfering with D’s power of recognition would prevent harm to some distinct individual E. We ought to think C’s interference with D to protect E is not wrong since it fits the standard Millian line that interference with an individual to prevent harm to others is, ceteris paribus, justified. However, building on Parfit’s work, I am suggesting that the moral distinction between preventing harm to E and preventing harm to D-future is not so clear.⁷ My claim is that these points pose a particular problem for the commitments of Cholbi’s rational will view. The power of recognition was tied to the most wrongful kinds of paternalistic interventions. Yet here we have a case where interceding in the power of recognition does not appear to be wrong. If that’s right, then interceding in the power of recognition will not necessarily be as determinate a wrong-making feature of paternalism as the rational will view maintains. This then will have knock-on effects on the capacity of such a view to discern which cases of paternalism are generally more wrongful than others, which is critical to its ability to adhere to the view that paternalism is presumptively wrong.

To be clear, Cholbi might accept that cases like these affect our judgment of when interceding in the power of recognition is wrong, but I think that undermines the distinctive appeal of connecting this power to the wrongness of paternalistic intervention. The exercises of our power of recognition that are going to be very important are likely to be the ones that deeply impact our future selves. There could be a non-negligible number of cases where those exercises impact our future selves in ways that leave us pretty badly off, where an intercession in this power will not be as wrong-making as Cholbi thinks.

However, one might argue that for Cholbi’s view to be undermined it is not sufficient to simply identify a counterexample to his claim that the most wrong paternalistic interferences are those that affect the power of recognition. Rather, we would need to upset the hierarchical order he sets up by identifying cases where interferences with the power of recognition are just as wrong as interferences with the power of discrimination or satisfaction. How might we go about doing this? It might be that the end the agent is attempting to pursue is so important to them – in terms of constituting their practical identity – that interferences with how they recognize it as an end or are capable of satisfying that end could be equally problematic. Consider the case of a Jehovah’s Witness who is given a (nonlife-saving) blood transfusion against their will.⁸ Imagine a variant of the case where we could either hypnotize them by blocking any thoughts they might have about blood transfusions (corresponding with an interference with recognition) or simply hypnotize them by paralyzing them and making them unable to escape the procedure (corresponding to an interference with satisfaction). Assuming that either option leads to a blood transfusion, is it clearly the case that interfering with the power of recognition is more wrong than interfering with the power of satisfaction?

To my mind, the interference with satisfaction in this case seems more repellent than the interference with recognition but it might be just as intuitive to say both seem problematic.⁹ But if that’s the case, then we continue to have no real distinctive reason for worrying about paternalistic interferences with the power of recognition. Furthermore, both kinds of interferences could equally shape one’s future self in practical identity-affecting ways. To me, this suggests once more that there may be other factors affecting our judgment about what makes paternalism differentially wrong across cases. In Section 5, I suggest what might be relevant factors for the question of what makes paternalism wrong which may also help Cholbi avoid the problems I have raised regarding future selves. Nevertheless, I provisionally conclude that these problems upset the tidiness of Cholbi’s Rational Will View and its attempts to simply discern which cases of paternalism are more wrong than others.¹⁰

4. Objections

Let’s consider several objections that Cholbi or an adherent of the rational will view might attempt. Consider this response first. One might deny that there is anything wrong with letting D-present make choices that make D-future significantly worse off. They might argue that D is the owner of their successive selves, of D-present, D-later … D-future, etc. This ownership of one’s successive selves is an important part of possessing a practical identity. When someone paternalistically intercedes in D’s power of recognition, they still wrong them significantly because they implicitly are adding ends that D does not recognize as choiceworthy, namely, the end that ‘one ought not make one’s successive selves worse off’. C wrongs D qua their rational power of recognition because it should be up to D to shape not just their current practical identity but their future practical identities too. Therefore, the paternalizer who appeals to this Parfitian argument would seek to justify ‘colonizing’, as Cholbi puts it, all of a paternalizee’s practical identities, present and future. So, this problem of future selves leaves the rational will view unscathed and, better yet, sharpens the importance of a practical identity being up to the paternalizee.

While I can see the appeal of such a response, I do not think it is sound. We can see why when we consider which D is and ought to be the owner of their successive selves. The above response leaves it entirely ambiguous which D we are talking about. When D is making a choice, this is a choice that occurs at a certain time, and so this D is actually D-atsomepointintime. There might be a D over and above all of the D’s in time, but this over-and-above D never really decides; rather, certain time-indexed selves of theirs make decisions over the course of D’s life. Assuming then that every choice D makes is actually a choice by D-atsomepointintime, why ought D-atsomepointintime have their chance at constituting a practical identity which can have serious ramifications for D-future’s ability to fashion their practical identity? If D-atsomepointintime takes up alcoholism, then they affect D-future’s power of recognition in problematic ways (assuming an alcoholic way-of-life affects what ends D-future can recognize). It would be ‘chronologically unfair’ as Rawls (1999, 254) put it (in a different context) to privilege D-atsomepointintime’s chance at constituting their practical identity. Despite D’s ownership of their selves, a closer look reveals it does not appear to be wrong to interfere with D’s power of recognition if only to prevent harm to D-future. Another way of putting it is that, although it is wrong to interfere with someone’s rational powers for their own benefit (even for the sake of preserving their own rational powers), it might not be wrong (to prevent someone from compromising the autonomy of another individual (future D)).¹¹

Another kind of response might argue that, if Parfit is right, we cannot really refer to C’s intervention with D-present to protect D-future as ‘paternalistic’ at all. If D-future is really so different as to be comparable to someone like E, then C’s intervention is not a case of paternalism. It is just ‘straightforward’ prevention of harm to others. Typically, the concept of paternalism is thought to range over choices that cause only harm to themselves (Dworkin, 2013; Mill, 2010). So, for C to count as paternalistically interfering with D, it must be the case that the person interfered with is the same person that C intends to benefit. Alternatively, the reason C interferes with D is to benefit D. But, by stipulation, D-future could be barely the same person as D-present, so when C interferes with D-present in order to protect D-future that is not strictly a case of paternalism. As such, the future selves of D do not actually throw up concerns for Chobi’s rational will view of paternalism. This is because they are actually cases where we must think about the limits of preventing harm to others, where the moral rationale for doing so is non-paternalistic in nature (e.g. preventing unconsented-to harm).

This response assumes a particular conception of paternalism which is problematic. C’s reason for interference with D could be that it would be good for D but that might not exhaust her reasons for action. Even if it were only one of C’s reasons, it would not follow that C’s other reasons for action render her action no longer paternalistic in nature (Bullock, 2015). Indeed, some philosophers go so far as to say that it need not even be the main reason for C’s action to count as paternalistic (Bullock, 2015; Grill, 2007). Consider if C’s interference with D had the consequence not just of preventing harm to D but also preventing harm to E. Indeed, with a lot of cases of in public health, for example, there might be a mix of paternalistic and non-paternalistic rationales but it might still be plausible to call an instance of such policy-making ‘paternalistic’. A paternalizer can have many reasons for action, some that are paternalistic and others that are not. So, we can say about the above response that C’s actions still count as ‘paternalistic’ even if one of the reasons is paternalistic (to prevent D-present from harming themselves) and the other is non-paternalistic (to prevent D-future from being harmed by D-present). In fact, if such non-paternalistic reasons are also quite strong then that can further tip the scales toward permissibly interfering with D’s rational power of recognition. Nevertheless, such an action directed at D-present can be motivated by different rationales and still be considered ‘paternalistic’ in nature.¹² Again, it seems that intercessions in the rational power of recognition will not always be indicative of the wrongfulness of a paternalistic intervention.

A third response accepts the problems posed by future selves but argues that these are not uniquely problematic for the rational will view’s account of paternalism’s wrongness. It is worth noting that this is a quite conciliatory response and so should arguably be pursued as a last resort. I also do not think that it is even true. There are many accounts of what makes paternalism wrong and examining them all would take focus away from the rational will view. Some accounts hold that paternalism is wrong because it violates personal autonomy or because autonomy or one’s will morally excludes acting for the reason that a paternalistic intervention would be good for the paternalizee (Arneson, 1980; Birks, 2018; Enoch, 2016; Feinberg, 1986; Groll, 2012). These accounts have their issues, but they do not make claims about an individual’s practical identity in the way that Cholbi’s does. Instead, they appeal to notions such as exclusionary reasons or certain kinds of nonwelfarist values compromised by paternalistic intervention. The particular wrong we do a paternalizee according to these accounts does not necessarily refer to their practical identities and so points about these individual’s future selves, while interesting, do not target them specifically. By contrast, Cholbi’s view must deal with the fact that there are multiple practical identities an agent might have over time. As such, I am tempted to conclude that these problems particularly implicate Cholbi’s account.

5. Casuistry and future selves

It seems then that some plausible assumptions about a paternalizee’s future practical identity throw a spanner in the works of Cholbi’s Rational Will View. Cholbi’s two theoretical ambitions were to (1) offer an account to explain why paternalism is presumptively wrong and (2) relate the relative wrongness of paternalistic interventions to the rational powers that such interventions target. The fact that C might justifiably interfere with D’s rational power of recognition to prevent harm to D-future provides a direct counterexample to (2). If (2) is undermined, this puts pressure on Cholbi’s ability to provide an account that explains which instances of paternalism are more wrong and thus makes it unclear whether it can pick out paternalism’s presumptive wrongness. Where to go from here?

In the paper, Cholbi (2017, 125, 151) discusses the possibility that his account could become part of a casuistry of paternalism, that is, it could help us determine which cases of paternalistic intervention are permissible and which are not. His ultimate aspiration is not in providing such a casuistry although being able to provide guidelines for how one ought to reason about various cases is theoretically desirable. It is desirable especially if one wants to provide an account that claims paternalism is presumptively but not always all-things-considered wrong. Cholbi’s paper can be understood as trying not just to account for paternalism’s presumptive wrongness, but also for trying to account for some but not all of the things we ought to consider to determine its wrongness. The appeal to the different ways paternalism can intercede in our rational powers was an attempt to conduct this latter task of what other things we ought to consider. Ultimately, I do not think the appeal to such powers is sufficient since the problem of future selves and practical identities should make us rethink how we reason about these cases.

However, part of the way forward might be to develop a more sophisticated casuistry that incorporates concerns about future selves to aid our judgment about paternalism’s wrongness. By ‘sophisticated casuistry’, I simply mean a method that brings the tools of practical reasoning to bear on cases of paternalistic intervention. Consider the idea of appealing to ‘modifiers’ in the context of paternalism (Dancy, 2004; van Oosterum, 2024).¹³ Normally, when trying to figure out what we ought to do about a particular case we might appeal to reasons either for or against a particular course of action. For example, let’s say I am deciding whether to go to the café. The fact that I like coffee provides a reason for me to go to the café. By contrast, the fact that I am trying to save money provides a reason against buying a coffee (because I could make it at home). However, the fact the coffee is 50% off today might strengthen my reason to go to the café. This fact alone is not a reason to go to that café. After all, if I didn’t like coffee, the fact that the coffee would be cheaper would have no effect on my reasons to go to the café. This suggests that there are some facts, which though not themselves reasons for action, can nevertheless strengthen or diminish said reasons for action.

How does this relate to Cholbi’s proposal and work on paternalism? ‘Simple casuistry’, the kind that both Cholbi and I think is not worth pursuing, simply weighs up reasons for paternalistically interfering (providing certain quanta of benefit to the paternalizee) and reasons against doing so (interceding in our rational will) and then algorithmically spits out a verdict about whether a case of paternalism was wrong. Cholbi’s appeal to the distinct rational powers we can intercede in was intended as a modification of this simple casuistry. Moral reasoning about paternalism ought to be sensitive to more considerations than a simple balancing of reasons.¹⁴ I agree and I offer the suggestion that we appeal to considerations like modifiers to bolster Cholbi’s account and accommodate the challenge from future selves.

In Cholbi’s original presentation, interceding in certain rational powers might be more wrong than interceding in others. Another way of interpreting that claim is that we have a presumptive reason against interfering with someone because of how we intercede in their rational will. The fact that we interfere with their power of recognition modifies or strengthens a reason against interfering. So, there is nothing about interceding in the power of recognition that makes it wrong per se. Rather, we have a reason against paternalistically interfering that is grounded by the wrongness of interceding in someone’s rational will. This reason in turn is strengthened by virtue of the fact that we intercede in an agent’s power of recognition. Modifiers thus fit quite naturally into Cholbi’s account. We can also put it to the task of dealing with the future selves challenge.

Recall Parfit’s claim that psychological continuity comes in degrees, our future selves can be more or less continuous with our present and past selves. This feature of psychological continuity can be useful here. The more a paternalizee’s present choice harms themselves, in such a way that the future self becomes different from the present paternalizee, the weaker our reason becomes against paternalistically interfering. In other words, we can maintain that interceding in the power of recognition is wrong (because it strengthens a pre-existing reason not to intercede in a paternalizee’s rational will). All we need to add is that the impact of a paternalizee’s present choices on their future self can modify the strength of this reason. In the case of C’s potential alcoholic adult child, D, our reason against interfering gets weaker by virtue of the harm that might befall D-future if we let D-present become an alcoholic. Our casuistry for paternalism becomes more sophisticated when we consider not just the reasons for and against interfering, but the facts that strengthen or weaken such reasons. The challenge can perhaps be absorbed by accounts like Cholbi’s so long as we recognize the modifying role that facts about our future selves can play in determining the reasons justifying paternalism’s presumptive wrongness.¹⁵

What about trickier cases such as the Jehovah’s Witness where it is less clear which power is more wrong to interfere with? I would argue that the significance of the end to the paternalizee might also work as an additional modifier, further strengthening our reasons against paternalistically interfering. In that case, it need not be a particular rational power that drives our judgment about the wrongness of paternalism but the significance of the end to the agent that makes any kind of intercession in their rational powers very hard to justify.¹⁶

In short, if we abandon (2) and instead tie the wrongness of paternalistic intervention to this sophisticated casuistry, we might be able to move forward in providing an account of paternalism’s presumptive wrongness. There is much to be gained for debates on paternalism by invoking the tools of practical reasoning and the larger literature on weighing reasons. I hope that this suggestion is a sophisticated way out of challenges for Cholbi’s view that nevertheless originally provides an interesting account of paternalism’s wrongness.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by Institute for Ethics in AI.

Notes

1 I thank an anonymous reviewer for pointing out this important difference in the way these philosophers speak about identity and for suggesting a way of discussing these in the context of Cholbi’s work on grief.

2 Another way of expressing the distinction between these two senses of identity is provided by Parfit (2012, 5) himself. He refers to Cholbi’s Korsgaard-inspired view as a person’s character whereas he is interested in the question of numerical identity. Thanks to an anonymous reviewer for bringing this paper to my attention and pointing out how it mirrors the distinction I drew above.

3 There is skepticism about whether Parfit’s view really leaves out discussion of practical identity. For instance, Jonathan Anomaly (2008, 342) argues that, for Parfit, the idea of practical identity is not ignored; though his work focuses on metaphysical descriptions of identity, part of its target would be the idea that there is a practical identity over and above ourselves that is truly continuous over time.

4 Parfit’s account of why identity does not ‘matter’ generates answers to metaphysical questions about numerical identity as well as questions about what justifies prudential or egoistic concern (McMahan, 2002, 42).

5 Strictly speaking, the prudential unity relations for McMahan (2002, 7) include psychological connectedness, psychological continuity, and capacity for consciousness.

6 I thank an anonymous reviewer for helping me distinguish between the two senses of ‘identity’ being used and pressing me to identify the relation between them that is crucial for my argument to go through.

7 Other philosophers have cast doubt on the moral asymmetry that justifies prevention of harm to others but not harm to oneself (Birks, 2018). However, their doubts are not motivated by temporal considerations and future selves.

8 The clarification that the blood transfusion is nonlife-saving is useful to keep the discussion focused on life-changing scenarios opposed to the risk of death typical in the philosophical literature on Jehovah’s Witness cases (for example, see Hanna (2018, Chapter 4) for discussion).

9 One might think that what is driving the intuition that this interference with power of satisfaction is more repellent is that we can imagine the horrifying nature of being operated on against our will. One could imagine an alternative case where the procedure is arguably less horrifying, perhaps something like cutting one’s nail off because there’s a risk of the nail infecting the whole person and seriously compromising their health. Yet the reservation about cutting the nail off might be that their cultural or religious beliefs seriously proscribe bodily modification. Ultimately, my example aims just to show that it is unclear that interferences with the power of recognition are necessarily worse and I believe one could modify the details of the example to pump that intuition more precisely. I thank a reviewer for pressing me on the example.

10 I am once more indebted to an anonymous reviewer for suggesting this point to sharpen my argument against Cholbi’s view.

11 Note that this argument is merely a reply to the objection regarding the relevance of one’s ownership of their successive selves. My argument does not rely on the thought that paternalistic interferences are permissible only when they protect that future person’s rational powers. If my argument depended on this, one might worry that the scope was limited to extreme cases like addiction where a future self’s rational powers might be severely affected where more common cases might not affect rational powers. Instead, I am making a more general claim about the permissibility of paternalism to affect a future person’s well-being writ large.

12 For philosophical sticklers, we could say that this kind of action is “weakly paternalistic” (Bullock, 2015, 11).

13 For a grounded discussion and defense of reasons, modifiers and particularist practical reasoning see Schroeder (2011).

14 See Shafer-Landau (2005) and Malm (2005) for some recent discussion of the merits and defects of the balancing strategy.

15 In this respect, I agree with Wilkinson (2023). In it, he argues that ‘future self’ problems ought not decisively count in favor of intervention. Instead, he claims that it may provide a “stronger ethical reason” for intervention. My language invoking modifiers is more precise and targets the reasons against interfering as opposed to the reasons in favor of interfering. Those latter reasons are likely to be modified by considerations about the degree to which a paternalizee’s well-being is diminished by their choices (see Groarke, 2002; Birks, 2018; van Oosterum, 2024) for suggestions.

16 For more on the idea that the significance of an agent’s end can affect the strength of our reasons to avoid paternalistic interference, see Groll (2019: 67) and van Oosterum (2024: 565). Cholbi (2017: n 11) also seems friendly to this idea when he discusses the relative importance of the end as a factor complicating his framework.