Paternalism in Friendship

Abstract

1. Introduction

Is paternalism more often justified in friendship? Are we sometimes permitted to treat our friends paternalistically, while a similar kind of interference would be impermissible or at least much more problematic toward strangers? In short, does the fact that two people are friends play a distinct role in the justification of paternalism? This question has recently attracted some attention. George Tsai, for example, argues for the claim that “the closer, more intimate the relationship between the relevant parties, the stronger the overall moral justification for paternalistic intervention” (Tsai 2018, 249). And Ellen Fox contends that “certain kinds of [paternalistic interventions] are justified by the bonds of friendship and intimacy which connect particular people” (Fox 1993, 594).

Common to these accounts is that they assign friendship an irreducible role in the justification of paternalism (see also Raz 2001; Sullivan and Niker 2018]). They thereby put pressure on extant accounts of paternalism that have mostly been developed in the context of institutional paternalism or interpersonal paternalism toward strangers. On one prominent account, paternalism is considered to be pro tanto wrong because it violates the autonomy of persons (Dworkin 2020). On a more recent account, paternalism is regarded as pro tanto wrong because (or if) it manifests an insulting or disrespectful attitude of distrust in the other’s judgmental and agential capacities (Cornell 2015; Shiffrin 2000; Sheintul 2023).¹ The wrong-making properties of paternalism that these accounts target are relationship-independent: “they make no mention of anything about the target’s relationship to the one who is paternalistically interfering, or who might be in a position to do so” (Tsai 2018, 349). By contrast, those who argue that paternalism is more often justified or even morally required in friendship appeal to relationship-dependent reasons for paternalism. More specifically, they hold that norms of friendship and/or specific relationship goods characterizing the nature and value of friendship—such as care for one’s friend, vulnerability, intimacy, or trust—play a distinct role in the justification of paternalism.²

The aim of our paper is to develop a more nuanced and systematic view of how friendship affects the justification of paternalism.³ In a first step, we will identify three general challenges that any account of “benign paternalism” in friendship faces. We then introduce the Modifying View according to which relationship-independent reasons determine the (im-)permissibility of paternalism, while relationship-dependent reasons affect and modify the moral quality of paternalism, both positively and negatively. In a last step, we identify cases where the Modifying View seems to issue false verdicts. It cannot do justice to the fact that paternalistic interventions in friendship that qualify as permissible in individual instances can be more problematic when they occur in a pattern over time. Taking into account the dynamic and diachronic nature of friendship, we demonstrate how paternalistic interventions can become morally objectionable, because they express attitudes that are incompatible with the norms of friendship, and ultimately undermine it.

Before we begin our discussion, a few clarifications are in order to explicate the central concepts that we will use. First, we work with a rather broad definition of paternalism, according to which paternalistic interventions are characterized by: (i) an intention to help the person intervened with (in our case, the other friend) avoid making a mistake; (ii) a kind of belief or judgment that the other person is unlikely to avoid this mistake on her own; (iii) an other-regarding motivation to promote the interests or well-being of the other person for her own sake; and (iv) the absence of explicit consent or an (implicit) previous agreement to such interventions.⁴ Such a definition is more adequate in the context of intimate relationships than narrower definitions of paternalism that require direct interferences with a person’s freedom or autonomy (e.g. Dworkin 2020), because the definition captures well cases that often occur in friendship (such as persuasion or non-cooperation) and that are clearly paternalistically motivated.

Second, we conceptualize autonomy in a ‘content-neutral’ or ‘procedural’ way. Accordingly, there are no moral or reason-based (and thus substantive) constraints on what qualifies as autonomous. Instead, it is sufficient for autonomy that some procedural conditions are met that account for an action or decision to be an agent’s own. Respect for autonomy, on this view, requires that a person’s actions, decisions or ways of life should be prima facie respected if they satisfy these conditions, even if we have good reasons to believe that there are moral or welfare-based reasons to intervene. Such a content-neutral or procedural conception matters here for methodological reasons: after all, to even be able to raise the question whether friendship distinctly affects the justification of paternalism, it must be possible that respect for the friend’s autonomy and the duty to care for the well-being of one’s friend may come apart.

Third, and with regard to the concept of friendship, we hold that for two people to qualify as friends, they need to respect each other as agents with autonomous ends, have sufficient care and concern for each other, and trust that they both have a robust interest in their respective well-being for their own sakes (Helm 2021). In our view, these are the constitutive norms of valuable and normatively relevant friendships that always need to be fulfilled, at least to some degree. There is room for interpretation and negotiation within friendship regarding both the balance of these constitutive norms and other norms that characterize more substantive conceptions of friendship (Cocking and Kennett 1998; Keller 2022; Keller 2023). For this reason, our discussion can only provide a general orientation in terms of what to look for regarding the justification of paternalistic interventions in particular friendships.⁵

But note that to sensibly pursue our question, we need to rule out that friendship conceptually entails paternalism. After all, it is question-begging to ask whether friendship plays a distinct role in the justification of paternalism if friendship implies paternalism. Even if one accepts, as we do, that friendships are “marked by deep and particularized care for each other” that “is understood as practical concern for the friend’s good” (Koltonski 2022, 475), it does not follow that paternalism is required or justified if friends seem to do something that goes against what is good for them. Instead, it remains an open question what the alleged interest in the friend’s well-being precisely involves, and how it is to be balanced against other relevant norms, such as respect for the friend’s autonomy.

2. Relationship-Dependent Reasons in Favor of Paternalism—A Critical Review

In this section, we critically engage with extant accounts in favor of the view that relationship-dependent reasons affect the justification of paternalism. This discussion is meant to help us formulate three challenges that need to be met if one is to convincingly defend such a view.

2.1. Shared projects and ‘meshed identities’

According to a first explanation as to why the fact that two people are friends affects the justification of paternalism is to point to important projects that friends share and the success of which would be in danger without paternalistic intervention. As Tsai argues, furthering the success of a project that one shares with one’s friend “may involve acting in a way that involves circumventing the friend’s agency, yet promotes the well-being of both friends” (Tsai 2018, 356). On a related and even more encompassing view, friendship is said to involve the (partial) meshing of identities. After all, friends form some kind of ‘we’ and exhibit shared agency. Since this meshing of identities manifests itself “in part by the liberties that one friend takes in guiding the life of the other” (Fox 1993, 575), paternalism, so this view holds, is more often justified.

An obvious objection to both of these views is that they do not describe cases of paternalism. The main aim of the intervention is not directed at the well-being of the other friend for that friend’s sake. Instead, it is the success of the shared project or the upholding of the meshed identity that the intervention primarily targets. To illustrate, let’s suppose that two friends used to play tennis together. Much of their friendship surrounded this shared project, which included going to a bar after the match and chatting about all kinds of matters of mutual concern. They had even come to see themselves as a ‘we’ as they shared a project that shaped, in large part, their identity. If one of the friends stopped showing up for their tennis matches, the other friend would talk him into coming again, nag him if he still did not attend, or come and pick him up at his house at the usual time, even though the friend had not indicated that he wanted to be picked up. Both the shared-projects view and the view concerned with the meshing of identities thus claim that, provided shared projects and meshed identities are what friendship is about (partly at least), friends need to treat each other paternalistically (sometimes at least) if their projects or identity can no longer be realized.

But on both accounts, the friend and his well-being merely serve as means to the success of a project or the maintaining of their identity, such as playing tennis together. It thus seems more plausible to assume that it is the friend intervening who has an interest in rescuing the shared project or maintaining that meshed identity (given that the other friend has withdrawn from it). As a result, on both accounts, the friend intervenes primarily for self-regarding reasons, and therefore does not act paternalistically at all. Shared projects or meshed identities may, in fact, provide reasons for interventions, but on closer inspection such interventions do not qualify as paternalistic.⁶

Proponents of these views could concede that such interventions do not target the well-being of the other friend for their sake alone but raise doubts about the distinction between ‘I’ and ‘we’ that the objection assumes. More specifically, they could argue that it does not make sense to say that the intervention is not primarily concerned with the well-being of the other friend since the two friends conceive of themselves, partially at least, ‘as one’. And if this is the case, every intervention is, partly at least, motivated by a concern for the other friend.

But this response is rather unconvincing, because it rests on a substantive and controversial conception of friendship. For example, it does not allow that friends can disengage from a shared project, while still conceiving of themselves as friends. It might still be true that particular friendships between two persons take the form of ‘meshing identities’. But the arguments presented in favor of paternalism are not restricted to such cases; they are rather formulated as general claims. As such, they can reasonably be contested.

As a preliminary result, neither shared projects nor meshed identities can vindicate that friendship plays a distinct role in the justification of paternalism: either because the interventions on behalf of the shared projects or the meshed identities do not qualify as paternalistic, or because they rest on a controversial conception of friendship. What is more, there are more plausible ways to explain the (controversial) intuition that shared projects or meshed identities render paternalistic interventions less morally problematic or even required. Friends could be thought to have consented or agreed to paternalistic interventions when they embarked on a shared project or when they were about to forge a ‘we’.⁷ But, again, in such a case, the intervention at issue does not qualify as paternalistic.

2.2. Vulnerability and intimacy

A second approach to justify paternalism on account of constitutive elements of friendship appeals to vulnerability and intimacy. The fact that friendship involves vulnerability is taken to be non-accidentally related to the fact that paternalism is often a requirement of being a friend. We value friendship, in part, because we can “relax the default self-protective strategies that we usually have in life” (Tsai 2018, 356; 2016). That is, part of being a friend is that friends open themselves up and share intimate details about their lives, thereby rendering themselves vulnerable to betrayal and exploitation. But to the extent that friends trust each other qua friends, and present themselves as vulnerable, paternalism is thought to be the “expression of the kind of intimacy we cherish” (Tsai 2018, 357). In a similar vein, Fox maintains that “intimacy without kibitzing is hardly intimacy at all” (Fox 1993, 593).

What this approach seems to get right is that relaxing one’s default self-protective strategies and realizing intimacy in friendship can be relevant for the justification of paternalism. If we are aware of the vulnerabilities and intimate details of our friends, we may often know better what is good for them, what they would autonomously want, and what their weaknesses are. We thus also know when their agency is impaired or when their decisions will have a severely harmful impact on them (Mason 2020). In our view, such considerations can provide reasons for paternalistic interventions. But importantly, these reasons do not ultimately derive from the vulnerability or intimacy characteristic of friendship, but rather from relationship-independent reasons for (soft) paternalism. After all, it is the impaired agency of the respective friend or the severe harm that might accrue that ultimately licenses paternalism. The fact that two people are friends provides epistemic access to these relationship-independent considerations, but it does nothing to show that friendship itself (at least partly) justifies paternalism. To regard vulnerability and intimacy as the sources of relationship-dependent reasons for paternalism seems to be a kind of category mistake, confounding the access to reasons with the source of these reasons. It thus might seem to us that paternalism is more often justified in friendship because of the vulnerability and intimacy characteristic of friendship, but, in fact, both vulnerability and intimacy only put friends in a position to see and better understand relationship-independent reasons for or against paternalism.⁸

Tsai and Fox seemingly want to argue for a different view: They hold that vulnerability and intimacy can only be realized if (or require that) friends paternalistically intervene in each other’s affairs. But there is no conception of vulnerability or of intimacy on offer that entails paternalism (Tsai 2016; Gunkel 2024). To the contrary, it seems that vulnerability and intimacy provide us with reasons to abstain from paternalistic interventions based on the friends’ intimate knowledge about each other’s vulnerabilities. This is because such interventions might lead friends to shy away from each other and to withdraw as they no longer feel safe and respected in their autonomy.

In sum, vulnerability and intimacy can play a role in the justification of paternalism but only because they enable epistemic access to relationship-independent reasons for paternalism. They do not explain why the fact that two people are friends plays a distinct role in the justification of paternalism. More importantly, if vulnerability and intimacy are to be regarded as important friendship goods, they may ground reasons that move friends to refrain from paternalistic interventions. This is because such intervention may undermine a respectful and trusting relationship.

2.3. Duty to care

The last and perhaps most straightforward argument to justify paternalism in friendship appeals to the robust concern for the well-being of friends for their own sakes that is widely regarded as being constitutive of friendship. Accordingly, we simply have a stronger duty to care for our friends (and other intimates) than we do to care for strangers. David Enoch gives voice to this idea:

But she [my wife] is also under special obligations, compared to strangers, to promote my well-being […] This is why, whatever you may think about her hiding the candy (in order to promote my well-being and protect me from my own akrasia), surely you think it would be much more morally objectionable for a distant colleague of mine to do so. (Enoch 2016, 34; see also Tsai [2018, 355])

We readily accept the idea that friends have a (stronger) duty to care for their friends and be concerned with their well-being for their friends’ sake. After all, it is the kind of special duty that many accounts of partiality highlight.⁹ We also agree that the duty to care provides the right kind of reasons (in contrast to shared projects, meshed identities, vulnerability and intimacy) to affect the justification of paternalism. However, it clearly does not follow directly that a duty to care renders paternalistic interventions always better justified or even required in friendship. To see this, consider Enoch’s example more thoroughly: His wife rightly believes that candy is not good for him, and she knows that his desire for candy is a result of his weak will. There is thus a relationship-independent reason to interfere (impaired agency) and a relationship-dependent reason that derives from the duty to care.

Fox also builds her discussion on a case with a similar structure, in which a man, Steve, ends his relationship with his partner, Cathy, by leaving a message on her answering machine:

Cathy is frantic to find Steve’s abrupt dismissal of the relationship waiting for her when she walks in the door. Agitated and overwhelmed, she impulsively starts throwing clothes into an overnight bag in preparation for the two-day drive to the city where Steve lives. This sudden decision will have drastic consequences for her career; her new supervisor and coworkers have already observed a certain emotional fragility in her manner, connected with the affair with Steve. She has an important presentation to give in the morning; to abandon them will spell the end of her career. (Fox 1993, 576)

Cathy’s friend Elizabeth (who happens to visit Cathy in this moment), after some discussion, takes Cathy’s purse which has her car keys in it, as well as other essentials, and hides them from her. According to Fox, Elizabeth is justified to intervene, thereby preventing Cathy from ruining her career. But again, the case crucially involves relationship-independent reasons (impaired agency and severe harm). It thus remains an open question what role the duty to care exactly plays.

To illustrate the importance of this observation about the cases, consider the following variations: Let us suppose that Cathy has decided to drive to see Steve not in an impulsive way but in a calm and autonomous manner, and that eating candy is a deliberate lifestyle choice on the part of Enoch (rather than a sign of his weak will). It surely seems much more controversial to claim that a paternalistic intervention is vindicated in these cases on account of the intimate relationship in question.

There are two explanations that go to show why the duty to care cannot by itself establish that paternalism is justified. First, the initial set of examples points to relationship-independent reasons for paternalism, such as the impaired agency of the friend or partner, and/or the severe harm that can be prevented by intervening. Here, relationship-independent reasons clearly play the relevant role in justifying paternalism, and it remains elusive whether there is any role to play for the duty to care for one’s friend. Second, and as the variations of the initial cases highlight, in cases in which relationship-independent reasons are not in play (e.g., the friend’s agency is not impaired), the duty to care seems to come into conflict with the duty to respect the autonomy of one’s friend. And it is far from clear whether the duty to care should win out. Bengtson and Midtgaard, for example, claim that “precisely because you are in an intimate relationship, you have a stronger duty to respect the autonomy of the other” (Bengtson and Midtgaard 2023, 8).

Whatever the exact relation between the duty to care and the duty to respect autonomy is (see below), our discussion brings to light more general questions that need to be addressed: How exactly do relationship-independent and relationship-dependent reasons interplay, and what exactly is the role of relationship-dependent reasons? Do they provide a stronger justification for paternalism in friendship, when paternalism is already morally permissible due to relationship-independent reasons? Or do relationship-dependent reasons render paternalism morally permissible when it is not permissible for relationship-independent reasons? Or is the relation of the two kinds of reasons of yet another kind?

Before we address these questions, let us summarize our discussion by formulating three general challenges that proponents of the view that relationship-dependent reasons change the deontic status of paternalism must meet:

(Challenge 1) Any such view must show that interventions in friendships are instances of paternalistic interventions, and not other kinds of interventions.

(Challenge 2) A plausible and non-question-begging conception of friendship must not rest on a one-sided focus on certain norms of friendship (such as the duty to care), while neglecting other norms (such as respect for autonomy).

(Challenge 3) An account that defends relationship-dependent reasons for paternalism must explain why relationship-independent reasons (i.e., impaired agency or severe harm) do not suffice to account for the justification of paternalism, and why friendship independently affects the justification of paternalism (beyond the mere fact that it provides epistemic access to relationship-independent reasons).

3. The Role and Source of Relationship-Dependent Reasons—the Modifying View

Building on our previous discussion, we now wish to defend a view that meets these challenges. According to what we call the Modifying View, relationship-independent reasons (such as impaired agency or severe harm) determine the permissibility of paternalism in friendship, while relationship-dependent reasons affect the overall justification of paternalistic interventions in the following ways: If paternalism is permissible, our duty to care may transform the deontic status of paternalism from permissible to morally laudable or even morally required; but if paternalism is impermissible (either because it violates the duty to respect the friend’s autonomy or because no severe harm is at stake), the duty to respect the autonomy of our friends makes the paternalistic intervention in question even more contemptible (and undermines the friendship itself).

To bolster this view, consider the case of William who is very upset by the way in which his supervisor dealt with him during a meeting and decides to immediately call him to quit his job. He runs into his friend Sarah during the lunch break and tells her about his plans. Sarah knows that William is generally happy with his job, and that he will not easily find a new or better one. Crucially, she also knows that William, who in general is very calm and considerate, tends to make impulsive decisions when he is not taken seriously, decisions that he later regrets. For that reason, Sarah tries to persuade William to wait another day. But William is really upset and reaches out for his phone to call his supervisor. For a second, Sarah is paralyzed by the situation, but then she just takes William’s phone out of his hands and calmly explains: “I would never do this unless I was sure that you are not thinking straight and will come to regret your decision. But I must do this, as your friend.”

We believe that many would agree that such a paternalistic intervention is permissible or even morally required. But what role do friendship-dependent reasons play in the justification? To begin with, there are strong relationship-independent reasons in favor of paternalism. William faces severe harm by losing his job, and his agency is impaired because he is susceptible to acting or deciding impulsively when not taken seriously. Because Sarah knows William well, she has access to these reasons, and she particularly emphasizes his impaired agency in her explanation. But this is not the whole story, since Sarah also claims that she “must do this” as his friend. Plausibly, we think, this emphatic statement refers to her duty to care for William that clearly applies in this case: Sarah should not be indifferent to William’s decision when she is certain that he is not thinking straight and will come to deeply regret his decision. If she were to just let William act on his impulse, he could rightly complain later, “Why did you not stop me?” The duty to care thus plays a distinct role in the justification of paternalism in friendship, but only in the following way: it transforms the deontic status of the paternalistic interventions from permissible to morally required (or, at least, to morally laudable).

But if this is the correct view, it seems clear to us that friendship can also affect the justification of paternalism in the opposing direction and provide additional reasons against a paternalistic intervention. Consider the following variation of the example:

William meets Sarah for lunch and tells her about his meeting with his supervisor. He calmly explains to Sarah that he does not want to endure the disregard of his supervisor and wants to immediately confront him. “I just do not want to be treated in this way, and I want to make this clear to him right now,” he asserts. Sarah believes that William is generally happy with his job and that the confrontation might impair the relationship with his supervisor; she also believes that William places too much weight on his supervisor’s behavior. While she acknowledges that it is William’s own autonomous desire to no longer be treated like this, she thinks that he would be better off if he were to let go of his grievances. Sarah tries to argue with William, but he has made his decision and attempts to pick up his phone. Sarah takes the phone away and explains: “I would never do this unless I was sure that it is in your best interest not to do this. I must do this, as your friend.” William is outraged by Sarah’s behavior and confronts her: “How can you do this to me as my friend?!”

We believe that many would understand William’s reaction and would regard such a paternalistic intervention as impermissible. For one thing, there are relationship-independent reasons against the paternalistic intervention: Sarah’s assessment of the situation is questionable. After all, merely confronting the supervisor may not yield the harmful consequences that she anticipates. More importantly, Sarah disregards William’s own perspective on the situation and does not respect his autonomous decision. Sarah’s intervention thus appears to be impermissible for relationship-independent reasons. But, again, this is not the whole story, as indicated by William’s complaint—“How can you do this to me as my friend?!”

In our view, this emphatic statement refers to Sarah’s duty to respect the autonomy of her friend William. While we generally owe respect to the autonomy of all persons, we suggest that this duty takes a specific meaning and shape in the context of friendship.¹⁰ As relational theories of autonomy stress, it takes self-trust and self-esteem to develop and maintain one’s autonomy (Govier 1993; Benson 1994; Anderson and Honneth 2005). And it is trust in intimate relationships in particular, as well as interpersonal relations of equality more generally, that provide the necessary conditions for self-trust and self-esteem (Scheffler 2014). If this is on the right track, then respect for a friend’s autonomy requires acknowledgment of what that friend values and trust in the friend’s autonomous capacities. The duty to respect the autonomy of a friend thus requires more than respect with regard to the autonomy of strangers.¹¹ It is a positive duty that involves acknowledgment, trust, and support.

Such a respect-based duty provides an additional and distinct relationship-dependent reason against Sarah’s paternalistic intervention. More precisely, the intervention is impermissible for relationship-independent reasons, but it is even more objectionable due to relationship-dependent reasons that derive from our special duties to respect the autonomy of our friends. As a result, the special duty of respect can change the status of paternalism in friendship from impermissible to particularly contemptible.

The Modifying View captures intuitions about paradigmatic cases in an elegant way and provides a systematic explanation of the interplay of relationship-independent and relationship-dependent reasons. Thus, it can accommodate many cases discussed in the literature and, in our view, it provides an important perspective on how to think about paternalism in friendship. Furthermore, we suppose that even those who are skeptical about the claim that friendship distinctly affects the justification of paternalism might want to accept the Modifying View, because it leaves the general structure of how to morally evaluate paternalistic interventions intact:¹² if we are only concerned with permissibility or impermissibility, we need only consider relationship-independent reasons; but if we are concerned more generally with the moral value of paternalistic interventions and their status in friendship, we may look at relationship-dependent reasons that can strengthen the overall justification for or against paternalism. More importantly, the Modifying View is able to meet the three challenges mentioned above. It clearly considers cases of paternalistic interventions. It does not one-sidedly focus on certain norms of friendship, and, most importantly, it explains in what ways friendship can affect the relationship-independent reasons for paternalism (beyond the mere fact that it provides epistemic access to those reasons).

So far, we have discussed the source and role of relationship-dependent reasons focusing on single instances of paternalism. Our proposed Modifying View is a promising account to explain the justificatory role of friendship within this framework, and it correctly determines its boundaries in paradigmatic cases. On closer inspection, however, there remain borderline cases where there is reasonable disagreement about relationship-independent reasons for the (im-)permissibility of paternalism. As we will now show, the Modifying View does not have the resources to provide sufficient orientation in such cases. More precisely, it does not do justice to the fact that paternalism in friendship is often embedded in a diachronic and dynamic structure and can carry a distinct meaning that transcends the meaning of particular instances of paternalistic interventions at single points in time.

4. Paternalism and the Dynamic and Diachronic Nature of Friendship

Borderline cases of paternalism are cases in which relationship-independent reasons do not clearly settle the question whether paternalism is justified. Such cases occur quite often, since impaired agency or harm come in degrees.

Consider, for example, the case of Simon who regularly provokes other people during team meetings. His friend Stephen works at the same company and is worried that Simon might take things too far and face severe consequences. For that reason, he regularly intervenes in such situations by deflecting or mediating. For some time, Simon does not oppose this behavior. He is aware of his tendency to get carried away. However, after several incidents, Simon gets annoyed by Stephen’s interventions. He feels that Stephen does not respect his way of handling things. He confronts Stephen: “Why do you always intervene when I push controversial ideas or upset others during meetings? I thought we were friends.” Stephen is perplexed and truthfully replies that he is concerned about Simon’s career and, as a friend, he really wants to protect him from consequences that will be detrimental to his well-being: “I was just trying to be a good friend!”

We assume that there are no compelling relationship-independent reasons for or against a paternalistic intervention, partly because the intervention is rather mild (albeit clearly paternalistically motivated), and partly because Simon’s agency is neither severely impaired nor does he clearly face immediate harmful consequences. When considering a single instance of Stephen’s paternalistic interventions, we may reasonably disagree whether a paternalistic intervention is called for. Perhaps friends should intervene in milder cases of impaired agency or in light of bad yet not so terribly harmful consequences, because they are friends and have a duty to care for each other. But perhaps they should respect each other’s autonomy. It seems that we cannot properly assess the case when looking at a single instance of Stephen’s paternalistic intervention.

However, when considering the pattern of paternalism that Stephen’s behavior exhibits, one might start to understand why Simon is upset: To constantly intervene in a friend’s behavior can be considered as disrespectful and as meddling with the friend’s affairs. We generally expect our friends to trust in our agency and our capacity to take care of ourselves. After all, if Stephen regularly intervenes on Simon’s behalf, he clearly disrespects Simon’s autonomy. He thereby compromises their friendship, turning it into some kind of custodian or therapeutic relationship.

The example suggests that patterns of paternalistic interventions that occur over time may be especially problematic in friendship. Friendship thus affects the justification of paternalism in a distinct way. To illustrate the point, consider another borderline case in which such a pattern is absent: Ron is a very helpful person. He often offers his help too quickly or takes up requests for support by others too easily. He is aware of this tendency and has shared his struggle with his friend Sam. Sam is in a difficult situation because he is moving out and needs to clean his apartment before the weekend. Ron is the only friend who is likely to come to help if Sam were to ask him. But as a friend, Sam decides not to do that, because he knows that Ron would be unable to refuse the request and is already overcommitted. After the weekend, Ron learns from another friend of Sam’s that Sam has been working all night and nearly did not manage to finish the cleaning. He asks Sam, with a slightly indignant tone: “Why did you not ask me? I would have helped you. That’s what friends are for.” Sam retorts that he did not ask because he knew that Ron would come to help, despite his other commitments. “But you could have asked,” Ron responds, and Sam answers: “Yes, but knowing you, I thought it would be better for you if I did not ask. As a friend, I want you to do well, and I know that you are already overcommitted.”

Again, whether Sam’s behavior is a permissible form of paternalism cannot be settled by appeal to relationship-independent reasons alone. There might be reasonable disagreement, even between Ron and Sam, whether this is a case that calls for a friend’s care or whether it calls for respect of the other friend’s autonomy. But what speaks in favor of care is that Sam’s paternalistic intervention is bound to a specific situation and communicated as such. His behavior thus does not express that Ron is generally unable to exercise his own agency. If, by contrast, Sam were to regularly not ask Ron for help anymore, things might be different, and his paternalistic behavior may be regarded as offensive by Ron because it exhibits a disrespectful attitude, belittling Ron as being incompetent to stand up for himself.

So far, we have highlighted the problematic nature of ongoing patterns of paternalism in friendship that has hitherto been neglected in discussions about paternalism in friendship.¹³ But the more general point, according to which paternalism can carry a distinct and morally relevant meaning in friendship, can also be illustrated with regard to single instances of paternalism that affect both the history and future of the relationship in question.

Consider Claire who decides to take a long holiday, although she has much work to do and risks falling even further behind. Her friend Christine believes (with good reason) that Claire would be better off if she were to first finish at least some of the work. However, when Claire tells Christine about her travel plans, Christine deliberately resists arguing with her and trying to convince her to change her plans. When Christine talks about the situation at home, her partner asks: “Why do you not try to convince her that it is a mistake?” Christine replies: “I want to support Claire in this difficult situation. I do not want to express to her that she is unable to take care of her affairs. She is already in a difficult situation and has limited trust in her own capacities.” We think that many would well understand such an answer. Sometimes not to intervene carries a distinct meaning, namely, to empower a friend or support a friend’s autonomy. Friendship can thus provide distinct reasons against paternalistic interventions.

To summarize, our examples manifest borderline cases because the relationship-independent reasons do not determine whether paternalism is permissible. At the same time, there can be reasonable disagreement as to which norm of friendship should have more weight. What is more, these cases show that while it may be justifiable to let the duty to care prevail in individual instances that justifiedly change the deontic status of paternalism, patterns of such care may have an opposing effect. Provided that friendship is a relationship that is to be guided on the basis of equality, patterns of care risk undermining the friendship. This is why friends are well advised to negotiate their joint understanding of where they want to draw the line between different norms of friendship.

We thus hope to have shown that friendship-dependent reasons can tip the balance in favor of paternalism when relationship-independent reasons are not decisive, either because problematic patterns of paternalism negatively affect the friendship, or because a single instance of paternalism may carry a distinct meaning for the friendship as a whole.

5. Conclusion

According to the view developed in this paper, friendship can affect the justification of paternalistic interventions in two ways. First, in paradigmatic and clear cases in which relationship-independent reasons render paternalism permissible or impermissible, friendship strengthens such reasons and affects the moral quality of the intervention. Second, in borderline cases where relationship-independent reasons do not clearly determine whether paternalism is permissible or not, friendships can provide distinct reasons for or against paternalistic interventions that derive from competing norms of friendship (such as the duty to care and the duty to respect the autonomy of one’s friend) and the specific meaning of paternalistic interventions. Specifically, patterns of paternalistic interventions risk compromising the friendship even if single instances of paternalism may be rather unproblematic.¹⁴

REFERENCES

Footnotes