Friendships Need to Go Wrong in Order to Go Right

Abstract

Companion friendship is a paradigm example of a trusting relationship and is a central good in human life. These friendships are also complex; navigating this complexity carries risk. Philosophical work has largely overlooked questions about how friends might navigate this morally risky space in ways that protect and develop their relationship over time. More specifically, although it is generally accepted that friendship involves acting to promote the well-being of one’s friend, ethical analysis of such interpersonal action has not addressed questions such as: How does acting for a friend’s well-being follow from and affect the trust within these relationships? What are the risks of acting for a friend’s well-being? Do genuine but unsuccessful attempts to promote a friend’s well-being, that bring about a rupture to the trust, necessarily cause lasting damage to trusting relationships? If not, why not? We argue that getting it wrong when acting for a friend’s well-being can provide an opportunity to protect and develop the trusting relationship, even while it causes harm to one’s friend and temporarily damages the relationship.

“You can’t stay truly connected without some level of misunderstanding or conflict…”

— Aminatou Sow and Ann Friedman

1. Introduction

In the HBO show Insecure, the characters Issa and Molly are best friends in their late twenties who have been close since college. In season three, Issa is in an intimate relationship with Nathan, but then Nathan abruptly “ghosts” Issa—ignores her attempts at communication—for a whole month. In episode eight, Molly is outside Issa’s apartment on Issa’s 30th birthday when Nathan shows up with flowers. Molly tells him to leave: “Issa’s in a good place right now,” she says, and tells Nathan that he’s not allowed to ruin Issa’s birthday. Molly has planned a perfect birthday for Issa. Issa does not discover that Molly sent Nathan away until the end of the day. While Molly and Issa discuss the great job Molly did planning a perfect birthday for Issa, Molly mentions that she even sent Nathan away because she knew he would ruin Issa’s day. Issa is mad: Molly had no right to send Nathan away, she says, and Molly just projected her own negative perspective onto the situation. Molly has started seeing the worst in people, Issa says. They part on a bad note. From Issa’s perspective, Molly has clearly done something wrong in their friendship. In season four, Molly and Issa reconnect and their friendship continues, albeit perhaps on slightly different terms.

The question of just what Molly did wrong, and how to understand this conflict within the context of her friendship with Issa, is of philosophical interest. It cannot be the case that Molly was a bad friend simply because one friend should never act for another, since a commonly held understanding of intimate friendship is that it is a relationship in which both parties are motivated, where appropriate, to act for each other’s benefit for their friend’s own sake (Badhwar 1987, Blum 1980, Cocking and Kennett 1998, Thomas 1987, Friedman 1989 and 1993).

This issue has recently captured philosophical interest. George Tsai argues that some paternalistic treatment may be justified by the constitutive features of intimate relationships, which include “joint identification and shared projects, trust and vulnerability, mutual understanding, and shared history” (Tsai 2018: 358). Yet other philosophers disagree with Tsai that behavior which is wrong outside of friendships may be justified within them. Andreas Bengtson and Soren Flinch Midtgaard (2023), for example, raise a series of objections which purport to show that the constitutive features that Tsai identifies do not license paternalism. Furthermore, while not addressing paternalism directly, Elizabeth Brake (2025) points out that these same constitutive features create opportunities for wrongs that can only occur within friendships, due to the characteristic vulnerabilities which friendships enable. While philosophers generally do not dispute the characteristic features of friendship—including promoting one another’s well-being through the unique understanding, care, and trust that distinguishes friendship from other relationships—there is normative disagreement about which types of interventions count as permissibly “promoting one another’s well-being.”

In this article, we are not directly concerned with the issue of which acts within friendships count as paternalistic and whether such acts are justified. Rather, we consider a related, but overlooked issue: given that intimate friends will often attempt to act in ways that promote one another’s well-being due to the trusting nature of companion friendship, what happens when these attempts go poorly? In other words, while Tsai (2018) and Bengtson and Midtgaard (2023) focus on the permissibility of particular acts within friendships, we align with Brake (2025) in our shared interest in how behavior within friendships can accumulate to characterize those relationships over time. Unlike Brake’s focus on the distinctive wrongs made possible by the constitutive features of friendship, we ask whether interpersonal errors in attempting to act for the good of a friend can aid in the promotion of these constitutive features of friendship.

Specifically, we explore the idea that companion friendships need to “go wrong in order to go right”: that friends must experience and weather misunderstandings and conflicts (two terms we explain in more detail below) to maintain the intimacy that is almost universally agreed to be a key characteristic of companion friendship. A few clarifications are in order at the outset. First, in using the language of “need,” we do not mean to imply strict necessity. That is, we are not making the strongest claim that all possible close friendships must, in some moral sense or otherwise, go through this cycle of rupture and repair. Rather, we propose that, in the kinds of close friendships where people do act for each other, and given the inevitable epistemic conditions of human relationships, this cycle of rupture and repair will be experienced. We are specifically interested in the relational benefits of this cycle. We do argue that friendships in which some (reparable) things “go wrong” are often better than friendships in which there are no such ruptures and their attendant opportunities for repair. Second, in using the language of “going wrong,” we do not intend to imply the philosophical concept of wrong, understood as the violation of a persons’ moral rights. Rather, we mean “going wrong” in the colloquial sense—that there has been a disturbance in the trajectory of the friendship, which may or may not have been caused by a moral infraction.¹ Third, since our aim is to defend a counter-intuitive thesis—that ruptures in friendships can strengthen them—we spend comparatively little time discussing the intuitive thesis that ruptures damage friendships. While we agree that some ruptures may provide good reasons to end the relationship, we draw attention to the opportunities that ruptures provide in the context of a committed friendship—opportunities that are available, but ultimately not taken, even in cases when someone reasonably decides to end the relationship. Lastly, while we refer to “friendships” throughout the text, we do not intend to exclude other kinds of intimate, symmetric relationships from our argument. Rather, we are following philosophical convention in referring to such relationships as “friendships,” while acknowledging that there may be diversity in the types of relationships with the qualities identified as central to friendship (in section 2).

Our interest in the way that particular acts contribute to the nature of a friendship over time is captured by our use of the metaphor of a “relational ecology” (Niker and Specker Sullivan 2018). As with natural ecology, which is the study of patterns of relationships among organisms within their environment, a “relational ecology” is the pattern of interactions between two or more people over time. Within the relational ecology of friendship, particular interactions have significance for the relationship as a whole and over time, even as they may also be meaningful as isolated events in themselves. The lens of “relational ecology” thus allows us to consider the dynamic and cumulative effects of patterns of interactions over time within a friendship. Here, we argue that making mistakes when acting for a friend’s well-being can provide an opportunity to develop and protect the ongoing health of the relationship, even while it may harm one’s friend and (temporarily) damage the relational ecology. In previous work, we have used a version of this relational ecology account to argue that certain purported moral wrongs are not in fact wrong when certain relational conditions are met (Specker Sullivan and Niker 2018). As with the idea of “ecological health,” our reference to the “health” of a relationship is intended to pick out relationships with the qualities of stability and dynamicity, two qualities we explain in more detail in the next section. As we will also explain in section 2, we join other philosophical accounts in considering trust to be central to healthy interpersonal relationships.

By attending to the normative, diachronic space of intimate friendship, we bring to the fore overlooked questions about the protection, maintenance, and repair of these valuable relationships. These include: How does acting for a friend’s well-being follow from and affect the trust within these relationships? What are the risks of acting for a friend’s well-being? Do genuine but unsuccessful attempts to promote a friend’s well-being, that bring about a rupture to the trust, necessarily cause lasting damage to trusting relationships? If not, why not?

Our argument proceeds as follows. In the next section, we describe what we mean by companion friendships. In the third section, we clarify the risks that accompany a relationship in which friends act for each other’s well-being. We distinguish two different ways—“misunderstandings” and “conflicts”—in which these risks can “go wrong” and thus damage the relational ecology of the friendship, and we show how a dynamic understanding of the friendship allows for a clear distinction between errors that are opportunities for relational growth and errors that signal a need to end the friendship. In the fourth section, we explain how understanding a friendship as a trusting relationship helps to make sense of the opportunities that errors create as well as the repair that must happen for these errors to play a role in cultivating healthy friendships.

2. Friendships as Trusting Relationships

We follow the philosophical literature in using the term “companion friendships” to refer to lasting, intimate friendships marked by “reciprocal deep affection, well-wishing, and the desire for shared experiences” (Cocking and Kennett 1998: 502) and in which friends act for each other’s good. Our goal is not to defend any single account.² Our proposal that companion friendships need to experience ruptures for development extends across all accounts that accept that companion friendship is marked both by a desire to support and promote each other’s well-being and by interventions that act on mutual understanding in order to fulfil that desire.

Two features of this broad philosophical account are especially important for our purposes. These features are largely taken for granted within the existing philosophical discussion about friendship and so require some elucidation. These are the fact that companion friendships are trusting relationships and, relatedly, the fact that friends are usually both expected and motivated to act for one another’s well-being, which carries some risk. The interaction between these two central features is relevant for our argument about cases where actions taken for a friend’s well-being misfire and so affect the grounds of the trusting relationship.

2.1. Trusting Relationships

In our view, trusting relationships have two defining features, both of which arise from their fundamental mutuality (Niker and Specker Sullivan 2018). The first is attitudinal: trusting relationships are based on mutual care and concern. Companion friends care deeply for one another and desire that the others’ aims in life be fulfilled. Friends also care for the friendship—a deep source of value for each of them. The second feature is epistemic: trusting relationships are based on a mutual understanding, sometimes described as epistemic access or intimate knowledge, that develops primarily through the friends’ shared personal history. Companion friends know a lot about each other; sometimes they know things that no one else knows. This knowledge is essential to being able to care for each other well—one friend cannot act to promote the well-being of another without knowing what counts as benefitting that friend and how they might need help to attain it. This doesn’t mean that the two friends know each other perfectly well, just that one of the distinctions between companion friends, on the one hand, and acquaintances and non-companion friends, on the other hand, is intimate knowledge of each other.

We capture the fundamental mutuality of trusting relationships in these two defining features by conceptualizing trust as a property of the relation. This means that trust is applicable primarily to the relations between A and B—their mutual understanding and care—rather than to A’s and/or B’s psychological states taken individually (Niker and Specker Sullivan 2018). This is not to say that trust is not also an attitude captured by each individual’s psychological state. Rather, it is to emphasize that individual, attitudinal trust is different from participation in a trusting relationship, where the relationship is more than the mere sum of individual instances of trust. A rupture to the relationship is not simply something that needs to be fixed in order to ‘return to normal’. Rather, a rupture to the relationship creates an opportunity for repair such that the relationship becomes healthier—i.e., creates a more stable foundation for trust—than prior to the rupture.

The relational ecology of trust within companion friendships, when it is rooted in these two features, is simultaneously stable and dynamic—a state which friends themselves wish to continue. It is stable because there is more than a minimal level of mutual confidence in each person’s commitment to the other and to the friendship. Confidence, here, is another term for each person’s belief that the other has made a set of commitments which they will keep. It is dynamic because it is constantly changing, if only in small ways. The relational ecology develops through the interactions between the friends; these interactions constitute their relationship. For instance, if one of the friends experiences a hardship then this may provide an opportunity for her friend to care for her. Care, in this case, may include acting on her behalf in ways that support her well-being and her autonomy (Specker Sullivan and Niker 2018). This may allow trust to extend into other domains or situations previously not included within the relational ecology of their trusting relationship. Conversely, if one of the friends oversteps, this could lead to a contraction in the relational ecology of their trusting relationship. We describe these errors, and their ramifications, in more detail in section 3.

Due to the existence of these two key features (mutual care and concern and mutual understanding), companion friendships are paradigmatic trusting relationships. Many examples of these types of friendships can be found in popular culture, at varying ranges of intimacy: Frog and Toad, Sherlock and Watson, Friends’ Monica and Rachel, Insecure’s Issa and Molly. Unlike other trusting relationships premised on shared familial histories, biological ties, or socially recognized commitments (and the institutions that uphold them), such as parents and children or spouses, friendships are unique; their level of intimacy rivals that of the others while nevertheless always being voluntarily “chosen.” In other words, whereas in rough patches the bonds of family or the social institution of marriage may hold people together, in a friendship such external bonds may exist but are perhaps less common. Rather, both members of the friendship must in some sense recognize the relationship as good and worth continuing, even during conflict.

The expansive nature of the trust shared between them is what enables friends to access the unique and precious goods of companion friendship. But it is also what makes friends vulnerable to both experiencing and bringing about a range of distinctive and deep harms. This is a result of the expectation that friends will, as and when appropriate, act for each other’s well-being—out of the concern and care they have for one another and using the intimate knowledge and understanding they have of each other.

2.2. Acting for One Another

It is commonly acknowledged, both culturally and philosophically, that central to what it means to be in a committed friendship is that friends will act to promote their friend’s well-being. This is part of what it means to care about and for one’s friend (Helm 2021).

Acting for a friend’s well-being means not merely offering gifts and invitations that one knows their friend will enjoy, but also taking steps that they believe will be good for their friend. The simplest and least risky mode of acting for a friend is helping a friend when they ask for assistance, or explicitly offering help and only acting if the offer is accepted. While such transparent communication in friendships is laudable, it does not take advantage of the mutual understanding characteristic of intimate friendships. If one of the features of close friendships is that friends know each other intimately, then a benefit of such a relationship is surely that one does not always need to explicitly ask for help or support to receive it. It seems intuitively plausible that a person may feel let down if they need to tell a friend explicitly each time they need help. We want our friends to be attuned to us such that we do not always have to ask for help. This attentiveness, as much as the action that issues from it, matters when promoting a friend’s well-being.

Actions rooted in attentiveness need not be extremely consequential. Acting for a friend without invitation or permission most often happens through conversation: based on the mutual understanding in a trusting relationship, friends can encourage each other to reconsider or to change patterns of thought that are not serving them, or to reevaluate decisions that may have been made in error or haste (cf Tsai 2014). Yet these actions can be more overt and risky, such as when one friend plans a surprise birthday party for the other, tells a third mutual friend to initiate a difficult conversation with the friend, or prevents a former intimate of the friend from reinitiating contact.³

Such acts are based in the discretionary power that the trusting relationship affords to friends and which friends are expected to use if they are to live out and grow their friendship. The language of “discretionary power” here refers to Annette Baier’s idea that trust involves a special kind of vulnerability, which comes from handing over power to another person to look after something you care about according to his or her own discretion (Baier 1986: 239). Trust is thus the accepted vulnerability to possible (though not expected) harm that comes with “leaving it to them” (Baier 1991: 137).

Moving beyond Baier, each friend not only accepts the vulnerability that comes with trusting the other; they also accept the risk (and its attendant vulnerability) that comes with using their discretionary power to act for the other’s good. Our relational ecology account of trusting relationships captures this idea that companion friends accept and simultaneously live out these two roles. It allows us to see the bidirectional vulnerability of each friend. For example, Issa makes herself vulnerable to being hurt or having her trust betrayed through Molly’s actions as Molly exercises her discretionary power in acting for Issa’s well-being. But Molly makes herself vulnerable in a different way—by putting this very valuable good of hers, the friendship, at risk—when she takes on the role of acting to promote Issa’s well-being. And vice versa.

As a psychological point, many friends won’t experience using their discretionary power to try to promote their friend’s well-being as risk-taking in the moment, unless the action is a very weighty one. Indeed, on some views the action carrying the risk is taken to follow naturally from the relational ecology of the friendship. On Nehamas’ view, for example, “Our friendships permeate our personality, they structure our perceptions of the world, and in many circumstances enable us to act in a particular way without a second thought: they are part of the background that allows us to perceive directly that we must do something for a friend that we wouldn’t do for someone else” (Nehamas 2016: 149). But even if we accept a view that naturalizes the actions we take for our friend, this does not negate the risk-taking nature of the action. It may offset it in various ways: when things are functioning well in a relationship, there is an attunement that follows from the attentiveness, communication, intimacy, and so on that marks the character of the friendship’s relational ecology. But there are still several ways in which an action, whether the result of some direct perception or not, can misfire and so damage the trust that licensed it, as we discuss further in the next section.

Furthermore, while our relational ecology model of trusting relationships differs from Baier’s three-place model of trust (in that relationships themselves are characterized as “trusting,” and trust does not necessarily require an explicit object which is entrusted; see above and Niker and Specker Sullivan 2018), we agree that implicit within trusting relationships is the constant negotiation, explicit and implicit, of each friends’ ability to use their discretion in acting for the well-being of the other. This is part of what it means for a friendship, as a trusting relationship, to be dynamic: the boundaries of this discretionary power are actively and continually negotiated.

Actions taken at the boundaries of the relational ecology of trust are the most difficult and risky due to reasonable doubt as to whether a particular action would be welcomed or not. In many cases, the friend’s welcoming of the action will be affected by contextual factors, which is why these types of relationships require a moral competence to be attentive, perceptive, and responsive to each other’s signals (Jones 1996). One of the conditions for trusting relationships is that generally parties are able to have a good understanding, not only of the contours of when acting for the other is appropriate, but also of the textured nature of what acting for the other’s benefit best (or at least appropriately) looks like in the particular situation, from the other’s point of view.

The risk inherent in acting for a friend’s well-being is often worth taking—the friendship will be strengthened if the act goes right, where this means that the act does indeed support the well-being of the friend and that the friend welcomes the action. But friends’ attempts to act for each other, however well intentioned, can also misfire. When there is such a misfiring of discretionary power, this can be described as a way in which an act of friendship—and possibly something deeper about the trusting relationship—has gone wrong.

3. How Friendships Can Go Wrong

There are two ways in which actions taken for a friend’s well-being can go wrong, which are the inverse of the ways such acts can go right. First, an act may defy expectations and be detrimental to the friend’s well-being—this failure is rooted in either an epistemic misunderstanding of a friend, or a practical misunderstanding of how to act for one’s friend. Second, even though the act may support the friend’s well-being, the friend may nonetheless reject the act—this failure is rooted in an inapt expression of attitudes, not a mistaken understanding. Adopting Sow and Friedman’s (2021) terminology in the epigraph, we refer to these as “misunderstandings” and “conflict,” respectively. In this section, we explain how misunderstandings and conflict affect the relational ecology of trust, rupturing the trust shared between friends. Furthermore, we explain that such ruptures to the relational ecology of trust can be more or less severe: some will be relatively minor and fairly easy to navigate, while more extreme instances may provide grounds for ending a friendship.

One clarification is in order. While we use the language of “friendships going wrong” in this article, by so doing we do not intend to imply that misunderstandings and conflicts are best understood as violations of moral rights (as Brake (2025) does). Nor do we intend to imply that misunderstandings and conflicts can only be classified as harms. This is because, while friends’ well-being can be harmed by misunderstandings and conflicts, we do not think this is the only way in which acting for a friend’s well-being can “go wrong.” Due to the fuzzy boundaries of discretionary power within intimate relationships as well as to friends’ unique expectations about consent-seeking and permission-giving practices, it is possible that misunderstandings and conflicts will include both failure to meet interpersonal obligations and pernicious consequences. Yet to separate the two would be to unnecessarily simplify the complex nature of the interpersonal space of intimate relationships, and so we do not distinguish them here.

3.1. Misunderstandings

The type of rupture generated by “misunderstandings” involves a failure of understanding that leads to an act being unexpectedly detrimental to the friend’s well-being. Standardly, failures of understanding will fall into one (or both) of the following categories: one friend misunderstanding what their friend needs or wants (and hence what would support their well-being) in a given situation; and/or, even with a good grasp of their friend’s needs or wants, one friend misunderstanding something important about the context in which they are acting in support of those needs or wants.

A failure of understanding can come about for different reasons. First, it could simply be a well-intentioned mistake. Even when friends are communicating well and the relational ecology of trust is healthy, a friend’s efforts at care can go awry because, for instance, they may not know—through no fault of anyone involved—some relevant contextual fact (such as if Issa had invited Nathan to her birthday, without Molly’s knowing).

Second, a failure of understanding may be the result of an epistemic failure of the friend taking the action. We understand this broadly to include any failure that affects the friend’s ability appropriately to form and maintain (accurate) intimate knowledge about their friend and hence which is likely to generate one or more misunderstandings. Examples include inattentiveness, complacency, lack of epistemic effort or imagination, and blind spots. Blind spots are more likely when friends have different backgrounds, influenced by class, race, nationality, gender, etc. But, more broadly, friends may have blind spots about their own motivations for some action they are undertaking on behalf of their friend.

Such epistemic failures are often bound up with failures of care and concern (discussed in more detail below). For instance, not paying adequate attention to changes in a friend’s life is an obstacle to intimate knowledge and can affect your interpretive and imaginative abilities with respect to what they might need or want from you in certain situations. Issa accuses Molly of making a similar type of mistake.

Third, the friend taking the action might misunderstand what is included within their discretionary power at a particular time. One set of reasons for this relates directly to the epistemic failures and related failures of care and concern outlined below (e.g., a lack of attention to the shifts in intimacy and what this might mean for the appropriateness of their actions). But a more distinctive set of reasons concerns communicative atrophy. For instance, a friend may act in a way that constitutes a misunderstanding because their friend has not shared certain details with them about what is going on in their life. This may have to do with changes in the friend’s life that are unrelated to the friendship itself. Nevertheless, it affects the relational ecology of the friendship because intimacy declines (at least temporarily), and their friend’s intimate knowledge fails to track reality in one or more important ways. Sow and Friedman, for example, explain that the rough patch in their friendship “happened, like a lot of relationship breakdowns, because of what we were going through individually and not communicating about” (2021: 151). In some such cases, the responsibility for some misunderstanding may fall more with the acted-for friend, and the friend who acted for them may have been acting appropriately given the intimate knowledge and contextual information available to them at that time.

3.2. Conflict

The second kind of rupture, for which we use the shorthand term “conflict,” is generated by some kind of attitudinal failure. This occurs when an action is received as unwelcome by the friend, even when it is not based on any misunderstanding about such an act supporting their well-being. There are two distinct categories of attitudinal failures: one relating to the appropriate expression of concern (by the acting friend) and another relating to appropriate receipt of the other person’s concern (by the acted-for friend).

Appropriately expressing one’s care and concern often comes naturally for intimate friends (Nehamas 2016). Yet even when a friend is acting with the best intentions, the expression of their care can go wrong. Most innocuously, friends—who are fallible, imperfect human beings—cannot be expected to be maximally compassionate all the time; sometimes issues such as exhaustion can, completely reasonably, affect a friend’s ability to gauge how to express their care and concern appropriately.

More troublingly, a friend’s expression of care and concern can be inappropriate for reasons to do with the timing of the act. For instance, a friend may misjudge their action out of their—let’s assume, genuine and enthusiastic—desire to support their friend’s well-being, with the result that they prematurely appropriate their friend’s agency. Expressing care by “leaping-in” for a friend typically involves one friend taking over in concerning or troubling situations (Heidegger 1927/1962: 122). According to some, such actions are paternalistic (see Tsai 2014). Expressing one’s care and concern in this way misfires because it represents, and is also often received as, an expression of disrespect for their friend’s agency. Although it is conceptually possible that paternalism can occur even when it is not received as such, the “conflict” that we discuss here will only result when paternalistic actions are received by the acted-for friend as unwelcome, usually because they are received as disrespectful.

In other cases, a friend may fail to appropriately express care and concern in a given situation by waiting “far longer than [they] should to help, out of a misplaced sense of respect and deference” (Vallor 2015: 119). Being aware of the risk of paternalism can block appropriately timed actions and cause conflict within a trusting relationship. In such cases, the action itself is not rejected by the friend, but the timing of the action is unwelcome. The acted-for friend would prefer that their friend had acted earlier and interprets their period of inaction as a temporary failure of care which they may attribute to, for example, a lack of attentiveness, attunement, or responsiveness in their friend.

There are also attitudinal failures relating to how the acted-for friend receives the other person’s concern. Such attitudinal failures may be marked by an inappropriate reaction by the acted-for friend. For example, a friend might receive an action as unwelcome in the moment, experiencing a psychological reaction about their friend seeing a vulnerability that they themselves are not quite ready to fully acknowledge. After some reflection, the acted-for friend recognizes that their friend used their discretionary power in not only an appropriate but a virtuous way, and they retrospectively welcome the act. Nonetheless, their defensive or even retaliatory reaction is likely to have been hurtful to their friend and, at least temporarily, damaging to the relational ecology of their trusting relationship because it frustrates their friend acting in this helpful way in the future.

3.3. Severity

We have differentiated between “misunderstandings” and “conflicts” as two different types of relational rupture within friendships. Yet there are also parallels across these types of rupture with respect to their range of severity. This has implications for the opportunities that ruptures present for relational growth (as we argue in section 4).

Most simply, misunderstandings and conflict can be the unintended consequence of a friend’s well-meaning attempt to act to promote their friend’s well-being, but where the risk taken backfires. We label these “innocent ruptures” to capture mistakes that do not indicate anything more generally about the state of the trusting relationship. While these types of misunderstandings and conflict can have a negative impact on the friends’ relational ecology, the resultant rupture is likely to be both relatively minor and temporary, if one occurs at all. Nonetheless, innocent ruptures can produce downstream effects for the relational ecology of trust and thus have a causal link to more severe relational ruptures.

Some misunderstandings and conflicts issue from an underlying, possibly unrecognized, damaging dynamic in the relationship. A rupture can indicate that something else has gone wrong or is in the process of going wrong in the relationship. These are “harbinger ruptures” in our terminology because they signal that the underlying state of the trusting relationship has shifted or is in the process of shifting in a negative direction. These changes may have been faintly perceptible to one or both friends but usually haven’t been experienced in any concrete sense before the rupture.

Depending on friends’ responses at this time, harbinger ruptures may instigate a process of relational growth (more on this in the next section) or feed into and accentuate the damaging dynamic that caused them. Harbinger ruptures are standardly caused by, and hence signal, a subtle change to the relational ecology of trust and thus to the boundaries of the discretionary power held by friends within the relationship. Navigating these subtle, often unspoken shifts is difficult; the stability which friends have taken for granted may be shaken by the occurrence of a harbinger rupture and what it reveals (or is taken to reveal) about the friend or the state of the friendship, and further misunderstandings and conflict can occur as a result. At its most extreme, this can lead to a dynamic that produces the most severe form of harbinger rupture which signals the potential end to the friendship.

On our relational ecology account, although it is possible that a severe rupture can occur “out of the blue” within a trusting relationship, it is much more likely that these severe, even friendship-ending ruptures are the result of a dynamic and cumulative process in which trust and its bases are eroded over time. The linguist Deborah Tannen, in her work on the language of women’s friendships, puts the point thus: “Even if a cutoff can be traced to a single moment—a cruel thing said or outrageous thing done—that supremely tellable violation usually is the climax of frustrations and disappointments that had been building over time” (Tannen 2018: 172). We see this dynamic at play in the Issa and Molly case: true, Molly’s action of sending Nathan away caused a rupture because she acted on her misunderstanding of Issa’s priorities with respect to her romantic relationship; but the series emphasizes that this rupture is rooted in a more subtle process of Molly and Issa slowly growing apart, and coming about as a misplaced attempt on Molly’s part to reestablish the friendship’s intimacy.

Finally, some ruptures are so severe that describing them as misunderstandings or conflicts seems like a mischaracterization. Our characterization of misunderstandings and conflicts assumes that these are well-meaning missteps which are due to the often murky and unstable space of intimate interpersonal relationships. Misunderstandings and conflicts occur when friends are attempting to act for each other’s well-being but misunderstand what a friend needs or mis-express their care. In other words, these missteps are mistakes, not intentional harms or wrongs. If a friend were to intentionally harm another friend, or to wrong them by deceiving them, violating their privacy, or breaking a promise towards them, then we would agree with Brake (2025) that these constitute “friendship wrongs”: wrongs facilitated by the ongoing proximity and intimacy which characterizes these relationships.

For instance, if in the Insecure episode, Molly had sent Nathan away because she knew how much Issa wanted to see him and she wanted to ruin Issa’s birthday, this would not be a misunderstanding or a conflict, and it would not offer an opportunity for relational growth. Rather, it would be a friendship wrong in that Molly would have taken advantage of her intimate knowledge of Issa and her access to Issa’s life to hurt her. Because our article is intended to defend what we believe is a counterintuitive thesis—that ruptures in friendships can strengthen them—we set aside this comparatively intuitive thesis that “friendship wrongs” categorically damage friendships.

In the next section, we consider how actions which are intended to benefit a friend, yet misfire into misunderstandings or conflicts, may nevertheless present essential opportunities for friendships to deepen and grow.

4. Ruptures as Opportunities for Relational Growth

We claim that the kinds of ruptures to the relational ecology of trust described above can benefit friendships. This seems like a counterintuitive idea because the ethics of interpersonal actions are often considered synchronically, where their normative features are fixed to the point in time at which they occur (cf Brake 2025). Yet acknowledging the diachronic nature of intimate relationships can render the ethical analysis of single acts incomplete.⁴ A more dynamic approach allows us to explore the possibilities for relational growth—which we understand as an increase in mutual understanding and mutual care—that otherwise would be overlooked by a synchronic ethical analysis. This is not to take away from a discussion and a verdict about whether some act was harmful or morally wrong within the context of a trusting relationship; rather, it is to situate this act within its broader context, both in terms of the dynamics that may have brought it about and in terms of the further dynamics that it may set in motion, so that a more complete ethical analysis can be offered.

We agree with Annette Baier in holding that “a complete moral philosophy” should be able to “tell us how and why we should act and feel toward others in relationships of shifting and varying power asymmetry and shifting and varying intimacy,” not only how we should act and feel when things are equal and stable between the two actors (1986: 252) and in holding that trust plays a key role in understanding the ethical landscape in these (more complex, but also more realistic) kinds of cases.

More specifically, we argue that the processes of bringing to the surface underlying problematic dynamics within the relationship, addressing ruptures, and engaging in relational repair have important roles to play in growth within trusting relationships. Although the two parties do need to reassess the basis for their friendship in terms of knowledge and care in the wake of a rupture, their relationship need not be permanently damaged by it. In fact, as Sow and Friedman have grasped, friends “can’t stay truly connected without some level of misunderstanding or conflict” (2021: 175; emphasis added). A relationship in which these ruptures did not occur would be naïve, because intimate knowledge and care are never perfect and because trusting relationships are, as we observed in section 2, at once stable and dynamic.

While trusting relationships are based on a bedrock of mutual confidence in one another’s epistemic and attitudinal commitment to the relationship, it is an unavoidable fact that both parties are continually changing, to a lesser and greater degree, and that the nature of their relationship—and hence the contours and subtle dynamics of their relational ecology of trust—is continually changing, to a lesser and greater degree, too.⁵ These facts create the possibility of misunderstandings and conflicts. A commitment to maintaining the mutual understanding and caring practices constitutive of companion friendship under these dynamic conditions will require some trial and error.⁶ This is the unavoidable work of maintaining and developing a valued connection with a friend. Doing this ‘work’ is morally risky: as the friends act for each other, they take on risks and will make mistakes (as we discussed in section 2), and the ruptures brought about by misunderstandings and conflict affect the relational ecology of trust negatively—at least to some degree and for some amount of time (as we discussed in section 3). But doing this ‘work’ is also the only way to “stay truly connected.” Even when the risk backfires and a friend gets it wrong when acting for their friend’s well-being, the resultant rupture can be part of the process of not only maintaining, but also developing the connection and trust between friends by providing an opportunity for friends to better calibrate their understanding of the other in the future and to demonstrate their commitment to the relationship.

To explore this idea further, in this section we first describe how the severity of the rupture plays a role in the types of opportunities for relational growth offered by a rupture and then we examine how the responses of both friends to a given rupture affect whether they take up the opportunities presented by ruptures.

4.1. Opportunities Created by Ruptures

The opportunities for relational growth presented by ruptures depend, in part, on the type of rupture. In terms of a rupture’s severity, misunderstandings and conflict that produce what we’ve called innocent ruptures provide opportunities to maintain and update intimate knowledge and to respond in ways that stave off any potential negative downstream effects for the relational ecology of trust that might otherwise develop if these relatively innocuous ruptures are left unacknowledged and/or unaddressed. However, innocent ruptures do not offer significant opportunities for relational growth; they do not say anything about the state of the relationship more generally, and hence less can be learned from them. Even so, the kinds of relational skills and virtues that friends practice in the much lower-stakes case of innocent ruptures—such as humility, attentiveness, seeking and granting forgiveness, and so on—feeds directly into their ability to act on the opportunities presented by other, more severe ruptures such as harbinger ruptures.

Misunderstandings and conflict that produce harbinger ruptures offer genuine opportunities for relational growth because, by signaling that the underlying nature of the trusting relationship is shifting or has already shifted in a negative direction, they bring to the surface problematic dynamics within the relationship. This process can be important for protecting the friendship’s relational ecology of trust from further potential decline, as it gives friends a chance to first recognize and then respond to the dynamic that has been highlighted by the rupture. The rupture can lead to a pause in the ‘normal’ proceedings of the friendship, allowing friends some time to reflect on and reexamine the nature of their current relationship and the health of the dynamics with it. This process may lead to one or more of the following happening as a result: (i) further discussion and increased (or at least adapted) mutual understanding relating to the changes experienced by the friends and the effects of these on their relationship, (ii) breakthroughs with respect to friends’ vulnerability with each other, (iii) additional clarity on the scope and contours of the discretionary responsibility granted within the trusting relationship, (iv) a restoration of appropriate epistemic and moral humility, (v) a resetting of unhealthy dynamics, and (vi) a confirmation of commitment levels.

For example, when responded to appropriately (more on this in the next section), misunderstandings offer opportunities for friends to recalibrate their intimate knowledge of each other and to reevaluate any incorrect assumptions they had or mistaken inferences they had drawn from their friends’ behavior. This involves demonstrating, first, an epistemic humility and a willingness to admit that one was not acting on a correct understanding about their friend’s needs and wants and, second, a commitment to revising one’s current understanding in line with what has been revealed by the rupture and their friend’s reaction to it. Here, we understand epistemic humility to mean an acknowledgment that your interpretation of the other person could be or has been wrong. This is different from open-mindedness, which seems to convey a willingness to be proven wrong, which does not have the same relational quality. Given that understanding of the other is a basis of acting for the other, epistemic humility that one’s understanding could be wrong is, in this context, fundamentally relational. Exercising this humility means being cautious about acting for other another.

In a similar fashion, conflicts allow friends to reevaluate the degree of vulnerability they have been and currently are willing to express in their relationship and to reassess their willingness to take risks to care for each other and to receive care. In section 2, we explained that while friendships offer unique benefits, they also come with unique risks. By allowing another person the intimate knowledge and proximity of a friend, along with the possibility of acting for your well-being, you make yourself vulnerable to the potential harms delineated in section 3, of your friend misunderstanding what is good for you or exercising inappropriate care for you. Likewise, to be in the position of a friend is to be close enough to another person to act on necessarily imperfect knowledge of the other. To act for another person as a friend is to be vulnerable to “getting it wrong” and potentially rupturing the friendship—a source of well-being in your life—in an irredeemable way.

As with misunderstandings, the benefit of a conflict is the opportunity it provides for friends to reevaluate the types and levels of vulnerabilities that each friend is willing and able to express in the relationship. For instance, perhaps Issa is willing to be vulnerable to Molly in a variety of ways—for advice about her career, help with housing, and care for her emotional states. But Issa may not be willing to allow Molly the discretionary power that would make Issa vulnerable to Molly’s actions in her romantic relationships. Or, perhaps Issa is willing to discuss her romantic relationships with Molly— making herself vulnerable to harm through Molly’s input or advice—but is not willing to allow Molly to intervene in her romantic relationships directly. The first point refers to the type of vulnerability, while the second point refers to the degree of vulnerability. As with Issa and Molly, such conflicts between friends allow them to examine and, if needed, to redraw the boundaries each sets around each other’s discretionary powers within the relationship and their attendant vulnerabilities to each other. This strengthens the trusting nature of the relationship by clarifying the dimensions of the mutual understanding and mutual care to which each are committed in the friendship, thus restoring one another’s confidence that they have each made commitments which they can keep.

But what about the very severe cases of harbinger ruptures that, all other things being equal, signal the potential end of the friendship? Can these offer the same kinds of opportunities for relational repair and growth? In principle, yes, but they are unlikely to do so in practice. This is because, as we noted above, such severe ruptures are likely to be the result of a cumulative process in which friends have been presented with opportunities for relational growth from innocent and less severe harbinger ruptures but have failed to take up these opportunities at earlier points. Such friends are in a weaker position to take up any present opportunities, for at least two reasons. First, their trusting relationship is likely to be in a critical condition, with previous ruptures having eroded trust, and the mutual understanding and caring practices that go with this, over time. Second, it is likely that they have forgone opportunities to practice the skills needed for the required relational repair.⁷ Yet just because this task is difficult does not mean that it is impossible. Its success depends on how the two parties respond to the rupture, to the opportunity it offers, and to the relational repair efforts of the other friend, as we address below.

In this section, we have argued that—within the context of trusting relationships—opportunities for relational growth are provided primarily by standard harbinger ruptures, as opposed to either innocent ruptures, on the one hand, or very severe forms of harbinger ruptures, on the other. Yet innocent ruptures provide opportunities for low-stakes relational repair that play a formative role in enabling friends to take up the opportunities for relational growth that may be offered in more high-stakes situations of misunderstanding or conflict. And, although it is more difficult to take up and successfully navigate the opportunities offered by severe harbinger ruptures, such opportunities are still available to friends. We discuss this issue of taking up opportunities and thus their successful realization in the form of relational growth in the next section, along with this growth’s dependence on the bidirectional, interactive responses of both friends.

4.2. Right Response and Relational Growth

Whether these opportunities are taken up or not depends on how each of the friends responds both to the rupture and to the efforts of the other to repair the harm done by it. When friends successfully engage in relational repair efforts, this can reestablish and even strengthen the confidence in each other’s commitment that was knocked by the occurrence of the rupture.

Clearly, the friend who got it wrong when acting for their friend’s well-being must respond in the right, or at least an appropriate, way to their error. Most fundamentally, this involves acknowledging the fact that they have overstepped some moral boundary within the trusting relationship and issuing a form of apology. Importantly, however, the friend who made the mistake is not the only person who must respond in the right way. Echoing the key insight of the relational ecology account, Sow and Friedman note that “[o]ne person can’t unilaterally decide to fix a friendship. Repair is a choice that two people have to make” (Sow and Friedman 2021: 158; emphasis in original); making this choice will involve relational work for both parties.

In some instances, this can involve the acted-for friend taking the initiative. For example, the friend who got it wrong may have failed initially to recognize that their behavior caused, or was responsible for causing, a rupture. In such (most likely, relatively minor) cases, the acted-for friend may need to find a way of communicating the impact of this action on them and, thus, on the friendship’s relational ecology of trust. This is an additional relational burden, especially when one is feeling hurt not only by the unwelcome action but also by their friend’s lack of awareness about its effects. This communication carries its own risk, as one cannot always know how one’s friend will react and, thus, whether communicating will lead to a process of relational repair and deepened understanding or to a deeper rupture. But this is a way in which the acted-for friend can demonstrate their commitment to the friendship, since a consistent failure of friends to engage in explicit repair can lead to a situation of accumulated bad feelings that erode the mutual understanding and mutual care and concern of the trusting relationship, leading to more severe ruptures.

Instigating a discussion about the rupture may also highlight to the acted-for friend that the misunderstanding was not necessarily or solely due to their friend’s inattentiveness, complacency, or a lack of care and concern. For example, it may highlight to the acted-for friend that there were in fact some reasonable grounds for the friend’s epistemic or attitudinal failure in this instance, which may have something to do with things that they themselves are going through and about which they have not communicated.

Whether by their own recognition or via the communicative process outlined above, the friend who got it wrong responds appropriately by exercising epistemic humility and using the misunderstanding or conflict as a chance to learn something about their friend and/or a dynamic (or set of dynamics) affecting their friendship. In the case of a misunderstanding, the friend can use the rupture, and discussions that follow from it, to better calibrate their intimate knowledge of their friend and, in so doing, to deepen their friendship by enriching one of its constitutive features. And yet, the repair is unlikely to be completed by discussion alone. In many cases, the friend who got it wrong may also need to express their attitude of epistemic and moral humility by asking for permission when attempting to act for their friend in the future. We suggest that seeking permission is an important way in which a person can demonstrate their recognition that their action overstepped a moral boundary and has altered the bases of trust within the relationship.

Permission-seeking is a significant step in repair because it recognizes that the relationship has changed. Whereas prior to the rupture, each friend was confident in each other’s commitment to the relationship and believed that they both understood the terms and bounds of that commitment, a rupture upsets this confidence. Friends in this context may wonder if the other person really knows them and cares about them. This uncertainty renders null the de facto discretionary power that each friend enjoyed by virtue of the friendship, as the characteristics of the friendship are now under reevaluation. Just as a bank account may be frozen while significant changes are made, so is the discretionary power of friendship “frozen” during this reevaluation process. To seek permission when acting for a friend is thus to recognize that one’s understanding of and care for the friend are now under question—if not regarded with outright suspicion—and so one’s discretionary power to act without permission has been suspended. Furthermore, in recognizing that their discretionary power has been temporarily suspended and in asking permission, the friend restores trustworthiness by displaying their understanding of the situation and their willingness to step back from their discretionary power as the relationship is reevaluated.

Permission-seeking is thus one of the ways in which the friend can demonstrate their trustworthiness to the other person. On our view, such permission-seeking practices are remedial: they restore the friendship to the kind of trust in which they are no longer required (for those actions that fall within the discretionary power granted them by the trusting relationship).

While ultimately, these “right responses”—of apology, forgiveness, epistemic and moral humility, and permission seeking—improve the friendship and allow both friends to grow within it, friends might also resist the self-revision that is necessary to become a better friend in the wake of a rupture. This is because acknowledging a misstep and being willing to change as a result of it, requires vulnerability—arguably, a greater degree of vulnerability than existed in the friendship initially. As we noted previously, friendship is a trusting relationship, and one of its constitutive elements is vulnerability. In Baier’s language, “we must allow […] other people to get into positions where they can, if they choose, injure what we care about, since those are the same positions that they must be in in order to help us take care of what we care about” (Baier 1986: 236). It is due to this essential risk of friendship—allowing others into intimate proximity—that friends might resist vulnerability after a conflict. Yet doing so is essential to friendship, and while a conflict may highlight that one of the friends had been resistant to vulnerability, that friend must be willing to exhibit vulnerability if the friendship is to continue stronger than before.

Indeed, it is easier to be a friend when things are going well than it is to navigate a friendship when things are going poorly. Following a misunderstanding or a conflict, both parties in the friendship have a role to play in reestablishing the trust that comes from mutual vulnerability. For example, the friend whose misstep in providing care resulted in a conflict must exercise humility and a willingness to change, taking a risk and thus exposing themselves to the uncertainty of their friend’s response. But the friend who was harmed must not close themselves off to their friend’s attempts at care. They may need to exhibit a greater willingness to be vulnerable—to allow other people to enter the intimate sphere of interpersonal life in which harm is possible.

This picture fits nicely with Kim Atkins’ account of the mutuality of forgiveness. On standard accounts of forgiveness, Atkins writes, “It is the forgiver who exercises his or her will or compassion in overcoming resentment; the forgiver who decides what deserves forgiveness; and the forgiver who manages to change his or her perception of the agent into an appropriate object” (Atkins 2002: 125). While it is true that, for instance, the misunderstood friend has an important role in the repair and growth of the friendship insofar as they must forgive the misunderstanding, it is also true that recognizing the agency of the friend who made the error requires acknowledging the role they must play in bringing about forgiveness by enabling the resumption of trust.

In this section, we have focused on the opportunities presented by relational ruptures and the types of responses that can strengthen the relationship, rather than allowing the rupture to consume and destroy the relationship. To put our argument in terms of the relational ecology metaphor once more: just as forest fires are a natural and necessary part of the ecosystem, working to support overall health and growth, so must there be similar periods of “rupture” in trusting relationships. In these periods, friends are provided with an opportunity to reexamine the nature of the relationship and the health of the dynamics within it, so that, for instance, any unhealthy dynamics can be reset, commitment levels can be confirmed, appropriate humility can be restored, and additional clarity on the scope of discretionary responsibility gained.

Somewhat ironically, the healthiest kinds of trusting relationship are not those in which breaches of trust do not occur. In fact, we suggest that such “ruptureless” friendships are lacking a central part of the maintenance and growth of healthy trusting relationships that is provided by the periods of instability and reexamination that ruptures can bring about. Approaching the ethical analysis of such ruptures diachronically brings a new perspective on this: even while a rupture damages the relational ecology of trust, it can provide opportunities to protect and develop the health of the friendship, and hence ultimately have a positive effect on the relationship overall.

5. Conclusion: Going Wrong in Order to Go Right

In this article, we have argued that two classes of missteps within friendships—misunderstandings and conflicts—are important to the relational growth of friendships’ constitutive features of mutual understanding and care. Some may disagree with our proposal. One might argue that friends should only act for each other’s well-being with explicit permission and full transparency, lest they risk unjustified paternalism. However, as our explanation of what it means to be a friend highlights, drawing such a firm boundary seems to foreclose the possibility of friends being in true trusting relationships, in which both parties accept vulnerability to the feelings, ideas, and actions of their friend. A relationship in which neither person retained any kind of discretionary power to act for the other would not be an intimate or trusting relationship, on our account. While this invariably comes with risks—some friends do act paternalistically—it also is the source of the unique, constitutive good of friendship identified by other scholars. Furthermore, even when these risks are borne out—when friends do misstep in acting for each other—we have argued that these missteps create significant opportunities for relational growth. The relationship just described, in which neither friend acted for the other without explicit approval, is one with few risks, but also with few opportunities. As the saying goes, growth happens through two steps forward, and one step back. In this paper, we have identified a significant way in which a step back in a friendship as a trusting relationship facilitates the growth of two steps forward.