What’s Bad about Friendship with Bad People?

Abstract Is there something bad about being friends with seriously bad people? Intuitively, it seems so, but it is hard to see why this should be. This is especially the case since some other kinds of loving relationship with bad people look morally acceptable or even good. In this paper, I argue that friendship inherently involves taking one’s friends seriously, which involves openness to their beliefs, concerns, and subjective interests. Deeply immoral views and attitudes ought not to be taken seriously or considered as options, and I argue that this explains why being friends with bad people is itself morally problematic. I go on to contrast this with Jessica Isserow’s (2018) explanation of what’s bad about friendship with bad people, and I suggest that my account is better placed to explain why friendships with bad people are morally problematic but some other loving relationships with bad people are not.


Introduction
It's not unusual to think badly of individuals who are friends with seriously bad people. Such friendships, we tend to feel, reveal something unappealing even about otherwise good individuals, and we feel uneasy about them. 1 More specifically, there seems to be something morally troubling about friendships with bad people; it's common to feel that those who are friends with bad people have thereby made questionable choices in entering into the relationship, or perhaps that there's something morally bad about simply being friends with the bad person. For example, news coverage of the disgraced financier Jeffrey Epstein, a sex offender revealed to have been involved in extensive sex trafficking, has suggested that his close friends were tainted too. Such reports implied that there was something bad about their being friends with him. Some of those cases are complicated by concerns that his friends were complicit in his wrongdoing or even active wrongdoers themselves, but even those who were not complicit in his wrongdoing seem somehow sullied by their friendship with him. Similarly, think about those who are friends with viciously racist people. Such friendships seem to reflect badly on them. Intuitively, the mere fact of being close friends with someone deeply immoral makes us uncomfortable.
Philosophers thinking about friendship recently have been interested in the relation between friendship and morality. Many have been intrigued by Aristotle's suggestions that true friendship requires the participants to be virtuous, and more or less equally so, 2 and there has also been much discussion about whether friendship can require us to do immoral things. 3 Despite this, the question of whether there's something bad about friendship with bad people has been almost wholly overlooked. This is surprising, since a full ethic of friendship seemingly requires us to ask not only what friendships require of us as friends, but also to consider who we ought to be friends with at all, and why.
Our unease about friendships with bad people is particularly noteworthy given that we do not tend to think that all other positive relationships are ruled out by one party being morally bad. For example, we do not usually think that there is anything bad about a parent continuing to love their child even if their child is deeply immoral. On the contrary, we might even be morally concerned about the parent who failed to love their child in such a situation. 4 Similarly, we would perhaps not be worried about a mentor who was deeply invested in and cared about their mentee even if the mentee were a seriously bad person. Our concern, then, does not seem to be about close or loving relationships in general, but rather is fairly specifically about friendship (as well, perhaps, as some other relationships). 5 Why should this be the case? Initially, it may seem somewhat unfair to judge someone negatively for someone else's (i.e., for their friend's) moral failings. Indeed, one might even worry that this is an irrational tainting of our views about the good friend, with our disapprobation unjustly spilling over from the appropriate target (the bad person) to an innocent one (the friend) through their mere association. However, I will argue that the intuition that the friend of a bad person is, as such, doing something bad (in a certain respect) is well grounded; friendships with bad people are in a certain respect morally bad. What grounds this intuition, I will argue, is that friendship involves taking one's friends seriously, a term I will use to refer to practises such as taking one's friend's beliefs and attitudes as prima facie valuable and choiceworthy. Taking one's friends seriously, I will suggest, is partly constituted by openness to the other person's subjective concerns and perspective. Since some moral views and perspectives should not be taken seriously or considered as options, friendship with bad people is itself morally bad in that respect (though not necessarily wrong overall). 6 In this paper, I will begin by considering an explanation of the badness of having bad friends based on the popular idea that friendship involves openness to having one's character shaped by the other. I will then offer reasons to think that this cannot be a general account of the badness of friendships with bad people. In section 2, I will draw from it the idea that friendship involves taking one another seriously and show how this feature of friendship explains what is bad about friendship with bad people. In section 3, I discuss a potential tension between the good of loyalty and the badness of friendship with bad people, suggesting that these are compatible. Finally, in section 4, I compare my account of the badness of having bad friends to Isserow's (2018) account. I suggest that my account is better placed to explain why it is friendship with bad people, rather than positive or loving relationships in general, that is ethically questionable.
Before I begin to argue for this, I want to make four clarifications about my argument. Firstly, I am thinking about friendships with individuals who have serious moral deficiencies, not merely minor flaws-unrepentant racists and misogynists, for example, rather than those whose humour may sometimes err towards the uncomfortable. 7 I take it that it is only in such cases that the friendship seems intuitively concerning. Secondly, I am talking exclusively about friendships that are deep and close, not merely casual friendships or relationships with acquaintances such as those with whom one occasionally chats at social gatherings. Thirdly, I am assuming that such deep friendship involves significant knowledge of the other person, and thus that the friend to a bad person is aware of the objectionable views and attitudes that the bad person holds. Finally, I am thinking about friendships where the bad person is an otherwise good friend to the other person (they are kind, supportive, pleasant to be with, and so on). I am not exploring instances of friendship where the bad person's negative traits impact their treatment of their friend. Such relationships are problematic for obvious reasons: no one ought to be treated in these ways. 8 My claim, however, is that there can be something morally bad about a friendship even if the participants treat one another well.
My conclusion is thus limited to the claim that there is something bad about deep friendships with seriously bad people, though I will not continue to make these clarifications. Again, I take it that this claim is intuitively plausible. The challenge, I take it, is to explain why this should be the case, and I turn to do so now.

Friendship and character-shaping risks
In asking why friendship with bad people might be problematic, a good place to start seems to be our general conception of friendship. Friendships are characteristically close, intimate and trusting relationships, and this has led various recent philosophers to claim that friendship involves openness to having one's character shaped by one's friends. Dean Cocking and Jeanette Kennett, for example, write: [A]s a close friend of another, one is characteristically and distinctively receptive to being directed and interpreted and so in these ways drawn by the other. (2000, 503; emphasis added) Alexander Nehamas makes a similar claim: Friends have a privileged role in this lifelong process of self-construction … [W]e give them the power to lead us to need, desire, and plan things that we can't possibly anticipate. (2010,288) On these philosophers' conceptions of friendship, friendship constitutively involves being "drawn" or "constructed" by one's friends, or at least openness to this. 9 Their thought is that friends play an important role in the construction of the self: we give our friends' interpretations of us a certain authority over us and use them not only as passive descriptions of who we already are but also as guides to direct the process of constructing the person we will become. This idea could be used to ground one kind of worry that we might have about being friends with bad people. 10 If friends shape or construct each other, being friends with bad people involves bad people shaping our future selves (or at least, being in a position to do so). Allowing bad people to shape our future selves would seem unwise given their bad character, and one might worry that they would lead one morally astray. This worry is gestured at by Nehamas: [N]othing ensures that our relationship won't harm or degrade my judgement itself, making me feel happy to have become someone I would have hated to be, perhaps rightly, had [the friend] not come into my life. (2016, If this is correct, friendships with bad people risk one's moral character. They entail risking one's character being shaped in immoral ways in accordance with the friend's mistaken attitudes and ideals. 11 Such a view does highlight one concern that we might have about friendships with bad people. We may worry that their friends' callousness, casual misogyny or cruelty will "rub off" on erstwhile good people, leaving them morally worse for the friendship. Nonetheless, this response seems inadequate to explain the intuition that being friends with bad people is itself morally problematic. Even if friendships involve significant shaping of one friend by the other, the morally good friend might reasonably hope that the influence will instead work the other way around: they might hope that the bad person's character will be improved by the friendship, rather than that their own will deteriorate. Perhaps, then, the moral risk would instead be a noble thing to take on-bad people's friends might risk moral degradation for a morally worthy cause (morally helping the bad person). Alternatively, the risk of moral degradation might be something that one could offset by also being friends with many good people: one might judge that by being friends with many virtuous people, their influence on one's character would outweigh the pernicious influence of the vicious friend. The moral risk one accepts in such a friendship could thus be minimised by offsetting it with various other influences. The idea that being friends with bad people is bad because it involves risking one's moral character thus seems unsatisfactory as a general account of the badness of being friends with bad people. Though this is undoubtedly a worry we sometimes have in such scenarios, the risk can be neutralised in various ways.
Importantly, this explanation of what's bad about having bad friends also seems questionable on a deeper level; it depends on an implausibly strong interpretation of the extent to which our friends shape us. Friends, as such, need not be granted free rein to shape one another's character. It is not plausible that friends, as such, are always open to having their moral character shaped by one another, since it seems to be at least possible to have significant moral disagreements with one's friends. Friends-even longstanding friends-may plausibly even appreciate and value their differences from one another. The idea that merely because someone is one's friend, one risks becoming similar to them with regard to goodness seems to be an overstatement, and plausibly a risk that one could guard against. Individuals do retain some control over their character despite their friends' influence.
Still, there is an idea in this general region that seems right and that offers the beginning of an account of what is bad about friendships with bad people. A weaker interpretation of the kind of "shaping" mentioned above seems highly plausible. Cocking and Kennett, for example, state that one is characteristically receptive to being shaped by one's friends, which seems to allow that there remains a significant agential role for individuals to play in their own character development within friendships. As I noted above, whilst friends do often influence who we become, they are not simply given free rein to shape our characters; rather, in becoming friends with someone, their beliefs and attitudes become salient to us as possibilities. The idea central to my account, then, is that our friends shape the options which we take seriously. I will now turn to examine this idea further, and I will argue that it explains what is bad about being friends with bad people.

Friendship and taking one another seriously
The striking feature of friendship on which I'll focus, then, is that friends take one another seriously; friends take one another's interests, projects, and beliefs as at least prima facie reasons to believe, care, and act in similar ways. 12 If, for example, one's friend has a deep love of opera, that gives one reason to attend a performance and to be at least open-minded about its value. One might attend the opera and decide, on balance, that it's not to one's taste. But the friend's caring about it gives one reason to consider it as a potential option, and to need reasons to reject it: their caring gives one reason to see it as a choiceworthy option. It also gives one reason to remain open to the possibility of opera being valuable even if one has disliked it, and to acknowledge the respects in which one might have failed to appreciate it properly: one might, for example, feel put off by its melodramatic tone, but recognise that the friend's practised ear is able to discern subtleties that one missed out on. 13 Similarly, if a friend has firm political convictions that matter a lot to them, their conviction itself gives one reason to take that political position seriously, and to need a reason to reject it. Failing to take a friend's beliefs and attitudes seriously in this way would seemingly be a failure of friendship: a friend who simply dismissed one's perspective, attitudes, and beliefs out of hand would seem to be a very poor friend, if a friend at all. 14 This conception of friendship seems highly intuitive, and it has been assumed by various philosophers writing on friendship. Alexis Elder, for example, connects the requirement that friends take one another seriously with the idea of "closeness." Closeness, she claims, is partly constitutive of friendship. She understands it as follows: [C]loseness involves taking the other's perspective and subjective interests as prima facie valuable, even where one would ordinarily be left cold by these considerations. (2014,89) That is, Elder understands closeness with one's friends as a matter of taking their perspective and subjective interests (i.e., those things they conceive of as their interests) seriously. And, again, she regards this closeness, and thereby taking one's friends seriously, as a constitutive part of friendship itself. Elder is not alone in this view; Marilyn Friedman, for example, suggests something very similar: [F]riendship provides us with an inclination or invitation to "take our friends seriously," and to take seriously what our friends care about. Friendship opens us up to considering new 12 The phrase "taking one another seriously" is used regarding friendship by both Friedman (1989) and Cocking and Kennett (2000). Both take it to be an important element of friendship. 13 Being a good friend to the opera lover thus need not entail being willing to attend the opera if one has previously disliked it, especially if one has attended numerous times. But taking the friend's interest seriously might involve seeing it as a potential option (albeit a potential option ruled out by negative experience), regarding it as a choiceworthy activity, acknowledging that other people may discern more in the music than one does, and so on. 14 I am stopping short of the claim made by some epistemic partialists that our actual beliefs ought to actually be influenced by our friends' beliefs. Rather, my claim is merely that friendship involves at least considering those beliefs (attitudes, values, etc.) that are important to one's friends. possibilities, and to considering old possibilities in a new light-admittedly without predetermining what will result from the consideration. (1989,5) Again, Friedman here suggests that friendship involves being open to new possibilities through taking seriously those things that our friends care about. Whilst she allows that this does not entail that one will simply adopt the friend's cares and interests, she nonetheless suggests that friendship involves the "inclination" or "invitation" to do so. 15 Taking one's friends seriously, then, is a matter of taking their projects, beliefs, and attitudes as prima facie valuable or as prima facie choiceworthy. This is not to say that one needs to actually take on one's friends' beliefs or adopt their attitudes. But it does involve taking them to be prima facie valuable, needing a reason to reject them, and being open to the possibility that there might be something in them.
A background reason to think that taking another's attitudes and perspective seriously is crucial to friendship is the idea that friendships are paradigmatically equal relationships. This is unlike love, which can be one-sided, or a relation like parenthood, where the parent's relation to the child (namely, parenthood) can differ radically from the child's relation to the parent. As Laurence Thomas notes, friendship demands a certain kind of equality; it requires that "neither party to the relationship is under the authority of the other" (1987,221). 16 Friendship seemingly requires viewing the other party as someone who has fundamentally equal standing to oneself, and thus as someone whose general perspective is prima facie trustworthy and worth taking seriously.
The equality requirement in friendship can be contrasted with other close relationships such as parent-child relationships. In those relationships, the parent can seemingly have a fully loving relationship with the child without assuming that the child's perspective is trustworthy or worth taking seriously. A parent need not take seriously, for example, their child's fondness for bright clashing colours or their love of highly processed, sugary foods; they do not need to work with a presumption that their child's preferences are revealing of anything important. This does not seem to be specific to familial relations involving children either: one can plausibly have loving familial relationships with cousins, aunts, grandparents, and so on without taking their perspective seriously. But friendship seems very different, and taking one's friends to be one's equals (again, a constitutive part of friendship) plausibly does require taking one's friends' perspectives to be prima facie valuable. Dismissing one's friend's love of opera in the way one might be justified in dismissing one's child's love of clashing colour would seem to be a failure of friendship, and a friend who was dismissive of one's general perspective would seem like no friend at all. 17 One might, of course, regard a friend as unequal to oneself in some limited domain such as gardening or cooking and either defer to them or reject their views there. Friendship does not require taking every belief or attitude of one's friend seriously. But what is important for my 15 Koltonski also suggests something similar with his discussion of caring about friends as agents (2016, 475) and sharing friends' ends (2016, 476). Cocking and Kennett, whose work on friendship has been very influential, also write: The interests of the other in friendship, whether serious or slight, are not, in general, filtered through one's antecedent tastes and interests or subjected to rational or moral scrutiny before they acquire action-guiding force. (2000,285) Though their focus is on how friendship affects the friends' behaviour, this also seems like an affirmation of the idea that friendship involves taking one's friends seriously and lending a special weight to their perspective that it would not otherwise have. It is, however, a significantly stronger claim than those put forward by Elder and Friedman, and a stronger claim than is necessary for my argument. 16 Thomas is not alone in this view. Though little-discussed in the literature, the idea that ideal friendship involves some kind of equality between the participants has been unanimously endorsed by those who mention it. See Marilyn Friedman (1989Friedman ( , 1993, Joseph Kupfer (1990), Daniel Koltonski (2016), and Uri Leibowitz (2018). 17 Of course, respect for others' choices and projects should be extended to all regardless of our relationships with them. However, the notion of taking seriously that I am working with here is considerably more demanding than respect. argument is that one must view one's friends as equals overall, or as people with equally valid overall perspectives. Friends' overall perspectives and their deeply held or overarching beliefs and attitudes cannot be dismissed or rejected out of hand. Those features of one's friends that are important aspects of their overall character and perspective should be taken seriously. Crucially, this importance can be objective or subjective: things can matter because they matter to the friend or because they matter objectively. It might be acceptable, for example, to dismiss a friend's passing interest in opera, but not to dismiss their deep, sustained care about it and commitment to it. Similarly, gardening skills do not usually affect the way in which one views a friend overall. However, it is not plausible that moral character could be hived off in this way, particularly when the moral views and attitudes at issue are serious ones. Moral character seems an objectively important part of one's overall perspective, and thus does not seem to be the kind of thing that friends could simply dismiss. More specifically, any trait sufficiently bad to count as a serious deficiency seems to thereby be an important part of a friend's character, and therefore must be taken seriously by their friends. 18 It thus seems plausible that friendship requires that friends take one another seriously. This in turn entails that the friend's moral attitudes, beliefs, and overall perspective are taken as prima facie valuable, choiceworthy, and as things which one needs a reason to reject. Like the previously considered claim, this claim that friendship requires taking one's friends seriously does entail that our friends characteristically shape who we become morally. However, this weaker version does not entail that the shaping is direct or that the person who is influenced by the other need be passive in the process. On this conception of friendship, being friends with someone involves taking their beliefs, attitudes, and projects seriously, which might lead to their being adopted but need not do so. Indeed, one might take a friend seriously whilst ultimately rejecting a great many of their attitudes, beliefs, and projects, and thus without any actual shaping occurring at all.
Friendship, then, requires that one take one's friends' perspective, attitudes, and subjective concerns seriously. And this has very concerning implications regarding friendships with bad people. Relating to bad people as equals whose views and attitudes are to be taken seriously entails that the friend takes the bad person's immoral views and attitudes seriously, at least initially. That is not to say that they ultimately accept them or are shaped by them (as per the argument outlined above). However, it does mean that the immoral views and attitudes are salient to them as options, that they consider them carefully, and so on. This is a problem because considering deeply immoral views and attitudes in this way is itself a moral failing.
Recall that at issue is whether it's bad to be friends with people with serious moral deficiencies, not merely minor flaws. The kinds of beliefs, attitudes and projects that the mere friend "takes seriously," then, would not be matters on which there is genuine and sincere moral disagreement, but beliefs, attitudes, and projects that are seriously morally repugnant, such as deeply racist attitudes. Taking such views, attitudes, and actions seriously seems itself to be morally bad; such views should not be among one's moral options even if one ultimately rejects them. At no point should one regard those beliefs and attitudes as choiceworthy or prima facie valuable: they should never be among one's moral options at all. Imagine, for example, someone who takes seriously some deeply misogynistic beliefs, such as the belief that "women are not deserving of respect." They ponder this as a possibility and weigh the appeal of the claim before ultimately rejecting it. Despite their rejecting the belief, we would think the worse of them for it; such beliefs simply should not be among one's options. The very willingness even to question women's moral status seems like a kind of misogyny. Elizabeth Anscombe memorably expresses a similar point about the significance of our moral beliefs: [I]f someone really thinks, in advance, that it is open to question whether such an action as procuring the judicial execution of the innocent should be quite excluded from consideration-I do not want to argue with him; he shows a corrupt mind. (1958,17) Here, Anscombe suggests that there is something deeply morally awry with taking seriously certain morally objectionable views. Taking such deeply immoral views and attitudes seriously, she suggests, would itself be morally bad.
I am not claiming here that a desire to understand why people hold certain views is morally objectionable, or that enquiring into why certain attitudes are objectionable is a bad thing. Such enquiry can be helpful and sometimes morally desirable. 19 In one sense, then, it is permissible (and often good) to take immoral views seriously. For example, one might seriously ask why it is that procuring the judicial execution of the innocent is wrong, or why it is that so many people hold misogynistic attitudes. These seem like valuable lines of enquiry. However, these questions need not involve taking the claims seriously in the sense I intend here, since they need not involve taking "procuring the judicial execution of the innocent being good" as a real possibility or taking misogynistic attitudes to be potentially valuable. If, in exploring why misogyny is widespread, one took it as a real possibility that it was in fact a good attitude to hold, then I take it that doing so would be seriously morally defective. Fortunately, the former kind of enquiry can easily proceed without taking immoral views and attitudes seriously. Friendship, I have suggested, requires taking one's friends' perspectives seriously in the thick sense of taking them as a real possibility rather than merely wishing to understand their views and attitudes. It is this thicker "taking seriously" that is morally bad.
In being a friend to a bad person, then, the friend considers deeply objectionable views, attitudes, and projects as real possibilities, and this is a moral failing. But taking the immoral perspective seriously, as discussed above, is plausibly an important constitutive part of friendship; it would not be possible to have a (deep) friendship with a bad person without taking their objectionable views seriously, and thus without doing something morally bad. Both becoming friends with a bad person and sustaining a friendship with a bad person are thus pro tanto bad, since entering or sustaining a morally bad relationship is plausibly itself morally bad. Being friends with a bad person is thus bad because it inherently involves taking seriously views and attitudes that are deeply morally wrong.

2.a A Comparison with the character-shaping risk view
At this point, it will be useful to compare the account I have offered to the character-shaping risk view. In section 1, I suggested that the character-shaping risk view (that friendships with bad people risk the good person becoming corrupted) failed to offer a general explanation of the badness of friendship with bad people. In particular, I suggested that such risks can plausibly be offset and thereby neutralised, and that the account depends on an unduly strong notion of the shaping involved in friendship. The view I have just offered clearly depends on a weaker interpretation of the shaping involved in friendship, and therefore avoids the latter critique. What about the former? Might the fact that one needs to take seriously the bad person's objectionable views be "neutralised" by the fact that friendship would require that they, too, must take one's own views seriously?
Unlike the character-shaping risk view, the account offered in section 2 does not allow for the badness of being friends with bad people to be "neutralised." On the risk view, risks can be neutralised because an influence in one direction could be counterbalanced by a risk in a different direction. Such counterbalancing could completely offset the original risk. However, this is not the case with the account I offered in section 2; whatever good effects such a friendship might bring about, the mere fact of taking deeply objectionable views seriously remains bad. This is because the risk view suggested that friendships with bad people are troubling only because they might bring about bad consequences. If one offsets this risk, there remains nothing to be troubled by. But on the account I offered, friendships with bad people do not merely risk bringing about bad consequences. Rather, there is something inherently bad about such friendships, and thus there is no possibility of neutralisation eliminating such badness (though other features of the friendship might render it morally acceptable overall). My account is thus well placed to explain our general intuition that being friends with bad people is bad. 20 3. Friendship and loyalty I have argued that friendship with bad people is a moral failing because friendship involves taking seriously the beliefs, attitudes, and projects of the other person, and some beliefs, attitudes, and projects ought not to be taken seriously. But one might worry here that this does not speak to one common reason why people are friends with bad people, namely that the friendship predates the person's becoming morally bad. One might therefore worry that the intuition that there is something bad about friendships with bad people is incompatible with the intuition that loyalty is an important aspect of friendship. Is a good person making a moral error if they remain friends with a person who becomes bad? Would dropping the friend when they morally decline not seem disloyal? Can these two intuitions be made compatible? 21 The first thing to note here is that not all kinds of loyalty seem equally valuable. Loyalty to one's friends throughout their low points and times of struggle seems important and valuable, and someone who dropped their friends as soon as their company was not pleasurable would seem to be a very bad friend indeed. In fact, it is plausibly an intrinsic requirement of friendship that one stick with one's friends in such circumstances. But loyalty to one's friends when they become bad people who are cruel or deeply disrespectful to others seems somewhat different to this. Loyalty to those who are struggling and in need is plausibly very different to loyalty to those who are seriously immoral. 22 Whereas loyalty in the former context is clearly valuable, it is far from obvious that loyalty in the latter context is valuable. So, it may be that our intuitions regarding loyalty are more fine grained than has been acknowledged thus far, and that we, in fact, value loyalty only in some circumstances.
However, there are other responses available that can accommodate the intuition that loyalty in friendship is generally important. The first is that it is possible both that there would be something 20 At this point, a worry might arise from reversing the character-shaping risk view, namely: moral change or progress can sometimes seemingly depend on friendships between bad and good people. That is, bad people sometimes morally improve through their friendships with good people, as in the case of Megan Phelps-Roper discussed in Chen (2015). Does what I have said condemn or cut off this avenue for moral progress? I think not. Firstly, I have said there is something bad about such friendships, not that they're necessarily wrong. So, it's possible that such friendships could be permissible overall despite there being something bad about them. Secondly, one might think that moral progress is possible without deep friendship with bad people: perhaps other kinds of caring relationship or weaker friendships could bring about moral change. Finally, this might simply be a tragic dilemma that we face. (One might also doubt that these were genuinely friendships as opposed to merely friendly relationships). 21 Many have thought that loyalty or commitment to one's friends is central to true friendship. Cocking and Oakley (1995) and Wilcox (1987) argue that commitment to one's friends is essential for friendship, and suggest that it is ruled out by commitment to consequentialism of any kind. This seems compatible with the claim I make here, since one could be highly committed to a friend whilst nonetheless being willing to end the friendship in some circumstances. Jeske (1997) similarly suggests that duties of continuing special concern for one's friends are based on the friendship relationship itself, but allows that some moral reasons outweigh reasons of friendship. 22 Though it should be noted that these two can be connected: personal struggles can influence moral views and attitudes. Whether it is acceptable to remain friends with a bad person who is also personally struggling may ultimately depend on the degree to which one considers them to be morally responsible. bad about their being friends with the bad person, and that there would be something bad about their disloyally dropping the bad person. That is, so long as one holds that tragic dilemmas are possible, there may be something bad (and something good) about both courses of action. Perhaps there is simply no good line of action here, or no action that will not involve sacrificing something of great moral importance. Moreover, my claim has not been that it is always wrong or even that it is overall a bad thing to be friends with bad people. Rather, my claim is simply that there is something bad about such friendships. This is consistent with there being other values that are realised by the friendship, perhaps even values that are more significant than the badness of taking bad views and attitudes seriously. The idea that there is something bad about such friendships does not therefore directly entail anything about whether such friendships ought to be ended.
In addition to this, ending a friendship need not entail disloyal abandonment of the friend, because friendship is not the only kind of loving relationship there is. As I have explained it, the badness of having bad friends arises specifically because the equal relationship of friendship involves taking one's friends seriously. But friendship is not the only valuable relationship we can stand in to other people, and there are loving relationships of other kinds that do not constitutively require taking one another seriously, such as the familial and mentoring relationships mentioned above. As such, there are loving relationships which do not involve openness to the other's pernicious beliefs and attitudes: one could continue caring about the other person, trying to help them, and so on. Not all loving or meaningful relationships involve taking the other person seriously, and thus not all loving relationships with bad people are morally problematic. Even ending a friendship, then, need not amount to disloyal abandonment, but might instead herald the beginning of a different kind of loving relationship.

A comparison with Isserow
I want to end by comparing the account I have offered to the only other recent treatment of this topic in Jessica Isserow (2018). Isserow offers an alternative account of what is wrong with friendship with bad people that focuses on forming friendships with bad people. On her account, making friends with a bad person reflects poorly on the friend because it suggests a flaw in their moral priorities. She notes that friendship is a relationship we voluntarily enter into: it involves a degree of choice in that we have the capacity to reflect on our attraction to possible friends and evaluate whether or not we should be thus attracted. 23 As such, she argues that the choice to enter a friendship requires evaluating the other person positively overall, and thus that befriending a bad person suggests that one regards serious moral vices as forgivable, or as overridden by other charms (3111).
Isserow's account is consistent with my own; there might be various things that are bad about being friends with bad people. 24 However, there are some independent reasons to doubt the account she offers. Isserow's account depends on the idea that we choose our friendships. On one interpretation, it also depends specifically on the idea that this choice is based upon an overall positive evaluation of the other. On another, it depends merely on the idea that some bad qualities should be a bar to friendship. 25 The former would seem implausible; friendship does not seem to require overall positive evaluation, and we don't seem to choose friends on this basis. It seems plausible that we make friends with people in whom we see something appealing or attractive, but it 23 She writes: "Friendship is not merely something that 'happens' to us. We can choose whether or not to forge these relationships. Though it is perhaps not so easy to avoid being drawn to someone, we do have the capacity to reflect upon just what it is that attracts us to them, and whether or not it ought to" (Isserow 2018, 3111;my emphasis). 24 Additionally, there are important parallels between the two. For example, on both accounts the problem with being friends with bad people is the moral complacency it entails, thought these are understood differently. Isserow suggests that the complacency consists in the fact that the friend "excuses that which ought not to be excused" (2018, 3112). On my account, the moral complacency concerns the friend's willingness to take seriously views and attitudes that are immoral. 25 Many thanks to an anonymous reviewer for this journal for pointing out the possibility of the latter interpretation.
is not obvious that we need regard our friends in an overall positive light. It seems perfectly possible for friendship to coexist with a high degree of disappointment in one's friends' tastes, beliefs, or choices, for example. Friends will often be aware of one another's flaws, and it is not obvious that those flaws need be outweighed or excused by positive traits in order for the two individuals to be friends. Moreover, on either interpretation my account seems better placed than Isserow's to explain why it is that it is specifically friendships with bad people that seem troubling, rather than loving relationships in general. This is because Isserow's account of what's wrong with friendships with bad people turns on the idea that friendship involves choice. But having a loving relationship of any kind seems to involve a choice, at least to the same degree that friendships do. 26 For example, we may have historically embedded and developed relationships with our parents, but the same is true of many friendships. Simply standing in a parent-child, cousinly, or mentor-mentee relationship as such may not always be a matter of choice, and feeling love for those others may also not be under our control. However, having a close or loving relationship of those kinds would be a matter of choice. It therefore seems as though Isserow's account risks overgenerating cases of problematic relationships, entailing that there's something wrong with many kinds of loving relationship with bad people (i.e., any that involve a significant degree of choice) that seem intuitively unproblematic. By contrast, the account I have offered ties the badness of friendship with bad people to a feature of friendship which is not common to all kinds of loving relationship. It is therefore better placed to explain the initial intuition-the intuition that there is something specifically bad about friendships with bad people-than Isserow's.

Conclusion
To conclude, I began by noting that deep friendship with seriously bad people is intuitively problematic. I then argued that the character-shaping risk view does not yield a satisfactory general account of the badness of such friendships. Instead, I argued that in deep friendships the friends take one another seriously, and that this involves taking their projects, beliefs, and attitudes as prima facie valuable options. Taking deeply immoral projects, beliefs, and attitudes seriously is itself morally bad, and therefore friendship with bad people is morally bad. I then noted that this is compatible with the intuition that loyalty to one's friends is also morally valuable. Finally, I noted that my account thus seems to have an advantage over Isserow's since her account does not tie the badness of friendship with bad people down to features of the relationship that are sufficiently specific to friendship.