1. Forgiveness in the broad

It is not possible in this article to do justice to the live debates on forgiveness. Yet, giving an account of epistemic forgiveness is most usefully done against the backdrop of a clear picture of what forgiveness is, and so it will be helpful to articulate that picture explicitly.

In the interests of space, we will begin by piggybacking on an essay by Roberts (Reference Roberts1995) on the character trait of being a forgiving person, what he calls “forgivingness.” As is well known, Roberts has it that emotions are a kind of “seeing as,” akin to the way the same drawing might appear to one as a duck or a rabbit depending on how one attends to it. An emotion is a “concern-based construal,” a way of cognitively relating to the object as relevant to one's values in a particular way. For Roberts, forgiveness involves moving from an anger-based construal of the wrongdoer to a more benevolent one.

The importance of being a forgiving person, for Roberts, is that viewing another person as a wrongdoer, continuing to construe him through the eyes of anger, precludes the ability to relate to the other person as a person, that is, someone with strengths and weaknesses, successes and failures, someone capable of wrongs but also capable of much more. This dynamic explains why we do and at least sometimes should find genuine repentance moving and why appreciating the other's story from their perspective might move us to forgive and not just excuse.

Roberts is concerned with describing a virtuous disposition to forgive. Suppose that it is possible for the virtuous to become the kind of people who see those who commit wrongs against them in benevolent terms. One might accept that while doubting that such benevolent construing is strictly necessary for forgiveness. It is instructive, though, that a similar thought shows up in other treatments of forgiveness, even if not in Roberts’ terms.

This is supposed to be a point that is generally applicable. To hold a wrong against someone involves thinking of someone in certain affectively laden ways, which then can frame how one relates to them generally. To forgive someone requires re-framing how one thinks about the other person. This brings us to a point well worth highlighting for the purposes of this paper, namely, that what Roberts and Griswold here point to without explicitly noting it is that there is an epistemic dimension of forgiveness. Forgiveness requires a change in the affectively loaded way one represents the wrongdoer to oneself and perhaps also to others.Footnote ²

Within the literature, there are various thoughts on what the functional role of forgiveness might be within one's psychic economy and whether it is unique either in its ends or its means to its ends. Yet, there is a fair amount of agreement that forgiveness at least moderates resentment, that it changes either the victim's right to blame the wrongdoer or their reasonable exercise of that right, and that it promotes reconciliation other things being equal.

Resentment can mean different things in the forgiveness literature,Footnote ³ but it can be taken as a useful proxy for the suite of negative reactive attitudes one holds towards a wrongdoer in virtue of a wrong. Taken as a genus in this way, resentment is a personally invested emotional disapproval of the other person. Suffering a wrong can provoke in one a disinclination to interact with the other person, a negative characterological assessment of the wrongdoer, a hostile and uncharitable interpretation of actions outside of the wrong done, a vested interest in others coming to see the wrongdoer in the same negative light, and a disgust at the wrongdoer's own attitude towards the event. The emotional force of any of these attitudes towards the wrongdoer could be expressed in the language of resentment. One does not simply disagree or disinterestedly disapprove. A sense of being wronged, rather, often comes with a disapproval that one either would not have or would not feel so acutely were it not for the wrong at issue.

One could hold that a true act of forgiveness would involve forswearing or giving up resentment towards the wrongdoer in all of its manifestations.Footnote ⁴ But that is the extreme. It seems reasonable to allow that someone has successfully engaged in an act of forgiveness even if one involuntarily experiences negative appraisals of the other person, if, for instance, these emotions are less than they may have been and one resists the temptation to stoke one's resentment against the other.Footnote ⁵ What seems even more clear, however, is that if one's feelings of resentment are unaffected by a putative act of forgiveness that is reason to think that the act was unsuccessful. Someone who does nothing about a desire to see the wrongdoer suffer intensely, no matter what else she may do, does not manage to forgive him. Hence, forgiveness moderates resentment.

Resentment and blame are related. If one resents another person for an action that is not blameworthy, it becomes harder to see how that resentment could be justified. At the least, resentment in the absence of blameworthiness does not seem like a justified moral emotion. Likewise, even though it is possible to imagine dispassionate blame, it is hard to imagine the emotional tenor of blame as we typically experience it without invoking some form of resentment.Footnote ⁶ Scanlon defines blame in terms of judging another person to have damaged his or her relationship with oneself and relating to the other person accordingly.Footnote ⁷ Insofar as we are invested in our relationship and care how others treat us, then one should expect blame in Scanlon's sense to be accompanied by resentment. Whether one accepts Scanlon's definition of blame as a conceptual matter or not, he is pointing out our tendency to think of blame within the context of our relationships. In fact, one might think that the interpersonal point of expressions of blame is to give voice to one's resentment in a way apt for producing a matched emotion of remorse in the wrongdoer.Footnote ⁸

Moderating one's feelings of resentment gives one the psychological space to relate differently to blame. Yet, how exactly one relates anew to blame seems to depend on the context of forgiveness. In the paradigmatic interactive case, a wrongdoer owns their guilt, expresses contrition, and makes a good faith commitment to do better going forward. The wrongdoer in this case enacts the aptness of blame. Given that blame has found its mark, continued blame behavior on the part of the victim can strike one as a refusal to accept the repentance of the wrongdoer. Forgiveness, in this context, has to do with accepting the relational gambit of the wrongdoer. Forgiving a good faith apology creates the social expectation that one will no longer engage in blaming for this event.

There are, of course, other cases and scouting the terrain a bit will prove helpful when we turn to specifically epistemic forgiveness in the next section. The wrongdoer might disagree about the nature of the wrong, about the assessment of blame, and about what future changes are appropriate. Thus, a wrongdoer might only approximate the ideal request for forgiveness. And, of course, they might not make the request at all in which case some might hold that forgiveness either is not possible or not appropriate.Footnote ⁹ Nonetheless, what is certainly clear is that renegotiating blame is possible and sometimes appropriate outside of the paradigmatic interpersonal case.

Most straightforwardly, one might decide to accept a real but less than ideal apology as good enough. One might, in effect, take a satisficing approach to the repentance of the other, perhaps factoring in the importance of the relationship, the triviality of the wrong, and the other's past track record. To even the unapologetic wrongdoer whose moral trajectory nonetheless is in a positive moral direction, one could allow the other's track record to count towards the worthiness of the other to be forgiven or the safety of doing so. One could forswear one's right to blame such a person as a sacrificial act even though one did not, as it were, get the opportunity to have one's day in relational court. Alternatively, one might not forswear one's right to blame so much as decide not to exercise it at present in a way that ends up facilitating relational repair in the final analysis, functioning as a kind of de facto forgiveness. In these less than ideal cases, our social norms create less pressure on us to forgive and future blaming behavior does not violate our expectations as markedly.

More controversial, of course, is the malefactor who thinks they have done nothing wrong and shows no sign of changing moral course. One might be concerned, for instance, with whether forgiveness in such a situation is consistent with healthy self-respect (cf. Murphy Reference Murphy1982: 505). Even in this case, one might choose to moderate one's resentment and renegotiate one's relationship to blame and, indeed, might do so as an expression of self-respect. One might, for instance, forswear one's need to engage in blaming behavior as an outlet for one's feelings of resentment or even cultivate an openness to the future repentance of the wrongdoer. One might do this in no small part because one does not want to be the kind of person who is caught up in resentment but rather someone who embraces hope for relationships like this one. One's self-respect, in effect, trumps one's desire to hold onto one's resentful feelings.Footnote ¹⁰ Many would not think of such a stance as one of forgiveness, but at a minimum, there would be a family resemblance.Footnote ¹¹

The upshot of our discussion of blame is that remorse and forgiveness are ways of expressing a pro-attitude towards a relationship within the context of a relational rupture. They are ways of adopting such a pro-attitude not only despite the aptness of blame but in ways that depend on the aptness of blame. To act like a broken trust merits intentional, emotionally costly redress is to express the value of the trust that was broken, which itself goes a fair way to creating conditions for trust to flourish moving forward. Expressing blame and receiving blame can play an important role in repair, but a continued stance of blame precludes the healing of relational rupture. Thus, moderating resentment and renegotiating blame is related to promoting reconciliation.Footnote ¹²

Not every wrongdoer should be re-accepted into whatever level of relationship the victim enjoyed with them prior to the wrong, and, of course, it is possible to wrong someone with whom one never had a relationship. Creating conditions friendly to reconciliation, however, even if not always achieving it, is already implicit in the moderation of resentment and renegotiation of blame. In removing obstacles to relationship and taking an interpersonal stance more conducive to interpersonal trust, forgiveness promotes reconciliation. There are many routes towards the moderating of resentment and loss of a disposition to blame the other person. Forgiveness, however, distinguishes itself partly because it is the moderating of what one takes to be legitimate resentment and not clinging to a genuine right to blame because one thinks that, to whatever extent it can be realized, it is both possible and desirable to repair the relational wound caused by the wrong. Though certainly not inevitable, this act of repair makes starting or renewing a relationship possible.

It should also be noted that successfully navigating the forgiving of a wrong can be a foundation for greater intimacy and cooperation in the future. Forgiveness, in promoting reconciliation, does not simply promote the resumption of whatever relational state was had prior to the wrong. Rather, one's relational history now includes not only something new that could push oneself and the other apart (aka the wrong committed) but also something new that militates in favor of intimacy, trust, and cooperation. To experience oneself as someone for whom the other is willing to forego rightful negative emotions in the hope of repair is to be given a gift that should cause one to think better of the giver and to value more the relationship in which it was given.Footnote ¹³ Likewise, to experience the other as able to take ownership of their missteps, extend empathetic care towards those affected by them, and embrace corrective action has the potential to show one new grounds for trust that could even outweigh those causes for distrust latent within the wrong forgiven. Where there was no prior relationship, forgiveness can sometimes create a foundation for one.Footnote ¹⁴

Thus, putting it all together, I take the following to be a good approximation of forgiveness in the broad. Forgiveness is an integrated cognitive-affective stance towards both the wrongdoer and one's feelings of resentment towards the wrongdoer that includes a commitment, even if a conditional one, to move from a stance of blame to an openness to relational repair and reconciliation as appropriate given the relational histories of the parties.

2. Forgiveness and the epistemic

With more of a sense of what forgiveness is in hand, we will turn to its epistemic applications after addressing a worry one might harbor about the project as a whole. In so doing, we will gain a better understanding of what repairing epistemic trust would amount to which will put us in better staid to account for epistemic forgiveness.

One might object that there is a confusion latent within thinking of epistemic trust as the sort of thing that is a candidate for repair. Either one has reason to trust someone that p on an occasion or one does not. One might think that epistemic trust is not interpersonal and thickly relational in the way other kinds of trust are. Rather, one might think it a causal notion, perhaps one closely associated with the basing relation. As such, it would be just as capable of picking out tools (e.g. trusting a thermometer) and mental faculties (e.g. trusting vision) as other persons.

I readily grant that epistemic trust can be asocial. When it is another person that one is trusting epistemically, however, episodic trust is highly parasitic on trust in particular persons with whom one has a relationship or on roles that other persons occupy in situations defined by stable social norms (Green Reference Green2014). After all, if one did not already trust certain particular people in a more generalized way or trust that certain kinds of people in certain social situations were worthy of trust, then it would be difficult to find reasons that justified episodic trust events.Footnote ¹⁵

In that vein, Hieronymi (Reference Hieronymi2008) points out, in an essay on trust, that we are not moved to trust by general facts about the value, utility, or safety of trust in the abstract, but, interestingly, require trust to be grounded in the particular features of what (or who) we are trusting that make it trustworthy. One might think, then, that, though the three place relation of Agent-Source-Object of Trust might be the right logical formula to use in an analysis of trust, perhaps adequately capturing the phenomenon requires stressing the second variable, one's relation to the person or thing trusted. Moreover, when it comes to epistemic trust in a person, one needs to be flexible enough in how one thinks about the third place in the relation to account for not just believing what another person believes but trusting someone epistemically to give one pushback on one's beliefs, help one state one's beliefs more precisely, bolster one's confidence when one doubts one's abilities, brainstorm together, and so on.Footnote ¹⁶ Distinctively epistemic trust needs to cover the various ways in which one person can rely on another when exercising one of a number of activities in the overlap of the epistemic and the social.Footnote ¹⁷ In short, then, insofar as our epistemic relations with others are diachronic rather than episodic, are situated in contexts mediated by the epistemic and moral agency of the other party, or involve epistemic activities where one does some of one's thinking with the other person, it is appropriate to think of epistemic trust as the sort of thing that is an apt candidate for repair.Footnote ¹⁸

Let us turn, then, to the topic of forgiveness and the epistemic by considering some cases. These cases illustrate the way in which all acts of forgiveness have epistemic implications as well as some of the diversity of the distinctly epistemic wrongs that can become candidates for forgiveness.

In each of these three cases, we must ask how forgiveness would have to manifest itself in order to address the epistemic dimension of each case and thereby foster renewed epistemic trust. The wrong in each case intersects the epistemic arena. It does so by making it harder to believe the perpetrator, by creating conditions more aversive to collaborate with them epistemically, or by otherwise negatively affecting the legitimate exercise of epistemic agency of the wronged party.

To claim that a wrong can have epistemic dimensions is not to, by itself, take a stand on whether or not there is a distinctively epistemic kind of blame. Boult (Reference Boult2020), for instance, argues that there is. If blame just is a modification of one's intentions, expectations, and attitudes vis-à-vis another person in virtue of their falling short of some normative ideal, then surely, so the argument goes, one can and should sometimes engage in such an adjustment in virtue of the other's falling short of an epistemic normative ideal. Epistemic blame would be a straightforward candidate for a state that could be a candidate for epistemic forgiveness. Likewise, Rettler (Reference Rettler2017) argues that we can be blamed for our beliefs in virtue of certain types of reasons-responsiveness. If we can be blamed for our beliefs, perhaps we can be forgiven for having them.

To the extent that one thinks there is a distinct kind of epistemic blame, there should be an easy road to affirming that there is such a thing as epistemic forgiveness. Yet, my thesis is not restricted to those cases. Even if one morally blames a knower for a wrong that is distinctively epistemic in nature, then, as we shall see, the epistemic dimension of the wrong will enter into what it means to forgive in this case. Likewise, as I will show, moral wrongs routinely have epistemic dimensions that require something of us qua epistemic agents if we are to forgive them.

A wrong can cause negative emotions that radiate out from the wrong at issue to other potential objects. In fact, one might think that epistemic wrongs and the epistemic dimension of general harms are especially susceptible to having effects that are not localized solely to the particular action of the wrongdoer. The reason has to do with rational updating and our lives as social knowers.

If a non-social doxastic source did something analogous to wronging one, giving bad information or not providing the information it was supposed to, then the rational thing to do would be to become less inclined to trust that source and perhaps less trusting of sources like the failed one. If a thermometer gives the wrong reading, get a new thermometer and think twice before buying the same brand. Part of being an epistemic agent is not just accepting information at a time but accepting information in a way reflective of one's past experiences and future informational needs. Thus, discovering that a trusted source may not be trustworthy causes not only an assessment of the isolated failing but of past and potential future inquiries in light of that failing. It affects how one maps the epistemic terrain.

What goes for non-social sources applies that much more for social ones because people, unlike tools, are not only epistemically reliable or not simpliciter, they are typically only reliable sources if, among other things, one has their attention and good will. Moreover, we are not interested in the truth of just any question nor should we be. Rather, our epistemic needs and interests are shaped by our non-epistemic values and goals. Our mapping of the epistemic terrain prioritizes where we can get information on matters we care about and who will collaborate with us epistemically and otherwise on projects we are invested in.

A wrong, whether of a distinctly epistemic sort or not, signals a mis-alignment of values, goals, and intentions such that one is given reason not to rely on the other person whatever their baseline of competence. Martin, for example, might firmly believe that the affair is over and that Madeleine intends to make it work while finding it unnatural to just assume a harmony in their values and goals in the wake of the relational rupture. This feeling changes what shows up as an epistemic risk and what defeaters occur to Martin for relationally relevant beliefs. It changes what epistemic projects he is willing to engage in and what non-epistemic projects he will fail to commit to, in part, for epistemic reasons.

Individual victims and wrongdoers not only affect each other on an individual basis but in virtue of the social networks in which each are placed and the cognitive niches partly composed by them. When Miles is dismissive of Maya, he models an attitude towards her that could easily be picked up by others who witness the event. Miles helps to deny Maya status as a knower within and possibly even access to his social network. This diminishes Maya's exercise of her epistemic agency and jeopardizes her access to the goods, epistemic and otherwise, that this network could mediate. Maya acquires new reason to think that the social networks Miles represents will be inhospitable to, if not unavailable to, her because of, not just Miles, but people like Miles or who tolerate behavior like his.

A mere change in emotion or a mere decision to treat the other person as if the wrong were forgiven are no more sufficient for forgiveness on the epistemic level than they were for forgiveness in general. If, for instance, Marie's father tried to smooth over the wrong done her by asking Marie to take charge of the distribution of her deceased mother's things and, if, furthermore, this caused a moderating of her feelings towards Mona (e.g. by symbolizing Marie's close connection with her mother), this change is not equivalent to forgiveness. Even if it does not trace the complete path of the ill feelings engendered, forgiveness requires a cognitive-affective stance towards the wrong and wrongdoer that commits to mapping the epistemic and relational terrain differently. A mere diversion of attention, even for good reason, is not forgiveness, and diverting one's attention is consistent with the persistence of a perspective on the wrongdoer that is defined by her wrong that can be activated in the future. To forgive a wrong in any but a quite minimal sense requires an attempt to re-map the emotional terrain that's been affected by having been wronged. Just so for epistemic wrongs and the epistemic dimension of non-epistemic wrongs. To forgive another person on the epistemic level is to represent their merits as a potential epistemic collaborator in a way that is not dominated by the wrong they have done.

What then does resentment and the moderating of resentment look like through an epistemic lens? Resentment, we said in the previous section, functions most principally in the forgiveness literature as a proxy for a suite of negative reactive attitudes levelled principally at the wrongdoer. It is a personally invested disapproval. This disapproval guards against misplaced trust and the personal costs that can come with depending on undependable people. Clearly, it is possible to harbor resentment towards another person qua epistemic agent.Footnote ²⁰ Just as resentment in general fixes one's attention on reasons not to easily trust another person, so epistemic resentment fixes one's cognitive and affective attention on reasons to downgrade or eliminate one's epistemic dependence on the wrongdoer or people like the wrongdoer.

Specifically epistemic resentment of the wrongdoer might take many forms. Martin now finds it easy to give a negative appraisal of Madeleine's sincerity as a testifier. Martin feels alienated from the goals and values which shape Madeleine's epistemic interests. Maya acquires a vested interest in down-grading Miles’ competence and might enjoy a little schadenfreude if he endured a public silencing of his own. She acquires a wariness of Miles’ sensitivity to defeaters such that even if Miles testifies on a matter within his competence, Maya does not feel confident that he has recognized all the relevant higher-order evidence because his biases might have insulated his perspective. Maria might feel repulsed by the way that Mona's action shows a willingness to let selfish interests, by Maria's lights, encroach on her epistemic behavior. Maria might suspect that Mona's ability to rationalize her behavior implies that she is prone to self-deception. And, of course, all of the above is reason for Martin, Maria, and Maya to be less inclined to treat their epistemic others as peers when they disagree, not because they have discovered their respective wrongdoers to be incompetent per se but simply because treating someone as an epistemic peer often requires a great deal of epistemic trust.

To moderate one's distinctively epistemic resentment and renegotiate blame in an act of forgiveness, the victim must reject the centrality of the wrong as a lens for viewing the desirability of epistemic dependence on and collaboration with the wrongdoer. This affects the moderation of, but not necessarily the elimination of, epistemic resentment towards the other person. In its place, forgiveness requires an increased openness to the wrongdoer demonstrating that they can be worthy of epistemic trust.Footnote ²¹ The specific means of moderating one's resentment can be various but will need to address both the epistemic and the moral level. It is a truism of the testimony literature that trusting a testifier requires being able to trust not just the testifier's competence but also her sincerity. Just so, repairing epistemic trust across the diverse ways in which one can partner with another person is not a matter of simply restoring one's faith in the cognitive horsepower of the other, though it may involve that. One's faith in their goodwill, sincerity, and attention matter too.

As with forgiveness in general, one might moderate one's epistemic resentment by actively being mindful of mitigating circumstances, of the positive qualities of the other person, of one's own failings, of the sincerity of their contrition, or of the ways they have made changes in their life for the better. One might plan for ways of expressing one's emotions that have less potential to damage the relationship or to overstep the bounds of what counts as a reasonable reaction and so on. The over-arching point, however, is that, akin to the way one might adopt a mental frame to facilitate a “seeing as,” just so forgiveness requires an implicit recognition and repudiation of the way a wrong can provide a mental frame that defines one's relationship with the wrongdoer, including one's relationship to the wrongdoer as a knower. “Seeing as” can, of course, occur independent of one's agency. One can, however, facilitate it by choosing to be mindful of certain pieces of background information rather than others, of recalling some episodes rather than others, of fixing on some aspects of a scene rather than others, and so on. Alternatively, one might simply seek to recognize and disengage from ways of allocating one's attention or interpreting behavior that are dominated by a prior “seeing as” with which one is seeking not to identify.

Just as with forgiveness in general, there are relevantly different cases that shape how epistemic blame should be renegotiated. One is when the wrongdoer owns his or her blame, for example, by apologizing. When Miles apologizes to Maya, he owns that he has wronged Maya in committing an epistemic injustice against her and pledges to make relevant changes. If Maya chooses to forgive him, one natural way that relating differently to blame might look for her is to use his apology as the relevant lens through which to see their relationship and thus their future epistemic collaborations. Rather than simply expunging the wrong from her memory, using the apology itself as the relevant lens honors the relevance of the wrong to their relationship while nesting it within a good faith acceptance of his contrition and an openness to his making good on his commitment to change. Instead of simply writing off Miles as someone likely to commit more epistemic injustices, Maya could instead choose, in accepting his apology, to see him as someone who is now more capable of relating well to her for having repented of an act he recognizes as having harmed her as a knower.Footnote ²²

What, though, of the case where the wrongdoer does not apologize? Again, there is a relevant distinction to be made. Suppose the wrongdoer's track-record and current trajectory is such that it is open to the victim to view them as someone who is becoming or already is a person who is worthy of more epistemic trust than the wrong committed would imply on its own. In that case, the victim might take that fact as reason to adopt the same stance towards the wrongdoer that they would adopt in the face of a good faith apology. For example, suppose that though Madeleine confessed that she had the affair, she did not present it as a wrong committed against Martin, and thus, Martin does not feel like she owned any guilt or took on blame for the way she hurt him. Yet, over time, Madeline shows by her actions a newfound commitment to the relationship, and Martin is able to open himself to trusting her more, both epistemically and otherwise. Though sometimes tempted to see an ambiguous action through the lens of the past relational rupture, Martin chooses to prioritize her track-record since the relational rupture as the relevant lens and in so doing enacts a forgiving stance towards her vis-à-vis the affair.

Our social norms do not pressure us to apologize for our epistemic wrongs as frequently as for other kinds of wrong, at least not as yet. Moreover, when we do so apologize, our norms are often satisfied without the demonstration of strong emotion and thus without a “costly signal” that helps us discern when an apology for an epistemic wrong should be accepted. Thus, one might think that adopting a forgiving stance towards an epistemic wrong based on one's assessment of the overall character and characterological trajectory of another person may be especially important for us, perhaps more important than it is for non-epistemic wrongs.

We are left, of course, with the case of the wrongdoer who provides no behavioral clue as to their worthiness for renewed epistemic trust. Indeed, they may give cause the other way. Suppose, for example, that Mona not only does not admit her wrong to Marie but bristles in their interactions, allowing the same psychological dynamics that led Mona to lie to her sister to poison their interactions in their time of bereavement. Mona is not a reliable or just testifier in such a state and to avoid having to tell the truth about what happened with their mother, Mona continues to lie and evade. Does it make any sense to speak of forgiveness here, epistemic or otherwise? It is not hard to imagine something like the following. Even though Marie does not take Mona to have owned her guilt and even though the wrong done is indicative of Mona's past and continuing character, Marie might nonetheless adopt a compassionate frame in how she thinks of Mona. She might, for instance, choose not to let the fact that she feels wounded by Mona's actions act as the frame through which she views Mona but instead effortfully foreground in her mind the ways in which Mona's actions themselves come from a wounded place and do not damage the reservoirs of competence and goodwill that Mona is capable of displaying in other situations. This allows Marie to hold onto the fact that Mona is worthy of epistemic trust on many occasions and that it is reasonable for her to hope that epistemic cooperation with Mona in the future is worth investing in now.

Moreover, one might think that forgiveness in general is helpful to one's life as a knower quite apart from the state of the wrongdoer. Marie, for example, might recognize that adopting a cognitive-affective stance towards Mona that foregrounds the wrongs Marie has suffered would have a jaundicing effect on Marie's own outlook that Marie does not want for herself. She may not want it for herself in part for an epistemic reason, namely, that allowing a wrong to be the lens through which she sees the world would damage her ability to represent that world fairly. Unforgiveness is an obstacle to knowledge. The result, if it is not an act of forgiveness, must at least be acknowledged as a forgiving outlook that predisposes Marie to hold up her end if an opportunity for reconciliation with Mona arises.

Highlighting the interactions of forgiveness with the epistemic is partly important because repair is an essential activity for social knowers. Anyone that one epistemically relies on is someone that sometimes commits epistemic wrongs and wrongs that have epistemic implications. The overlap of forgiveness and the epistemic is also important for understanding forgiveness itself. It is easy to fall short of full forgiveness by neglecting the epistemic. The first and most obvious reason why this might happen is that it is comparatively easy for an epistemic wrong to escape notice, especially for the perpetrator of one. That is not to say that epistemic wrongs are not keenly felt by the victim, but it was only relatively recently that epistemic injustice, for example, became a mainstream category in philosophy. The material imprint of a wrong in the world can be hard to deny, likewise the voicing of hurtful words. The damaging of epistemic trust often comes with both but is subtly in the background.

Second, an easy way for forgiveness to fall short is for a discrepancy to arise between the forgiving of an act and the forgiving of the person. One might, for instance, forswear taking vengeance or demanding restitution for a wrong. One may pledge, to oneself if not to others, not to dwell on the particular episode or bring it up again. One may be fully convinced within one's own mind that one has forgiven the wrong, and yet, represent the character of the other person moving forward in a way that has been definitively altered by the wrong done. In effect, one is holding the wrong against the wrongdoer in the way one represents the character of the other person, which in turn influences one's future interactions and the possibility of reconciliation. Moreover, even if one is live to the possibility of holding onto a wrong in how one represents the character of the other person, one is more likely to spot the relevance of moral than of epistemic character. After all, we commit wrongs qua moral agents, but one cannot divorce the moral and epistemic quite so neatly.

Finally, as concerns one's expectations of others not directly involved in the incident, it is again easy to not notice the way a wrong might color one's perspective. If a bad relationship causes one to expect one's next partner to betray one, then one has been harmed epistemically because one's way of representing the world has been poisoned. One has become less reliable in an important range of beliefs. Even when a jaundiced representation of the world ends up being accurate, a wound-dictated detection of bad actors is not likely to be reliable in the sense of a “sensitive” belief, a belief such that if it had been false, one would not have formed it.

Another way in which a lack of forgiveness can cause harm to one's future interactions, though, is by making one more hesitant to engage in epistemic trust in new situations. This damaging of one's willingness to trust need not take the form of an inability to trust. It can take the form of affecting one's risk-taking thresholds, effectively turning situations where trust should be easy into the equivalent of high stakes situations in which trust is hard to earn and act on.Footnote ²³ Consider, for instance, that if Miles personally sees to it that restitution is made to Maya by way of actively attempting to facilitate her inclusion in social networks that Miles inhabits, the epistemic injustice committed against Maya might still make it harder for her to use them effectively because of the effect of the wrong at issue on her capacity to exercise epistemic trust. Even in the presence of maximally cooperative and penitent partners, epistemic forgiveness for Maya might be crucial not just for her future interactions with Miles but for Maya's exercise of her epistemic agency in general. Thus, Maya might have a reason to embrace a forgiving way of relating to Miles that is at least partly independent of how worthy he is or is not of being forgiven.

Thus, we see that far from being a curious extension of the concept of forgiveness, epistemic forgiveness is actually central to forgiveness in general. Being wronged pushes one in the direction of epistemic distrust, both when the wrong at issue is distinctively epistemic and when it is not. Joint action and the sharing of a common life have an epistemic dimension to them, and epistemic distrust presents a formidable obstacle to their achievement. Yet, a forgiveness that makes possible neither joint action nor the intimacy of a shared life brings forth a rather shabby sort of reconciliation.

In this essay, I have shown that acts of forgiveness in general have epistemic implications and that the forgiveness of epistemic wrongs is important in its own right. To be a social knower, one must not only identify experts, reason with peers, and avoid bad actors. One must engage in the hard work of repair – owning blame, moderating resentment, negotiating reconciliation. The social knower cannot skip repair. This is not because we cannot do without a few brilliant misanthropes and thus must forgive a class of people we would prefer to cut off. Rather, though not all wrongs are equal, the people we depend on individually, in groups, and through institutions are going to disappoint us morally and epistemically, and we will damage the trust of others in turn. We are fallible. Without the ability to repair trust and the frequent exercise of that ability, our common life, including our epistemic lives, will be severely impoverished. Forgiving and knowing, at least for us, are deeply intertwined.Footnote ²⁴