Avishai Margalit offers an interesting justification for the duty to forgive. The obligation to forgive the other’s crimes for one’s own sake “stems from not wanting to live with feelings of resentment and the desire for revenge.” (Margalit 2002, 207) Where Immanuel Kant holds that we owe it to ourselves as rational beings to be free from the pathological heteronomy of malice, Margalit associates this obligation to forgive with the obligation to preserve one’s own health: to fail to forgive is to imbibe “poisonous attitudes and states of mind.” (Margalit 2002, 207) In either case, the obligation to forgive is a special case of self-preservation and self-care.

Part of his reasoning is that Margalit wants to preserve the role of regret and to distinguish forgiveness from forgetting. His principle concern is memory and the obligation to remember past transgressions, so Margalit argues that we need to find a way to deal with past transgressions in a way that does not completely blot out their memory. As a religious practice, forgiveness models the divine absolution between the Creator and his creation and so requires two actors: the penitent, who remorsefully reports her crimes to the priest or to his Creator who already knows them, and the confessor, who judges her contrition and offers penance and the absolution of her sins. It is the transgressor’s act of remorse that ‘covers up’ her crimes. In undoing his past acts, the remorseful transgressor not only makes himself worthy of forgiveness, but creates a positive obligation in his victim. Quoting Maimonides, Margalit chides us that “it is forbidden to be obdurate and not allow oneself to be appeased.” (Margalit 2002, 194) But in converting the religious vocabulary of divine prohibition into a humanist dialect of ordinary duties and rights, he preserves the claim that forgiveness is a duty rather than supererogatory.

Focusing on the root ‘give,’ Margalit articulates the problem of an obligatory gift, but, pointing to the work of early anthropologists, concludes that some gifts are “intended to form or strengthen social ties between the original giver and the one who returns the gift.” (Margalit 2002, 195) On this reading, the original giver is the transgressor, who offers his remorse.

“I am claiming that the obligation to forgive, to the extent that such an obligation exists, is like the obligation not to reject a gift—an obligation not to reject the expression of remorse and the plea for forgiveness.”(Margalit 2002, 196)

To refuse to forgive is to refuse a gift, not an exorbitant gift, but an ordinary one, as when gifts are exchanged in what eventually appears as an economic transaction similar to any other commercial deal. The victim is obligated to respond in kind by granting forgiveness, in the name of “social ties.” Compare this to Hannah Arendt’s judgment of Eichmann: that no one, not even she, could be asked to “share the world” with him. For Margalit, the forgiver must exclude the forgiven act from all future judgments of her transgressor. The transgressor’s remorse covers over the past, but it still falls to the forgiver to resist peaking underneath the cover of forgiveness.

The metaphor of the palimpsest helps to illustrate this task: though the transgression is indelible and ineradicable, the forgiver nonetheless scratches it out. Forgiveness leaves an overwritten mark that can be deciphered, especially in light of further transgressions.

What does it mean to claim that such a pure gift is possible? To give remorse in expectation of forgiveness is like giving a gift in expectation of the return of a gift of equal or greater value. It is the economy of the gift, the obligatory mutuality of this “mutual release,” that troubles Arendt when she encounters it in Auden’s letter. My suspicion is that it cannot be obligated in the way that Margalit suggests: if forgiveness is to retain its capacity to begin anew, it cannot be subjected to this sort of calculation.

Margalit’s account mirrors that of another recent theorist of forgiveness, Jacques Derrida, who pushes this tension into a now-famous paradox to the debates on deliberation based on a similar paradox in the gift. (Bernstein 2006, Derrida, et al. 2001) Derrida juxtaposes obligatory forgiveness with what he calls impossible forgiveness. On the one hand, there are acts of forgiveness required by the regular and ordinary relations of friends or fellow citizens. We ask and expect forgiveness for lateness or when we brush past someone in a crowded space. To refuse to forgive in those situations appears as a provocation or an attack. When strangers deny each other this petty reconciliation, they are declaring hostility. When friends refuse to make allowance for each others’ foibles, they effectively dissolve the friendship. Insofar as we wish to avoid hostilities or preserve the friendship, we are obliged to trade remorse for this ordinary and expected gift of forgiveness. As such, ordinary reconciliation is hardly real forgiveness, just as an exchange of equally valuable goods isn’t really giving.

On the other hand, there is forgiveness that Derrida labels “impossible.” This is the skandalon, the unforgiveable act over which our efforts to forgive not only stumble but are absolutely incapable of making headway. Systematic injustices, massacres, torture, and acts of genocide all present themselves to us as candidates for forgiveness, but Derrida argues that these acts are not really within our jurisdiction to forgive.  Either because we cannot represent the dead victims or because the act itself is too unimaginably atrocious that it resists our efforts even to understand or the criminal’s efforts to encompass in meaningful remorse, forgiveness in these situations is impossible.

For Derrida, this is an aporia: forgiveness is impossible, because it is only really required when we face an unforgiveable transgression. Yet those unforgiveable transgressions are the only times when forgiveness is necessary. Anything less than the unforgiveable need not be forgiven, since such negligible acts can be simply reconciled, overlooked, or embraced as peccadilloes if friendship is to be possible at all. Since I must forgive minor transgressions, the only tests of forgiveness are precisely those acts that are beyond my capacity to forgive. Is it obligatory? Is it possible? We cannot know this a priori: we must wait and see.

As I see it, the limit of forgiveness is not within our voluntary power, an act of will, but rather in developing the capacity to imagine the act that we are trying to forgive. Thus the skandalon of forgiveness is an imaginative challenge, we stumble over it when acts are unimaginable, and we overleap it when our imagination succeeds. We make these imagined acts meaningful for others through poiesis: we create a world of meaning in which they are imaginable by marking exemplars, noting commonalities, and creating spaces of remembrance. The product of our work thus makes these meaningless deaths and thought-defying atrocities meaningful and thinkable. If you think about it from the perspective of un-consolable resentment, this is a crime akin to justification or exoneration.

This fundamentally creative act is ultimately what made Arendt’s work on Eichmann so troubling: not that she herself made him appealing or granted him mercy, but that by imagining the kind of character that could have helped commit genocide, it made that genocide forgiveable as it ought never to be. Whether Arendt forgave Eichmann or condemned him to die is then irrelevant: insofar as she created the conditions for forgiveness, she deserves repudiation and hatred. She used her imagination to make the impossible-to-forgive possible.