Anyone who chooses to question the idea that both individuals and societies have an absolute moral obligation to remember the past, and above all, never to let its horrors be forgotten, has an obligation whilst making such a case to recognize fully the hurt and offence that doing so is likely to cause, and how understandable it is that it should do so. Nanjing, Dresden, Hiroshima, Auschwitz – how could any person of conscience even fleetingly entertain the thought that any one of these paradigmatic twentieth-century catastrophes is better off forgotten? Surely, if there is a problem with contemporary attitudes toward remembrance and commemoration, it is not that we have become too obsessed with memory, but rather that we are not focused on it enough. But with apologies to Freud (and the political and ethical objections to forgetting are not without links to the psychoanalytic insight associating forgetting with the psychic dangers of repression), remembrance too has its discontents. It is all very well to speak of a duty to remember, and to correlate it with the exhortation “Never Again” – but to remember what, remember how, and remember for how long?Footnote ¹

There is something unhelpfully Manichaean about the discourse of the partisans of “memory at all costs” in that for all intents and purposes they present remembrance as the party of the light struggling against the party of darkness that is forgetting, which is seen as memory's immoral anti-principle, best shunned by all decent human beings. This sacralization of memory, whether in the literal sense or in the Kantian one of a categorical moral imperative that must never be evaded under any circumstances, serves as an effective prophylactic against thought. And as long as it persists, it is difficult to see how the current focus on the moral demands of memory could be enlarged so as to include a serious examination of the moral dilemmas of memory. To say this is emphatically not to claim that those who champion the duty of memory are blind to the difficulties inherent in their project. To the contrary, the best among them – Ricoeur, Nora, Todorov, Margalit – have understood perfectly that, no matter how important it is for a society to collectively remember,Footnote ² there can be no guarantee that such remembrance will be appropriate and will not be an abuse, or at least a misuse, of the duty to remember.

Tzvetan Todorov was especially alert to this danger. In his pamphlet “The Abuses of Memory”, he warned that “in the contemporary world, the cult of memory does not always serve good causes”, and pointed out that an obsession with commemoration had been a hallmark of Fascist Italy, Nazi Germany and Stalin's Russia.Footnote ³ But Todorov argued that it was still possible to devise some sort of general moral paradigm that would allow one to distinguish the proper uses of memory from abuses of it. It is the weakest part of his argument. But in an email he sent to me in 2016 in response to my own book on collective memory and commemoration, In Praise of Forgetting, which contains a long discussion of his work, Todorov ended his lengthy critique of many of the book's key propositions with these sentences:

It was a characteristically generous concession on Todorov's part, but at the same time, from my perspective it raised more questions than it answered. Above all, who was this “one” or this “we” who was going to do the judging? It is not that Todorov was unmoved by this problem. In The Abuses of Memory, he himself offers the example of the Serbs during the wars in the former Yugoslavia, who used history as one of their most important justifications for “their aggression against the other peoples of ex-Yugoslavia”.Footnote ⁵ And Todorov cited two other emblematic cases of the dangers of memory weaponized: Northern Ireland and Israel-Palestine. On the latter, Todorov quoted approvingly the terms of reference of a meeting in 1988 between Israelis and Palestinians which had included these words: “in order to begin to talk, the past must be put in parentheses”.Footnote ⁶

Todorov's case-by-case basis turns out no better than his general rule in determining who will have the legitimacy to decide when what is taking place constitutes a morally licit use of historical memory and when it is a morally impermissible one. The problem is as obvious as it is intractable. Todorov is certainly not suggesting that governments can be relied upon to adjudicate wisely, as is clear from his stern critique of the Loi Gayssot in France criminalizing any public denial of the reality of the Shoah.Footnote ⁷ But if it is not to be governments, it is even less likely that the pronouncements on a given case by a group of philosophers, ethicists, historians or activists would enjoy any substantial legitimacy with the parties concerned. Was the Irish Republican Army (IRA) a legitimate military force fighting for the just cause of the unification of Ireland against an illegitimate colonial power, or was it a terrorist group made up of unlawful combatants whose cause was unjust and whose methods in waging their armed struggle breached the Geneva Conventions in innumerable ways? And the same sorts of questions can be posed with regard to the former Yugoslavia, Israel-Palestine, Kashmir or Sri Lanka, to name only a few of the conflicts whose opposing parties have not just conflicting but utterly incompatible visions of the rights and wrongs both of the past and of the present.

To be sure, there will always be circumstances in which a determination of who was in the right and who was in the wrong can be imposed on a community by a hegemonic force, be it internal or external, forcing that community to lay down not only its weapons but its narratives. An obvious example of the former was post-Second World War Germany in which the victorious occupying powers were in a position to in effect implant a repudiation of Nazism into the DNA of the two German successor States, the Bundesrepublik and the DDR.Footnote ⁸ As for the latter, contemporary Rwanda provides an example of how the victorious side in an ethnically driven civil war succeeded in imposing on the Rwandan people as a whole an anti-ethnic vision both of Rwanda's past and of its future. And legitimation crises tend to dissipate or at least go very deep underground when a State or an occupier not only has a virtual monopoly of force but also demonstrates its ability and willingness to use it.

But this “total victory” model is only one of several. This is why it is so profoundly misleading to generalize about post-war settlements, including the moral settlement that the duty to remember is supposed to lead to. For it tells us virtually nothing about what is to be done after a war in which there has been neither a clear victor nor a clear loser, a conflict that ends with a messy, more often than not unjust peace accord that leaves both sides seething, unreconciled, nursing their grievances, cultivating their hatreds. This is what one sees when one crosses from Loyalist East to Republican West Belfast, from Tel Aviv to Hebron, or from Sarajevo to Pale. And in these contexts, no outsider's judgments, no matter what their merits, are likely to carry much weight, for they share neither the same starting point nor the same ethical basis. Insiders are a different matter, of course, but the problem is that in most intractable conflicts the cosmopolitans, the activists of “enlightened” civil society, find themselves politically marginalized and too often dependent on outside institutions for funding, which is itself delegitimizing in the eyes of many of their fellow citizens for whom they hope to offer humane alternatives to the rule of the gun, the plutocrat, or the populisms of either the right or the left.

It is in these places – the Northern Irelands, the Israel-Palestines, the Bosnias and the Colombias – that the moral superiority of memory over forgetting which Ricoeur and Margalit largely taken for granted needs to be questioned.Footnote ⁹ To make this claim is not at all an effort simply to invert George Santayana's celebrated contention that “those who cannot remember the past will be condemned to repeat it”.Footnote ¹⁰ To the contrary, where it is feasible during wars and dictatorships and in their aftermaths, it is clearly preferable to remember (though the advantages and risks of commemoration in such periods are less clear-cut). The reasons for this are obvious, even to this sceptic of a categorical imperative to remember. To begin with, remembering thwarts the efforts of totalitarian and other tyrannical regimes guilty of crimes against their own people to conceal what they did.Footnote ¹¹ It provides victims of these regimes, as well as the family and friends of those victims who were murdered, with both information about what happened and State acknowledgment that it happened, which is of benefit not only to those immediately concerned but to society as a whole. Finally, it opens the possibility that those who committed these crimes will be tried for them.

The experience of the countries of the Southern Cone of Latin America – Chile, Argentina and Uruguay – shows clearly that although the process was in none of these three instances a smooth or a linear one, it is absolutely possible to secure truth, justice and peace without making any significant concessions to either the leaders or the foot soldiers of their respective dictatorships (or, for that matter, the very substantial part of their citizenries that supported them).Footnote ¹² In short, everything else being equal, remembering is to be preferred both because memory facilitates criminal justice (whereas forgetting facilitates impunity for the perpetrators of crimes or injustices) and because forgetting stands in the way of the sorts of political and economic change that are needed if there is ever to be social justice.

But note the caveat: everything else being equal. For that is precisely the point – when is everything else equal? Take South Africa, for example. At first glance, the defeat of the Apartheid regime in 1994 and the establishment of a democratic government for the first time in the country's history should have been at least as promising a context for truth with justice as that which obtained in Chile, Argentina or Uruguay. For unlike the countries of the Southern Cone, where, depending on the period, a majority or near-majority supported the dictatorships or were at a minimum ambivalent about them, the overwhelming majority of South Africans were jubilant at the end of Apartheid.Footnote ¹³ In political and judicial terms, the victory of the democratic forces was a crushing one, but in military and constitutional terms, it had been nothing of the kind. Loyalists of the dictatorship still controlled the army, while the new South African Constitution was the product of a negotiation between the National Party that had ruled under Apartheid and the African National Congress (ANC). These were the realities that informed the lacerating conclusion drawn by Nelson Mandela's new government that it would be impossible for it to impose either transformative economic change or criminal justice towards the butchers of the former regime. As critics of the Truth and Reconciliation Commission (TRC) argued at the time, the process was one that traded truth for justice. A torturer could present himself before the Commission and, provided he admitted what he had done and replied truthfully to the commissioners’ questions, he would be immune from prosecution no matter how grave his crimes.Footnote ¹⁴

Mandela's reasoning for why this was necessary was quite straightforward: the risk of an uprising by diehard whites for whom amnesty for the crimes of Apartheid was the sine qua non for acquiescing to a democratic South Africa was simply too great.

As a result, Mandela found himself faced with the worst-case scenario that Ricoeur had warned against in Memory, History, Forgetting, one in which what he called “the thin line of demarcation separating amnesty from amnesia” would be crossed, and in which “private and collective memory would be deprived of the salutary identity crisis that permits a lucid reappropriation of the past and of its traumatic charge”.Footnote ¹⁵ This certainly does not mean that the TRC served no useful purpose. To the contrary, a central moral purpose of the duty to remember is to enable the relatives of the victims of criminal regimes to finally learn what actually happened to their loved ones. But those who champion the moral imperative of remembrance make a wider claim, which is that collective memory serves the goal of human emancipation and not further enslavement.Footnote ¹⁶ Given the realities of present-day South Africa, it is difficult to see how the TRC did anything of the sort.

And at least some accountability, however insufficient, was a realistic expectation in the South African case; the victory of the ANC had made that possible, if little else. But when one shifts focus to other post-conflict societies such as Bosnia or Northern Ireland – that is, to cases in which there is neither a clearly victorious nor a clearly vanquished side – even lower expectations are required regarding what remembrance can achieve. If in South Africa the duty to remember had failed to live up to the claims made for it, in Northern Ireland and the former Yugoslavia the duty to remember had been viewed as having been incompatible with securing a peace agreement in the first place, and as remaining immiscible with the continuation of that peace. In Bosnia, the dominant view in each of the three principal communities – Bosniak, Serb and Croat – about what happened during the war is quite irreconcilable with those in the two others. In theory, it's possible that had the European Union, the United Nations or NATO chosen to maintain Bosnia-Herzegovina indefinitely as some sort of colony, a single version of the past could eventually have been imposed on all three groups. But there was certainly no willingness on the part of the major outside actors to make such a commitment. Nor was success assured, and if the last fifteen years have taught us anything, it is that occupations are wickedly difficult to get right.

The case of Northern Ireland has been even less promising. Ever since the Good Friday Agreement of 1998 brought the fighting to an end, it has been clear to everyone but the most diehard human rights fundamentalists that there was no realistic possibility in the near or medium term of arriving at a consensus about what happened during the thirty years of low-intensity guerrilla war that is rather euphemistically known as “The Troubles”.Footnote ¹⁷ If in Bosnia the uneasy Bosniak–Croat condominium got one statelet and the Serbs the other, in Northern Ireland the price of peace was the handing over of political control of the province to the belligerents themselves, with the political leader of the leading Unionist party becoming first minister of Northern Ireland and the former commander of the IRA becoming his deputy. To emphasize this point is in no way to belittle the civil society initiatives aimed at fostering reconciliation, almost all of which have been constructive and some of which have been absolutely heroic. But even today, a full two decades after the Good Friday Agreement, and at a moment when the Loyalist– Republican coalition has rarely been more dysfunctional, few people believe that the time is yet right to break the de facto public silence about the past. To do so would simply be too great a risk to what social harmony has been forged in post-war Northern Ireland.

This is certainly the view of the belligerents themselves, and that of a wide cross-section of the civil society out of which they emerge and on whose loyalty their legitimacy is based.Footnote ¹⁸ But the conviction that peace is best served by forgetting rather than remembering is grounded in history as well as ideology. In his 2010 book The Imperative to Forget and the Inescapability of Remembrance, the German classicist Christian Meier looked at post-war settlements going back to the conclusion of the Peloponnesian War, and passing through the unsuccessful effort by Henri IV of France to end the Wars of Religion in sixteenth-century France through the issuing of the Edict of Nantes in which the French people were commanded to forget, and the Treaty of Münster in 1648 that successfully ended the Thirty Years’ War by decreeing “perpetuaoblivio et amnestia”, up to and including our own era. What he found was that it has been an imposed public silence about the past and not public remembering that has promoted the peace and kept renewed conflict at bay.Footnote ¹⁹ To believe otherwise either presupposes that war has become an atavism destined to disappear – which, given that historically war and not peace has been the constant in human history, requires one to believe that our era represents the most radical break imaginable with the species’ entire past – or is just wishful thinking, pure and simple.

But it is precisely this view of our own time as being sui generis that informs much of the contemporary human rights movement, with its adherence to the view best expressed by Michael Ignatieff that the post-1945 era had been one in which a largely successful “revolution of moral concern” had taken hold of humanity's collective conscience. On this view, it is anything but magical thinking to insist that Martin Luther King was right when he said that “there can be no justice without peace and there can be no peace without justice”.Footnote ²⁰ But even Human Rights Watch, which has advanced the most uncompromising case for refusing to forego justice in the name of peace, has acknowledged that “there is not one formula that is suitable to all situations”, and that “well-known counter-examples do exist”.Footnote ²¹ The example that Human Rights Watch cites is Mozambique, which it concedes never meted out any form of justice for the crimes committed during the civil war but has remained stable since the peace agreement of 1992 that ended the conflict. My own view is that Northern Ireland and Bosnia are equally powerful counter-examples. But even if I am wrong about both, the fact that counter-examples exist – as, per Human Rights Watch, they undeniably do – radically undermines the case for the duty to remember being regarded as a categorical moral imperative.

Meier's work largely vindicates this view, but with the notable exception of Auschwitz, which he presents as an exceptional case.Footnote ²² In this, he approaches the view expressed by Hannah Arendt when she laid down as a necessary prerequisite that “[i]n fear and trembling finally to grasp all that humanity is capable of – that is indeed a precondition of modern political thought”.Footnote ²³ Implicit in this view is the idea that it is a precondition of any decent modern political order as well. This Auschwitz exception is one that many people of conscience and historical sophistication continue to insist upon without any illusions, contra the Human Rights Watches of this world, about what the immediate effects of memory on peace are likely to be.Footnote ²⁴ And on one level, the claim is unarguable in that it is a metaphysical assertion, and as such immune to falsification. A moralist is perfectly within his or her intellectual rights to say: “It doesn't matter whether remembering will help communities formerly at one another's throats to reconcile; remembering is a moral imperative whatever its practical effects, even if those effects historically have often been negative for the at-risk polities in question.” Indeed, while I do not share it, it seems to me that only the moralist's claim and not the pragmatist's, let alone the legalist's,Footnote ²⁵ supports the assertion that there is a blanket duty to remember.

Where does all this leave those of us who are not confident about what is an absolute moral duty and what must be thought of in admittedly more ethically constricted pragmatic and empirical terms? The short answer, I think, is that one should strive for remembrance where possible, but accept that there are times and places where more forgetting – by which, again, I mean silence in the public spaces and contexts of the State and of civil society – is the only safe choice to make. One may hope that at some point in the future the need to remember will sweep away a prudential decision to forget. But while we are within our moral rights to hope that, in a given case, forgetting itself will outlive its usefulness (something that seems to have been taking place in Spain over the course of the past decade with regard to the memory of the Civil War), conflating our wishes, however well-intentioned, with teleological certainties is an exercise in hubris, not morality. So no, no sacred duties, no categorical imperatives – either for memory or for forgetting.