1. Questions About the Past

Let me start by distinguishing the question I shall be concerned with here from other questions about the past. The Greek term ἱστορία, historia, is usually translated as meaning ‘inquiry’ or ‘knowledge acquired by investigation’. Thus, it is already ambiguous between knowledge about the past and the kind of inquiry that seeks to obtain it, an inquiry now led by the discipline of history. We can also distinguish between such knowledge—something that we can now lack or possess—and the objects of that knowledge—past events stretching back many thousands of years into the past and the numerous long deceased people who participated in these events. These people and events are contained in the past, a stretch of time that ultimately extends all the way to the Big Bang.

So there is (1) the past itself, (2) the events it contains, (3) our knowledge of these events, and (4) the inquiry that produces such knowledge. We can ask questions about each of these as well as about their interrelations. For example, we can ask how we can know about the past or whether historical events can be explained and how. I shall not be concerned here with these familiar questions in the philosophy of history. I will be interested in how the past matters, a broadly ethical question. By this I do not mean the question of whether it matters that there is some past behind us. I will be concerned with another question: given that the past is what it is—not an empty stretch of time, but a tangled human history—should these historical events matter to us and in what way?

This is a question about the significance for us now of past events themselves. I will not be directly concerned with the significance of the present remnants of past events—for example, with questions about the value of burial grounds or archeological finds. But an account of the significance of the past will obviously also bear on these further questions (for important recent work addressing questions about the value of historical artifacts, see Matthes Reference Matthes2013 and 2015; James Reference James2013; Young Reference Young2013; Korsmeyer Reference Korsmeyer2016, Reference Korsmeyer2018; and Bicknell et al. Reference Bicknell, Judkins and Korsmeyer2019).

On the account I will propose, it is primarily the significance inherent in these past events, in the things past people did or experienced, that explains why we care and ought to care about the past. Past value explains the value of historical knowledge and inquiry. By contrast, the competing accounts I will first consider try to explain our concern for the past, or specifically the value of historical knowledge and inquiry, in a more direct fashion, whether by focusing on their benefits or by ascribing to them brute intrinsic value.

Before I turn to these competing accounts, let me clarify three things about what I will be trying to do. First, our concern about the past raises both justificatory and explanatory questions. My primary aim will be normative: to justify our concern about the past. While I believe that this justification also helps explain why we care about the past, this is a further, broadly empirical question that I will not be able to address fully here. Second, I will be largely concerned with our collective reasons for caring about the past. I will only later consider whether and how these reasons apply to us as individuals. Finally, throughout this paper I will speak about reasons to know about the past as shorthand for reasons for a wider range of cognitive states and activities such as finding out about past events and then maintaining and disseminating that knowledge (or, when knowledge is impossible, the best supported beliefs).

2. Instrumentalist Accounts

Historians do not need to be persuaded that the past matters. But the question of why it matters tends to receive only passing mention in their voluminous writings. When historians do consider this question, they offer a wide range of answers (see, e.g., Stearns Reference Stearns1998). These answers usually appeal to the alleged instrumental value of historical knowledge, to how it might bear on us and on what matters to us now. Since such answers typically mention only instrumental considerations and are clearly meant to capture our core reasons for caring about history, they are best interpreted as assuming

Utilitarianism and most consequentialist views are obviously committed to a strong version of this claim. But this is a widely held view.

Historical inquiry has been claimed to be useful in multiple ways. Most common is the idea that understanding the past is the key to understanding the present and predicting the future. Thucydides already suggested that exact knowledge of the past is useful as ‘an aid to the interpretation of the future’ (Thucydides, Reference Thucydides1950: 15), and Hobbes wrote that ‘the principal and proper work of history being to instruct and enable men, by the knowledge of actions past, to bear themselves prudently in the present and providently towards the future’ (Molesworth Reference Molesworth1839: 6). Macaulay went so far as to write that ‘no past event has any intrinsic importance. The knowledge of it is valuable only as it leads us to form just calculations with respect to the future’ (Macaulay Reference Macaulay1828: 155).

This remains a dominant view. E. H. Carr writes that ‘the function of the historian is neither to love the past nor to emancipate himself from the past, but to master and understand it as the key to the understanding of the present’ (Carr Reference Carr1961: 29). Hayden White even more bluntly asserts that ‘the contemporary historian has to establish the value of the study of the past, not as “an end in itself,” but as a way of providing perspectives on the present that contribute to the solution of problems peculiar to our own time’ (White Reference White1966: 125). White equates the value of history with its usefulness for life or, as he puts it, its ability to answer the practical question ‘what is to be done’ (White, Reference White2010: 15). White is here deliberately echoing, not only Lenin but also Nietzsche who wrote that ‘we want to serve history only to the extent that history serves life’ (Nietzsche [Reference Nietzsche1874] 1997: 59).

A more elevated variant of this view regards history as a source, not of practical lessons but of moral ones—both by offering us moral exemplars to follow and by alerting us to the dangers of evil (Stearns Reference Stearns1998). Defending the moral evaluation of historical acts, Waldron similarly suggests that ‘our moral understanding of the past is often a way of bringing to imaginative life the full implications of principles to which we are already in theory committed’ (Waldron Reference Waldron1992: 5). History provides, he writes, ‘a lesson about what it is like for people just like us—human, all too human—to face real moral danger’ (5). Waldron further argues that history can ‘sustain the moral and cultural reality of self and community’ (6). While this claim does not need be understood in purely instrumental terms—see below—this is how it is often understood, as in the common idea that history promotes patriotism and instills a sense of national identity.

I could continue the list of history's supposed benefits. But there is no need to consider closely each instrumentalist account because such accounts all fail for similar reasons. Consider, as an example, the common claim that the study of history is the key to understanding the present or a source of lessons. It is obvious that what we do is continuously informed by our knowledge of the past. It is hard to show that such knowledge is employed in ways that regularly lead to better outcomes, but it is wildly implausible that total ignorance of the past would improve decision-making. And when our knowledge of the past does not improve our decisions, it at least improves our ability to understand what is happening around us.

Still, most of the knowledge of the past that is potentially useful in this way is fairly specific and recent. A contemporary president has plenty to learn from the Cuban missile crisis. But it would be far-fetched, to put it mildly, to hold that current policy can be usefully informed by a study of the function of Stonehenge, Mayan chronology, or the life of Ramses II. Current social science is a far greater resource for plotting policy and for understanding the present than most historical work.

In any event, even when the study of certain historical events can plausibly be held to teach important lessons it is still false that these lessons are our main reasons for studying these events. Knowing about the atrocities of the past can no doubt do some good and may help reduce the likelihood of future atrocities. But it would be preposterous to suggest that our only or even main reason to know about the Holocaust or slavery is that they teach us such lessons. As though if we had concluded that no such lessons would ever be learnt—or even that the wrong lessons would be learned—then it would be just fine to forget about these horrors.

Parallel points apply to other instrumentalist accounts of the value of history. Such accounts at best justify a contingent concern with a narrow range of historical events. Even here, they put history in a subsidiary role to other sources of knowledge that can achieve that instrumental goal more directly and effectively. Finally, and most important, we would still have reasons to be concerned about the past if these instrumental considerations were absent or even reversed.

To see this, consider the following scenario:

This proposal is different from the Orwellian rewriting of the past associated with totalitarian regimes. Such rewriting involves mendacity and manipulation. Its aim is to control the masses and undermine their sense of reality. But no mendacity is involved here, and I assume the aliens are genuinely benevolent and may even be right that it would be better for us to give up the past.

The aliens’ bargain would remove our instrumental grounds for holding on to our concern with the past. The various instrumental benefits associated with such a concern would be either rendered irrelevant or obtained in a superior manner in another way, or otherwise outweighed by vastly greater benefits of other kinds. For example, we would not need the lessons of history in the future offered to us—either because that future would be so different or because the aliens would offer us better advice as to how to manage our affairs.

Should we accept the offer? This may be tempting to some, but I at least find the idea appalling. Accepting the offer would be a profound mistake—those who order the shutting down of museums or the destruction of precious artifacts and archives should feel deep shame. If you agree, then you must reject instrumentalism.

3. Non-Instrumentalist Self-Centered Accounts

Collingwood held that history matters because it provides us with ‘human self-knowledge’. This is because ‘knowing yourself means knowing what you can do; and since nobody knows what he can do until he tries, the only clue to what man can do is what man has done. The value of history,’ therefore, ‘is that it teaches us what man has done and thus what man is’ (Collingwood Reference Collingwood1956: 10).

This can still be understood in instrumental terms: we can find out about ourselves by learning about past others. Understood in this way, this account is vulnerable to the same objections that undermine other instrumentalist accounts. There are better ways to find out about ourselves and what we can do than by studying Neolithic carvings.

But the relation between historical knowledge and self-knowledge need not be instrumental. Perhaps knowing our origins, including our historical origins, is itself a form of self-knowledge. Moreover, returning to the idea of understanding the present via knowing the past, that too need not be instrumental. Knowing what happened in the recent past may be constitutive of understanding many present events. Waldron's (Reference Waldron1992) remarks about the relation between history and identity can also be understood along similar lines. If our history is constitutive of our collective or individual identities, then knowledge of that history may be required to sustain our identity.

According to

On such views, our core reasons for looking backward are located in some noninstrumental value, but that value resides here, in us. It is natural to understand that value as prudential, but it could also be claimed to be moral.

Such accounts make no claims about the consequences of historical knowledge and concern. They can therefore offer reasons for rejecting the extraterrestrials’ offer in blank slate. The extraterrestrials would be asking us to give up knowledge that is constitutive of our identity and that is needed if our projects and commitments are to make sense. What good would the ultra-advanced technology be for us if to enjoy it we must give up what makes us us?

In assessing the self-centered account, it is important not to conflate it with

This is a claim about agent-relative reasons to care about aspects of the past. Self-centrists, however, assume not only that we have such partial reasons, but also that they are our main reasons for caring about the past. And self-centrists also make a further claim about why we have such partial reasons—basing these reasons in present value, in value to us.

Self-centrism may seem attractive because historical partiality offers a plausible account of some of our backward-looking reasons. Some past events are more important for us to know about because we relate to them in some distinctive way, most commonly because they involve past people to whom we bear a special relation because they are our ancestors or predecessors. A Jew, or a German, may have special reasons to know and care about the Holocaust that others do not have.

However, they also have reasons to know about chattel slavery or about the 1942 Bengal famine or about numerous other past horrors and glories that in no way bear on them, their origins or identity or their projects. And if we have such weighty impartial reasons for caring about some past events, then historical partiality does not capture key reasons we have for caring about the past. And since self-centrists assume that our reasons for caring about the past are partial, that is enough to show that self-centrism is false.

It could be replied, echoing Collingwood, that we do have something in common with all past people. We are all human. Writing of the ‘patrimony of humanity’, Young (Reference Young2013: 36) similarly proposes that ‘humans and human achievements have value—cosmopolitan value—to everybody precisely because they are human achievements’, while Matthes suggests that ‘simply being human provides one with a reason to value the prehistoric places and artifacts that shed light on, or are associated with, our development as a species.’ (Matthes, Reference Matthes2015: 1016). But besides emptying the idea of a special relation to aspects of the past of its content, this suggestion still fails to capture that source of our reasons for caring about the past. Past horrors would merit our attention even if their victims were not human. We have reasons to know and even care about the demise of the Denisovans whether or not humans were involved. Or imagine that we discover that millions of years ago a failed extraterrestrial expedition attempted to establish a colony on Earth. We would still have reason to assemble the story of that doomed colony.

Moreover, self-centrism also fails even as an account of the partial reasons we undoubtedly do have for caring about aspects of the past. There is something disturbing in the idea that, say, African Americans have special reasons for caring about slavery only in so far as this is required for them to understand themselves or sustain their identity. Rather, having a certain identity may give one agent-relative reasons to be especially concerned with what is independently significant. This is how partiality works in more familiar contexts. I should worry about my friend's grave illness not because our friendship is part of my identity and therefore something I need to maintain for my own sake, but because, as a friend, my friend's agony deserves my special attention. The ultimate source of these partial reasons is my friend's agony, not our friendship, let alone my identity—what our friendship does is link these reasons to me, and it also amplifies them (for the similar point that ‘personal’ value depends on prior ‘impersonal’ value, see Raz [2001]).

4. Disinterested History

We need an account of the value of history that is not instrumentalist or present-centered. Historians who reject such views often assume that the alternative must be a view of history as disinterested and value-free, aiming solely at historical knowledge for its own sake. Geoffrey Elton, for example, held that the study of history is ‘autonomous and justified within itself’, and it is therefore a ‘cardinal error to reverse this process and study the past for the light it throws on the present’ (Elton Reference Elton1967: 66). This is a view of history that is often associated with the professional historians of the nineteenth century and encapsulated in Leopold von Ranke's famous remark that his aim is to ‘simply show how it really was’(cited in Braw Reference Braw2007: 47).

Elton was one prominent twentieth-century defender of the disinterested view. Another was Michael Oakeshott, who sharply distinguished the ‘historical past’—the concern of the true historian, who is interested in the past for its own sake—from the merely ‘pragmatic past’, which regards history in relation to the present (Oakeshott Reference Oakeshott1962). Oakeshott went so far as to assert that history must be ‘wholly without relevance to practical life’ (Oakeshott Reference Oakeshott1933: 157) and that the true historical masterpiece ‘releases us from the burden of history as the intellectual and moral preface to the contemporary world’ (cited in Dray Reference Dray1968: 33).

One critic of this view ridiculed it as ‘own-sakism’ (Jenkins Reference Jenkins1995: 70). But it is a fitting label, and I will use it without any derogatory connotations. Although the own-sakist view is often presented as a rejection of value claims, the claim that we should seek historical knowledge for its own sake is nevertheless best understood in value terms:

Own-sakism seems to me superior to instrumentalist and other present-centered accounts. Own-sakism focuses on the past qua past. It identifies reasons for concern with the past that are independent of contingent utility or our identity—reasons we would have regardless of who we are.

But own-sakism is nevertheless mistaken. First, own-sakists assert that historical knowledge is valuable in itself, but they usually offer little in defense of this claim. Now that is fair enough. If historical knowledge is intrinsically valuable, perhaps there is nothing more to say. But the force of this reply depends on there not being more to say; and as we shall see, there is.

Second, own-sakism can only account for our cognitive interest in the past. It cannot explain the many other ways in which the past matters to us noninstrumentally. We can see this when we consider blank slate. To be sure, to accept the extraterrestrials’ offer would be to give up a hard-earned store of valuable knowledge. However, this loss would surely be balanced by the great gains in other kinds of valuable knowledge (which also happen to be vastly more useful). And own-sakism cannot explain why accepting the extraterrestrials’ offer would be a moral failure.

Finally, the claim that historical knowledge is intrinsically valuable is presumably not the claim that any knowledge about the past has such value or at least not that all such knowledge is equally valuable—what Heydrich said at the Wannsee Conference and what he had there for breakfast. Own-sakists will reject as a caricature Nietzsche's characterization of the ‘antiquarian’ historian as an obsessive hoarder of facts and artifacts who lacks the ‘discrimination of value and… sense of proportion which would distinguish between the things of the past in a way that would do true justice to them’ (Nietzsche [Reference Nietzsche1874] 1997: 74). But own-sakists owe us an account of why certain kinds of historical knowledge are more important than others, and if historical inquiry is completely autonomous, they cannot do so by appealing to external considerations.

The old debate between own-sakism and instrumentalist and self-centrist views typically assumes that either we purge historical discourse of all valuation, explicit or implicit, or our interest in the past must be entirely pragmatic, or self-centered. Oakeshott, for example, thought that ‘the categories of “right” and “wrong”, “good” and “bad”, “justice” and “injustice” etc. relate to the organization and understanding of the world in respect of its relationship to ourselves’ and therefore ‘to inquire into the moral value of past conduct is to relapse into a practical attitude toward the past’ (Oakeshott Reference Oakeshott1962: 144, 164).

Oakeshott therefore held that historians should refrain from making statements such as ‘It would have been better if the French Revolution had never taken place’ or even ‘He died too soon’ because such statements reflect ‘contemporary values’ (1962: 163).

But this is a false dichotomy. Our value judgments about the past need not express a ‘practical attitude towards the past’, and although our value judgments about the French Revolution reflect our ‘contemporary values’, that is no different than saying that our judgments about Caesar's reasons for crossing the Rubicon or, for that matter, a past sighting of a comet reflect our ‘contemporary’ psychology and astronomy.

The important point, though, is that we should distinguish value from evaluation and evaluation from practical interests. We can then see that a rejection of instrumentalist or self-centrist accounts need not force us to conceive of history as a value-free and purely cognitive enterprise. Conversely, recognition of the fact that our thinking about the past is pervaded with value need not force us to think of it as merely the projection of our shifting practical interests. A third option opens up: our concern with history and the value of historical knowledge is grounded in prior value without being instrumental or otherwise self-centered. The rest of this paper will sketch such an approach.

Just to be clear, I have no intention to deny that instrumental and self-centered considerations and even curiosity can be reasons to look backward. What I deny is that such reasons—whether considered separately or pooled together—can capture our most fundamental ground for caring about the past.

5. Back to the Beginning

Herodotus begins his Histories by telling us that he engaged in this inquiry (ἱστορία)

And Tacitus later explained that his purpose ‘is not to relate at length every motion, but only such as were conspicuous for excellence or notorious for infamy. This I regard as history's highest function, to let no worthy action be uncommemorated, and to hold out the reprobation of posterity as a terror to evil words and deeds’ (Tacitus Reference Tacitus1942: 137).

These are some of the earliest accounts of the aim of history: history as concerned with the great deeds of the great men of the past. This may seem, well, a touch out of date. But I think Herodotus and Tacitus were closer to the truth than many later and contemporary authors.

Let us first set aside the instrumental element that is admittedly present in these remarks: the aim of explaining the (then present) conflict between the Greeks and the Asians in Herodotus or that of deterring future evil by the promise of posterity's reprobation in Tacitus. The main idea, though, is simply that we should preserve the memory of those past actions and events that were ‘conspicuous for excellence or notorious for infamy’, or otherwise ‘great and wonderful’. And we should not merely remember those actions, but remember them as glorious or notorious—and admire or disapprove of them accordingly.

Commenting on Herodotus’ opening words, Arthur Danto (2003) wonders whether they suggest a noninstrumental account of the value of history. Perhaps, Danto tentatively proposes, we should ‘shift ground entirely from present to the past itself’ and consider the possibility that we have ‘some kind of duty to the past itself to find out its truth’:

This suggestion is in the right direction. We should shift from the present to the past itself. But not because we have duties to past persons as we might have a duty to keep a promise to someone who died (though see Smilansky Reference Smilansky2021) or because we should shift from a focus on value to us to focus on value to the long dead—as if we benefit them by keeping a record of their deeds. Rather, we should shift to a focus on the value of these past people and of the remarkable things they did.

What Herodotus and Tacitus are saying is that there are past people and actions that stand out as remarkable, whether positively or negatively, and these people and actions should therefore be remembered. But we can abstract away from this focus on great acts and the glory or infamy they deserve. What we then get is the idea that some things in the past—people, acts, events, artifacts—possessed value, whether good or bad, and some possessed significant value. We should remember those past things because of their value and in ways that fittingly respond to that value. Herodotus and Tacitus got this right, even if they developed this idea in terms of an axiology we no longer share.

6. The Significance of the Past is Past Significance

According to what I will call valuism, the past matters because it contains a vast amount of things that matter: because it contains suffering and death and injustice and virtue and achievement and all that.

A large-scale war, for example, contains a vast amount of evil—think, if you need an illustration, of the tens of thousands of Tommies being shredded to pieces by machine gun fire and artillery in the first few hours of the Somme Offensive. Or think of the horrors of chattel slavery— of, for example, the African slaves who were thrown overboard by the crew of the Zong in order to claim insurance. That past disvalue calls for a response. It matters that all those people died so horribly. And it should therefore matter to us that they died. This is why we should know and care about the Great War and the transatlantic slave trade.

Events in the past matter because they possess or contain value, as a war does, but they can also matter by causing or making a significant difference to value—think of a declaration of war as well of the sequence of events and other factors that led to such a declaration. Our concern with the past primarily, I believe, reflects our recognition of such past value. The value of historical knowledge and of the inquiry that seeks it is largely derived from this prior significance. We have reason to find out about (and to preserve the memory of) significant past events—and to know of them as significant—because that is a fitting response to that past significance. Knowledge of such past events is also, of course, a necessary condition for further fitting responses.

Because these backward-looking reasons are not self-interested, because they often have a strong impartial character, and because ignoring or violating them often calls for a recognizably moral response, it seems appropriate to describe them as moral reasons. I concede that it is not clear how such reasons can be accommodated by familiar ethical theories. They are, for example, incompatible with most consequentialist views. But this seems to me a problem for these theories, not for valuism. (One consequentialist view that is compatible with valuism is Hurka's (2000) recursive account, on which it is itself good to love the good and hate the bad—though Hurka himself holds that such attitudes to value in the distant past have only little value.)

7. Making Sense of Our Concern with the Past

Valuism, I believe, offers the best account of our sense that we ought to care about the past. Valuism can explain why it would be wrong to turn our back on the past even when we remove or dramatically outweigh any of the instrumental benefits of backward-looking concern and even if we were offered considerable cognitive benefits in exchange, such as deeper knowledge of the working of the universe. Valuism can also explain why turning our back on the past would be morally wrong rather than merely a prudential error.

Valuism is also best positioned to explain the noncognitive aspects of our concern with the past: memorial services held for the fatalities of now distant wars or the monuments and museums dedicated to past atrocities or struggles against injustice as well as our emotional responses to the past (for discussion of such impartial emotional responses, see Smilansky Reference Smilansky2013; Kahane Reference Kahane2019). These are all responses to significant instances of past value. But valuism does not merely capture an optional overlay on top of an independent cognitive interest. It also explains the character and focus of that cognitive interest. In particular, it offers a straightforward account of what about the past is worth knowing about—an account that seems to me largely to track what historians and the rest of us deem to be historically important: historians focus more on large-scale wars than on isolated outbursts of violence, more on the sources of prolonged periods of prosperity than on, say, changes in grammar or decor. And while more recent historiography has shifted from the traditional focus on diplomacy and conflict to pay more attention to the ‘microhistory’ of the daily lives of ordinary people (see, e.g., Zemon Davis Reference Zemon Davis1984), this seems to me to reflect the realization that it is in such lives that most value in the past actually resides.

Finally, valuism is better placed than its competitors to answer the skeptical challenge. Unlike instrumentalism, it does not appeal to self-interest or rely on vague claims about supposed benefits. Unlike own-sakism, valuism does not simply assert that historical knowledge is intrinsically valuable—a value the skeptic about history obviously cannot see and will not accept. Moreover, since own-sakism understands that value in purely cognitive terms, it is hard to see how it could defend our backward-looking practices against competing moral pressures. By contrast, valuism offers moral reasons to be concerned about the past and, moreover, reasons that do not appeal to new and contested values. It appeals to the very same values that the skeptic would already accept—namely, the badness of suffering and injustice, goodness of happiness and achievement, and so on. The skeptic may deny that, say, suffering and injustice still matter when they move to the past. But this is a different challenge, a challenge that, as I will now argue, we can also answer.

8. Value and Time

Valuism says that we should care about the past because the past contains value, and we should recognize and fittingly respond to that value. Now there is nothing remarkable about the idea that value calls for a fitting response. On some views, for something to have value just is for there to be reasons to have certain fitting attitudes towards it (e.g., Scanlon Reference Scanlon1998), but we do not need to accept such a view to agree that when something possesses value then there are such reasons to respond fittingly to it.Footnote ¹

Now if we thought that the only way to respond to value is to promote it, then the very idea of caring about value in the past would make no sense because nothing we can do can change the past. But such a view is implausible. By contrast, if fittingly responding to value primarily involves having certain attitudes towards it (Scanlon Reference Scanlon1998; Scheffler Reference Scheffler2011), then it is obvious that we can, and routinely do, respond to past value: just think of regret, gratitude, grief, and similar backward-looking sentiments. Even when these attitudes are focused on recent past events in our own life, they are still responses to past value. And we often enough have such attitudes to properly historical events even if they typically focus on the fairly recent past.

A skeptic might object: ‘Suppose we do have reasons to care about what has value. But what has value are things in the present. Things in the past are things that had value. Present suffering is bad, and matters. Past suffering was bad, and no longer matters. We need not care about the past.’

It is bad when someone experiences agony. When that agony is in the past, it is, of course, true that it is not bad now. But it would be odd to say that for that reason this agony is no longer bad. No: it is a bad thing that is now in the past. It is a bad thing that was, not a thing that was bad. Suffering is bad in virtue of what it feels like. And that does not change when it moves to the past. And that past badness (badness in the past) continues to pull its weight in familiar ways, for example, by making a difference to how good or bad a life is overall. In the same way, the badness of suffering experienced in the distant past is not erased: it continues to make a difference to how good or bad the world is, overall. Moreover, the idea that we should forget the past and focus instead on the present and future is self-defeating. Do we really genuinely value something if we also believe that it would be just fine to completely forget all about it as soon as it is gone?

In thinking about these questions, it might help to briefly consider possible views about the metaphysics of time. On eternalist and growing block views of time, the past still exists. Past pain really is there, just in a different temporal location. On such views, it would be hard to see why we would not have reason to care about past value, however temporally distant— just as a catastrophe does not stop mattering just because it is many miles away—as if you could erase its significance by driving away.

Presentists do claim that only the present exists. Homo erectus, the Lisbon earthquake, and Kublai Khan are not still there, somewhere in the past. They are all gone. Now such a view may support the view that the present is especially important in some sense (Schopenhauer Reference Schopenhauer2007: 19). But I do not see how presentism could plausibly support the view that nothing in the past matters. If only what exists matters, then for presentists nothing in the past or future could matter in itself—not even if it is just a moment ago or a moment from now. We would have to give up not only all backward-looking attitudes, such as gratitude and grief, but also forward-looking ones, such as fear and hope. This seems absurd. And even if I feel relief that a pain is now gone, it still makes my life overall worse, as does the misery I felt as a child and other misfortunes I had experienced.

Past value in my life continues to matter, and therefore past value can matter. And past value relating to others clearly does matter. We are horrified to read about civilians being gassed. We should be horrified. But even if we are well-informed, we will find out about such an evil hours or days after it happened. And if it makes sense to be horrified about evil that took place hours ago, then at least so far as presentism goes, it should also make sense to be horrified about the Zyklon B that the SS's so-called Hygienic Institute delivered by ambulance.

Value in the past still matters. This does not mean that it matters in the same way as present and future value. We can act to prevent future pain, but nothing we can do will erase past pain (nor, for that matter, present pain, narrowly conceived). Past pain may also matter less. We certainly care more about the present and the near future than about the distant future, and we care more about the future than about the past. We prefer, as Parfit pointed out, a pain in our past to a pain in our future (Parfit Reference Parfit1981). Valuism is compatible with such a bias. It claims that the past matters, not that it matters as much as the present and future.

We have a further bias. We care more about the recent past than about the more distant past (Hurka Reference Hurka2000; Orsi Reference Orsi2013: 692). We engage in backward-looking temporal discounting. It is not easy to justify such discounting. Certainly presentism will not help here because it cannot distinguish between the near and distant past. Discounting in the forward direction is often justified in epistemic terms—typically, the further an outcome is in the future, the less confidence we have that it will actually occur. Such considerations obviously do not apply in the backward direction. While we often know less about the distant past, this is not always true. And what we do know about the past we often enough know with something close to certainty. Those impressed by such considerations might want to understand valuism in temporally neutral terms (Greene and Sullivan Reference Greene and Sullivan2015). But, again, valuism is perfectly compatible with the view that the more distant past matters less.

9. Too much to value?

From the objection that the past contains nothing for us to value, we can turn to the objection that it contains too much value. After all, history is crammed with a mind-boggling number of individuals, acts, events, and other things that matter. We could not possibly care about or even begin to know this immensity: all those massacres and intellectual breakthroughs, the joys and ecstasies, and, especially, all those people. And we cannot have reasons to do what we cannot do.

It hardly follows from this, however, that we have no reason at all to respond to the value the past contains or even that it is entirely at our discretion what in the past to respond to. What we can do is break down the totality of past value into its significant parts—to broad patterns or causally linked clusters of value—and respond to their value. We can start with those parts that contain the greatest amount of value and then, working our way down, move either to smaller parts or to parts within parts and so forth. How far down we can go would depend both on the actual distribution of value in the world and on our own limits. We cannot cover everything, but we can cover plenty.

That history cannot be and need not be an exhaustive chronicle of everything that has happened in the past is a truism of historiography. We need a criterion of selection. Valuism offers such a criterion: it distinguishes between the valueless and the valuable and between what has (or affects) more or less value (recall Herodotus singling out ‘great and wonderful acts’). People need to know about the Holocaust—about Mesoamerican butterfly sacrifice, not so much.

It might be objected that the problem is not that we could not possibly know about and value every past event of value, but that we need not. Yes, to value something is, roughly, to care about it, but we have no reason whatsoever to value most of the things we judge to have value (Raz Reference Raz2001; Scheffler Reference Scheffler2011; Matthes Reference Matthes2015; Bulgarian history is one of Scheffler's examples of things that he regards as valuable but need not value). We can recognize the value of entomology or origami without feeling the slightest normative pull to engage in these activities—and not just because we do not have the time. Shifting back to the context of the past, to hold that we should each respond to as much past value as we can is to construe the normative pull of the past on us in an implausibly demanding way.

This objection would have considerable force if valuism claimed that each and every one of us has a reason to care robustly about and engage with each and every thing of value in the past, at least to the best of our ability. But valuism makes a much weaker claim.

First, recall that valuism is in the first instance only a claim about the reasons that we have together. Some of these reasons apply to us as individuals. While few of us are constantly looking backward, ignoring what is around us, we do expect people to have a fairly broad knowledge of significant historical events. History is allocated its space in the curriculum, archives and museums are given their budgets. And we are all expected to pause, on certain days or weeks, to mourn or celebrate aspects of the past—such events are precisely designed to bracket pressing present concerns, allowing us to turn our attention backward.

But our reasons to respond to past value more often take a collective form. We respond to the value in the past when some of us fittingly respond to it. That is to say, we also engage in a kind of division of cognitive and evaluative labor. The core division is between a majority largely focused on present concerns and historians and others who are dedicated to backward-looking projects. Then there is the internal division of labor within the discipline of history into the study of different periods and areas or a focus on diplomacy, economy, or culture, and so on. In this way, we collectively respond to reasons to care about the past in ways that not only greatly transcend our individual ability to do so, but also prevent that mass of past value from making oppressive demands on our individual lives.

Second, valuism does not claim that we need to value past events of value, only that we should respond to them in fitting ways. In plenty of (though not all) cases, the fitting response to past events is a much paler thing than the degree of emotional investment that is involved in robust valuing. It will often involve instead pro tanto preferences that, for example, things had gone differently than they did (Kahane Reference Kahane2019). And at the far end, the response may just involve knowing about a past event so long as one knows it because of its value and as involving that value.

It might be objected that knowing and similar cognitive attitudes are not pro or con attitudes and therefore cannot be fitting attitudes to past value. In reply, consider the following: (i) In order to respond fittingly to past value, we need to know enough about it; to lament past suffering without actually knowing anything about the past is not a fitting response to that past suffering. (ii) I did not say that merely knowing about value in the past is sufficient. I said we need to know about past value as possessing that value and because of that. Now, such value judgments are typically accompanied by fitting pro or con attitudes. These can involve a felt affective response even if a faint one. However, as I just wrote, they can also involve unfelt backward-looking preferences. On some internalist views, there may be a conceptual link between value judgments about the past and such further attitudes. But even without such a link it still seems that, as a matter of human psychology, such attitudes typically do accompany our value judgments even when they are about the past. (iii) Even if or when such judgments about the past are not accompanied by such further attitudes, recognition of that past value can still be expressed by acceptance of norms about what attitudes should not accompany such judgments. Giggling at a memorial service is a fault, as is remembering a past catastrophe only as the punchline to a crude joke. (iv) The cognitive states may themselves serve as fitting responses to past value.Footnote ² The premature death of a close friend may no longer move me now or not move me as much. But it is important that I remember that friend and her death—and if I just forgot all about her, perhaps even her name, that would be a serious failure. This is, in part, why many people have a horror of being forgotten once gone (Margalit Reference Margalit2002). As I wrote above, such cognitive states are valenced because they either involve value judgments or are formed, and maintained due to judgments about value. Knowledge of the past, however precise and exhaustive, that is completely severed from any evaluation is not a fitting response and, in my view, of little value. And if, say, the last record of some historical atrocity could be preserved only in order to serve as sadistic entertainment, then I think it would be better if that record was lost.

Third, valuism only makes a claim about the existence and source of a range of backward-looking reasons. The strength of these reasons compared to the many other reasons we recognize is a further question. As I develop valuism here, it offers an account of reasons we already recognize—reasons reflected in a range of institutions, practices, and attitudes. These are already demanding in the sense that they involve a cost, as we saw earlier, because they come at the expense of other things we can do or invest in. But this is a cost most of us are willing to bear. Valuism explains why we are not wrong to do so.

A related objection is that my use of examples of past atrocities such as the Holocaust allows me to overstate the case for backward-looking moral reasons to care about the past. Such horrific instances of moral evil may indeed generate impartial reasons to look backward. But much of the past—the terracotta figurines of the Nok culture or the etiquette of the Heian period—falls outside the moral domain and does not generate such reasons.Footnote ³

I did largely focus on examples of events that involve great value or disvalue. These, however, also include major intellectual breakthroughs and cultural achievements—events whose value is not clearly moral but that, I believe, still generate impartial backward-looking reasons. Such past events involving immense value are ones where these backward-looking reasons seem to me to apply to all or at least to many of us and where they also call for something close enough to robust valuing.

Other historical events—including many that have a moral dimension—involve less value and thus generate weaker reasons. These reasons call for a largely cognitive response and apply to us only as a collective rather than to each of us. And in some cases, responding to them may be only supererogatory. Even here, however, the normative pressure exerted by the past can potentially constrain our individual actions. I do not need to know anything about medieval rug weaving techniques. But it would be wrong for me to knowingly use our only evidence about them in some papier-mâché construction—thereby perhaps forever severing our link to a valuable aspect of the past.

And we can agree, finally, that there are numerous past events that involve too little value to merit anyone's attention—think of a Phrygian's bad breakfast.

Suppose, though, that this last objection succeeds, and it is only events of moral significance that generate the relevant impartial backward-looking reasons. Since Gibbon only slightly exaggerated when he wrote that history is ‘little more than the register of the crimes, follies, and misfortune of mankind’ (Reference Gibbon1846: 85), valuism would still justify and potentially also explain a great deal of our concern with the past. Moreover, to understand past moral evil we would need to know about and understand much else about the past. This narrowing of the view would therefore not significantly affect its implications.