A liberal realist answer to debunking skeptics: the empirical case for realism

Abstract

Debunking skeptics claim that our moral beliefs are formed by processes unsuited to identifying objective facts, such as emotions inculcated by our genes and culture; therefore, they say, even if there are objective moral facts, we probably don’t know them. I argue that the debunking skeptics cannot explain the pervasive trend toward liberalization of values over human history, and that the best explanation is the realist’s: humanity is becoming increasingly liberal because liberalism is the objectively correct moral stance.

1 Debunking arguments for moral anti-realism

1.1 Three skeptical accounts of moral belief

Can we ever know what is objectively right or wrong, good or bad? Moral realists answer yes. Anti-realists answer no: they believe that either there are no objective moral truths, or we have no knowledge of these truths.

Anti-realists have often defended their position by appealing to one or another debunking explanation for moral beliefs. According to debunking explanations, our moral beliefs are chiefly or entirely produced by psychological mechanisms that are not suited to arriving at objective truths; hence, even if such truths exist, we probably don’t know them. In principle, indefinitely many kinds of debunking theories are possible. For instance, if it turned out that your moral beliefs were implanted in your mind by a capricious hypnotist, those beliefs would thereby be ‘debunked’. In practice, however, the debunking explanations seriously advanced have generally been of just three sorts.

First, some hold that an individual’s moral beliefs are entirely a function of that individual’s emotions and desires, understood as purely non-cognitive states. Thus, David Hume states that “morality is determined by sentiment” and that “to have the sense of virtue, is nothing but to feel a satisfaction of a particular kind from the contemplation of a character.”¹

Second, some say that our moral beliefs are chiefly or entirely the product of our culture. For example, those raised in strongly Christian or Islamic communities today often judge homosexuality to be morally wrong. But had they been born in ancient Greece, they would likely have accepted homosexuality.² Furthermore, many theorists deny that our culture reflects any objective evaluative facts; it is just the set of practices that we happen to have adopted, no better or worse, objectively speaking, than any other set of practices.³

Third, in recent years, evolutionary explanations of moral attitudes have grown in popularity.⁴ For example, on the assumption that our genes influence our moral beliefs, we can understand why most people believe in a strong moral obligation to care for one’s own children, but no parallel obligation to care for unrelated persons: in our evolutionary past, ancestors who accepted such an obligation tended to leave behind more surviving offspring than those who rejected any such obligation. Enthusiasts for evolutionary psychology claim that there are many other cases in which common moral attitudes are best explained by natural selection.

Perhaps the most plausible debunking theory is a combination of the above three accounts: perhaps moral judgments are caused by emotions, desires, or other non-cognitive states, and these non-cognitive states, in turn, are products of both genes and culture. Some recent work in psychology lends credibility to this hypothesis. Psychologist Jonathan Haidt has found evidence that moral judgments are largely a product of gut reactions, with moral reasoning mostly functioning as post hoc rationalization.⁵ The dramatic differences in moral beliefs across societies lend credence to the claim that moral beliefs are largely produced by culture. That there is also a large genetic component is suggested by recent research finding that the heritability of political orientation is approximately 0.53.⁶

1.2 Why the debunking theories engender skepticism

Debunking accounts of moral belief could not show that there are no objective moral facts. They might show, however, that our belief-forming mechanisms are ill-suited to identifying objective moral facts.

Suppose that our moral beliefs are solely or chiefly produced by emotions. Emotions are typically not reliable guides to objective facts. Given the plausible assumption that knowledge requires a reliable belief-forming mechanism, this supposition would tend to suggest that our moral beliefs do not constitute knowledge of any objective facts; hence, that either there aren’t any objective moral facts, or there are but we don’t know them.

But suppose one thought that (some) emotions are evaluative representations.⁷ For example, perhaps to feel anger is, among other things, to feel that an injustice has been done. On this view, it would not be strange that emotions should be reliable guides to evaluative facts even though they are poor guides to non-evaluative facts. For this reason, it will not suffice merely to observe that emotions are typically (in descriptive matters) poor guides to objective facts. We must rather consider the causes of our relevant emotions (the emotions that are plausible sources of moral beliefs) and ask whether these causes make it plausible to view the resulting emotions as reliable guides to the objective evaluative facts, if such facts exist.

Thus, proponents of the first kind of debunking explanation for morality are naturally driven to elaborate their theory by appeal to one or both of the other debunking explanations: perhaps our moral emotions are caused by our culture or our genes. If so, there is no good reason to suppose that our moral beliefs will reflect the objective moral facts, even if such facts exist. Begin with the case of culture: there is so much variation in moral beliefs across cultures that culture, in general, cannot be a reliable guide to objective moral truths. Furthermore, there is no independent reason to think that our culture in particular should have the correct moral beliefs.

Consider next the case of beliefs explained by natural selection. What nature selects for is reproductive success. If correctly identifying the objective moral truths does not contribute to reproductive success, then there is no reason why evolutionary processes should have endowed us with the capacity to identify those truths. And there seems to be no reason independent of our current moral beliefs to suppose that knowing the objective moral facts would have contributed to reproductive success. Indeed, because moral properties seem to have no causal impact on the physical world, it is hard to see how moral reliability could impact reproductive success.⁸

The debunking explanations for morality thus lead us to doubt that our moral beliefs could plausibly constitute knowledge of objective facts.

2 A modest liberal realism

2.1 Three realist accounts of the source of morality

Most moral realists would dispute the anti-realists’ characterization of the source of moral beliefs. The realist need not and should not claim that no moral beliefs are caused by emotions, culture, or genes. We need not make that claim, since realism is not the view that all moral beliefs constitute knowledge of objective facts; realism holds only that some moral beliefs constitute knowledge of objective facts. Therefore, it need only be held that some moral beliefs derive from reliable belief-forming mechanisms.

Realists have advanced at least three accounts of the nature of these reliable mechanisms. First, some have held that we have empirical knowledge of moral facts, through observation or inference to the best explanation.⁹ Second, some have held that there is a dedicated moral sense, that is, a faculty that functions specifically to cognize moral facts (or evaluative facts more generally) and nothing else.¹⁰ I shall not discuss these first two views further, however, because I find them improbable.¹¹

The third account, by far the dominant one among intuitionists over the last century, is a rationalistic intuitionism.¹² This account holds that our moral knowledge is of a kind with our other a priori knowledge, such as our knowledge of mathematics and of necessary truths of metaphysics.¹³ The nature of this other a priori knowledge is a matter of controversy. Fortunately, we need not resolve that controversy here. Rationalist intuitionism simply needs the assumption that there is some substantive, a priori (non-evaluative) knowledge. Knowledge requires a reliable belief-forming mechanism (as the debunking arguments of Sect. 1 suppose), so there must be a reliable mechanism that produces these non-evaluative a priori beliefs. Whatever that mechanism is, the rationalist intuitionist maintains, that mechanism is also capable of producing some moral beliefs. This is why it is plausible to think that some moral beliefs might be sufficiently reliable to qualify as knowledge.

There are some who deny the existence of a priori knowledge altogether.¹⁴ Others admit only analytic a priori knowledge, which is supposed to be explicable purely in terms of one’s understanding of the meanings of words.¹⁵ Here, I assume that these empiricist views are wrong; I assume, that is, that there are genuine cases of synthetic, a priori knowledge.¹⁶ The target of this paper is the theorist who thinks there is something specially problematic about objective moral knowledge, such that, even if synthetic a priori knowledge in general is possible, our knowledge of morality would not be an example thereof. This is precisely the sort of position supported by the debunking arguments discussed in Sect. 1.

2.2 Liberal realism

I am not only a realist but a liberal realist: I think that the objectively correct values are liberal values. When I speak of liberalism, I intend, not any precise ethical theory, but rather a certain very broad ethical orientation: liberalism (1) recognizes the moral equality of persons, (2) promotes respect for the dignity of the individual, and (3) opposes gratuitous coercion and violence. So understood, nearly every ethicist today is a liberal. But while this broad orientation is mostly uncontroversial today, this does not render the category of liberalism uninteresting, for as we shall see, human history has been dominated by highly illiberal views.

The three aspects of liberalism named above are not simply three unrelated moral commitments. Liberalism is a coherent ethical perspective. The idea that individuals should be treated with dignity fits together with the idea that individuals are moral equals, and that one should eschew violence and coercion against the individual. This helps to explain why it is a priori plausible to think that, if there are objective values, liberalism might be the objectively correct ethical orientation. This is not to deny that there might be other reasonable candidates for correct ethical orientations; it is only to say that liberalism ought to be counted high on the list of initially plausible candidates.

2.3 Modest realism

My realist view is modest in at least three respects. First, I do not hold that all or most moral beliefs constitute knowledge; I leave it open that only a small minority of moral beliefs might constitute genuine knowledge of objective moral facts.

Second, though I deny that culture and genes provide a complete explanation for our moral beliefs, I do not doubt that culture and genes play an important role in explaining moral beliefs. Even if we have rational, ethical intuitions that are sometimes reliable, these intuitions are not anything close to the whole explanation for our moral beliefs.

Third, though I insist that moral knowledge is possible, I do not claim that it is easy. Some moral knowledge might require careful reflection and skill in judgment. Some might emerge only from a long and difficult process. Our society may need to accumulate its moral wisdom over a period of centuries, and a great deal of moral knowledge may yet elude us. All of this is compatible with the claim that moral knowledge rests ultimately on intuition.

3 The phenomenon of moral progress

I shall contend that certain empirical facts are difficult to explain on any of the debunking theories mentioned in Sect. 1, and that by contrast, the modest liberal realist can offer a plausible account of the data. Roughly, the data in question concern the development of moral values over the course of human history.¹⁷

3.1 War and murder

In most societies throughout history, killing has been far more common than it is in our society today. The trend toward lower rates of violence is visible on the scale of decades, centuries, and millennia, it is consistent across countries, and it applies to both murder and warfare. These facts have been extensively documented elsewhere.¹⁸ Here, I will just mention one striking fact: in primitive societies, it is estimated that between ten and thirty percent of all deaths come at the hands of other humans, with most of these being deaths of men in war. Figure 1 shows estimates for the percentage of deaths due to war in seven contemporary primitive societies studied by anthropologists.¹⁹ Figure 2 shows estimated deaths due to war and murder in sixteen prehistoric primitive societies; these estimates are based upon sixteen archeological sites where human remains were found and examined for signs of death at the hands of other humans.²⁰ In each figure, the death rate for Europe and the United States in the twentieth century is shown at the bottom for comparison.

Fig. 1

Deaths by war in contemporary primitive societies

Fig. 2

Deaths by violence in prehistoric societies

There are many factors that may have contributed to the decline in violence.²¹ But one is of particular interest here: there has been a dramatic shift in human values over history.²² In primitive societies, including our own society in earlier centuries, physical combat was often regarded as glorious, honorable, and manly. Those who conquered others through violence were therefore honored—witness Alexander “the Great” and Peter “the Great”. Today, such leaders would more likely be reviled as criminal aggressors. Consider these sentiments from prominent thinkers of the past²³:

You say it is the good cause that hallows even war? I say to you: it is the good war that hallows every cause. (Friedrich Nietzsche, 1885)

If war made men brutal, at least it made them strong; it called out the qualities best fitted to survive in the struggle for existence. […] War, with all its horrors, could purify as well as debase [….] (Henry Adams, 1891)

Would not the end of war be the end of humanity? War is life itself. Nothing exists in nature, is born, grows or multiplies except by combat. (Emile Zola, 1891)

Murder has shown a marked decline over the centuries. In Europe, the murder rate has declined from about 35 per 100,000 population per year in 1300 A.D. to about 3 per 100,000 today.²⁴ Again, many factors may have contributed to the decline—among them the changing attitudes toward murder. Men of the past perceived many more things as reasons for killing.²⁵ Consider that in 1804, former American Treasury Secretary Alexander Hamilton died in a duel with sitting Vice President Aaron Burr. The duel was fought to settle a dispute over some disparaging remarks Hamilton had allegedly made about Burr.²⁶ Such behavior on the part of respected men would be unthinkable today.

3.2 Torture and execution

Governments of the past would execute citizens in gruesome ways at the drop of a hat. In the middle ages, capital offences included sodomy, gossip, stealing cabbages, picking up sticks on the Sabbath, talking back to one’s parents, and of course witchcraft.²⁷ Execution methods included burning at the stake, drawing and quartering, boiling, and sawing. The last of these methods is depicted in Fig. 3.²⁸

Fig. 3

A medieval execution

Torture was accepted as a method of investigation. A suspected witch might be tortured until she (1) confessed to witchcraft, and (2) named the other witches whom she presumably knew about. The other accused witches could then be tortured to verify their guilt. Happily, over the past 400 years torture has been abolished throughout Europe and most of the world.

3.3 Slavery

Slavery has been accepted in many societies throughout human history and was often endorsed by the moral authorities of the day.²⁹ Aristotle, considered by many the greatest philosopher of all time, endorsed waging war to capture slaves:

But the art of acquiring slaves, I mean of justly acquiring them, differs both from the art of the master and the art of the slave, being a species of hunting or war.³⁰

The Bible, long considered by many a font of moral wisdom, advised readers on how severely one may beat one’s slaves:

If a man beats his male or female slave with a rod and the slave dies as a direct result, he must be punished, but he is not to be punished if the slave gets up after a day or two, since the slave is his property.³¹

No sane author could write such passages today. Evidently these passages came across very differently in the societies for which they were originally written.

Over the past 200 years, the world saw a wave of abolitions, so that today, slavery is illegal in every country in the world. Figure 4 shows when slavery was abolished in 49 selected countries.³²

Fig. 4

Timeline for the abolition of slavery

3.4 Racism and sexism

Norms of the past were often unapologetically inegalitarian. In earlier decades, even in democratic societies, women were literally prohibited from voting, due in part to their supposed inferiority. That situation started to change around 1920 (see Fig. 5).³³ Though it remains true today that many desirable professions are disproportionately occupied by men, in earlier ages women were completely barred from most professions.

Fig. 5

The spread of women's suffrage

Even after slavery was abolished, Americans continued for decades to impose policies that were severely and explicitly racist. Black citizens were expected to ride at the back of the bus, use separate drinking fountains and restrooms, and attend separate schools—all to prevent contamination of whites by blacks. These laws ended in the United States in the 1960’s.³⁴ Since that time, attitudes have shifted dramatically: if someone today were to advocate the sort of laws that actually existed 50 years ago, listeners would take it as a tasteless joke at best.

3.5 Democratization

The vast majority of governments in history have been dictatorial. In the year 1800, there were, by modern standards, no genuine democracies. Since then, democracy has spread to about half of all the world’s countries and appears poised to take over the globe (see Fig. 6).³⁵ In the last 30 years, democracy has spread to about as many countries as it had reached during the previous 200 years.

Fig. 6

The spread of democracy

3.6 Decolonization

Throughout most of human history, building an empire through conquering other peoples has been viewed as a great achievement for a leader. The twentieth century witnessed two of history’s greatest empires: the British Empire and the French Empire. Both of these empires also collapsed in the last century, as nearly all the conquered peoples gained independence. In some cases, independence was gained through violence; in others, it was gained through non-violent protest movements.³⁶

The latter phenomenon is striking. Why didn’t the people conquered by Genghis Khan think to oppose Khan’s rule with non-violent demonstrations, general strikes, and civil disobedience? The answer is that if there had been a Gandhi in the time of Khan, he would most likely have been summarily executed and his head put on a spike by the conqueror whose moral legitimacy he’d sought to call into question.

Equally striking are those cases, almost all occurring in modern times, in which a dominant power freely granted independence to a weaker nation as a result of a popular referendum in the latter country. Thus did Guinea achieve independence from France in 1958, Malta from the United Kingdom in 1964, and Micronesia from the United States in 1986.

3.7 Summary

There has been enormous moral progress over human history. This progress is not just a matter of changing practices but of changing moral beliefs. Mainstream illiberal views of earlier centuries are shocking and absurd to modern readers. The trend is consistent across many issues: war, murder, slavery, democracy, women’s suffrage, racial segregation, torture, execution, colonization. It is difficult to think of any issue on which attitudes have moved in the other direction. This trend has been ongoing for millennia, accelerating in the last two centuries, and even the last 50 years, and it affects virtually every country on Earth.

This is among the most striking and important phenomena in human history. How are we to explain all this?

4 The failure of irrealist accounts

4.1 Liberal genes

In one trivial sense, an evolutionary account of ethics must be correct: human beings evolved; therefore, however our capacity for moral judgment works, that capacity is “a product of evolution,” in the same sense that our capacity for any sort of judgments is a product of evolution. This thesis of “an evolutionary origin for ethics” poses no threat to moral realism.

The kind of evolutionary thesis that poses a challenge to realism holds that our specific moral values are adaptations. For example, if the tendency to judge that adultery is wrong was selected for because that moral judgment promoted our ancestors’ reproductive fitness, this would cast doubt on the reliability of that moral judgment.

Recall that realism is committed only to the view that some moral beliefs constitute knowledge, whereas anti-realists hold that no moral belief constitutes knowledge of an objective fact. Thus, to support anti-realism using a debunking account of moral beliefs, the anti-realist must hold that all moral beliefs are adaptations or have some other presumptively unreliable source. In particular, then, the anti-realist must give a debunking account of liberal moral beliefs. Can such an account plausibly be given?

Suppose, first, that one holds that there is a gene or set of genes that produce a tendency toward liberal values, and that liberalism is adaptive. (Why? Perhaps because it enables one to get along with other people, and peaceful cooperation is usually more beneficial than attempted exploitation and conquest.) This might seem to explain, in evolutionary terms, why liberalism has spread over time.

But this does not account for the recency of liberal values. All of recorded history occupies only a few thousand years. It strains credibility that the adaptive set of values should have evolved during this brief period, having failed to evolve during the preceding two hundred thousand years that humans existed or the millions of years during which our primate ancestors existed.

To accommodate this, one might hypothesize that liberal values only recently became adaptive, at which point the liberal genes started to spread. During pre-civilized times, perhaps, liberalism was maladaptive. This theory, however, is extremely implausible empirically. One problem is that the shift in values has been far too rapid to be explained by biological evolution.³⁷ For example, the Jim Crow laws in the United States were abolished only in the 1960’s; before that, explicit racism was perfectly socially acceptable.

Another problem is that an adaptationist account of liberalization would have to work via the supposition that those with liberal values have in recent times had greater reproductive success than their ideological opponents. But there is no reason to think, for example, that in the 1960’s racists started having fewer children than non-racists and thus failed to pass on their racist genes, or that during the last 200 years, people who supported democracy started having more children than those who supported dictatorship.

4.2 Genes with variable expression

For the reasons given, the idea of accounting for ethical liberalization through genetic change is unpromising. The same genes, however, can sometimes be expressed differently in different environments. Thus, here is a second hypothesis to account for liberalization: perhaps we have a gene or set of genes with both of the following properties: (1) it inclines one toward illiberal beliefs if resources are scarce and survival uncertain, but (2) it inclines one toward liberal beliefs if one is well-off and secure. In that case, as a society advanced economically and its members became more prosperous and secure, the values of those members would become increasingly liberal.

While this hypothesis would conveniently solve the adaptationist’s problem, there doesn’t seem to be any reason to believe it. To begin with, it is not clear, on a theoretical level, why such a gene or set of genes should have been selected; it is not clear why liberal values would promote reproductive success for a well-off and secure person but not for one who is poor and insecure. In addition, in humanity’s illiberal past, those who were wealthy and secure (aristocrats) were typically not at all liberal; they were the ones oppressing the rest of their societies.

Here is a slightly different hypothesis: perhaps there is a gene that inclines one toward illiberal beliefs if one’s society as a whole is primitive and poor, but inclines one toward liberal beliefs if one’s society is advanced and prosperous. Again, it is unclear why such a gene would be especially advantageous, as compared with a gene that causes one to be liberal in all conditions, or illiberal in all conditions. Even if such a gene would be advantageous, there has not been sufficient opportunity for it to be selected, since for almost all of the history of the species, human beings have lived in poor, primitive societies. Humanity has not had enough experience with shifting between poor and prosperous, or primitive and advanced societies, for evolution to have designed special instructions governing what to do in an advanced, prosperous society.

4.3 The expanding circle

4.3.1 Singer’s account of moral progress

Consider next the account offered by Peter Singer, on which moral progress consists in an expansion of the circle of moral concern:

The idea of a disinterested defense of one’s conduct emerges because of the social nature of human beings and the requirements of group living, but in the thought of reasoning beings, it takes on a logic of its own which leads to its extension beyond the bounds of the group.³⁸

The basic idea seems to be that evolution selected for a tendency to consider impartially the interests of other members of one’s society (but not those outside one’s society), because this enables one to cooperate harmoniously with those others. However, this tendency, when combined with the faculty of reasoning, later led us to recognize that we must consider impartially the interests of all people, and perhaps it will in the future lead most to realize (as Singer would also argue) that we must consider impartially the interests of all sentient beings. This account does not seem to require any commitment to moral realism; the expansion of the circle of moral concern, on this account, might take place even if there is no objective requirement to consider the interests of anyone.

But this theory, if proposed as a general account of moral progress, appears to require two dubious assumptions: First, that there is some sort of logical or quasi-logical obstacle to embracing equal concern for all members one’s own tribe while being indifferent or hostile to outsiders. Second, that moral progress consists solely in the expansion of the range of beings whose interests we consider.

4.3.2 The coherence of partiality

Begin with the first point. How could reasoning lead one from concern for the interests of other members of one’s tribe to concern for the interests of individuals outside the tribe? Exactly what would the argument be? Perhaps Singer has in mind something like this piece of reasoning:

• 1. It is morally obligatory to consider the interests of other individuals within the tribe.

• 2. There is no morally significant difference between those in the tribe and those outside the tribe.

• 3. Therefore, it is also morally obligatory to consider the interests of other individuals outside the tribe.

But where would premise (2) come from? Not logic alone; there is no incoherence in holding that tribal membership is morally significant—or more precisely, that the relation of belonging to the same tribe is a morally significant relationship between persons.

Compare the view, held by many today, that familial relations are intrinsically morally significant, such that individuals bear special moral obligations to their family members that they do not bear to non-family others. Few thinkers would try to convict this view of incoherence; at any rate, the argument demonstrating such an incoherence has yet to be made. But if it can be coherently maintained that family relations generate differential obligations, then why could one not coherently maintain that the relationship of being co-members of a social group similarly generates differential obligations, such that one might be obligated to consider the interests of other members of one’s social group even though one is not obligated to consider the interests of anyone else?

To be clear, I am not disagreeing with Singer and Jamieson about the truth of principles like (2). I agree that in fact co-membership in a social group is either not morally significant, or not significant enough to explain why one might be justified in ignoring the interests of members of other groups. What I am challenging is the implicit suggestion that one can adequately explain one’s acceptance of principles like (2) merely by citing “reasoning”. (2) is not a truth of logic; it is a substantive, evaluative proposition. Singer has given no indication of how this proposition might be inferred from other propositions, leaving us with the suspicion that (2) is an ethical intuition.

Singer’s theory seems intended to explain the rise of morality by appeal to (1) natural selection and (2) reason. But it is not at all plausible that natural selection explains the tendency to endorse (2). If anything, natural selection should oppose any such tendency. If one claims, on the other hand, that reason alone leads us to endorse (2), then—given that (2) is a substantive moral commitment—one has joined the ranks of the ethical rationalists, defeating the main thrust of debunking accounts of ethics. One also, of course, incurs the daunting dialectical obligation of supplying the reasoning by which (2) can be derived.

4.3.3 A fuller view of moral progress

I turn now to my second major objection to Singer’s account. Singer seems to assume that morality is exclusively or almost exclusively concerned with promoting others’ interests, and that moral progress consists in the move from considering the interests of a small group to considering the interests of ever larger groups. That is indeed one important aspect of the moral progress that we have seen. No longer, for example, is it viewed as acceptable to make war on a neighboring society for no reason other than to capture their resources. But the notion of an expanding circle of moral concern is far from capturing all of the moral evolution that we observe over human history.

Consider, for example, the shift away from traditional moralities in which premarital sex is considered morally wrong. No doubt, the blanket condemnation of premarital sex was and is benighted—but not because it represents an overly narrow understanding of the group of people whose interests must be considered. It is not as though a prohibition on premarital sex promotes the interests of our social group as against others.

Or consider the reformation of views about capital punishment. It was once considered that execution was an appropriate punishment for adultery, theft, and numerous other crimes. What was wrong with this view, again, was not that it rested on an overly narrow circle of moral concern; what was wrong with it was simply that the punishment would be disproportionate to the crime.

Lastly, consider the practice of dueling, by which men were supposed to defend their honor. The men who were permitted or required to engage in duels were most assuredly within the circle of moral concern—that, in a sense, was precisely why they might need to duel, to defend their honor. The people who were excluded from the practice—women and children—were the people of lesser status. Thus, mere expansion of the circle of moral concern, one might think, would have entailed extending the practice of dueling so that women and children, too, were encouraged to defend their honor through duels.

It would not be open for Singer to plead that because utilitarianism is in fact the correct moral theory, moral progress consists solely in a move toward more universal and effective promotion of interests. This would not meet the challenge, because the challenge is not to explain what “moral progress” consists in. The challenge is to explain why the actual changes that have occurred over human history have occurred. These changes have not simply consisted in expansion of the circle of moral concern. Singer’s account thus fails to explain a large portion of the data.

4.4 Cultural evolution

Let us now turn from biological to cultural accounts of the source of morality. Of course, it is mainly a shift in culture that explains why a given individual today is much more liberal than most people in the past. We do not all independently figure out that slavery is wrong as we grow up; we are taught that slavery is wrong, as part of our culture. The interesting question is: why has our culture evolved in the way that it has?

Perhaps cultures simply change over time in unpredictable ways, so that there is no point in asking for an explanation of why a culture incorporates certain values, or why it has changed in a certain way. But in the present case, this would be not only an unsatisfying but a deeply implausible attitude to adopt. Note that the development with which we are concerned comprises a set of changes in attitudes on multiple different issues—slavery, war, torture, women’s suffrage, and so on—that all fit together; all the changes are consistent with a certain coherent ethical standpoint. Furthermore, the change has been proceeding in the same direction for centuries, and the changes have affected nearly all societies across the globe. This is not a random walk; this calls out for an explanation.

Perhaps we can provide piecemeal explanations for liberalization with respect to different issues and different countries. For example, why was slavery abolished in the United States? Abraham Lincoln prosecuted the American Civil War for the purpose of preserving the union. During the war, Lincoln issued the Emancipation Proclamation, freeing most of the slaves, as a measure to help the Union win the war: the Proclamation increased the morale of Union troops and induced some slaves in Confederate states to escape from their masters, which weakened the South. As the Union defeated the Confederacy, the Proclamation became enforceable. In some sense, this is a correct explanation for why most of the slaves in the U.S. were freed (the rest being freed later by the Thirteenth Amendment). But we should also find this unsatisfying, because slavery was not just abolished in the United States. Starting from a point when slavery was widely accepted, it came to be abolished in every country in the world over the past few centuries. Are we to believe it is coincidence that in all these countries some concatenation of events leading to abolition transpired, like those that occurred in the United States? Surely there is some further explanation; surely something needs to be said about why slavery in general was not a stable practice in modern times.

Of course, there are other kinds of explanation for the abolition of slavery. Perhaps slavery was best suited to an older, more agriculturally centered economy but was ill-suited to modern, industrial economies. This explanation is of course speculative. Moreover, it again requires that we accept a great coincidence. For it is not just that slavery was abolished. It is that liberalism triumphed on many different issues over the past few centuries. Are we to believe it is coincidence that, at the same time that slavery was becoming economically inefficient, some other trend was leading women’s suffrage to become more popular (perhaps women’s suffrage also becomes more economically advantageous in industrial societies?), another trend was causing democracy to spread across the world, another was causing war to seem less glorious, another made torture seem less beneficial, and so on? This is not just a series of unrelated changes; they are all changes in line with a certain coherent ethical perspective: all the changes fit together, in one way or another, with the value of equal respect for the dignity of persons.

I of course cannot anticipate every possible explanation for the shift in values over time. What I will do at this point, then, is simply to present my own realist account, leaving it to the anti-realists attempt to devise a better explanation.

5 A realist account of moral progress

Why was slavery abolished? Because slavery was unjust. Why have human beings become increasingly reluctant to go to war? Because war is horrible. Why has liberalism in general triumphed in human history? Because liberalism is correct. These, I suggest, are the most simple and natural explanations.

But how could such explanations be correct? Moral properties—injustice, horribleness, moral correctness—have no causal powers; therefore, they could not have any role in explaining any observable facts, could they?³⁹ In the remainder of this section, I explain why these moral explanations are plausible and how they are possible.

5.1 Societies progress toward the truth over time

Critics of moral realism often appeal to the argument from cultural variation: it is said that the truth of moral realism predicts that there should be broad agreement on moral values across the world; since in fact there is a great deal of disagreement, we have reason to doubt realism.⁴⁰

The critics are partly wrong and partly right. In a wide variety of fields whose objectivity is hardly in doubt, human beliefs have varied enormously across cultures. For example, in our society, the Earth is thought to have been formed as a result of gravitational accretion from a solar nebula. But according to an ancient Egyptian account, the Earth was the offspring of a mating between Sky and Moisture.⁴¹ In our society, it is thought that the continents were formed by plate tectonics. But according to the Iroquois of North America, the continents were formed as a result of a muskrat piling mud on the back of a turtle.⁴² Western societies, too, in earlier centuries had radically different views from those we hold today. Today, we believe that most diseases are caused by microscopic bacteria and viruses. But in the middle ages, it was thought that diseases were caused by imbalances of the four bodily humors, namely, black bile, yellow bile, blood, and phlegm.⁴³

For this reason, it should not be particularly surprising to us, even if moral realism is true, that the moral beliefs of our society should differ greatly from those of primitive societies, and from those of our own society in earlier ages.

Nevertheless, there is an insight in the argument from cultural variation: if there are objective ethical truths to which human beings have some epistemic access, then we should expect moral beliefs across societies to converge over time, if only very slowly. This prediction is not tied to any specific theory of how moral knowledge is gained. Convergence is observed in a priori fields such as mathematics, in experimental sciences such as physics, and in (partly) historical sciences such as astronomy, although in some cases the convergence has occurred over the course of centuries or millennia. As long as human beings possess some reliable belief-forming mechanism, whether the mechanism be a priori or empirical, it should be possible eventually to attain convergence on approximately true theories. Thus, if in fact we see no convergence in ethics, even over the course of centuries or millennia, then we have reason for doubting, either that the field of ethics contains objective facts, or that human beings have epistemic access to those facts.

But by the same token, if convergence does occur in ethics, then we have reason for believing that ethics does contain objective facts to which human beings have epistemic access. And that is precisely what the trends discussed in Sect. 3 above suggest: over the long term, societies across the world are converging on a liberal value system. Anti-realists can’t have it both ways: if divergence would be evidence against realism, then convergence would be evidence for realism.

5.2 Liberalism results from overcoming biases

Some moral beliefs give the distinct appearance of culturally induced biases—for example, the idea that members of one’s own social group are better or more important than those of other social groups; that it is good if one’s own society conquers others, but bad if another society conquers one’s own; or that those who disagree with one’s own religious or cultural beliefs are corrupt and should be punished and forced to convert.

And some moral beliefs give a distinct appearance of evolutionarily designed bias—for example, the idea that one’s own offspring are more important than other people, the idea that sexual promiscuity is good for a male but bad for a female,⁴⁴ and again the idea that killing others to seize their territory or resources is good.

Liberal moral beliefs, however, belong to neither class: they do not have the appearance, prima facie, of biases induced either by one’s culture or by one’s genes. In most cases, they have the opposite appearance of rejecting biases induced by culture or genes. Notably, the idea of the moral equality of all persons stands opposed to precisely those biases that human cultures have traditionally inculcated and that evolutionary psychology would lead us to expect.

Not all of our biases have as yet been overcome, but those that remain appear to be softening. Consider, for example, the idea that adultery, especially on the part of a woman, is a terrible crime. This belief can be explained by evolutionary psychology: sexual infidelity creates a risk of causing the woman to become pregnant with the offspring of the partner who is not her husband, which in turn greatly reduces the husband’s expected reproductive success. We can thus understand why a tendency to feel negative reactions toward such infidelity would have had survival value.

This evolutionarily-induced sentiment remains with us today; as a result, adultery is still regarded as seriously wrong in our society. But consider the change over history: today, adultery is considered grounds for divorce; but in traditional Judeo-Christian and Islamic doctrine, adultery was considered grounds for execution.⁴⁵ To the extent that evolutionary psychology explains the traditional hatred for adultery, the liberalization of attitudes about adultery suggests a move away from our biological programming.

5.3 How we know moral truths

In this section, I address an argument for moral skepticism. I begin in Sect. 5.3.1 by setting out the skeptic’s argument, which I will answer in Sects. 5.3.2 and 5.3.3.

5.3.1 Skeptical doubts about cultural beliefs

Do we really know, for example, that slavery is wrong, or do we merely (correctly) believe that it is? For our belief to count as knowledge, it seems (and I shall grant the skeptic this premise), the belief must have been formed by a reliable process, one that systematically tends to lead to true beliefs. Most of us believe that slavery is wrong because that idea is part of our culture. If we had been born 200 years ago in the American South, most of us would have believed that slavery was acceptable. It isn’t that we reason from “my society disapproves of slavery” to “slavery is wrong”. It is rather that our culture influences how things seem to us. If we’d lived in a slavery-practicing society, slavery just would have seemed a lot less bad to us.

But that suggests that our belief-forming mechanism is unreliable. Perhaps appearances in general are usually reliable, but appearances that are heavily dependent on one’s particular culture are not. If our belief-forming mechanism is “believe moral appearances caused by cultural conditioning”, that mechanism would have led to pro-slavery beliefs in most societies throughout history, in addition to many other false beliefs. So, even if we are right in condemning slavery, it seems that we do not know that slavery is wrong.

5.3.2 How I know scientific truths by trusting my culture

Before directly addressing the moral skeptic’s argument, it is worth considering the parallel reasoning as applied to scientific beliefs. In any normal context, if someone asks me, “Do you know how old the Earth is?”, the proper answer is yes: I know that the Earth is about 4.5 billion years old. But I did not observe any of the physical evidence on which that estimate is based. I could not even tell you how the age of the Earth was calculated. I am relying on the testimony of my society’s recognized experts. But in most places and times throughout history, such reliance would have led one to falsehoods more often than to truths. Trusting the recognized experts would have led me, in earlier times, to think that the Earth was only a few thousand years old; in another society, it would have led me to think that the Earth was produced by a mating of Sky and Moisture; and so on. So, even if I am right in placing the Earth’s age at 4.5 billion years, it seems that I do not know that this is the Earth’s age, since my belief forming process is unreliable.

I take it for granted that the foregoing argument is wrong; I really do know the Earth’s age. I also accept that knowledge requires a reliable belief-forming mechanism. Therefore, I think we must hold that my belief-forming mechanism in this case is reliable.

How so? I think we should say that the reliability of my belief-forming mechanism, when I rely on the accepted beliefs in my society, is properly assessed by reference to the reliability of my society. I am reliable because my society used a reliable method to arrive at its estimate of the Earth’s age—even if other societies in other times used unreliable methods. Perhaps, for example, my belief-forming method should be construed as something like “relying on accepted scientific beliefs, in a society that has an advanced scientific practice”. There are of course other ways of characterizing the belief-forming process; the important point is that the reliability of modern science should be implicated in the account of why my belief-forming process is reliable in this case, even though I did not myself perform any scientific investigations, because (other members of) my society performed scientific investigations to arrive at the belief about the age of the Earth, which they in turn transmitted to me.

5.3.3 How I know moral truths by trusting my culture

Now return to the case of moral beliefs. Why does my belief in the wrongness of slavery count as a belief “formed by a reliable process”? What I want to say here is analogous to what I have just said about scientific beliefs. My intuition is influenced by my culture, but this does not disqualify my intuition as a reliable source of moral guidance, because there was a reliable process by which my culture arrived at its current anti-slavery stance.

Of course, the process by which our culture developed its liberal values is not the same as the process by which scientists arrived at their estimate of the Earth’s age; liberalism did not triumph through scientific investigation. What matters is simply that there is some reliable process by which our values developed.

To explain the nature of that process, I must first mention some background assumptions:

• a. I assume that human beings possess a general capacity for a priori knowledge, sometimes called “reason”, “the understanding”, or “the intellect”. No very specific assumptions about this capacity are needed, beyond that it is capable of producing substantive, a priori knowledge.

• b. Ethics is among the subject matters to which that capacity can be applied. In other words, we can at least sometimes form a priori ethical beliefs through the same general mechanism by which we form other a priori beliefs.

(a) and (b) are standard assumptions of rationalist intuitionism. Note that they have not been devised ad hoc to help in the explanation of liberal progress over human history. Rather, these assumptions have seemed plausible to many moral realists, from Plato to W. D. Ross. But while I endorse (a) and (b), I also think they provide a very incomplete picture of moral thought. To better fill out the picture, we must include three more observations:

• c. Human beings experience non-rational influences on our moral belief-formation. Among these influences are emotions and desires, some of which may be explained by evolution and/or prevailing cultural practices. None of these factors strictly determines our moral beliefs, but each influences those beliefs.

• d. Because of the strong influence of culture, it is very difficult, and hence rare, for an individual to embrace a moral position that is radically at odds with the values of his own society, even if that position is objectively true. It is, however, much easier, and hence more common, for an individual to embrace a position that is slightly at variance with the values of his society. It is easier to move a small distance from one’s culture than to move a great distance.

• e. Individuals differ in their moral sensitivity and in their susceptibility to non-rational influences on belief-formation. Some are more biased than others, and some are better at apprehending moral truths than others.

None of these are ad hoc postulates. Assumptions (c) and (d) are independently supported by empirical evidence. Assumption (e) is a natural concomitant of moral realism: in every other area of cognitive performance, individuals show varying aptitudes. If moral judgment is a form of cognition, it would be amazing if some people were not better at it than others, or if some were not more susceptible to bias than others.

Now here is how the liberalization of values comes about. In primitive times, human beings begin with badly misguided moral beliefs. This parallels the widespread and severe error that primitive societies begin with in all other areas of inquiry. In the case of morals in particular, we have non-rational emotions and desires influencing our beliefs and hence leading us astray—the very sort of influences that the debunking skeptics advert to in their effort to impugn all moral beliefs.

Individual members of society differ in their ability to notice these moral errors. At any given point in history, there will typically be some individuals who see some of the moral errors of their society. If, however, the prevailing norms are very far from the moral truth, then it is likely that these individuals will not see all the way to the actual moral truth. Rather, they will likely embrace a position that is closer to the moral truth than their society’s prevailing norms, but that remains not too far from the prevailing norms. These individuals are not devoid of cultural or biological biases; they merely are somewhat less influenced by those biases than other members of their society.

Consider, for example, the case of John Locke, whose Letter Concerning Toleration is a classic in the literature of religious toleration. Locke was a great moral reformer, because he saw that it was wrong to persecute members of other religions. Yet he could not see his way to embracing tolerance for atheists; that was simply too far from the norms of his culture. Thus, after explaining the arguments for religious toleration, he adds, “Lastly, those are not at all to be tolerated who deny the being of God. Promises, covenants, and oaths, which are the bonds of human society, can have no hold upon an atheist.”⁴⁶ Though Locke here offers a superficially non-prejudicial, public-spirited rationale for intolerance of atheists, virtually no one today, even among social contract theorists, would find his argument at all plausible. The best explanation for why Locke endorsed that argument is that he was influenced by the prejudices of his day.

Frequently, those who see errors in their society’s prevailing norms will attempt to bring about reform. The reformers cause others who did not initially do so to question the misguided norms. Reformers are often effective, for at least two reasons. First, they tend to be more rational than their opponents who favor the status quo. This is because only those with an above-average tendency to form beliefs rationally, and a below-average level of bias, will see through the errors in prevailing cultural norms. As a result, in the ensuing debate, the reformers will tend to come across better than their conservative opponents.

Second, the reformers tend to be disproportionately influential members of society. They are more likely, for example, to be authors, professors, other intellectuals, or business or political leaders, as opposed to members of less influential professions. This is because the ability to see through errors in prevailing social norms will be strongly correlated with one’s degree of intelligence and reflectiveness, which itself is correlated with belonging to relatively socially influential professions.

Thus, when society has incorrect values, there is a systematic tendency for forces to arise that push society in the direction of more correct values. Once society has moved some distance in the right direction, a new generation of reformers may arise, realizing that society’s values still are not correct, and hence working to push society further along. For example, once Lockean toleration for all theistic religions was accepted, it was then possible for people to see that toleration for atheists should also be embraced. Over the long term, beneficial change can accumulate so that, perhaps after several centuries, a society has moved from horrific values to quite decent ones.⁴⁷

This is the sort of process through which our society has arrived at its current set of liberal values. Notice that for this process to work, no great cognitive virtue is required of any individual. It is not necessary that anyone eliminate cultural or biological influences on their thinking, nor need anyone be capable of seeing the moral truth entirely on their own. What is necessary is only that, at a given point in time, there be some individuals who are capable of seeing certain moral issues a bit more accurately than most of their contemporaries. It is even compatible with the story I have told that almost everyone’s moral beliefs be almost entirely determined by genes and culture—but not quite everyone, not quite entirely. Thus, empirical evidence showing that moral beliefs are often strongly influenced by genes and culture does not undermine my account.

The process of moral development is not at its end; further progress will undoubtedly be forthcoming in future generations. So I do not claim that we now know the precise or complete moral truth. I do claim, however, that we have a mechanism that systematically produces moral progress. Our current practices are, reliably, better than our past practices. As a result, while we perhaps do not quite know how we ought to act, we know at least some ways that we should not act. We should not hold slaves, for example. Or torture suspects to extract confessions. Or attack people to steal their land. Our condemnation of these sorts of practices is the product of a reliable belief-forming process.

5.4 The causal objection to the liberal realist explanation

We can now see how certain moral facts explain the phenomenon of liberalization over the course of human history. For example, here is why slavery was abolished:

• 1. There is a systematic tendency for human moral beliefs to become more accurate over time (as explained in Sect. 5.3.3).

• 2. Slavery is seriously unjust.

• 3. Therefore, it was probable that slavery would in time come to be generally regarded as seriously unjust.

• 4. Human beings tend to abolish practices that are generally regarded as seriously unjust.

• 5. Therefore, it was probable that people would in time abolish slavery.

This explains why the abolition of slavery was, though not inevitable, something to be expected.⁴⁸ Step 2 is a hypothesized moral fact: if we posit this moral fact, then we can explain why slavery was abolished. Similar posits can be made to explain other aspects of moral progress. The generalization would be that liberalism is the correct moral stance; this enables us to explain, in general, why trends toward liberalization are observed over history. Because these moral posits enable us to explain the historical facts, we have reason to accept these posits.

The above explanation takes for granted in step 1 that society has a systematic tendency to develop more accurate moral beliefs over time. My account of that, in turn, took for granted that human beings have a (fallible) capacity to apprehend moral truths. But, it might be said, the real problem is how human beings could have a faculty directed at the truth, in an area in which the facts are causally inert. If moral facts don’t cause anything (not even our beliefs about them), how could we be at all reliable at identifying them—how could there be a faculty that was any better than chance at identifying moral truths?

But this just raises the general problem of a priori knowledge. It is true of a priori knowledge in general that the facts to which it pertains do not cause anything, not even our beliefs about them. For instance, facts about abstract, mathematical objects do not cause anything to happen—this premise, at any rate, is as plausible as the premise that moral facts do not cause anything to happen. Similarly, necessary truths of metaphysics do not cause anything to happen. Yet we have knowledge of both mathematical and metaphysical truths.⁴⁹

As mentioned at the end of Sect. 2.1 above, my target herein is not the general empiricist who rejects all synthetic a priori knowledge. My target is the philosopher who thinks there is a special problem for moral knowledge (perhaps because moral beliefs can be debunked, in ways that mathematical and metaphysical beliefs cannot). Given that a priori knowledge in general is possible, the causal objection must be wrong. Either it is wrong because abstract facts of mathematics and metaphysics do have causal powers (in which case it is equally plausible that moral truths have such powers), or it is wrong because our ability to gain knowledge of something does not require the object of knowledge to have causal powers.

6 Conclusion

This paper has had two goals. The first is to show that debunking arguments fail to refute moral realism. They fail because they rely on general theories about the source of our moral beliefs that are not credible. These accounts of the source of moral beliefs lack credibility because they afford no explanation for the most important fact about the history of moral thought: the spread of liberalism across the world over the course of human history, especially recent history. Evolutionary theories cannot account for this fact because the spread of liberalism has been too rapid for biological evolution, and because the spread of liberalism has had no empirical connection to liberals’ somehow reproducing more than non-liberals.

Purely cultural accounts of the source of morals leave us at a loss to explain why the culture itself has moved in a given direction over time. At first glance, it may seem that many explanations are possible—for instance, perhaps changing technologies or changing forms of economic organization have somehow necessitated different values. But the list of potential explanations dwindles as we try to take into account the entire phenomenon: it is not just, for example, that slavery was abolished in the United States. It is that societies around the world have been liberalizing with respect to many different issues—slavery, war, torture, execution, democracy, women’s suffrage, segregation, and so on—and this has been going on for centuries. It is very difficult to come up with explanations for this broad phenomenon that don’t require us to posit large coincidences.

My second goal has been to suggest that we have positive evidence for a version of moral realism—a modest, rationalistic, liberal realism. This view holds that human beings have some limited access to objective values, by the same cognitive faculty or process that produces non-moral a priori knowledge; that the objectively correct values are in fact liberal values; but that culture, genes, and other forces may produce biases that pull us away from the purely rational (and liberal) moral beliefs.

Given this view, the trend toward liberalization can be explained. At any given point in history, there will be some individuals who are somewhat less biased and more morally sensitive than average. These individuals will push society toward what they, the sensitive individuals, consider morally correct, which will generally mean pushing society closer to the moral truth. Over the long term, beneficial changes accumulate and society’s value system approaches the moral truth. Since liberalism is the correct moral stance, society becomes more liberal over time.

No evidence cited herein logically entails that moral realism is correct. Nor can I canvass every possible non-realist explanation for the observed phenomena. But these caveats are standard ones for any inference to the best explanation. What one says of typical inferences to the best explanation applies here: having offered a plausible account of the evidence, which is superior to any account offered by the leading alternative theories, it is reasonable to endorse moral realism unless and until a better account appears.⁵⁰