Most people have moral beliefs, but few have anything as robust as a moral theory. A moral theory is meant to systematize and explain what makes actions right or wrong, states of affairs and motives good or bad, and traits virtuous or vicious. Moral theories are meant to explain rather than to guide: a moral theory explains how morality fits together. 1
That’s All, Folks!
It’s not clear that people need a moral theory so defined. Many people are good moral agents despite not knowing moral theory, or even despite accepting a bad moral theory. 2 The skill of doing is different from the skill of explaining. One might be good at doing even if one lacks the ability to explain what one is doing, or even if one has a bad explanation of what one is doing.
Jimi Hendrix was an excellent blues‐rock guitarist despite not knowing music theory. He couldn’t explain his own music as well as some music analysts, but he sure could play. Or consider that Tom Brady is an excellent football player, whereas Bill Belichick is an excellent coach. Brady can play better than Belichick, but Belichick has a better theory of football than Brady. In the same way, a moral theorist might have a better account of what morality is and how it all fits together than an average person, but that doesn’t necessarily make the moral theorist a better person. 3
In our daily lives, most of us get by just fine without invoking a moral theory. Suppose I’m in some difficult moral situation. I don’t need to appeal to a broad moral principle by asking, “Is acting on this maxim in this situation something I could rationally will the whole world to do?” Nor would I ask, “Does this action produce the maximal expected utility?” Instead, I’ve got a handful of commonsense moral principles, such as these:
Give people what they deserve.
Don’t harm others or aggress against them.
Respect people’s property.
Provide an appropriate amount of charity to help those in need.
Keep your word, and be honest to those who deserve it.
Reciprocate with those who have helped you.
Don’t take advantage of others’ misfortune.
Provide for those whom you owe a duty of care.
Don’t violate others’ rights.
I might quickly go down the list, and as long as I’m not violating these rules, I conclude whatever I’m doing is fine.
In commonsense morality, those norms strike us as useful rules of thumb. Sometimes, there are exceptions to the rules: for example, although I can’t kill you for fun, I can kill you to stop you from killing other innocent people. Sometimes, there are conflicts: for example, Jean Valjean might have to steal bread to feed his starving sister’s children. Sometimes, there are complexities: for example, it’s unclear what some job candidate deserves, or which applicant is more deserving, or just what counts as a basis for desert for any particular job. Still, in commonsense morality, wise people weigh competing principles, use their best judgment, make a decision, and move on. They don’t appear to make use of, or need, a deeper moral theory.
Most moral theorists, including deontologist Immanuel Kant and utilitarian John Stuart Mill, agree that this is what it’s like to be a moral agent. They agree that in our daily experience, it feels as if we are bound by and have to weigh a plurality of (often competing) mid‐ or low‐level moral principles. 4 But Kant and Mill think the mid‐ and low‐level principles are at most instances or approximations of one big, abstract, high‐level principle. Kant thinks that we have each of the duties on the list above, but he sees that list of duties as special applications of the categorical imperative—an abstract moral law that binds all rational agents of any species. 5 For Kant, in the end, there is just one fundamental moral principle. Mill agrees with Kant that in the end, there is just one fundamental moral principle, but he disagrees about what that principle is.
Kant and Mill both agree that commonsense moral thinking works the way I’ve described, but they see themselves as having discovered an underlying skeleton that holds morality together and gives morality its shape.
But what if no underlying skeleton exists? Moral pluralism, sometimes called Rossian pluralism after early 20th‐century philosopher W. D. Ross, claims just that. 6 Pluralism is in effect the thesis that the structure of commonsense moral thinking is all there is to morality. There is no unifying principle that explains all of morality.
Pluralistic theories hold that a multiplicity of basic moral duties and values exists, and these duties and values cannot be subsumed beneath one principle. For the pluralist, morality is not all one thing. Those principles can come into conflict with each other. Acting on them, applying them, and resolving conflicts requires good judgment, but no further theory gives precise principles about how to resolve the conflict or can substitute for good judgment. In any given situation, there’s a truth of the matter about how to resolve conflicts, but there’s no algorithm for determining that truth.
Against One‐Sentence Moral Theories
All moral theories are either monist or pluralist. 7 A monist theory of right action holds that exactly one fundamental feature of actions determines whether they are right and wrong. A pluralist theory holds that more than one fundamental feature determines whether actions are right and wrong. 8 A monist might agree that in commonsense morality, many features seem to count for and against the rightness of actions, but then holds that these features can be reduced to one deeper or more fundamental feature. The pluralist holds that many features count for or against the rightness of actions, but these features cannot be reduced to one deeper or more fundamental feature.
The best argument for pluralist moral theories is to see how inadequate all the monist theories are. Monist theories fail because they try to do too much with too little; that is, they try to explain all of morality with just one basic principle or basic idea.
Many of the moral theories discussed elsewhere in this book hold that morality can be summarized, systematized, and explained with just one sentence. So, for instance, Kantianism holds that an action is wrong if—and only if—it violates the categorical imperative. Ethical egoism holds that an action is right if—and only if—it is expected to contribute maximally to the agent’s welfare. Act utilitarianism holds that an action is right just in case it produces the greatest net aggregate utility.
But one‐sentence moral theories seem problematic. Most of them seem to have absurd counterexamples. Consider, for example, hedonistic act utilitarianism. This theory begins with the plausible thought that pleasure is good and pain is bad. It seems plausible that morality is about maximizing aggregate utility, here defined as pleasure minus pain. But this principle—“an act is right just in case it maximizes net aggregate utility”—has bizarre implications. For instance, it implies that I ought to break a promise whenever doing so produces an infinitesimal gain in utility. Worse, it implies that so long as the sadist enjoys watching others suffer more than he or she hates suffering, we’re obligated to submit to his torture. However plausible utilitarianism might be, it’s not plausible enough to justify biting these bullets.
Kant held that we had general duties to pursue our own perfection and to adopt others’ happiness as an end. He thought we have more specific duties (a) to avoid envy, ingratitude, malice, arrogance, defamation, ridicule, suicide, lying, servility, avarice, and intemperance; (b) to develop our natural and moral powers; and (c) to act upon dispositions of beneficence, gratitude, sympathy, and respect for others. 9 He thought that applying those principles in context was complicated and that philosophers could not provide any real algorithm for doing so. 10 Still, he believed that all of these lower‐level principles are just instances of and derivations from the categorical imperative.
But as anyone who has slogged through a class on Kant knows, Kant’s categorical imperative is notoriously difficult to apply. The universal law formulation—“An act is wrong just in case one cannot universalize the maxim of commission associated with that act” 11 —seems like a convoluted test. When Kantians try to unpack the formula, they often appear to jerry‐rig the theory to get whatever results they want. A Kantian who believes abortion is wrong always manages to “prove” the categorical imperative forbids abortion, whereas a Kantian who thinks abortion is permissible “proves” it does not. Both of their arguments seem equally good (or bad). It may be that on further philosophical investigation, we’ll find that the categorical imperative really does favor one over the other. But it may also just be a sign that the categorical imperative is too abstract to resolve this question.
The humanity formulation—an act is wrong just in case it fails to respect the humanity of each person as an end in itself—at first seems more promising than the universal law formulation. But upon further inspection, it looks vacuous and empty. Libertarian Robert Nozick, liberal John Rawls, conservative John Finnis, and socialist G. A. Cohen each agree we should treat the humanity in each of us always as an end and never as a mere means, but they dispute just what it takes to express such respect. When they debate each other, what does the work in the debate isn’t the generic idea of respecting others as ends, but instead it is reflections on mid‐level principles and intuitions about specific cases. 12
“Respect the humanity in others” seems almost as vacuous as “always consider and properly respond to the legitimate interests of anyone affected by your actions.” Well, yeah, every moral theory says that. It’s true, but it’s platitudinous. Again, perhaps a decisive Kantian resolution of this debate is forthcoming, but given how well Kantianism seems to fit with so many disparate views, perhaps we shouldn’t hold our breath.
David Schmidtz, a pluralist, suggests that what attracts us to one‐sentence theories is a misguided search for simplicity:
Would a monist theory be more useful? Would it even be simpler? The periodic table would in one sense be simpler if we posited only four elements—or one, for that matter—but would that make for better science? No. Astronomers once said planets must face circular orbits. When they finally accepted the reality of elliptical orbits, which favor two focal points, their theories became simpler, more elegant, and more powerful.… [W]hen a phenomenon looks complex … the simplest explanation may be that it looks complex because it is. We may find a way of doing everything with a single element, but it would be mere dogma—the opposite of science—to assume we must. 13
Moral philosophy faces a problem such as the one astronomers faced. When astronomers tried to cling to the view that all orbits are circles—and thus have just one focal point—they had two bad options. The first option was just to deny the phenomena—their observations—altogether. In moral theory, the equivalent would be a utilitarian insisting, “No, my theory is right, and most of the purported counterexamples to utilitarianism are actually just what morality requires.” The second option was to introduce arbitrarily complex epicycles into their theory to make the equations work. Soon, the theories became so vacuous they fit the phenomena, because they can fit all phenomena. In moral philosophy, the all‐too‐common equivalent is how it seems that every Kantian philosopher believes Kantianism tends to justify whatever political views he held before he discovered Kantianism.
For methodological reasons, it was good that philosophers repeatedly tried to systematize morality into one monist principle. After all, theoretical parsimony is a virtue. But it seems that we’ve continuously failed to produce a workable, plausible form of monism after repeated efforts. Perhaps it’s time to throw in the towel and go for a pluralist theory instead.
Kant’s categorical imperative is an absolute moral principle. To say a duty or a moral principle is absolute is to say that it can never be outweighed or trumped by a competing consideration. (Note, however, that although the categorical imperative is absolute, what the principle requires in any given context heavily depends on context. 14 So the principle is absolute, but contextual.)
In contrast, Ross doubted that any absolute moral principles existed. Instead, he thought all basic moral principles or duties were presumptive. There is a strong default presumption in favor of abiding by any of our basic duties, but other considerations could in principle outweigh or trump them.
Each presumptive duty is a consideration in favor of performing or avoiding some action. So, for instance, that doing x would keep a promise is a strong consideration in favor of x. That doing x would involve failing to rescue my children is a strong consideration against x. If those duties conflict in any way, then we would have to judge which duty trumped the other. Ross thought that in general, this involved weighing the duties against each other, and then acting on whatever principle was most weighty.
Ross defines our duty proper as what we should do, all things considered. If I have only one presumptive duty active in a given context, then that presumptive duty becomes my duty proper. If I have multiple conflicting duties, then my duty proper is whatever presumptive duty is the weightiest. 15
With that, consider a precise definition of presumptive duties, quoting Mark Timmons:
Definition: An action is a [presumptive] duty if and only if
It possesses some morally relevant feature that counts in favor of my doing (or not doing) the act, and
This feature is such that were it the only morally relevant feature of my situation, then the act (or not doing the act) would be my duty proper. 16
Pluralist moral theories provide an appealing account of what it’s like to be a moral agent making decisions on the ground. Commonsensically, it seems that there are multiple basic moral rules, that such rules can conflict, and that there’s no obvious “super‐rule” for resolving those conflicts. Instead, we have to use our best judgment.
Some people’s best judgment is better than others. Some people are better able to reason via analogy, to think through matters in a consistent and cool way, to note similarities among cases, or to be aware of what moral factors are at stake in a given situation. 17 Some are more prone to suffer from self‐serving or confirmation bias than are others. Indeed, much of contemporary moral psychology finds that people often act wrongly not because they have mistaken moral beliefs but because they simply fail to notice that they are in morally charged situations. 18
Which Presumptive Duties?
The ancient Greek philosopher Thales hypothesized that everything was water. We now know that water isn’t an element, but a compound, and that there is more than one element. Pluralists similarly hold that there are multiple moral elements. But one big question for moral pluralists, as for chemists, is just how many elements there are. Another big question is how those elements work or interact.
Earlier, I listed nine candidates for basic presumptive duties. Ross himself divided our duties into seven basic kinds. Following the periodic table metaphor, we might consider each of them as being similar to periods. Each period contains a number of duties within it, which play the role of moral elements. Thus, consider this Ross’s periodic table of moral elements:
Duties of fidelity • For example, duties to keep promises, to avoid deception
Duties of reparation • For example, duties to apologize for error, to accept punishment, to pay compensation for harms
Duties of gratitude • For example, duties to express thanks, to reciprocate favors
Duties of justice • For example, duties to give people what they deserve
Duties of beneficence • For example, duties to provide charity, to rescue those in great distress; certain duties of special obligation to loved ones (such as the duty to feed one’s children)
Duties of self‐improvement • For example, duties to improve one’s skills, to improve one’s character
Duties of nonmaleficence • For example, duties to respect rights, to avoid causing harm
There are disputes in physics about how best to characterize all the fundamental particles. Some chemists defend alternatives to the standard periodic table. It’s not that they dispute the basic elements, but just the best way to arrange them. In the same vein, different moral pluralists might disagree about what’s the best “periodic table” for moral elements. 19 They agree on a common list of duties but perhaps disagree on the details of the hierarchy. It won’t be important for us to debate that here.
One might think that if pluralists do not all agree on the best theory of pluralism, then pluralism is no better off than monism. But that approach might be similar to saying that because physicists dispute some of the fundamental particles, we’re no better off than Thales. To be a better theory than monism, pluralism just has to be better than monism. That is, it has to have more explanatory power, with less vacuity—and with fewer counterexamples or absurd implications—than the extant monist theories have.
Moral Dilemmas Are Real
One thing pluralists agree on is the fundamental structure of morality: morality is not all one thing, and different moral reasons can pull us in different directions.
On January 7, 2015, two armed terrorists threatened to kill Charlie Hebdo cartoonist Corrine Rey’s daughter unless she unlocked the office doors. She opened the door. They let her go but stormed the office and murdered 12 people, including 9 of her coworkers. Rey faced a difficult choice: Should she save her child or her coworkers?
Thus we have a classic moral dilemma. We see it frequently in fiction. In The Dark Knight, the Joker makes Batman choose between saving ace district attorney Harvey Dent and saving his childhood friend and love interest Rachel.
Consider another classic dilemma. Suppose I’ve promised to give you a ride to the airport. On the way to pick you up, I see a hurt child lying on the side of the road. On the one hand, it seems as if I have a duty to keep my promise. On the other hand, it seems as if I have a duty to help the child. But I can’t discharge both duties—if I help the child, I’ll be late and you’ll miss your flight, but if I keep my promise, I’ll abandon the child. It seems as though I have to weigh two conflicting moral duties and to determine which duty (in this case) is more important than the other.
A genuine moral dilemma is a situation in which we have conflicting moral obligations or duties. We have good reasons both to do something and not to do it. Batman has good reasons to save Rachel—she is his friend—but also good reasons to let her die and save Harvey Dent instead—Dent will probably help save Gotham City from further crime. In the hurt child case, I have good reasons both to keep my word and to break it. 20
Monistic theories hold that all apparent moral dilemmas are merely apparent. Monist theories hold that it is always possible in principle to adjudicate apparent conflicts of duties, because those conflicts aren’t real. Instead, such dilemmas seem real, because day to day we rely on helpful rules of thumb in making moral decisions. But, monists say, those rules of thumb are merely that. What we really ought to do in any given situation is whatever the one fundamental moral principle requires.
For instance, an act utilitarian would say that in the example above, Batman should do whatever produces the best consequences. If he were fully informed, he’d be able to determine whether saving Rachel or Harvey has better consequences, and he should choose accordingly. The utilitarian would recommend that I save the child, unless the consequences of your missing your flight are very severe.
Pluralist theories hold that the conflicts are real and that adjudicating them requires good judgment. What pluralists deny, though, is that to adjudicate between conflicting duties, we need to invoke a deeper, more fundamental moral duty or principle. Consider: It seems obvious that to rescue a drowning toddler, I may break my promise to meet you for dinner. In this case, the duty to rescue trumps the duty to keep promises. It also seems obvious that if I’m on the way to my best friend’s wedding, I don’t need to rescue a person whose car has broken down a quarter mile from the nearest service station. In this case, the duty to keep my promise to attend the wedding trumps the duty to rescue. (If the person were in severe distress, that would change.)
If someone objected, “How do you know that, without having a deeper moral principle?” I’d say the objector is overintellectualizing morality. The baseball player can catch a ball without knowing physics equations. Jimi Hendrix can play a melodious solo without knowing music theory. Any one of us can reliably distinguish cats from dogs without being able to give necessary and sufficient conditions for cathood or doghood. And so the average person can reliably choose among moral tradeoffs without having some fundamental principle in hand. For a more detailed account of how such moral knowledge is possible, see Michael Huemer’s chapter in this volume.
Going back to the periodic table metaphor: chemists learn over time how different chemical elements interact. Ross thought that as a person developed moral wisdom, so he or she would learn over time how different moral elements interact. But most of that wisdom remains tacit—we can act on it, but we can’t articulate it.
Some might be turned off by this metaphor. One can imagine, for example, Ayn Rand saying that Rossian pluralists are whimworshippers who lionize their arbitrary decisions by invoking a mysterious and otherworldly “moral insight.” Ross might respond that all moral theories—including Rand’s—involve making use of judgment and insight. Indeed, Rand’s theory relies more upon insight than does Kant’s or Mill’s. Mill thought he had a formula; Rand thought she had general principles that required virtuous judgment to apply.
When people try to produce a highly rigoristic theory that leaves no room for judgment, they typically hide their prior judgments inside their principles. We have a choice here. Either we can give useful general principles that require good judgment to apply, or we can try to give a principle that attempts to cover everything, but that isn’t all that useful. Ross’s theory is an instance of the former; Kant’s is an instance of the latter.
When I tell my students what makes for a good or bad term paper, I can give them general advice (e.g., “be original,” “respond to objections,” or “avoid BS”). But suppose I were to develop a metric with a set number of points for each bit of advice they follow. For example, they get five points per objection they consider. They lose two points for each extraneous sentence. That metric would give my grading the false appearance of rigor, free of subjective judgment, but it wouldn’t make my grading any better nor would it make it any less based on judgment. (After all, I can’t really say a priori that failing to consider an objection is always worth 5 out of 100 points.) Instead, it would inevitably lead to unfair grades. Similarly, monistic theories that try to dispense with judgment and insight inevitably lead to distortions and absurd counterexamples.
One might worry that Rossian pluralism has no way to resolve disagreements. Ross was more sanguine. He might begin by noting that disagreement is boring. Even when it’s indisputable that there’s an objective truth of the matter, we still see persistent disagreement. People disagree about all sorts of things—whether evolution happened, whether vaccines work or cause autism, or whether planet Earth is older than 6,000 years—about which we have overwhelming evidence for one side. The mere fact that people disagree tells us little about whether there’s an objective truth of the matter.
Part of the problem is that most of us don’t have consistent moral beliefs (or fully consistent beliefs about anything, really). We aren’t able to hold all of our beliefs in our conscious minds at once. Most of our beliefs are latent (or “nonoccurrent”). Because of that condition, we don’t notice conflicts and contradictions among them. Indeed, much of what philosophers do is point out those unnoticed contradictions and then work to resolve them.
The Demarcation Question and the Unconnected Heap
Consider the following two lists of norms:
Do not kill.
Do not cause harm.
Do not deprive of freedom.
Keep your promises and agreements.
Do not cheat.
Righty tighty, lefty loosey.
Use one finger per fret.
Place your pinky finger between the fourth and fifth laces on the football.
First depress the clutch; then shift.
Take Route 50 to the Key Bridge exit, and stay in the left lane.
Both lists contain various rules or norms. However, it’s clear to us that list A contains moral norms, whereas list B contains nonmoral norms. It’s wrong to kill and wrong to shift before pressing the clutch, but those are different kinds of wrong. Killing is morally wrong; trying to shift before pressing the clutch is a bad (and ultimately expensive) driving technique.
One thing a moral theory needs to do is explain what demarcates moral norms from nonmoral norms. We can see that everything on list A is a moral norm, but what makes those norms moral rather than nonmoral?
One might presume that this is an easier problem for monists to solve than for pluralists. After all, monists offer us one basic principle that is meant to encapsulate all of morality. For Kant, everything on list A is an instance of the categorical imperative. For Mill, everything on list A is an instance of rule utilitarianism. But because pluralists do not have one fundamental principle, it might seem that they have a more difficult time accounting for what distinguishes list A from list B.
This kind of reasoning leads to a common complaint about moral pluralism. It seems as if the principles are an unconnected heap. The trick for the pluralist, then, is to explain how all the presumptive duties are united in being moral principles, without thereby reducing everything on the list to a monist principle.
Those common complaints are about pluralist theories, but they are ultimately misguided. 21 To see why, consider that Kant and Mill are monists, but they disagree about what the fundamental principle of morality is. Still, though they disagree, they agree that list A and list B are distinct. When Kant and Mill disagree about what the fundamental moral principle is, they are not talking past each other but are instead talking about the same thing. So presumably Kant and Mill can agree on a theory‐neutral account of what demarcates moral norms from nonmoral norms.
Similarly, John Rawls says that although Marxists, libertarians, classical liberals, left‐liberals, communitarians, and others disagree about what justice requires, there’s a sense in which they all agree on what justice is. They have the same concept of justice but have different conceptions of it. Rawls says that assigning rights and duties and determining the proper distributions of benefits and burdens are built into the concept of justice. 22 Different conceptions (theories) of justice—utilitarian, liberal, libertarian, communitarian—disagree about what the various duties, rights, and distributions are, but they are all conceptions of justice because they all concern these same issues.
We could say that all of the moral theories discussed in this book are different conceptions of morality, but each of the theorists should share the same concept of morality. Whatever answer Kant or Mill gives to explain what demarcates moral from nonmoral norms is equally available to the Rossian pluralist.
In this vein, pluralist moral theorist Bernard Gert offers the following generic account of moral norms:
Morality is an informal public system applying to all rational persons, governing behavior that affects others, and has the lessening of evil or harm as its goal. 23
Gert sees this definition of morality as theory‐neutral, as describing what every major moral theory agrees on. Different moral theories provide different accounts of what the norms are or what explains them, but they each seem to agree on this definition.
Now, perhaps Gert’s attempt to provide a generic demarcation of morality from nonmoral norms does not quite succeed. After all, most moral theorists believe that we owe duties to ourselves, not just to others. Gert seems to think there would be no morality on a desert island, but Kant, Mill, and many other moral theorists would disagree.
So perhaps Gert’s definition is not generic enough. Without here trying to offer a superior definition, we might note that moral norms have the following features:
They are categorical. Moral norms bind us independently of particular desires we happen to have. For instance, if you don’t want to throw a spiral, you can just at whim opt out of rule 3 on list B. But you can’t just opt out of rule 3 on list A, even if you really want to. 24
They hold for all rational agents. Moral norms bind us in virtue of our being the kinds of creatures that (a) can understand right and wrong and (b) can act on this understanding.
They are not mere conventions. My extending my middle finger to express disrespect rather than respect is a social convention. We could have used the middle finger to mean what we mean by a salute, but we didn’t. Social conventions can—in some sense—just be modified. We could just agree starting tomorrow to switch the meaning of the middle finger and the military salute. Basic moral norms are different—a society may not decide not to respect rights. (Plenty do, but they shouldn’t.) They cannot modify moral norms by fiat.
They serve social cooperation. Moral norms make it possible for us to live together well. They help ensure that society is a positive‐sum game, where everyone benefits from social cooperation. (Gert was right that moral norms tend to reduce harm, but that’s not all they do.)
Pluralists and monists of all stripes can agree to this characterization of moral norms. Thus, even though pluralists (by definition) do not accept one fundamental unifying principle, they are not thereby stuck viewing moral norms as an unconnected heap.
Methodological Moral Pluralism
The prominent bioethicist and applied ethicist Peter Singer is a type of preference‐satisfaction utilitarian. Singer is famous for arguing for various controversial conclusions, for example, (a) that we shouldn’t eat meat, (b) that we should give most of our money to charity, or (c) that we should euthanize severely disabled newborns. 25 Those conclusions might follow from Singer’s controversial moral theory. But what’s interesting about Singer is that he doesn’t first try to convince you of his moral theory and then deduce those conclusions from it. Instead, he appeals to widely shared, commonsense, mid‐level moral principles and intuitions, principles and intuitions that Kantians, Objectivists, natural law theorists, and your moral theory–lacking grandma already accept.
Similarly, the Marxist philosopher G. A. Cohen has some controversial views about egalitarianism and justice. But when he wants to argue for socialism, he does not first try to convince you to adopt his version of egalitarianism and then show you that socialism follows from it. Nor does Cohen first try to convince readers to accept heterodox Marxist economics. 26 Instead, Cohen relies on widely shared moral intuitions, intuitions shared even by conservatives, free marketers, and libertarians. He tries to show readers that they themselves already accept moral principles and ideas that show they’re implicitly committed to socialism.
In contrast, consider economist Murray Rothbard, who dabbled in moral theory on the side. In The Ethics of Liberty, Rothbard first tries to convince us that one major moral principle—the nonaggression principle—is self‐evident. He then applies that principle dogmatically to every moral issue. The supposedly self‐evident principle leads to bizarre conclusions: for example, if my neighbors decide to let their newborn starve to death on their lawn, I must not take a single step onto their property to rescue the infant. 27
The reason Singer and Cohen are successful as applied ethicists, while Rothbard’s writings seem question begging and unconvincing to anyone who doesn’t already agree, is that Singer and Cohen appeal to widely shared mid‐level moral principles. They both have deeper moral theories that they believe justify, systematize, or explain the mid‐level principles. But they also recognize that the theories are themselves less plausible than are the mid‐level principles. To debate their interlocutors, they don’t start by invoking some highly abstract moral theory but instead start from common ground. They say to their debate partners: “You already accept A, B, and C. Don’t you see that A, B, and C together imply D?”
Good applied ethics seems to be committed to what we might call methodological moral pluralism. Methodological moral pluralism is the view that we should do applied ethics as if Rossian pluralism were true. A methodological pluralist might accept a monist moral theory or might be agnostic between monism and pluralism. However, the idea behind methodological moral pluralism is that although we might disagree about fundamental moral theories, we probably can each agree to a shared set of mid‐level moral principles. In trying to resolve debates about what to do here and now, we should try to appeal to those more obvious mid‐level principles rather than some less obvious fundamental theory.
For thinking about political philosophy, we see there is yet another reason for methodological moral pluralism. Most moral theories are highly abstract. Asking what Kantianism implies about distributive justice is a bit like asking what Einstein’s field equations tell us about the path of a falling feather. Einstein’s field equations describe the general ordering of space‐time. They are complicated and often cannot be used for direct calculations. They are highly abstract and devoid of specific empirical information. The equations are consistent with worlds radically different from ours, such as Gödel’s universe. 28 By themselves, the field equations don’t tell us much about a falling feather. To understand the falling feather, we use intermediary or mid‐level physical laws and models, and the laws and models we’d use are ultimately compatible with Newtonian or relativistic physics.
Kant’s theory is much the same. Kant sees his theory as highly abstract. It’s meant to apply to all rational beings of any species, including any possible rational aliens with highly different forms of life and biology from our own. Kant himself thinks that applying his theory to humans and human ways of life takes a huge amount of work, and depends on philosophical anthropology, the social sciences, and plain good judgment. 29 Kant ultimately grounds his political philosophy on the categorical imperative, but it takes him hundreds of pages of work to get there. If Kant’s political philosophy turns out to be mistaken, that finding might not be because his moral theory is wrong but because the intermediary work is wrong.
No Straight Path from Moral Pluralism to Libertarianism—or Any Other Political Theory
How might one argue for libertarianism on pluralist grounds? Frankly, it takes a lot of work, and that’s OK. That it takes a lot of work isn’t a flaw of pluralist moral theories or of libertarianism for that matter. (It would also take a lot of work to go from moral pluralism to most other plausible political theories.) The basic moral principles—avoid killing, avoid stealing, keep your promises—are obvious. But no particular political philosophy is obvious, and none follows straightforwardly from our basic moral principles.
If you’ve read the other chapters in this book, it should be clear that no one‐to‐one correspondence exists between libertarianism and any particular moral theory. A classical liberal or libertarian might accept any number of background moral theories, including any of those listed in this book, as well as others that didn’t make the cut.
John Rawls and I agree that we should regard people each as an end in themselves. We agree that we owe various duties of reciprocity, fidelity, beneficence, and nonmaleficence to others. But we disagree on how to apply many of our shared moral concepts. For instance, Rawls thinks that there’s a presumption in favor of an egalitarian distribution of wealth and that departures from equality have to be justified. But why does he think that? In my view, Rawls’s problem is that he finds what I consider a misleading metaphor illuminating. In his view, my problem is that I find what he considers an illuminating metaphor misleading.
Here’s the metaphor: suppose we simultaneously come across some resource that none of us have any prior claim to, such as a pie. 30 The most natural way to divide the pie—the way that would elicit the fewest complaints—would be to give everyone an equal share. But suppose it turned out to be a magic pie that would grow or shrink in size depending on how we cut it. In that case, if we were rational but not envious, we’d each prefer a bigger but unequal slice to an equally small slice.
Rawls thinks it’s illuminating to think of the “social product”—all the stuff we all produce while working together—as being like this unowned pie. I think it’s misleading. I think the “social surplus” is not like a pie that we all came across in the woods simultaneously and thus have an equal basic claim upon. Here, my political disagreements with Rawls aren’t disagreements about fundamental moral theory but about some of the intermediary intellectual tools that we use to apply our shared moral principles.
Or consider that G. A. Cohen and I share many of the same ideas about what a perfectly virtuous person would be like. However, Cohen and I disagree about what a perfectly just society would be like. He believes perfectly just angels would live under a kind of anarcho‐socialism, whereas I hold that perfectly just angels would predominantly live under a kind of cooperative, voluntaryist, anarcho‐capitalism as seen in the children’s TV series Mickey Mouse Clubhouse. 31
Still, our disagreement is not over fundamental values or moral theory. Rather, as I’ve argued elsewhere, Cohen made a simple mistake. He compared an idealized form of socialism (a socialist society inhabited by angels) with realistic capitalism (a capitalist society inhabited by real people, flaws and all) and concluded that the ideal form of socialism was better. But then he mistakenly concluded that this means socialism is better tout court, without his stopping to ask how a capitalist system inhabited by angels would work. Here, the problem isn’t that Cohen and I disagree about moral theory, but rather we disagree about how to apply that theory. We’re disagreeing not about the fundamental standards by which to judge things desirable but instead about how well different institutions would meet those standards, because we have empirical and conceptual disputes about what those institutions do.
Further, it’s implausible to think that one is going to derive libertarianism from a few moral premises without needing to consider empirical questions at great length. Part of what a political philosophy tries to do is determine the standards by which to judge social institutions. Social institutions—such as private property, democracy, or the nuclear family—are “the rules of the game in a society … the humanly devised constraints that shape human interaction.” 32 Every major moral theory, whether consequentialist or not, holds that at least part of what would justify or condemn various institutions is how well those institutions can be expected to work. For that, we’ll need economics, sociology, political science, and history.
Now, consider that the philosopher Joseph Heath and I disagree about what institutions would be best in the real world, given that people are not angels. Here, our disagreement stems not so much from differences in moral values, or even principles of justice, but rather from empirical disagreements. He thinks markets are more prone to failure than I do, whereas I think governments are more prone to failure than he does. We both agree more or less on what it means for governments and markets to work, but we disagree about how well they work.
Libertarianism as a Default
With those caveats aside, the most promising strategy for moral pluralists who want to justify classical liberal conclusions is the same strategy that Huemer takes in this volume. One could begin by noting that commonsense interpersonal morality seems to be libertarian. In our day‐to‐day dealings with one another, we seem bound by libertarian constraints. Even in illiberal countries, most people generally think they’re morally bound to leave others alone, so long as they’re not hurting others.
Consider: I am not allowed to use violence or threats of violence to get you to follow my religion, eat healthier food, or stop smoking cigarettes. I may not force you to fight my enemies or to give your money to worthy causes, no matter how worthy. I can’t force my neighbors to buy tomatoes from my garden rather than tomatoes from down the street or across the world. If my neighbor wants to sell or rent his house to people from outside the neighborhood, I can’t stop him. If I choose to spend my money educating my neighbor’s kids, I can’t then demand that the neighbors repay me with one‐third their income until they die. And so on. Again, in our day‐to‐day dealings with one another, we seem subject to libertarian constraints.
What makes libertarians unusual is that they think most or all of those constraints and prohibitions apply to government agents as well. For various reasons, people who hold ideologies believe government and its agents are exempt from some or all of those prohibitions.
They might be right! I’m not accusing the other side of having obviously absurd unnoticed inconsistencies. Rather, it might turn out that there is a philosophical justification for allowing governments to do things to us that we may not do to one another. Perhaps this justification even implies that the reason day‐to‐day morality is so libertarian is that the morality of state action is not. It’s an open question.
Still, from a pluralist perspective, the best way to defend libertarianism is to begin with the observation that day‐to‐day morality is libertarian. There is a presumption of liberty. By default, we presume people should be free to live as they see best, without having to ask permission from or justify themselves to other people. By default, all restrictions on liberty are presumed wrong and unjust, until shown otherwise. Coercive interference with others’ liberty must be justified. Political authority and all laws are assumed unjustified until shown otherwise.
One can then turn to nonlibertarians and say: “Look, I’m not an absolutist. I’m not pounding the table and saying the presumption of liberty can never be overcome. I’m just saying it has to be overcome. Can you explain to me why we should grant governments powers that we’d forbid private individuals?”
Nonlibertarians will happily oblige. They have plenty of arguments on offer. Remember, in their view, nonlibertarian conclusions are ultimately grounded in commonsense moral ideas as well.
Thus, the final step is to refute those arguments, starting from shared moral premises and relying on as uncontroversial or well‐established empirical premises as possible. 33 It’s unlikely you’ll arrive at one decisive argument for libertarianism. Rather, you can see nonlibertarians as making a series of separate arguments such as these: It will be a disaster if we don’t have government do x. It will be a disaster if we don’t have government do y. You’ll need to show, for each x and y, that empowering the government to do x and y doesn’t work, or isn’t worth the cost.
Rawls’s Argument for Social Justice
I’ll end by giving an example of a challenge to libertarianism from the Rawlsian camp. Libertarians are sometimes quick to say that taxes look like theft. Because everyone agrees that theft is wrong, it seems as if our basic shared moral intuitions forbid coercive taxation. Left‐liberals and even Marxists agree that people shouldn’t steal, so why then would they favor taxing Peter to pay Paul? In this section, I’ll explain why we cannot just derive libertarian politics straightforwardly or with ease from the widely agreed‐upon presumptive duty to avoid stealing. Consider this an illustration of my earlier point: going from moral pluralism to libertarianism is hard.
Libertarians often say that when the government taxes you and redistributes your income to others, this “is on par with forced labor.” 34 It’s as if I spend 700 out of my 2,000 yearly working hours working for other people’s benefit rather than my own. I don’t have a choice—the government won’t let me get paid for the remaining 1,300 hours unless I agree to pay it 700 hours’ worth of income. So it looks at first glance as if libertarians are right—taxation is theft, or, worse, a kind of moderate slavery.
Perhaps taxation does turn out to be a kind of theft. But it’s worth seeing that Rawls has a principled response to this accusation. To see why, we need to take a step back and ask, “What justifies the institution of private property in the first place?”
John Locke—though himself an ardent defender of the right to hold and use private property—notes that in the first instance, private property seems to limit other people’s property. To see why, imagine a world in which no one yet owns anything. Everyone is free to go where he likes and use what he wants. When the first person encloses a plot of land and declares it his own, he thereby in the first instance reduces other people’s freedom. They used to be able to go anywhere, but now there are 40 acres they can’t touch. And so, Locke realizes, we need to justify “original appropriation.” It won’t be enough to say that you earn a right to the land by “mixing your labor” with it. After all, when you privatize unowned land, you reduce others’ freedom. So you’ll need to compensate them in some way.
Locke thinks that everyone does indeed get compensated. Unowned land is not productive, whereas privatized land can be 10,000 times more productive. So, Locke thinks, when land is parceled and privatized, and when people are able to sell the products of their land on a market, the systematic effect is that everyone enjoys many times more wealth than they would under a system without private property. He’s absolutely right. The average American living today enjoys a standard of living about 60 times (yes, 60) higher than the average European colonist of 1600 AD. 35 Americans are thus better able to realize their conceptions of the good life and have more power to achieve their ends.
In effect, Locke thinks that what justifies the institution of private property is that it tends to leave more and better for others. But he’s not claiming that every individual transaction has to benefit everyone else. It’s not as though you can’t sell your guitar to your friend unless doing so helps literally everyone else on earth. Rather, Locke just means that the rules of private property as a whole should tend to make everyone better off.
Libertarian Nozick and left‐liberal Rawls agree. Part of what justifies the institution of private property is that people tend to have much better lives with it than without it. But this is where Locke and Nozick start to disagree with Rawls. They disagree about just how much the institution of private property must benefit everyone to be justifiable. For Locke and Nozick, it’s more or less enough that people do better with it than without it. Rawls has a stricter standard—he thinks that for a particular system of private property to be justified, it must tend to ensure that the representative member of the least advantaged working class does better than he would under alternative systems of private property. 36
Rawls thinks that meeting this standard will require having a series of strong, democratically controlled central governments, which (a) regulate the economy in various ways and (b) provide various forms of social insurance. Rawls’s argument is in effect this:
Normative claim. Any particular regime of private property is justified only if it satisfies the following principle: it should tend to ensure that the representative member of the least advantaged working class does at least as well as would be possible under alternative regimes.
Empirical claim. If we are to meet the standard in 1, it is necessary to have a liberal social democratic government, which taxes citizens to provide social insurance.
Implication of 1 and 2. When the government taxes citizens (in whatever amount is necessary to meet the obligations described in 1), the citizens are not entitled to the money it takes. Instead, the government is entitled to the money. Were the citizens to withhold taxes, they would be stealing from the government.
If Rawls is right, when the government taxes me, it isn’t necessarily stealing. Rather, it might just be doing what it takes to ensure that the system of private property is justified in the first place. Accordingly, if libertarians want to challenge Rawls, it’s not to declare that taxes are theft. They may be right, that’s a conclusion of their theory of property rights, not a premise. Thus, libertarians need to instead attack Rawls’s normative premise (1) or his empirical premise (2). That is, either they need to show that his standards for justifying the system of private property are too stringent—perhaps by defending a superior theory of the legitimacy of private property rights—or they need to show that a libertarian system can meet those stringent standards. 37
Rossian pluralism is a good theory to start and end with. It’s a good theory to start with, because it accurately describes what it’s like to be a moral agent. The other moral theories seem artificial, because they are indeed artificial. Rossian pluralism is a description of what we actually do, on the ground, as people making moral decisions.
It’s reasonable to hope for more (or, in a sense, less). It’s reasonable—it’s good philosophical methodology—to look for a simpler theory that reduces the number of basic moral principles as much as possible. It’s good methodology to try to find one unified explanation for what separates right from wrong, good from bad, virtuous from vicious. But this is good methodology only if we don’t end up producing a vacuous or absurd theory in the process. The problem with so many of our one‐sentence moral theories—be they Kantian, virtue ethics, or consequentialist—is that they do tend to be either vacuous or absurd.
Thus, after we repeatedly try but fail to make a monist theory work, we might want to end up back where we started. There are still plenty of questions moral theorists might try to answer, such as these: What kinds of truths are moral truths? What makes moral truths true? How is moral knowledge possible? And if morality is best depicted as being like the periodic table of elements, just what’s the best way to draw that table? In the end, we shouldn’t demand more precision from a theory than the phenomenon being studied admits.
Rossian pluralism doesn’t offer a 60‐second defense of libertarianism—or any other political philosophy for that matter. But that’s not a bad thing. It would be rather surprising if we could derive a political philosophy directly out of a basic moral theory, without having to first study economics and political science to learn how institutions actually function. Some libertarians are attracted to moral theories that let them bypass this difficult step, in much the same way that some bald men want to buy miracle hair‐growth formula. If we’re taking things seriously, though, we’ll have to admit that our basic moral ideas underdetermine on their own what politics should look like, and we’ll need to understand robust political economy to make a final determination about what justice requires.
For a thorough account of the theoretical goals of moral theory, see Jason Brennan, “Beyond the Bottom Line: The Theoretical Aims of Moral Theory,” Oxford Journal of Legal Studies 28 (2008): 277–96. ↩
I had a neighbor who accepted divine command theory, a theory refuted more than 2,000 years ago, but he was still just as good a person as anyone else I’ve met. ↩
For experimental evidence to that effect, see Eric Schwitzgebel and Joshua Rust, “The Moral Behavior of Ethicists,” Companion to Experimental Philosophy, ed. Justin Sytsma and Wesley Buckwalter (New York: Oxford University Press, forthcoming). ↩
Immanuel Kant, Practical Philosophy, ed. Mary J. Gregor (New York: Cambridge University Press, 1999 ), pp. 546–90; John Stuart Mill, Utilitarianism (Indianapolis, IN: Hackett, 2002 ), pp. 42–60. ↩
Mark Timmons, Moral Theory: An Introduction (Lanham, MD: Rowman and Littlefield, 2002), p. 161. ↩
W. D. Ross, The Right and the Good (Indianapolis, IN: Hackett, 1988 ). ↩
I’m treating moral particularism as an extreme instance of pluralism. ↩
Technically, a moral theory has a theory of the good and a theory of the right. It could be monist about one and pluralist about the other, or monist about both, or pluralist about both. Utilitarians are monists about the good and the right, whereas Kant is a monist about the right but a pluralist about the good. ↩
Timmons, Moral Theory, pp. 158–62; Kant, Practical Philosophy, pp. 546–90. ↩
Kant continuously warns readers that applying his principles requires knowledge that goes beyond a priori philosophical reasoning, and that knowledge cannot be codified. Kant, Practical Philosophy, pp. 546–90. ↩
For example, I have almost a line‐by‐line response to G. A. Cohen, Why Not Socialism? (Princeton, NJ: Princeton University Press, 2009) in Jason Brennan, Why Not Capitalism? (New York: Routledge Press, 2014), but neither Cohen nor I have to articulate a fundamental moral theory to have this debate. ↩
David Schmidtz, Elements of Justice (New York: Cambridge University Press, 2006), p. 4 [emphasis in original]. ↩
Kant makes this clear in Practical Philosophy, p. 584. ↩
Timmons, Moral Theory, p. 249. Because Timmons is discussing Ross, he says “prima facie” rather than “presumptive.” Philosophers today tend to prefer pro tanto rather than prima facie. I just skip the Latin terms here. ↩
See Michael Huemer, Moral Intuitionism (New York: Palgrave MacMillan, 2005). ↩
For one useful popularization of such research, see Max H. Bazerman and Ann E. Tenbrunsel, Blind Spots: Why We Fail to Do What’s Right and What to Do about It (Princeton, NJ: Princeton University Press, 2012). ↩
For examples, compare Ross’s Right and Good to Bernard Gert’s Morality: Its Nature and Justification (New York: Oxford University Press, 1999) or to Schmidtz’s Elements of Justice. ↩
A tragic moral dilemma is a scenario in which no matter what one does, one acts wrongly. Ross seemed to think that so long as you picked the weightier duty, you acted rightly. Some pluralists dispute that—they think morality might be unfair, and there might be times where through no fault of your own, the best thing you can do is still wrong. ↩
For example, Timmons Moral Theory, pp. 262–63; and Russ Shafer‐Landau, The Fundamentals of Ethics (New York: Oxford University Press, 2010), p. 235. ↩
John Rawls, A Theory of Justice (Cambridge, MA: Harvard University Press, 1971), pp. 5–6. ↩
Bernard Gert, “The Definition of Morality,” in The Stanford Encyclopedia of Philosophy, ed. Edward N Zalta (Stanford, CA: Stanford University, 2005). Gert uses “evil” in a nonmoralized way, so this definition isn’t question begging. ↩
Introductory ethics students often get tripped up on, and collapse, a number of distinctions. To say a norm is categorical is to say it binds you because you are a moral agent—you can’t just opt out of it at will. The contrast to categorical is hypothetical—a hypothetical norm (e.g., “Major in accounting, not art history, if you want a job”) binds you because of desires you happen to have. A different distinction is absolute versus presumptive. Absolute norms cannot be outweighed or trumped; presumptive norms can. A third distinction is noncontextual versus contextual. Noncontextual norms require the same behavior in every circumstance, whereas contextual norms require different behaviors in different circumstances. ↩
Peter Singer, Practical Ethics, 3rd ed. (New York: Cambridge University Press, 2011). ↩
G. A. Cohen, “The Structure of Proletarian Unfreedom,” Philosophy and Public Affairs 12 (1983): 3–33; here, p. 24, he accepts that “bourgeois economics” is basically sound. ↩
Bryan Caplan makes this complaint about Rothbard in “Thoughts on Jason Brennan’s The Ethics of Voting,” Reason Papers 35 (2013): 12. He cites Murray Rothbard, The Ethics of Liberty (New York: New York University Press, 1998), p. 100. ↩
Kurt Gödel, “A Remark about the Relationship between the Theory of General Relativity and Idealistic Philosophy,” Collected Works: Publications 1948–1974 (Oxford, UK: Oxford University Press, 2001 ), pp. 202–7. ↩
Kant, Practical Philosophy, p. 65; Robert B. Louden, Kant’s Impure Ethics: From Rational Beings to Human Beings (New York: Oxford University Press, 2000). ↩