Moral Philosophy, Seriously: The Four Big Theories and Where Each Breaks

Follows: The Trolley Problem & Why Ethics Is Hard and How Should I Live?

In this deep dive
1. What a moral theory is for, and four cases to test them on
2. Consequentialism: Bentham, Mill, rules versus acts, and the drowning child
3. Kant: the categorical imperative and the murderer at the door
4. Virtue ethics: Aristotle's return, Anscombe and MacIntyre
5. Contractualism: what we could justify to each other
6. When the theories run out: particularism, moral uncertainty and a scorecard

Part 1 — What a moral theory is for

The Trolley Problem ended on a humbling note. Your intuitions are strong, they're real, and they contradict each other. You'll pull the lever to save five, but you won't push the man off the footbridge, even though the arithmetic is identical. How Should I Live? offered four lenses (consequences, duty, character and care) and suggested running a hard choice through all of them.

Philosophers split ethics into three levels. Applied ethics asks about particular issues: euthanasia, eating animals, a just war. Metaethics asks what moral claims even are: facts, feelings, conventions? In between sits normative ethics, which asks what makes any action right or wrong in the first place. The four big theories are normative theories, and that middle level is where this piece lives.

This piece asks the harder question underneath. Can any single theory make sense of all our moral judgements? Or is ethics a toolbox with no master tool?

A moral theory tries to do three jobs at once:

1. Explain. Say why right acts are right. "Because it causes the most happiness." "Because it respects persons." "Because a good person would do it."
2. Guide. Tell you what to do when you're unsure.
3. Correct. Sometimes tell you that a strong intuition is wrong, the way physics overrules your sense that the Earth is still.

That third job is the uncomfortable one, and it's why theories matter. Your intuitions were shaped by evolution, upbringing and culture. Some of them, such as the conviction held for most of history that people of other tribes or races counted for less, turned out to be badly wrong. A theory is a way of testing intuitions against each other and against principles you can defend out loud.

A widely used method, set out by John Rawls in A Theory of Justice (1971), is called reflective equilibrium. You start with your considered judgements, look for principles that fit them, then go back and forth, revising a principle when it clashes with a judgement you're sure of, and revising a judgement when it clashes with a principle you find compelling. No theory gets to bulldoze every intuition. No intuition gets to veto every theory.

A good moral theory is like a good scientific theory: it explains the cases we're sure of, and it earns the right to overrule us on some of the cases we're not.

The test kit

To make this concrete, we'll run each theory through the same four cases. Two are new; two you've met.

Watch which cases each theory handles with ease, and which it strains to handle.

Part 2 — Consequentialism: the greatest good

Bentham's two masters

Jeremy Bentham's An Introduction to the Principles of Morals and Legislation (printed in 1780 and published in 1789) opens with one of the most famous sentences in ethics: "Nature has placed mankind under the governance of two sovereign masters, pain and pleasure. It is for them alone to point out what we ought to do, as well as to determine what we shall do."

Bentham was a legal reformer, and his theory was meant to be a tool. Morality and law should aim at one thing, the greatest happiness, and you could in principle measure it. His "felicific calculus" lists the dimensions on which a pleasure or pain counts for more or less: its intensity, duration, certainty, propinquity (how soon it comes), fecundity (whether it leads to more of the same), purity (whether it's free of the opposite) and extent (how many people it reaches).

The radical part is not the arithmetic. It's the impartiality. Everyone's pain counts, and counts equally. A lord's happiness weighs no more than a labourer's. In the 1780s, that was revolutionary.

Mill's upgrade

John Stuart Mill, raised as a Benthamite, wrote the classic defence in Utilitarianism (1861). His core principle: "actions are right in proportion as they tend to promote happiness, wrong as they tend to produce the reverse of happiness." By happiness he means "pleasure, and the absence of pain."

Critics had sneered that this was a philosophy fit for pigs. Mill replied that pleasures differ in quality, not just quantity, and gave a test: of two pleasures, the more valuable is the one that "all or almost all who have experience of both give a decided preference" to. Those who know both the pleasures of the intellect and the pleasures of the body, he argued, rarely trade the first for the second. Hence the line everyone quotes: "It is better to be a human being dissatisfied than a pig satisfied; better to be Socrates dissatisfied than a fool satisfied. And if the fool, or the pig, is of a different opinion, it is because they only know their own side of the question."

Mill also answered the obvious practical objection: nobody has time to calculate every action's consequences. His reply was nautical. Sailors don't calculate astronomy at sea; "they go to sea with it ready calculated." Likewise, humanity has accumulated rules of thumb (don't lie, keep promises) that record long experience of what tends to produce happiness.

Act versus rule

That reply opens the most important split inside the family.

• Act consequentialism: an act is right if its consequences are at least as good as any alternative's.
• Rule consequentialism: an act is right if it's allowed by the set of rules whose general acceptance would produce the best consequences.

Rule consequentialism is designed to handle our case 3. A rule allowing surgeons to harvest healthy visitors would be catastrophic. No one would go near a hospital. So the rule forbids it, even if one particular secret harvest might come out ahead.

The classic objection is a dilemma. Suppose following "don't steal" would, in some odd case, make things worse. Then the theory seems pushed to amend the rule ("don't steal, except when...") until it collapses into act consequentialism. If it refuses to amend, it seems to value rules over the very outcomes that justified them, a charge known as "rule worship." The Stanford Encyclopedia's entry describes the modern reply: the best rule consequentialism ranks codes by the results of teaching and accepting them, and a rule with endless exceptions is easier to rationalise your way around, so simple rules win. Scholars still argue about whether Mill himself was an act or a rule theorist; the SEP's entry on Mill discusses act, rule and even "sanction" readings.

Singer's pond

In November 1971, the philosopher Peter Singer sat down to write about East Bengal, where, as he put it, "Constant poverty, a cyclone, and a civil war have turned at least nine million people into destitute refugees." The essay, "Famine, Affluence, and Morality," appeared in Philosophy & Public Affairs in 1972 and became a classic.

Its principle sounds harmless: "if it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it." Then comes case 1: "if I am walking past a shallow pond and see a child drowning in it, I ought to wade in and pull the child out. This will mean getting my clothes muddy, but this is insignificant."

Everyone agrees. Then Singer argues that distance changes nothing: "It makes no moral difference whether the person I can help is a neighbor's child ten yards from me or a Bengali whose name I shall never know, ten thousand miles away." Nor does the fact that many others could also help. If that's right, then for most people in rich countries, a great deal of spending on non-essentials is morally comparable to walking past the pond. Singer even considers that we might be obliged to give up to "the point of marginal utility," where giving more would cost us as much as it saves.

Singer's argument doesn't need utilitarianism. It needs only one plausible principle and one undeniable case, which is what makes it so hard to wriggle out of.

The strongest case

Consequentialism's strength is that it takes seriously something no other theory can ignore: outcomes matter, and every person's suffering counts equally. It explains why we condemn cruelty, why public policy should weigh costs and benefits, and why the far-away child matters. It has also often been ahead of its time. Bentham wanted his penal code to include a section on cruelty to animals: "the question is not, Can they reason? Nor, can they talk? But, can they suffer?" That line, the SEP notes, lies at the beginnings of utilitarian arguments for the ethical treatment of animals.

The hardest counterexamples

1. The surgeon (case 3). Act consequentialism seems to say: kill the one, if you can keep it secret. Almost everyone thinks that's monstrous. Rule versions help, but only by appealing to bad side effects, which seems like the wrong reason. The wrong isn't that word might get out.
2. Jim and the villagers. In Utilitarianism: For and Against (1973), Bernard Williams imagined Jim, a visitor told that if he shoots one villager, twenty others will be spared. Utilitarianism says the answer is obvious. Williams's point wasn't that the answer is wrong. It was that the theory treats Jim's own agency as just another causal input, ignoring the difference between what I do and what others do, and so missing what he called integrity.
3. The separateness of persons. Rawls argued in 1971 that utilitarianism ignores "the separateness of persons": it lumps everyone together as if society were one person, so that one person's loss can be cancelled out by another's gain. But your pain isn't refunded by my pleasure.
4. Demandingness. If Singer is right, almost everyone is behaving very badly almost all the time. Some consequentialists bite this bullet. Others say a theory that demands this much has gone wrong somewhere.

Part 3 — Kant: duty, dignity and the categorical imperative

The good will

Immanuel Kant's Groundwork of the Metaphysics of Morals (1785) begins in the opposite place from Bentham. "Nothing can possibly be conceived in the world, or even out of it, which can be called good, without qualification, except a good will." Intelligence, courage and even happiness can all be used badly. A good will, even if it achieves nothing, would "like a jewel... still shine by its own light" (quotations from T. K. Abbott's public-domain translation).

What makes a will good isn't what it produces but the principle it acts on. And morality, Kant argues, can't rest on desires, because desires vary. It must rest on reason, which binds every rational being alike. Hence the categorical imperative: a command that holds unconditionally, not "if you want X, do Y" but simply "do Y."

The formulations

Kant states it in several ways, which he insists are versions of one principle:

Kant's own example of the first test is a lying promise. A man needs money, knows he can't repay it, and considers promising to repay anyway. Universalise the maxim, "when I need money, I'll promise falsely," and promising itself collapses. No one would believe a promise, so the maxim defeats itself. That's a perfect duty: you must never act on such a maxim. Other maxims can be universalised without contradiction but can't consistently be willed, such as never helping anyone in need. Those give imperfect duties: you must adopt the end of helping others, but have latitude about when and how.

The humanity formula is the one that does the most work in modern ethics. It explains case 3 immediately: the surgeon uses the visitor's body as a mere means. It also explains why the footbridge feels worse than the lever. It underlies modern ideas of informed consent, human rights and human dignity.

The murderer at the door

In 1797, the French writer Benjamin Constant objected that "a German philosopher" holds it would be a crime to lie to a murderer who asks whether your friend is hiding in your house. Constant's point: a duty to tell the truth is owed only to those who have a right to the truth, and the murderer doesn't.

Kant answered in a short essay, "On a Supposed Right to Tell Lies from Benevolent Motives." He did not back down. If you lie and say your friend is out, and unknown to you the friend has slipped out and meets the murderer in the street, then "you might with justice be accused as the cause of his death." Whereas if you tell the truth, perhaps neighbours catch the murderer while he searches the house. "Whoever then tells a lie, however good his intentions may be, must answer for the consequences of it." Truthfulness, he concludes, is "a sacred unconditional command of reason."

Most readers, including most Kantians, think this is a disaster. Defenders have tried several repairs. Some point out that Kant's essay is framed around legal liability and a "duty of justice," not the whole of ethics. Others argue that the humanity formula itself licenses lying to someone who is using deception to make you complicit in murder. Others again note that silence or refusal is always available. But the case lives on because it exposes a real feature of the theory: some duties are meant to be exceptionless, and exceptionless duties will eventually meet a case where keeping them looks insane.

Kant's theory protects people against being sacrificed for the greater good. The price is that it can tie your hands when the greater good is screaming at you.

The strongest case

Kantian ethics captures something consequentialism keeps missing: persons have a dignity that can't be traded. There are things you may not do to someone even to produce better outcomes. Kant also explains why morality feels like a demand rather than a preference, and grounds it in something everyone shares, the capacity to reason, so that a duty binds a king exactly as it binds a beggar.

The hardest counterexamples

1. The murderer at the door (case 2), above.
2. Conflicting perfect duties. If you've promised to keep a secret and you're asked a direct question, two exceptionless duties collide. Kant's system has trouble saying which yields.
3. The universalisation test gives strange results. Take an illustration of a familiar kind: "I'll buy used cars but never sell one" seems harmless, yet can't be universal, since someone has to sell. Meanwhile some wicked maxims can be phrased so specifically that they pass. Kantians have elaborate replies, but the test is clearly not a mechanical machine for producing verdicts.
4. The rocks (case 4). If each person is an end in themselves, why save five rather than one? Kant gives no obvious reason to count heads. John Taurek's 1977 paper "Should the Numbers Count?" pressed exactly this doubt. Most people find a coin toss between one and five hard to swallow.

Part 4 — Virtue ethics: the return of character

Anscombe's bombshell

By the 1950s, English-language moral philosophy was dominated by forms of utilitarianism and deontology. Then, in 1958, Elizabeth Anscombe published "Modern Moral Philosophy" in the journal Philosophy. The SEP summarises her three theses: that moral philosophy is useless until we have an adequate philosophy of psychology; that we should abandon the concepts of moral obligation, duty and the moral "ought," which make sense only within older conceptions of ethics, such as a framework of divine law, that no longer generally survive; and that the differences among the well-known English moral philosophers since Sidgwick are unimportant. In the same paper she coined a word for the view she was attacking: consequentialism.

Anscombe practised what she preached. In 1956 she publicly opposed Oxford's decision to give an honorary degree to Harry Truman, whom she considered a murderer for the deliberate killing of civilians at Hiroshima and Nagasaki. For her, some acts are simply ruled out, whatever the arithmetic.

Her positive suggestion was to go back to Aristotle: to think about virtues, about what it is to be a just, honest or courageous person, rather than about abstract "obligations." That suggestion, the SEP's entry on virtue ethics says, "heralded" the revival of the whole approach.

MacIntyre's diagnosis

Alasdair MacIntyre's After Virtue (1981) turned Anscombe's hint into a sweeping history of how modern moral philosophy failed. The IEP summarises his diagnosis: contemporary society is a "culture of emotivism," in which moral language is used "to manipulate attitudes, choices, and decisions," and objective-sounding moral rhetoric masks arbitrary preference.

His positive account centres on practices: complex social activities, such as chess, medicine, architecture, football or politics, with goods internal to them, which you can only get by striving for excellence in the practice itself. Virtues are the qualities that let us achieve those internal goods, give our lives a coherent narrative, and sustain the traditions that give practices their meaning.

What virtue ethics says about right action

The obvious challenge is: fine, but what should I do? The modern answer, developed by Rosalind Hursthouse and others, defines right action through the virtuous person. As the SEP puts it, a wrong act is "an act that the phronimos [the practically wise person] characteristically would not do... an act that expresses a vice." For guidance, there are "v-rules," rules framed in virtue terms: do what is honest, do what is kind; don't do what is cruel or cowardly. That turns out to be a lot of guidance.

Run the test kit. Case 3: a just and compassionate surgeon doesn't kill patients. Case 1: a generous and compassionate person doesn't walk past the pond, and probably gives more to distant strangers than most of us do. Case 2 is where virtue ethics shines. An honest person isn't a fanatic about literal truth-telling, and practical wisdom sees that honesty's point doesn't include helping a murderer. The virtuous person lies to him without hesitation.

The strongest case

Virtue ethics captures the parts of moral life the other theories flatten: motivation, perception and character. Bernard Williams made the point with a man who can rescue only one of two people from a catastrophe and chooses his wife. If his reasoning was "it's my wife, and in situations like this it's permissible to save one's wife," then, Williams said, he had "one thought too many." His wife might have hoped the thought was simply that it was her. Morality isn't just choosing the right act. It's being the kind of person who sees the situation rightly and acts from the right motives.

The hardest counterexamples

1. Guidance in real dilemmas. "Do what the virtuous person would do" is little help when you're not virtuous and the virtues conflict. Honesty says tell your friend the truth about her husband; kindness says don't. Practical wisdom is supposed to settle it, but that can sound like "use good judgement," which is what you were asking for help with.
2. Circularity. If right acts are what virtuous people do, and virtuous people are those who do right acts, we've gone in a circle. Defenders reply that virtues are anchored in an account of human flourishing, but that account is contested.
3. The situationist challenge. Philosophers such as John Doris and Harman have cited social psychology suggesting stable character traits may barely exist. In the famous "Good Samaritan" study (Darley and Batson, 1973), unhurried passers-by were about six times more likely than hurried ones to help a man in apparent distress (63% versus 10%, as summarised in the SEP). If whether you help depends on whether you're late, what is "compassion"? Virtue ethicists reply that genuine virtue is rare, so experiments on ordinary people don't refute it. And the evidence base itself is now under scrutiny: the SEP notes that the "recent prominent replication failures in social psychology" bear on how much weight these studies can carry. (Part 7 of How Do We Know Anything?, the next deep dive, looks at that crisis.)

Part 5 — Contractualism: what we could justify to each other

The idea

The fourth theory starts from a different question. Not "what produces the best outcome?" or "what would a virtuous person do?" but "what could I justify to each person affected?"

Its leading modern version is T. M. Scanlon's, in What We Owe to Each Other (1998). His formula: an act is wrong if its performance "would be disallowed by any set of principles for the general regulation of behaviour that no one could reasonably reject as a basis for informed, unforced, general agreement."

It helps to distinguish two social-contract traditions. Contractarianism comes from Hobbes and holds that self-interested people would agree to moral rules because everyone benefits from cooperation; David Gauthier is its best-known modern moral defender. Contractualism comes from Kant and holds that we owe each other principles that respect each person as a reasoner. Scanlon's is the second kind. The question isn't what self-interested bargainers would accept, but what principles no one could reasonably reject.

Why it's attractive

Contractualism has a feature that Parfit named the individualist restriction: in deciding whether a principle can be reasonably rejected, only "various individuals' reasons for objecting to that principle" count. You can't add up small benefits to many people and use the total to override one person's serious objection.

That explains case 3 cleanly. Could the visitor reasonably reject a principle permitting surgeons to kill healthy people for their organs? Obviously. Could any of the five reasonably reject a principle forbidding it? Each of them loses a chance at life, which is serious. But the visitor would lose his life by being used, and would live under a rule that makes everyone a potential organ bank. His complaint is stronger. Contractualism captures what Rawls called "the separateness of persons" without Kant's rigidity.

It handles case 2 easily too. A principle permitting lies to would-be murderers is one the murderer can't reasonably reject: his only complaint is that it frustrates his plan to kill. And it gives a natural account of the drowning child. A principle allowing people to ignore children drowning in front of them, to save their clothes, could be reasonably rejected by any child.

Contractualism's question is the one you'd want asked if you were the person on the losing end: could this be justified to me?

The hardest counterexamples

1. The rocks (case 4). This is contractualism's famous headache. The SEP lays it out using John Taurek's 1977 example. Each of the five on one rock objects to a principle that leaves them to die. So does the one on the other rock, for exactly the same reason. Under the individualist restriction, the five can't appeal to the fact that there are more of them, since that's not a fact about any one person. So the contest looks like a tie, and perhaps you should toss a coin. Many contractualists reply with a "tie-breaking" argument: once one person on each side cancels out, the remaining four still have unanswered claims, and treating five like one acts as if they weren't there. Critics, Otsuka among them, have challenged that argument, and the debate continues.
2. Redundancy. You can reasonably reject a principle because it permits wrong acts. But then the theory needs to know which acts are wrong before it can tell you which acts are wrong. Contractualists reply that "it's wrong" isn't the kind of reason that can be fed into the test; only effects on individuals and how they are treated count. Whether that reply works is a live question.
3. Those who can't take part. Animals can't reject anything, and future people aren't here to; the SEP lists both as groups contractualism seems to leave out. Scanlon's responses are to limit his theory to "what we owe to each other," leaving other moral concerns to other parts of morality, or to imagine trustees who argue on behalf of those who can't. Utilitarians, who count every suffering creature directly, find this a weak spot.

Part 6 — When the theories run out

Ross: several duties, no master rule

Not every moral philosopher thinks one theory must win. W. D. Ross's The Right and the Good (1930) proposed that we have several basic prima facie duties. On the SEP's reading there are five foundational ones: fidelity (keeping promises), reparation (correcting wrongs you've done), gratitude, beneficence and non-maleficence (not harming). Each is a genuine reason to act, but none is absolute. In a real conflict, you have to weigh them, and no formula tells you how. Ross was a deontologist, but a pluralist one, and his view is arguably close to how moral life actually feels.

Particularism: no principles at all

Jonathan Dancy pushes further, most fully in Ethics Without Principles (2004). His core claim, which the SEP calls the holism of reasons, is that "what is a reason in one case may be no reason at all in another, or even a reason on the other side." The SEP's example: that an act is against the law is usually a reason not to do it, but sometimes it's a reason to do it, as a protest against a law that has no business existing. If reasons flip like this, then no exceptionless principle ("lying is always wrong," "always maximise happiness") can capture morality. What's needed is the skilled perception of what matters in each case.

Particularism is the theoretical version of what the test kit keeps showing. Each theory gets some cases right by fitting their particular features, and breaks on cases where those features pull the other way.

Moral uncertainty: acting when you don't know which theory is right

Here's an honest predicament. Suppose you've read all this and you're unsure. You give some credence to consequentialism, some to Kant, some to the others. How should you act?

This has become a research field of its own. One simple answer, which Lockhart dubbed "My Favourite Theory," says follow whichever theory you think most likely. The SEP notes it faces "substantial difficulties": it ignores everything but one possibility. A more popular answer treats moral uncertainty like empirical uncertainty and says maximise expected choiceworthiness: weight each option's rightness under each theory by your confidence in that theory, the way you'd weigh risks under uncertainty. Versions have been defended by Lockhart (2000), Ross (2006), Sepielli (2009), and William MacAskill, Krister Bykvist and Toby Ord in Moral Uncertainty (2020).

The approach has a practical upshot called moral hedging. If one theory says an act is a grave wrong and another says it's merely suboptimal, you have strong reason to avoid it even if you lean toward the second theory. The SEP's own illustration: if you have 49% credence that some animal species has moral standing, it seems reckless to impose massive harm on those animals for a slight benefit to a human, just because the likelier theory permits it. Dietary choices are an obvious place this bites.

It also faces hard objections. How do you compare "how wrong" an act is across theories that measure wrongness in completely different units? This problem of intertheoretic comparison may have no solution. Weatherson argues that caring about "doing the right thing, whatever it turns out to be" is a kind of moral "fetishism," as opposed to caring about the people and things that actually matter. These are open debates.

The scorecard

Two lessons from the table. First, each theory's strongest case is another theory's hardest. The surgeon embarrasses consequentialism and flatters Kant. The murderer at the door embarrasses Kant and flatters virtue ethics. The rocks are easy for consequentialism and hard for both Kant and Scanlon. That pattern is evidence that each theory has grasped something real.

Second, the theories agree on almost every ordinary case. Don't torture for fun, keep your promises, help when it's cheap to help, don't use people. The disagreements live at the edges, which is why philosophers spend their time on trolleys and murderers. Those edge cases matter, because some real decisions, about war, medicine, policy and technology, live at the edges too. But it's worth remembering how much common ground the arguments are fought on.

The point of studying moral theories isn't to pick a team. It's to know exactly which of your values you're trading off when you choose.

The takeaway

Four big moral theories each capture something real. Consequentialism insists outcomes matter and everyone counts equally, and breaks on cases like the surgeon where maximising seems monstrous. Kantian deontology protects the dignity of persons, and breaks on the murderer at the door, where exceptionless duties look insane. Virtue ethics puts character, perception and motive back at the centre, and struggles to guide you through real dilemmas. Contractualism asks what could be justified to each person, and struggles to explain why the numbers count. Pluralists like Ross, particularists like Dancy, and theorists of moral uncertainty offer ways of living with the fact that none of them wins outright. Each theory's best case is another's worst, which suggests they're partial maps of the same territory.

Try this: take a real decision you're facing, or one from today's news, and put it through all four tests in writing, one sentence each. Outcomes: which option produces the most good for everyone affected? Kant: could I will my policy for everyone, and am I using anyone? Virtue: what would a person I admire do, and with what motive? Contract: could I justify this to the person who loses most? Then circle the test whose answer you most want to ignore. That's usually where the real moral cost is hiding.

Sources & Further Reading

Sources
- Jeremy Bentham, An Introduction to the Principles of Morals and Legislation (1780/1789), ch. 1 and ch. 4, as quoted and summarised in the Stanford Encyclopedia of Philosophy (SEP), "Jeremy Bentham."
- John Stuart Mill, Utilitarianism (1861), ch. 2 (Project Gutenberg). SEP, "Mill's Moral and Political Philosophy" (act, rule and sanction readings).
- SEP, "Consequentialism" and "Rule Consequentialism" (act/rule; collapse and rule-worship objections).
- Peter Singer, "Famine, Affluence, and Morality," Philosophy & Public Affairs 1 (1972), revised text as posted on utilitarian.net (archived copy).
- Immanuel Kant, Fundamental Principles of the Metaphysic of Morals (Groundwork, 1785), tr. T. K. Abbott (Project Gutenberg); "On a Supposed Right to Tell Lies from Benevolent Motives" (1797), in Abbott's Kant's Critique of Practical Reason and Other Works on the Theory of Ethics (1889; archive.org). SEP, "Kant's Moral Philosophy" (perfect and imperfect duties).
- Bernard Williams in J. J. C. Smart and B. Williams, Utilitarianism: For and Against (1973), and Moral Luck (1981), via SEP, "Bernard Williams."
- G. E. M. Anscombe, "Modern Moral Philosophy," Philosophy 33 (1958), via SEP, "Elizabeth Anscombe" and "Virtue Ethics." Alasdair MacIntyre, After Virtue (1981), via IEP, "Alasdair MacIntyre."
- SEP, "Moral Psychology: Empirical Approaches" (Darley and Batson 1973 figures; replication caveat).
- T. M. Scanlon, What We Owe to Each Other (1998), and John Taurek, "Should the Numbers Count?" (1977), via SEP, "Contractualism." SEP, "Contemporary Approaches to the Social Contract" (contractarianism versus contractualism).
- W. D. Ross, The Right and the Good (1930), via SEP, "William David Ross." Jonathan Dancy, Ethics Without Principles (2004), via SEP, "Moral Particularism." SEP, "Moral Decision-Making Under Uncertainty." SEP, "Reflective Equilibrium."

Go deeper
- 📘 Utilitarianism: For and Against — J. J. C. Smart and Bernard Williams (1973). A short book in which a utilitarian and a brilliant critic argue it out. Williams's half contains Jim and the integrity objection.
- 📘 Groundwork of the Metaphysics of Morals — Immanuel Kant. Short and difficult; read it alongside a guide. It's the source of everything in Part 3.
- 📘 The Life You Can Save — Peter Singer (2009). The pond argument turned into a book, with his replies to the main objections.
- 📘 After Virtue — Alasdair MacIntyre (1981). Sweeping, opinionated, and the book that made virtue ethics a public conversation.
- 📘 What We Owe to Each Other — T. M. Scanlon (1998). Demanding, but the source of the contractualist test and of most of the debates in Part 5.
- 🌐 Stanford Encyclopedia of Philosophy — Consequentialism. The standard overview, with all the variants.
- 🌐 Stanford Encyclopedia of Philosophy — Moral Decision-Making Under Uncertainty. The best map of the newest debate in Part 6.

For Your Vocabulary

Consequentialism (judging acts by their outcomes; a term Anscombe coined in 1958) · Felicific calculus (Bentham's dimensions for measuring pleasure and pain) · Rule consequentialism (acts are right if allowed by the rules whose acceptance would produce the best outcomes) · Categorical imperative (Kant's unconditional moral command) · Perfect / imperfect duty (a duty with no exceptions / a duty to adopt an end, with latitude in how) · Phronimos (the practically wise person of virtue ethics) · Reasonable rejection (Scanlon's test: could anyone reasonably reject the principle?) · Prima facie duty (Ross: a real but overridable reason to act) · Holism of reasons (the particularist claim that a reason in one case can be no reason, or the opposite, in another) · Expected choiceworthiness (weighting options across moral theories by your confidence in each)