1 What makes an action right or wrong?

The moral philosopher Derek Parfit spent thirty years writing a book called On What Matters.¹ Not “On What Is Interesting” or “On What Is Fashionable.” What matters. The ambition is almost embarrassingly earnest. Ethics is the branch of philosophy that asks how we ought to act, what kind of person we ought to be, and what makes an action right or wrong. Unlike mathematics or natural science, ethics is not primarily about describing what is the case but about what ought to be. That normative character makes it difficult — and makes the temptation to say “it’s all just opinion” understandable. The temptation can be resisted in different ways: by showing that ethical claims have a defensible logical structure (the realist’s route), by showing that even an expressivist reading does not entail “anything goes” (Blackburn), or by treating ethics as a tradition of practical reasoning rather than as a search for facts (the virtue-ethical route).

Disagreement does not by itself make all positions equally defensible — physics, history, and medicine involve disagreement too. The claim that some moral questions are easier than others (gratuitous cruelty is wrong; chattel slavery as practised in the Atlantic system was wrong) has to be argued for, not asserted; eighteenth-century defenders of slavery reasoned at length from premises they took to be obvious, and abolition overturned arguments rather than carelessness. The case for “some moral claims are more defensible than others” is itself a substantive philosophical claim, not a piece of common sense.

Three frameworks have dominated Western moral philosophy since the Enlightenment. They contradict each other on cases that matter.

1.1 Consequentialism

The rightness of an action is determined by its consequences. An action is right if it produces the best available outcome. The most famous version: utilitarianism, developed by Jeremy Bentham (Introduction to the Principles of Morals and Legislation, 1789) and refined by John Stuart Mill (Utilitarianism, 1863). Utilitarianism is the ethical theory of governments and policy-makers — the framework reached for when competing goods must be weighed across large populations. Its critics argue that the same machinery that makes it useful in policy contexts produces conclusions in individual cases that strike most people as monstrous.

“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.” — Jeremy Bentham, Introduction to the Principles of Morals and Legislation, Chapter I, §1³

The slogan, taken at face value, is underdetermined. “Greatest” how — sum total? average? minimum lifted (maximin)? “Greatest number” as a separate maximand, or absorbed into the sum? Henry Sidgwick already saw, in The Methods of Ethics (1874, Book IV, Chapter I), that these read off different verdicts in cases of population growth, sacrifice, and inequality.

R. M. Hare’s Moral Thinking (1981) responded with two-level utilitarianism: at the “intuitive” level we should rely on robust public rules (something like the slogan), reserving direct expected-utility calculation for “critical” reflection on the rules themselves. Direct calculation in the moment is, in Hare’s view, often counterproductive.⁴

1.2 Deontology

The rightness of an action is determined by whether it conforms to a moral rule or duty, independently of consequences. Immanuel Kant (Groundwork of the Metaphysics of Morals, 1785) is the central figure — a man who reportedly led such a regular life that his neighbours set their clocks by his daily walk, and whose moral philosophy has a similarly clockwork quality: precise, systematic, and occasionally infuriating.

“Act only according to that maxim whereby you can at the same time will that it should become a universal law.” — Immanuel Kant, Groundwork of the Metaphysics of Morals, Section II⁵

For Kant, lying is wrong not because it produces bad outcomes — but because the maxim “lie when convenient” cannot be universalised without contradiction. (If everyone lied whenever it suited them, the practice of truth-telling that makes lying effective would collapse.)

Harris’s account of his own state of mind during the Tham Luang operation — focused on what he owed to each child in front of him at each transfer rather than on the operation as a whole — is a Kantian description in form: not “the maximum lives saved across all transfers” but the duty owed to the particular person under the mask.

1.3 Virtue Ethics

The central question is not “what should I do?” but “what kind of person should I be?” Aristotle (Nicomachean Ethics, c. 350 BCE) developed the founding account: right action flows from virtuous character. A courageous person does the courageous thing — not because they have calculated it is right, but because courage is who they are.

The idea at the centre of Aristotle’s account is often rendered — via Will Durant’s paraphrase in The Story of Philosophy (1926) — as: “We are what we repeatedly do. Excellence, then, is not an act, but a habit.” This precise phrasing is Durant’s, not Aristotle’s.⁶ The actual passage reads:

“Excellence of character… is a result of habit.” — Aristotle, Nicomachean Ethics, Book II, Chapter 1⁷

Virtue ethics does not offer a decision procedure (unlike the utility calculation or the categorical imperative).

The Aristotelian framework has a non-Western parallel a century earlier in the Chinese philosopher Mencius (孟子, 4th c. BCE). Mencius 2A.6 introduces the four sprouts (si duan 四端) of moral character: hearts of compassion, shame, deference, and right and wrong — the seeds of ren (humaneness), yi (righteousness), li (propriety), and zhi (wisdom). Mencius’s locus classicus: anyone who saw a child about to fall into a well would feel alarm and compassion — not from calculation but from a sprout descriptively present in human nature, which must be cultivated by habit into the full virtue.⁸ Tham Luang fits the Mencian frame as readily as the Kantian one: Harris’s recoil at leaving the boys to wait out the monsoon, and his focus on each child as he reached them, are the four sprouts under operational pressure. Whether Mencius’s heart of compassion is evidence about the case (Position B’s defence of moral intuition) or about the recoiler (Position A’s debunking reading) is the question the fork below isolates.

1.4 The Trolley Problem

Judith Jarvis Thomson introduced the switch/footbridge formulation in “The Trolley Problem” (1985), building on Philippa Foot’s original case in “The Problem of Abortion and the Doctrine of Double Effect” (1967):⁹

Case 1 (the switch): A runaway trolley is heading towards five people. You can pull a lever and divert it onto a side track where one person stands. Do you pull the lever?

Case 2 (the footbridge): A runaway trolley is heading towards five people. You are on a bridge with a large man whose body would stop the trolley if he fell in front of it. Do you push him?

Most people say yes to Case 1 and no to Case 2. The consequences are arithmetically identical — in both cases, action kills one to save five. Yet our moral intuitions diverge sharply.

• A consequentialist seems committed to yes in both cases.
• A deontologist distinguishes: in Case 1 you redirect an existing threat; in Case 2 you use a person as a means (violating Kant’s formula of humanity).
• A virtue ethicist asks: what would a person of good character do? Is pushing a man to his death ever an act of virtue?

The Trolley Problem is not a parlour game. It is a controlled experiment that isolates the variables in moral reasoning — and exactly that isolation is what its critics (Anscombe, Williams) have called the source of its distortion, since real ethical life does not arrive with relationships, histories, and projects already abstracted away.¹⁰

1.5 Questions

1. The Trolley Problem produces different intuitions for Case 1 and Case 2. But the consequences are the same. What is doing the moral work in our different responses?
2. All three frameworks — consequentialism, deontology, virtue ethics — seem plausible in some cases and implausible in others. Does this mean none of them is right? Or that each captures part of the truth?
3. Is moral disagreement fundamentally different from factual disagreement? If two people cannot agree whether an action is right, is there a fact of the matter they’re both trying to track — or are they just expressing different values?
4. Aristotle says right action flows from good character. But character has to be built through choices. Which comes first?

2 Should we always maximise happiness?

Peter Singer’s drowning-child argument is the section’s argumentative crux: a demand on the affluent, derived from premises few will explicitly reject, that few who endorse them in fact follow. The path to it runs through three predecessors. Bentham gave the doctrine in its sharpest form (the felicific calculus); Mill’s higher/lower distinction is the qualification that almost saves it; Nozick’s experience machine is the thought experiment that suggests qualifications alone cannot save it. Singer’s contribution is the doctrine refused all internal qualifications and applied to a concrete demand: those who can prevent suffering at modest cost are obligated to do so.

Behind all three sits a question Dostoevsky put in the mouth of Ivan Karamazov to his religious brother Alyosha: would you consent to build the edifice of human destiny if it required torturing to death one innocent child?¹⁵ Ivan’s point is that the utilitarian arithmetic — one suffering, millions happy — is precisely what he refuses; Singer’s argument asks the affluent to perform that arithmetic anyway, on suffering they could prevent. Utilitarianism remains the most influential ethical theory in practical politics, public health, and policy-making, and its sophisticated defenders (Sidgwick, Hare, Singer, more recently Parfit and Railton) have replied to Ivan in ways that deserve engagement.

2.1 Bentham’s Felicific Calculus

Jeremy Bentham (Introduction to the Principles of Morals and Legislation, 1789) was an empiricist about morality: pain and pleasure are the only real moral currencies. Everything else — virtue, duty, honour — is meaningful only insofar as it connects to pleasures and pains. He proposed a felicific calculus: a method for quantifying pleasures and pains along seven dimensions (intensity, duration, certainty, propinquity, fecundity, purity, extent).¹⁶

Add up the pleasure, subtract the pain, across all affected people: the action with the highest net score is the right one.

Bentham’s formula, in his Introduction to the Principles of Morals and Legislation: it is the greatest happiness of the greatest number that is the measure of right and wrong. He was serious about the calculation. He spent much of the 1790s designing and lobbying for a prison he called the Panopticon — a circular building in which every cell would be visible from a central inspection tower, yet prisoners could never know whether they were being watched at any given moment.¹⁷ Bentham calculated that this design would maximise good behaviour at minimum cost, thereby producing the greatest happiness of the greatest number; he petitioned Parliament for nearly twenty years before the project was abandoned. That Bentham’s most sustained practical application of his own philosophy was a mechanism of total surveillance is not merely biographical: a system optimised for aggregate welfare can, by its own logic, justify comprehensive control over individual lives — the objection Ivan Karamazov presses. John Stuart Mill — raised by Bentham’s closest associate James Mill as a utilitarian educational experiment¹⁸ — later introduced the higher/lower pleasures distinction precisely to block the inference that a perfectly administered prison could constitute a good life.¹⁹

The problems with the calculus: Can pleasures really be added across different people? Can you sum your pleasure and my pain as if they were quantities of the same thing? Interpersonal utility comparison is one of the foundational problems of welfare economics.

2.2 Mill’s Refinement: Higher and Lower Pleasures

John Stuart Mill (Utilitarianism, 1863) noticed that Bentham’s calculus treats all pleasures as commensurable — a child’s joy in sweets and a philosopher’s joy in argument are just different quantities of the same thing. Mill thought this wrong.

“It is better to be a human being dissatisfied than a pig satisfied; better to be Socrates dissatisfied than a fool satisfied.” — John Stuart Mill, Utilitarianism, Chapter 2²⁰

Mill distinguished higher pleasures (intellectual, aesthetic, moral) from lower pleasures (bodily, animal). Higher pleasures are qualitatively superior — not just more intense but a different kind of good.

The problem with Mill’s refinement: it seems to abandon strict utilitarianism. If you’re allowed to rank pleasures qualitatively, you’re appealing to something other than the utility calculus. What determines which pleasures are “higher”?

Nozick’s experience machine sharpens the same pressure Mill saw — but with a thought experiment instead of a slogan.

2.3 Nozick’s Experience Machine

Robert Nozick (Anarchy, State, and Utopia, 1974) proposed a thought experiment that cuts to the heart of utilitarian hedonism:²¹

Imagine a machine that can give you any experience you desire. It can simulate, perfectly and indistinguishably from reality, a life of achievement, love, success, and happiness. You would not know you were in the machine — from the inside, it would be real. The machine can be programmed before you enter; you will then forget you chose to enter it.

Should you plug in?

Most people say no. They want to actually do things, not just have the experience of doing them. They want to be a certain kind of person. They care about their actual relationships, not simulations.

The utilitarian implication: if you say no to the experience machine, you are implicitly denying that life is only about maximising felt happiness. There are things that matter — achievement, authentic relationships, reality itself — that are not reducible to subjective experience.

2.4 Peter Singer and Effective Altruism

Peter Singer’s “Famine, Affluence, and Morality” (1972) was written in direct response to the 1971 East Bengal crisis: the cyclone that struck on 12 November 1970 followed by the war of Bangladeshi independence had displaced roughly nine million refugees into camps where, Singer argued, deaths from starvation and disease were continuing while affluent governments and individuals spent vastly more on items of comparable cost — concert tickets, new clothes — that they could not have considered as morally weighty. He generalised this to a principle:

If it is in our power to prevent something bad from happening without sacrificing anything of comparable moral importance, we ought to do it. This applies regardless of distance. The drowning child in front of you (Singer’s later, simplifying analogy) and the Bengali child dying in a refugee camp you’ll never visit have equal moral weight.

“I do not think I need to say much in defence of the view that we ought, morally, to give a great deal more than we do.” — Peter Singer, “Famine, Affluence, and Morality” (1972), section III²²

Singer’s argument implies that affluent people who spend money on luxuries while others die of preventable diseases are behaving immorally. He has himself given a substantial fraction of his income to effective-altruist charities for several decades, and distinguishes a “strong” version of the principle (give up to the point of marginal utility — until giving more would harm you as much as it helps others) from a “moderate” practical standard. In The Life You Can Save the moderate standard is calibrated to be reachable rather than deterring: at least 5% of income for ninety-five per cent of Americans, with a sliding scale rising sharply for higher earners.²³Effective altruism is the movement built on this principle, asking not just “should I give?” but “how can I do the most good with what I give?”

The objection: Singer’s argument seems to demand too much. It implies that middle-class people should impoverish themselves to the point of marginal utility — give until giving more would harm you as much as it helps others. Bernard Williams and Susan Wolf press the demandingness charge, arguing that the theory’s all-things-considered command leaves no room for personal projects and commitments.²⁴

2.5 Questions

1. If the experience machine provides everything hedonist utilitarianism values, and you still wouldn’t plug in, what does that tell you about the theory?
2. Singer argues there is no morally relevant difference between a drowning child in front of you and a starving child on another continent. Is this right? If you disagree, what is the morally relevant difference?
3. Mill’s distinction between higher and lower pleasures saves utilitarianism from the “pig satisfied” objection but introduces a qualitative ranking that the theory can’t fully justify. Is this a fatal flaw or a feature?
4. How much are you morally obligated to give? Is there a principled limit to utilitarian demands — or should we accept that the theory is demanding and adjust our lives accordingly?

3 What is duty — and can it override consequences?

Singer’s drowning child gave utilitarianism’s standing complaint: it demands too much — strictly applied, it would dissolve the personal life into a global welfare calculation. Deontology faces the mirror objection. Kant’s “murderer at the door” — the case Grüninger and Lutz lived through at the Swiss border and in Budapest in 1938–44, and the case played out at the Hôtel des Mille Collines in Kigali in 1994 — is the inversion: strictly applied, deontology seems to demand too little, forbidding the lie that would save the friend.

A worked example sits behind the doctrine. Between 6 April and mid-July 1994, the Rwandan genocide killed roughly 800,000 Tutsi and moderate Hutu in about a hundred days — the killing directed by radio broadcasts and government lists, carried out largely with machetes. Through those months, around 1,268 Tutsi and moderate Hutu refugees survived inside the Hôtel des Mille Collines in Kigali. Interahamwe commanders came repeatedly to the gate with names and room numbers; the refugees were not handed over; the hotel was never overrun. The popular account — Terry George’s film Hotel Rwanda (2004), the manager Paul Rusesabagina’s own An Ordinary Man (2006) — credits Rusesabagina with saving them through bribery and lies told to militia officers. That heroic-individual version is sharply contested, and his later prosecution in Rwanda complicates it further.²⁸²⁹³⁰ But the philosophical use of the case does not stand or fall with him. It requires only that someone at the hotel — the manager, the UN observers, the Congolese peacekeepers, the refugees coordinating at the gate — repeatedly chose to deceive armed men who had come for the names of those hiding upstairs, and that the deceptions kept over a thousand people alive.

Kant’s deontology is the most ambitious attempt to ground morality in pure reason, regardless of what works out in practice. It is also among the most counterintuitive: it seems to require telling the truth to a murderer who asks where your friend is hiding.

3.1 The Categorical Imperative

Kant’s central moral principle has several formulations. Two are most important:

Formula of Universal Law: “Act only according to that maxim whereby you can at the same time will that it should become a universal law.” (Groundwork, §421)³¹

Formula of Humanity: “Act so that you treat humanity, whether in your own person or in that of another, always as an end and never as a means only.” (Groundwork, §429)³²

The two formulas are supposed to be equivalent, though this is contested. Together they capture what Kant meant by the categorical imperative — an imperative that applies to all rational beings unconditionally, not contingently on what you want (that would be a hypothetical imperative: “if you want X, do Y”).

3.2 Testing Maxims

Kant’s method: extract the maxim of an action — the general principle the action expresses — and ask whether it could be a universal law. Can everyone act on this principle consistently?

Lying: Suppose I lie to someone to get money. My maxim: “Lie when it suits me.” Can I universalise this? If everyone lied whenever it was convenient, the practice of telling the truth that makes lying possible would collapse. (You can only successfully lie if the person you’re lying to expects truth-telling.) The maxim is self-defeating when universalised. Therefore lying is wrong — full stop.

Suicide: Suppose I am in despair and consider ending my life to escape misery. Maxim: “Destroy my life when its continuation brings more suffering than pleasure.” Kant argues this is self-defeating: life is the condition of all action; to destroy life to benefit oneself contradicts the principle that life is valuable as the ground of rational agency.

Helping others: Maxim: “Never help anyone; let everyone fend for themselves.” This can be universalised without contradiction — a world of pure self-sufficiency is not logically impossible. But Kant argues no one could rationally will it: we cannot consistently will a world in which no one would help us, since we know we will need help.

3.3 The Murderer at the Door

Kant’s most notorious commitment: in his essay “On a Supposed Right to Lie from Philanthropy” (1797), he argued that you must not lie to a murderer who asks where your friend is hiding. If you lie and the murderer somehow finds your friend anyway, you are responsible for the outcome. If you tell the truth and the murderer finds and kills your friend, the murderer alone is responsible.

“Truthfulness in statements that one cannot avoid is a human being’s duty to everyone, however great the disadvantage to him or to another that may result from it.” — Immanuel Kant, “On a Supposed Right to Lie from Philanthropy,” in Practical Philosophy, trans. Mary J. Gregor (Cambridge University Press, 1996), AK 8:426³³

This seems monstrous. Benjamin Constant and many later philosophers argued that duties apply only to people who have a right to the information — a murderer has forfeited his right to your honest answer.³⁴

Kant’s defenders argue: he is not saying “inform the murderer.” He is saying don’t lie. There is a third option: refuse to answer, misdirect without lying, etc. The strict deontological prohibition on lying does not require handing your friend over.

3.4 Can the Categorical Imperative be Repaired?

The question is not whether Kant produces the wrong verdict in extreme cases — most people agree he does. The question is whether the framework can be revised to produce the right verdict while remaining distinctively Kantian. Several repair strategies have been proposed.

1. The third maxim option. Kant himself suggested there is often a third option: you need not lie to the murderer; you can refuse to answer, misdirect without asserting falsehood, or ask why they want to know. At the Mille Collines, in a family hiding refugees from a regime, when a militia commander at the gate demands to know which rooms hold Tutsi, there is no time for philosophical evasion.

2. The rights-forfeiture approach. A murderer who intends to kill has forfeited their right to honest treatment. Benjamin Constant argued this against Kant directly; Christine Korsgaard has developed a neo-Kantian version.³⁵

But this requires importing a concept of rights forfeiture into the categorical imperative framework, which some regard as a fundamental revision rather than a repair. If we can forfeit our right to honest treatment by our intentions, then the categorical imperative is no longer genuinely categorical — it now has an escape clause that must be adjudicated in each case.

3. The conflicting duties resolution. W.D. Ross (The Right and the Good, 1930) argued that we have prima facie duties — duties that hold unless overridden by stronger competing duties.³⁶ In the Rwandan case, the prima facie duty of truth-telling is overridden by the prima facie duty to protect innocent life. This handles the case gracefully. But it also abandons the categorical nature of the categorical imperative, turning it into something much closer to a balancing test.

None of these three repairs preserves the framework intact. A repaired framework that allows rights forfeiture, or that treats duties as defeasible, or that requires case-by-case balancing, has changed something Kant himself thought essential. For Williams and the consequentialist tradition generally, that is evidence the framework fails in precisely the cases where moral theory should most help. Korsgaard, O’Neill, and Herman read the same evidence the other way: the murderer-at-the-door essay is a late and atypical text, the deeper Kantian commitment is to the Formula of Humanity, and the apparent rigidity dissolves when one attends to what the categorical imperative actually demands of a rational agent confronting an evildoer.

Of the three, Ross’s prima facie duties is the repair with the most lasting influence on contemporary deontology — and also the one that most visibly trades the categorical character of the imperative for a balancing test, which is precisely what Kant’s framework was designed to refuse. What gives Kant’s framework its continuing philosophical power is its insistence that persons have a dignity that cannot be traded away in a calculation. The Formula of Humanity captures something real.

3.5 Strengths of Deontology

Despite the hard cases:

• Deontology takes the separateness of persons seriously. Each person is an end in themselves, not a resource to be traded off against others. This blocks the utilitarian move of sacrificing one for many.
• It captures the intuition that some things are wrong regardless of consequences: torturing innocents, using people as experimental subjects without consent, punishing the innocent to appease an angry mob.
• It explains the concept of rights: something you hold against arbitrary interference, even well-intentioned interference.

The trolley case: a deontologist says you may divert the trolley (you are redirecting a threat) but may not push the man off the bridge (you are using him as a means). This is the agent-relative vs. agent-neutral distinction.

3.6 Questions

1. Can you really not lie to the murderer at the door? If not, is this a reductio of Kant’s theory, or a case where the theory’s rigour forces a hard but correct conclusion?
2. Kant says morality is grounded in reason, not emotion or desire. Does this make ethics more objective — or does it just disguise the fact that our “rational” intuitions are themselves culturally formed?
3. The Formula of Humanity says never treat someone merely as a means. But is it possible to never treat people as means at all? Taxi drivers are treated as means of transport; is that wrong?
4. Utilitarians and deontologists often reach the same conclusions in ordinary cases. What are the genuine, practical implications of the difference between them?

3.7 Interlude: What Kind of Claim Is a Moral Claim?

This section steps outside the unit’s usual shape. There is no info-box situation, no fork, no commentary — only a brief survey of the meta-ethical positions the rest of the unit assumes. Read it once before §4 and return to it whenever a fork tempts you to ask “but what is this argument actually about?”

You have now lived through one consequentialist disaster (Singer’s £12 coffee, the trolley body count) and one deontological one (the murderer at the door, Kant’s refusal of the lie). In each you watched a careful moral framework reach a verdict that some readers find right and others find monstrous. That recurring shape raises a deeper, prior question: what kind of claim is “you ought to push him” in the first place? When the consequentialist says it is required and the deontologist says it is forbidden, are they contradicting each other about a moral fact — or expressing incompatible attitudes — or both arguing inside a fiction?

Moral realism holds that moral claims are genuine truth-apt statements about an independently existing moral reality. When we say “gratuitous cruelty is wrong,” we are stating something objectively true, not merely expressing a preference. G.E. Moore (Principia Ethica, 1903) defended a version of this: moral properties like “good” are real, non-natural properties, not reducible to any natural facts about pleasure, evolutionary fitness, or social approval.³⁷

Expressivism holds that moral claims are not statements of fact at all — they express emotional attitudes or prescriptions. A.J. Ayer (Language, Truth and Logic, 1936) put the bluntest version: “If I say to someone, ‘You acted wrongly in stealing that money,’ I am not stating anything more than if I had simply said, ‘You stole that money.’ In adding that this action is wrong I am not making any further statement about it. I am simply evincing my moral disapproval of it.”³⁸ This is sometimes called the boo/hooray theory: moral statements are sophisticated expressions of approval and disapproval, not descriptions of moral facts. Simon Blackburn’s quasi-realism is a more sophisticated descendant: we can talk as if moral claims are true without being committed to moral realism, because the expressive practices are deeply entrenched in human life.³⁹

Error theory (J.L. Mackie, Ethics: Inventing Right and Wrong, 1977) holds a third position: moral claims do purport to state objective facts, but those objective facts don’t exist. Every moral claim is, strictly speaking, false. Mackie called moral objectivity a “queerness” — moral facts, if they existed, would be utterly unlike anything else in the natural world. On his view, morality is a useful and pervasive fiction.⁴⁰

Thick-concept theory holds that the entire debate above is mis-posed because the genuinely informative parts of our moral vocabulary are not the thin terms — “good”, “right”, “ought” — that the realism/expressivism/error-theory dispute is conducted in, but the thick ones: “cruel”, “treacherous”, “cowardly”, “brutal”, “promised”, “courageous”. Bernard Williams (Ethics and the Limits of Philosophy, 1985) is the central figure. These thick terms, Williams writes, “seem to express a union of fact and value. The way these notions are applied is determined by what the world is like (for instance, by how someone has behaved), and yet, at the same time, their application usually involves a certain valuation of the situation, of persons or actions.”⁴¹

To call an act treacherous is not first to describe it neutrally and then to add a private “boo” — as Ayer’s analysis would require — but to apply a single concept that already fuses observation and judgement. The expressivist programme has to claim that thick terms decompose into a descriptive complex plus a thin prescription; Williams argues this decomposition cannot in general be carried out, because someone outside the relevant social world cannot pick out which acts the term applies to.⁴²

Back to the trolley, the murderer, the £12 coffee. The four positions are not idle scholasticism. Whether the dispute between the consequentialist and the deontologist is a disagreement about a moral fact (Moore), about attitudes (Ayer/Blackburn), about a shared fiction (Mackie), or mis-posed because the genuinely operative concepts are thick (Williams), changes what would even count as resolving any of the cases the past three lessons have asked you to commit on.

4 What is virtue — and can it be taught?

Ask a child whether it was right to pull the trolley lever and save five people by killing one, and you will get an answer within thirty seconds. Ask the same question of the Bulger case — what should a community owe two ten-year-olds whose characters were still forming? — and the question stops being about a single act. Ask an Aristotelian, and the response begins by asking what kind of person you are, what habits you have built, what a person of good character would feel about being placed in such a situation. To a consequentialist this looks like evading the question; to the Aristotelian it looks like correcting a misframing. Aristotle’s ethics is not about rules or calculations. It is about character. The question is not “what should I do right now?” but “what kind of person should I become?” That re-framing shifts ethics toward the good life (eudaimonia) rather than a decision procedure for dilemmas.

4.1 Eudaimonia and the Good Life

Eudaimonia is usually translated as “happiness” but the Greek word is richer. It means something like human flourishing — living well, functioning well, actualising your capacities as a human being. It is not a feeling (you can feel happy while flourishing, but happiness is a by-product, not the thing itself).

Aristotle’s argument (Nicomachean Ethics, Book I): every action aims at some good. Goods are hierarchically ordered: we do X for the sake of Y, and Y for the sake of Z. There must be some final good — something sought for its own sake, not as a means to anything further. That final good is eudaimonia.

What is eudaimonia? Aristotle answers by asking what is distinctive about humans. All animals live; plants also grow and reproduce. What is distinctively human is rational activity. Eudaimonia is the excellent exercise of distinctively human rational activity over a complete life.

Aristotle takes virtue ethics to be tied to facts about human nature and what constitutes human excellence — facts he thought were discoverable through observation and reflection.

4.2 Virtue as Habit (Hexis)

A virtue (arete) is a stable disposition to feel and act in certain ways. Courage is not an isolated act of bravery — it is a settled character trait that reliably produces courageous behaviour in appropriate situations. Aristotle’s key insight:

“We become just by doing just acts, temperate by doing temperate acts, brave by doing brave acts.” — Aristotle, Nicomachean Ethics, Book II, Chapter 1⁴⁴

We are not born virtuous. We become virtuous through habit (ethos, from which we get “ethics”). Character is formed by repeated action. This is why education and upbringing matter: habits instilled in childhood shape character in ways that adult correction does not easily reach.

The virtuous mean: virtues lie between two vices, one of excess and one of deficiency. Courage is the mean between recklessness (excess) and cowardice (deficiency). Generosity is the mean between profligacy and miserliness. This is the doctrine of the mean (mesotes); the mean is relative to the person and situation rather than a mathematical average.

4.3 Phronesis: Practical Wisdom

The master virtue for Aristotle is phronesis — practical wisdom, or the capacity to reason well about what to do in particular situations. This is not the application of a rule; it is an acquired sensitivity to morally relevant features of situations.

Phronesis is what allows a virtuous person to navigate cases that rules cannot anticipate. A worked example: an emergency-medicine doctor faces a patient whose vital signs are within protocol but whose presentation is off in a way the protocol does not name. The doctor with phronesis recognises the wrongness and orders the further investigation the protocol does not require; the doctor without it follows the protocol cleanly and misses the early sepsis. Both have applied medical knowledge; only one has practical wisdom.

Aristotle therefore says ethics cannot be reduced to a calculation or a rule book. NE Book I: “it is the mark of an educated man to look for precision in each class of things just so far as the nature of the subject admits.”⁴⁵

4.4 Can Virtue Be Taught?

Plato’s Meno begins: “Can virtue be taught?” Socrates claims not to know even what virtue is, so how could he know whether it can be taught? The dialogue concludes — tentatively — that virtue comes neither from teaching nor from nature alone but from divine dispensation to rare individuals.⁴⁶

Aristotle disagrees: virtue can be cultivated through practice, habituation, and good models. But the process requires the right conditions: good upbringing, good political institutions, freedom from dire poverty.

Modern revival: Alasdair MacIntyre (After Virtue, 1981) argued that modern moral philosophy — dominated by utilitarian and Kantian frameworks — has collapsed, because it debates moral principles while having lost the tradition within which those principles made sense.⁴⁷

4.5 Questions

1. Aristotle says we become courageous by doing courageous acts. But how can you do a courageous act before you have courage? Which comes first — the act or the virtue?
2. Is eudaimonia the same for all humans, or does it differ between individuals and cultures? If it differs, is virtue ethics still universal — or does it collapse into relativism?
3. MacIntyre argues that modern ethics has lost the tradition within which virtues make sense. Does this mean virtue ethics is hopelessly tied to particular communities? Or can a secular, pluralist society support virtue ethics?
4. Aristotle’s doctrine of the mean suggests virtues are context-dependent. Is there a virtue corresponding to ruthlessness or deception in certain situations? (Machiavelli thought so.) Where does virtue ethics draw the line?

5 Does morality depend on culture?

When anthropologists visited societies with radically different moral practices — cultures where infanticide was normal, where theft from strangers was celebrated, where gender roles were utterly unlike our own — the obvious question arose: are these cultures wrong, or are moral standards simply different in different places? This question separates moral relativism from moral universalism, and the answer has enormous practical consequences. It is also, it should be said, a question that tends to make people in liberal societies feel pious on both sides: certain they cannot be judgemental about other cultures, equally certain that female genital mutilation is wrong.

5.1 Moral Relativism: Ruth Benedict

Ruth Benedict, the American anthropologist, argued in Patterns of Culture (1934) that moral norms are culturally constituted. What counts as normal or good varies by culture, and there is no standpoint outside all cultures from which to judge. She described cultures — including Western culture — as one pattern among many, each coherent on its own terms.

Benedict’s view is sometimes called cultural relativism: the descriptive claim that moral beliefs vary across cultures. This is uncontroversially true. But she sometimes slid toward moral relativism: the normative claim that there are no universal moral standards, and that a moral judgement is valid only relative to a cultural framework.

“Morality is a convenient term for socially approved habits.” — Ruth Benedict, Anthropology and the Abnormal (1934)⁵¹

If this is right: “female genital mutilation is wrong” is not a cross-cultural truth — it is the expression of one cultural framework’s norms.

5.2 The Problem with Moral Relativism

Moral relativism is appealing — it sounds tolerant and anti-imperialist. It has serious problems:

1. The reformer’s dilemma: If morality is just culturally approved habit, then moral reformers — abolitionists, suffragettes, those who resisted the Holocaust — were simply wrong by their culture’s standards. William Wilberforce, working from 1787 against the well-defended commercial and theological consensus of an empire whose Caribbean economy depended on enslaved labour, was on the relativist’s account wrong to oppose the slave trade until enough other people agreed with him to pass the Slave Trade Act of 1807 and (in the year of his death) the Slavery Abolition Act of 1833.⁵²

2. The tolerance argument is in tension with itself: Relativists often say “we should tolerate other cultures’ moral practices.” But tolerance is itself a moral value, and the relativist owes an account of its status. The sophisticated reply is that the relativist need not assert tolerance as a universal value but as a value internal to the liberal traditions that produced relativism — which makes the position more coherent at the cost of denying its own anti-imperialist appeal, since cultures whose internal traditions endorse intolerance cannot then be criticised on relativist grounds.

3. Disagreement doesn’t entail there is no fact: People disagree about historical questions, about the safety of vaccines, about whether quantum mechanics is complete. We do not conclude from this that there is no fact of the matter. The relativist needs a further argument that moral disagreement is different in kind — most plausibly that it persists under ideal conditions of reflection in a way that scientific disagreement does not. Mackie’s “argument from relativity” attempts to make this case.

5.3 Moral Universalism

Moral universalists hold that some moral claims are true regardless of cultural context. The strongest version of the case starts not from a list of doctrines but from a claim about the human creature: any being capable of suffering, capable of being deceived, capable of caring about the fate of others, capable of being humiliated or honoured, stands in moral relations to others by the brute fact of those capacities. Aristotelian eudaimonist universalists (Nussbaum’s central human capabilities), Kantian dignity-universalists, and theological natural-law universalists differ on the metaphysics but converge on the structure: the moral standing is grounded in features of human beings that cross cultures rather than in features of any single culture’s moral vocabulary. The arguments for universalism that follow articulate three forms of that case.

• Common humanity: all humans are capable of suffering, have basic needs, and stand in relations of mutual dependence. This is offered as ground for at least a minimal universal morality. The relativist’s standing rejoinder — that “what counts as suffering” and “what mutual dependence requires” are themselves culturally inflected — has to be answered, not waved away.
• Convergence: across radically different cultures, there are striking convergences — prohibitions on gratuitous cruelty, norms of reciprocity, care for children. Whether this is evidence of shared moral truth or merely of shared selection pressures on a cooperative species is exactly where universalist and relativist part company.
• Moral progress: the abolition of slavery, the extension of rights to women, the recognition of LGBTQ+ rights — these look to most modern observers like progress, not merely change. Two relativist escapes are available: deny that “progress” is anything but a value-laden judgement that begs the question; or grant that within a given culture some changes count as progress on the culture’s own terms (Bernard Williams’s route, distinguishing “real” from “notional” confrontations between value systems⁵³).
• Cross-cultural moral development: Lawrence Kohlberg’s research, building on Piaget (The Philosophy of Moral Development, 1981), traced a consistent developmental sequence — pre-conventional reasoning about punishment and reward, conventional reasoning about social conformity, post-conventional reasoning about universal principles independent of social rules.⁵⁴ The post-conventional stage represents reasoning that can criticise social norms from a principled standpoint, and Kohlberg found the sequence across many cultures. Carol Gilligan’s In a Different Voice (1982) raised the standing objection that Kohlberg’s schema valorised an abstract justice model characteristic of Western male research subjects, undervaluing the care-ethic more characteristic of women’s moral reasoning;⁵⁵ the dispute is not over whether the developmental sequence exists but over whether “higher” stages are neutrally identified or selected by criteria already friendly to the universalist’s preferred shape of reasoning.

5.4 The UN Declaration of Human Rights as a Test Case

The Universal Declaration of Human Rights (adopted 1948) asserts that certain rights apply to all human beings regardless of nationality, religion, or culture. It was drafted in the shadow of the Holocaust and adopted by the UN General Assembly — but not unanimously.

The universalist assumption: Article 1 — “All human beings are born free and equal in dignity and rights.” This is a substantive moral claim, not a cultural one.⁵⁶

The relativist challenge: Governments from various cultural traditions have argued that the UDHR reflects Western liberal individualist values and cannot simply be imposed on societies with different traditions of community, authority, or religious law. Asian values debate (1990s); Islamic states’ reservations about inheritance law, gender equality provisions.⁵⁷

The philosophical question: Is the UDHR a statement of universal moral truths, or a powerful particular culture’s values expressed in universal language?

5.5 Questions

1. Is the distinction between “cultural relativism” (descriptive) and “moral relativism” (normative) a stable one? Can you accept the former without being committed to the latter?
2. If moral norms are culturally relative, then no culture can be criticised by another. But many moral reformers within their own culture have appealed to principles that transcend local norms. What does this show?
3. Kohlberg’s stages imply that some moral reasoning is better than others. Does this require that some moral conclusions are objectively truer than others?
4. Is the Universal Declaration of Human Rights a universal moral truth — or Western moral imperialism? Is there a difference?

6 Can science tell us what is good?

Science and morality are at war in contemporary philosophy. The temptation runs both ways: if science can tell us about human psychology, neuroscience, and evolutionary history, why can’t it tell us how we should act? — and conversely, if morality has any grip on us, why should it not be answerable to the best account of what we are? Hume’s objection (Treatise, III.I.I⁶³) is a point about what an argument’s premises license. From purely descriptive premises — sentences of the form “X is the case” — you cannot validly derive a normative conclusion of the form “Y ought to be done,” because the conclusion contains a concept (“ought”) that is nowhere in the premises. No amount of empirical evidence supplies that concept on its own.

6.1 The Is-Ought Gap, Applied

Hume showed that you cannot derive an “ought” from an “is” — no set of purely descriptive premises entails a normative conclusion. Applied to ethics: no scientific facts, however comprehensive, automatically generate moral conclusions. You need an additional normative premise.

This doesn’t mean science is irrelevant to ethics. It means:

• Science tells you the consequences of actions — what will happen if you do X.
• Science tells you what people prefer, what causes suffering, what promotes welfare.
• But “this action causes suffering” does not, by itself, entail “this action is wrong” — without the further premise that causing suffering is (other things equal) wrong.

6.2 G.E. Moore’s Naturalistic Fallacy

G.E. Moore (Principia Ethica, 1903) argued that any attempt to define “good” in terms of natural properties — pleasure, survival, what evolution favoured — commits the naturalistic fallacy. His Open Question Argument: for any natural property N, the question “Is N good?” remains genuinely open. If good just meant N, the question would be trivially closed.⁶⁴ Moore’s argument is not a separate objection to Hume’s; it is the same logical pressure run through a different test — Hume’s about premises and conclusions, Moore’s about meaning and the open-question criterion.

This creates a puzzle: if good cannot be defined in natural terms, what is goodness? Moore thought it was a simple, non-natural property, directly apprehended through moral intuition. This view — moral non-naturalism — has its own difficulties.

6.3 E.O. Wilson’s Sociobiology

Edward O. Wilson (Sociobiology: The New Synthesis, 1975; On Human Nature, 1978) argued that human moral intuitions — altruism, kin preference, reciprocity — can be explained by evolutionary biology. We care for our children because kin selection promotes gene replication. We have a sense of fairness because it stabilises cooperative groups.

Wilson went further: ethics should be “biologised.” Our moral intuitions are adaptations, shaped by natural selection. Understanding this should inform our ethical thinking.⁶⁵ (Moore’s argument from 1903 predates Wilson by seventy years; Wilson’s project is one instance of what Moore’s general pressure attacks.)

The philosophical problem: even granting that our moral intuitions are products of evolution, this on its own does not show they are correct. Sharon Street and Richard Joyce’s evolutionary debunking argument is a dilemma for the moral realist: if our moral intuitions came from selection, and selection tracks reproductive success rather than moral truth, the realist owes an account of why those intuitions are reliable evidence about moral facts. Either show that selection pressures did track moral truth (a tall order, since reproductive success and moral truth are not obviously related), or accept a non-realist account on which there are no moral facts for selection to have tracked.

6.4 Sam Harris: The Moral Landscape

Sam Harris (The Moral Landscape, 2010) argues that Hume’s guillotine can be cut through: if we define morality as concerning the wellbeing of conscious creatures, then science can study morality — because wellbeing is an empirical phenomenon.

“There are right and wrong answers to moral questions, just as there are right and wrong answers to questions of physics, and such answers may one day fall within reach of the maturing sciences of mind.” — Sam Harris, The Moral Landscape (2010), Chapter 1⁶⁶

Harris’s move: accept one normative premise (wellbeing matters) as a foundational axiom, and then do science about how to maximise it. The foundational axiom is not itself derived from science, but everything else can be.

The objection: the foundational premise — “wellbeing is what morality is about” — is contested. A Kantian would say morality is about duty, not wellbeing. A virtue ethicist would say it’s about character. Harris’s foundational commitment is itself a moral stance in need of justification.

6.5 What Neuroscience Can and Cannot Tell Us

Neuroscience of morality (Jonathan Haidt and others) has shown:

• Moral judgements involve both emotional and rational brain circuits.
• Moral intuitions are often post-hoc rationalisations of gut reactions (Haidt’s “moral dumbfounding” experiments, first reported in “The Emotional Dog and Its Rational Tail,” Psychological Review 108.4, 2001).

Jonathan Haidt (The Righteous Mind, 2012): moral reasoning is largely motivated reasoning — we reach conclusions emotionally and then construct arguments to justify them. This is “the rider and the elephant”: reason (rider) sits atop moral emotion (elephant) and mostly goes where the elephant goes.⁶⁷

The epistemological implication: if much moral reasoning is post-hoc rationalisation, what work can moral arguments still do? Rawls’s reflective-equilibrium method takes intuitions seriously as data; Haidt’s findings put pressure on how much weight that method can bear.

6.6 Questions

1. Harris says accepting “wellbeing matters” as a foundational axiom allows science to study morality. But who gets to choose the axiom? And if the axiom is a moral commitment, hasn’t Harris just conceded Hume’s point?
2. Evolutionary psychology explains our moral intuitions as adaptations for fitness. Does this make our moral beliefs less trustworthy — since they were selected for reproductive success, not truth?
3. Haidt argues that moral reasoning is mostly post-hoc rationalisation. If that’s right, should we trust philosophical argument in ethics? Or does this point undermine the whole enterprise?
4. Is there a version of naturalism in ethics that avoids the naturalistic fallacy? Or does any attempt to ground ethics in natural facts face Moore’s Open Question?

7 What are human rights — and where do they come from?

“We hold these truths to be self-evident, that all men are created equal, that they are endowed by their Creator with certain unalienable Rights.”⁶⁹ The American Declaration of Independence declares its moral foundation to be self-evident — the one kind of claim that requires no argument. Jefferson was a philosopher, and he knew perfectly well that this was a rhetorical move, not a proof. The Universal Declaration of Human Rights (1948) asserts that every human being, by virtue of being human, has certain inalienable rights. Inalienable according to what, exactly? Where do they come from?

Three positions on grounding will be surveyed — natural rights (Locke), constructed rights through agreement (Rousseau, Rawls), and constructed rights through historical extension of sentiment (Rorty) — followed by the working problem all three must face: rights in conflict.

7.1 Natural Rights: Locke

John Locke (Two Treatises of Government, 1689) argued that humans have natural rights to life, liberty, and property — rights that exist in the state of nature, prior to any political arrangement. These rights are grounded in the nature of human beings as rational, created by God, with inherent dignity.

“The state of nature has a law of nature to govern it, which obliges every one: and reason, which is that law, teaches all mankind, who will but consult it, that being all equal and independent, no one ought to harm another in his life, health, liberty, or possessions.” — John Locke, Second Treatise of Government, Chapter II, §6⁷⁰

Locke’s rights are pre-political. Governments are legitimate only insofar as they protect these natural rights; on Locke’s account, systematic violation of natural rights gives subjects grounds to resist.

The problem: the claim that rights are grounded in God and nature requires the existence of God and a particular account of nature. For those who do not share these metaphysical assumptions, the grounding fails. Jeremy Bentham called natural rights “nonsense upon stilts.”⁷¹

7.2 Social Contract: Rousseau and Rawls

Jean-Jacques Rousseau (The Social Contract, 1762): rights are the product of a social contract — an agreement among persons to form a political community under shared rules. Rights are what rational persons would agree to in constructing a political framework.

John Rawls (A Theory of Justice, 1971) gives the most sophisticated modern version. The original position: imagine you must choose the basic principles of a just society from behind a veil of ignorance — without knowing your gender, race, class, talents, or conception of the good. What principles would rational persons choose?⁷²

Rawls’s answer: they would choose principles that maximise the position of the worst-off (the difference principle) and guarantee basic liberties for all.⁷³

The strength: Rawls generates rights without needing God or a fixed “nature.” The grounding is in rational agreement.

The problem: the original position is a thought experiment. Real social contracts were made by some people at particular times, typically excluding others. And rational choice behind the veil of ignorance requires idealising people in ways that may not reflect actual moral psychology.

7.3 Rights as Constructed: Rorty

Richard Rorty (Human Rights, Rationality, and Sentimentality, 1993) takes a deflationary view: human rights are not grounded in nature, God, or rational construction. They are historical achievements — cultural practices that extend the circle of moral concern progressively. Instead of asking “what are the philosophical foundations of human rights?” we should ask “how do we get more people to care about more human beings?”⁷⁴

Rorty’s answer: through literature, narrative, and emotional identification. When people read novels, watch films, and hear stories about the suffering of distant others, they extend empathy. Rights practices spread through sentiment, not philosophical argument. Rorty’s project is edifying rather than foundational: in Contingency, Irony, and Solidarity (1989) he proposed that the most useful figure for the rights-defender is not the philosopher with arguments but the novelist who can make a previously distant suffering recognisable — Harriet Beecher Stowe, Primo Levi, Toni Morrison — and that the historical extension of rights tracks exactly this kind of work.

The objection: if rights are just constructed cultural practices with no deeper grounding, what do we say to someone who rejects the practices?

7.4 Questions

1. If God does not exist, does Locke’s natural rights theory collapse? Is there a version of natural rights that doesn’t require a theological foundation?
2. Rawls’ veil of ignorance is supposed to eliminate bias from the choice of principles. But is it possible to choose behind a veil of ignorance — doesn’t choosing still require a conception of what matters?
3. Rorty says rights practices spread through stories and sentiment, not philosophical argument. Is he right that narrative is more effective than argument in expanding the circle of moral concern? And if he’s right, does that tell us anything about the grounding of rights?
4. Article 19 of the UDHR (“Everyone has the right to freedom of opinion and expression”) is regularly in tension with Article 12 (“No one shall be subjected to arbitrary interference with his privacy”).⁷⁵ When the two conflict — publication of private information, defamatory speech, incitement to hatred — how should we decide which takes priority? Is the answer principled or political?
5. The same question recurs in three working forms. Does freedom of expression include the right to incite violence against minorities (free speech vs. hate speech)? Do basic needs override property rights (food, shelter, healthcare)? Can states compel vaccination to protect public health (individual rights vs. collective welfare)? Each is a test of any account of rights, and rights frameworks do not resolve them automatically.

8 How far does our moral circle extend?

The history of moral progress can be told as the expansion of the moral circle — the set of beings whose interests count morally. It has expanded from the tribe, to the nation, to all humans, and there is ongoing argument about whether it should extend further: to animals, to future generations, to ecosystems, possibly to artificial minds.

8.1 Peter Singer on Animal Liberation

Singer’s argument on animals is the paradigm case for the expansion arguments that follow: a boundary defended by an apparently obvious property (in this case, species membership) is shown to track that property only as a proxy for the morally relevant one (capacity to suffer), and the proxy is shown to license the wrong verdicts. The other three expansions (future people, ecosystems, AI) work the same shape with different proxies.

Peter Singer (Animal Liberation, 1975; Practical Ethics, 1979) made the philosophical case for extending the moral circle to animals.⁷⁷ His starting point is the utilitarian principle: suffering counts morally, regardless of whose it is. The capacity to suffer is the relevant criterion for moral consideration, not intelligence, rationality, or species membership.

Jeremy Bentham had already put it:

“The question is not, Can they reason? nor, Can they talk? but, Can they suffer?” — Jeremy Bentham, Introduction to the Principles of Morals and Legislation (1789), Footnote to Chapter XVII⁷⁸

Singer extends this: factory-farmed animals suffer; their suffering counts morally; and on his view we give it less weight than equivalent human suffering for reasons he argues amount to species prejudice — what Richard Ryder named speciesism, by analogy to racism.⁷⁹ Singer’s claim is not that any preference for one’s own species is unwarranted; he targets the standard reasons offered for the preference — rationality, language, social complexity — and argues they do not track the morally relevant property, capacity to suffer.

The argument does not require that animals have the same moral status as humans. It requires only that equal suffering deserves equal moral weight, and that current practice gives animal suffering far less weight than equivalent human suffering for reasons Singer argues will not stand up. Critics (Cohen, Scruton, Cora Diamond) reply that the analogy with racism mischaracterises species membership, that a being’s place in a “form of life” is itself morally relevant, and that the calculus Singer requires for cross-species comparison is more contested than he allows.⁸⁰

8.2 The Moral Status of Future Generations

Do people who don’t yet exist have moral claims on us? This question bears directly on climate change, nuclear waste disposal, and long-term environmental policy.

The standard consequentialist answer: future people will be able to suffer and flourish; their interests count; therefore present actions that harm them are wrong. On this view, failing to address climate change is a massive moral failure — since the harms fall on people who have no say in present decisions.

The non-identity problem (Derek Parfit, Reasons and Persons, 1984): here is a puzzle.⁸¹ The specific people who will exist in 2150 depend on every decision made between now and then (different choices lead to different people meeting and having different children). If we had pursued a different energy policy, the specific individuals alive in 2150 would not have existed. Those individuals cannot be harmed by our policy, because without that policy, they would not exist to be harmed. Different (perhaps better-off) people would exist instead.

Parfit calls this the non-identity problem: when the number or identity of future persons depends on our decisions, standard harm-based moral reasoning has strange implications.

8.3 Environmental Ethics

Do non-human organisms, species, or ecosystems have intrinsic moral value — independently of their usefulness to humans or other sentient creatures?

• Anthropocentrism: only humans have intrinsic value. Nature has instrumental value — insofar as it serves human interests.
• Sentientism (Singer): all sentient creatures have intrinsic value.
• Biocentrism (Paul Taylor, Respect for Nature, 1986): all living things have a “good of their own” that deserves moral consideration. A tree has interests (in flourishing) even without consciousness.⁸²
• Ecocentrism (Aldo Leopold, A Sand County Almanac, 1949): the “land ethic” — ecosystems and species have value that transcends the value of individual organisms. “A thing is right when it tends to preserve the integrity, stability, and beauty of the biotic community.”⁸³

These positions generate different prescriptions. An anthropocentrist and an ecocentrist may agree that the Amazon rainforest should be preserved, but for different reasons; in cases where human interests clash with ecosystem integrity, the reasons start to matter.

8.4 The Moral Status of Artificial Minds

If a sufficiently sophisticated AI system can suffer — or can have something functionally equivalent to preferences, desires, and aversions — does it have moral status?

The question is not purely futuristic; it forces us to clarify what grounds moral consideration:

• Sentience (capacity to suffer): on Singer’s view, a suffering AI has moral status.
• Rationality: on Kant’s view, only rational agents who can give and respond to reasons have moral status — and whether an AI can genuinely do this is contested.
• Relational accounts (Mary Anne Warren’s six criteria of personhood; Annette Baier on relational moral status): moral status is partly constituted by relationships and social recognition — not purely intrinsic properties.

Nick Bostrom and David Chalmers are among the philosophers who have argued the question is now a serious one rather than a futuristic speculation.⁸⁴

8.5 Questions

1. Singer’s argument against speciesism rests on the analogy with racism and sexism: we can’t privilege our own species without an argument any more than we can privilege our own race. Is this analogy good? Are there morally relevant differences between species membership and race?
2. Bentham’s criterion — “Can they suffer?” — includes many animals but excludes plants, ecosystems, and future persons (who don’t yet exist). Is suffering the right criterion for moral consideration?
3. Parfit’s non-identity problem seems to imply that we can’t harm future generations by our choices if different choices would have led to different people existing. Does this argument work? What are its limits?
4. Leopold’s “land ethic” says ecosystems have intrinsic value. But how do we weigh ecosystem integrity against the basic needs of human beings who depend on those ecosystems for survival?

9 Media

Novels, films, and artworks that illuminate the questions above:

• Fyodor Dostoevsky, The Brothers Karamazov (1880) — Ivan Karamazov’s “Rebellion” chapter poses one of the most powerful objections to utilitarian theodicy: can any ultimate harmony justify the suffering of even one innocent child? Indispensable for thinking about consequentialism and the limits of moral calculation.
• Ursula K. Le Guin, “The Ones Who Walk Away from Omelas” (1973) — A short story. A utopia is built on the suffering of one child. Its citizens know this. Some accept it; some walk away. A masterclass in consequentialist dilemmas, written by a literary artist rather than a philosopher.
• Kazuo Ishiguro, The Remains of the Day (1989) — A butler reflects on a life of perfect professional duty and wonders whether he made the right choices. An oblique, devastating meditation on virtue, integrity, and what is owed to those with moral authority over us.
• Michael Sandel, Justice: What’s the Right Thing to Do? (2009) — Not fiction, but based on Harvard’s most popular course. Accessible and philosophically honest survey of consequentialism, deontology, and virtue ethics using real cases. The companion video lectures are freely available.
• Cormac McCarthy, The Road (2006) — Father and son surviving a post-apocalyptic world. Ethics stripped to its bare minimum: what do we owe each other when all social structures have collapsed? Which moral intuitions survive, and which dissolve?
• Ridley Scott, Blade Runner (1982) (and Denis Villeneuve’s Blade Runner 2049, 2017) — The moral status of artificial beings; what grounds personhood; the ethics of creating conscious life for exploitation. Connects to the final lesson on extending the moral circle.

10 Bibliography

11 Notes