Argument

Well-Being Cannot Ground Morality Defeater

Intro

"Morality is about the well-being of conscious creatures. Good is whatever makes conscious beings flourish, bad is whatever makes them suffer, and science can tell us which is which." That is the thesis of Sam Harris's The Moral Landscape (2010), and it has become the standard answer for atheists who want to say morality is real without saying God exists.

Harris deserves credit for one thing. He says some moral claims are objectively true. He is not a relativist. He thinks a culture that throws acid in girls' faces is wrong, not just different. Christians agree.

The problem is what he says makes things right and wrong. If right simply means "produces the most well-being," then the theory has to approve whatever produces the most well-being. And it is easy to build cases where the most well-being comes from doing something everyone knows is evil: killing one patient to save five, framing an innocent man to calm a mob, drugging a whole society into contented ignorance.

This page walks through ten of those cases. Each one is a test. If the well-being standard gives the wrong answer, then well-being is not what makes things right. It might be part of what is good, and Christians gladly say human flourishing matters. But it cannot be the foundation.

Then it names the deeper problems: whose well-being, measured how; why the individual gets sacrificed to the total; how you get from "this is what makes people happy" to "you must do it"; and why calling well-being "good" does not prove it is.

The Christian answer is not "well-being does not matter." It is that well-being matters because people matter, and people matter because they bear the image of God. That grounds the individual's worth in a way no sum can override.

◎ See this page's graph and discover linked topics →

Cheatsheet

The 30-second reply:

If right just means "maximizes well-being," then it is right to kill one healthy patient for organs that save five, right to hang an innocent man if it stops a riot, and right to keep a society happily drugged. You know those are wrong. So something other than well-being is deciding what is wrong, and that something is the real foundation. Science can tell you what hurts people. It cannot tell you that you must not hurt them.

The fast facts:

1. Harris is a realist. He agrees morality is objective. The dispute is only about its ground. Credit the common ground.
2. Ten counterexamples. Organ harvest, scapegoat, experience machine, drugged happy society, Omelas, sadistic majority, undetected violation, painless killing, broken deathbed promise, eliminating low well-being lives. In each, well-being approves what conscience condemns.
3. Adding up sacrifices individuals. Any standard that maximizes a total will spend a person when the total goes up.
4. The is-ought gap. "This lowers well-being" is a description. "You must not do it" is a command. The first does not produce the second (David Hume).
5. Harris denies free will. In Free Will (2012) he argues no one could have done otherwise. That sits badly with blaming anyone for anything.

The 3 strongest counter-moves:

• "Is it right to harvest one patient to save five?" Ask and wait. If no, well-being is not the standard. If yes, the audience has heard enough.
• "Who says we ought to maximize well-being?" Make them ground the first premise of their system. It cannot come from science.
• "If you add rights and justice, what grounds those?" The hybrid concedes that well-being is not the foundation.

Reciprocal concessions (grant the small point, then collect a bigger one):

• Grant: well-being matters, and science really can tell us a great deal about what helps and harms people. Now collect: then science is a tool for applying morality, not its foundation, and the question of what makes harm wrong is still open.
• Grant: most of the time, doing right also increases well-being. Now collect: then the cases where they come apart are exactly where we learn which one is the standard, and in every such case we side with right over the total.
• Grant: Harris is right that some moral claims are objectively true. Now collect: then he owes a ground that does not approve murder for organs, and a personal, good God whose image every person bears is the ground that fits.

The closing line:

Well-being matters because people matter. People matter because they bear the image of God. Take God away and you are left with a total, and a total will always be willing to spend you.

In full

A defensive defeater against welfarist moral realism, the view that moral facts are identical to, or wholly grounded in, facts about the well-being of conscious creatures, so that the right act is the one that best promotes aggregate well-being and moral questions are in principle empirical. Its most prominent contemporary statement is Sam Harris, The Moral Landscape: How Science Can Determine Human Values (2010). The view is a form of naturalist cognitivist realism, option (b) in the Grounding Argument from Cognitivism.

The defeat has three parts. First, a reductio by counterexample: ten cases in which maximizing well-being yields verdicts that are clearly wrong, showing that well-being is not extensionally equivalent to rightness. Second, five root problems that explain why the cases arise: indeterminacy of the metric, aggregation's indifference to persons, the is-ought gap, the open-question problem, and blindness to non-welfare goods (truth, faithfulness, justice). Third, a positive alternative: goodness grounded in God's character, and human dignity grounded in the image of God (Genesis 9:6), which secures the individual against aggregation. A coda notes the tension between Harris's moral realism and his hard determinism.

This page complements Harm-Reduction Cannot Ground Morality (Defeater), which attacks harm-based ethics on structural grounds (defining harm, Arrow's theorem on aggregation, symmetry-breaking). This page is built around concrete test cases and Harris's specific thesis.

Argument structure

Form

Reductio by counterexample (P1 to P3): a theory that identifies rightness with a property is refuted by one clear case where the property and rightness come apart; ten cases make the result robust. Diagnostic (P4): the root problems explain the pattern, so patching individual cases will not help. Concession trap (P5): the standard repair abandons the thesis. The positive case for theism is made in Grounding Argument from Cognitivism and Moral Argument; this page removes one secular alternative.

Common ground first

Before attacking the thesis, grant what is true in it.

• Harris is a cognitivist realist. He holds that moral claims can be objectively true or false and that relativism is wrong. That is P1 of the Grounding Argument from Cognitivism. Christians share it.
• Well-being is a real good. Scripture cares about bodies, health, food, and joy. Jesus healed and fed people. Nothing here says suffering does not matter.
• Science helps. Medicine, psychology, and economics tell us a great deal about what helps and harms.

The dispute is narrow: is well-being what makes acts right? The cases below say no.

P1 to P3, The ten failure scenarios

For each case: the setup, what the well-being standard says, and why that verdict is wrong. In every case you can stipulate secrecy, so the standard reply ("it would lower well-being in the long run") is blocked.

1. The organ harvest

• Setup. A surgeon has five patients dying for lack of organs: heart, liver, two kidneys, lungs. A healthy young man comes in for a checkup. He matches all five. The transplant case is made famous by Judith Jarvis Thomson in the literature on the trolley problem.
• Well-being verdict. One death, five lives saved. Kill him quietly and harvest.
• Why it fails. Everyone knows this is murder. The young man has a claim not to be used as spare parts that no total can override. The verdict shows the standard has no room for the individual as such.

2. The scapegoat sheriff

• Setup. After a crime, a mob is about to riot and many will die. The sheriff can stop it by framing and hanging an innocent drifter. No one will ever find out. The case is a staple of the utilitarianism literature.
• Well-being verdict. Hang the innocent man. One death prevents many.
• Why it fails. Punishing the innocent is injustice, whatever it prevents. Caiaphas reasoned exactly this way about Jesus: "it is expedient for you that one man should die for the people, and that the whole nation perish not" (John 11:50, ASV). Scripture records it as the logic of the cross's enemies.

3. The experience machine

• Setup. Robert Nozick (Anarchy, State, and Utopia, 1974) imagines a machine that gives you any experiences you want: success, love, achievement, all simulated, while you float in a tank. You will not know it is a simulation.
• Well-being verdict. If well-being is how life feels from the inside, plug in. Plug everyone in.
• Why it fails. Almost no one wants to plug in, and we would think it a tragedy if a loved one did. We want to actually do things, actually be loved, and actually know the truth. Nozick's conclusion is that something matters to us beyond how our lives feel. Well-being as experience leaves out reality.

4. The drugged happy society

• Setup. Aldous Huxley's Brave New World (1932): a society engineered for contentment. Citizens are conditioned from birth, sorted into castes, kept busy with pleasure, and given a drug, soma, whenever unhappiness threatens. Almost everyone reports being happy.
• Well-being verdict. If measured well-being is high and stable, this is a moral success.
• Why it fails. Huxley wrote it as a warning, and readers take it as one. People are kept from truth, love, art, and God. They are managed, not respected. A world can score high on contentment and be a moral horror.

5. Omelas

• Setup. Ursula K. Le Guin's story "The Ones Who Walk Away from Omelas" (1973): a city of joy and beauty whose happiness depends on one child kept in misery in a dark cellar. Every citizen learns of the child. Most stay. Some walk away.
• Well-being verdict. One child's suffering against a whole city's flourishing. The trade is a bargain.
• Why it fails. The story's power comes from the reader's certainty that the ones who walk away are right. The child's suffering is not offset by anyone's happiness. Jesus' warning about causing one of "these little ones" to stumble (Matthew 18:6) and the shepherd who leaves the ninety-nine for the one (Matthew 18:12-14) run directly against aggregate logic.

6. The sadistic majority

• Setup. A large majority gains real, intense pleasure from tormenting a small minority. The pleasure of the many outweighs the pain of the few by any measure.
• Well-being verdict. The torment is justified, because total well-being rises.
• Why it fails. Pleasure taken in cruelty does not count in favor of cruelty. It makes it worse. The standard cannot see the difference between good and evil pleasures without a prior moral judgment about which pleasures are fitting, and that judgment is not well-being.

7. The undetected violation

• Setup. A man secretly films a woman in her home without her knowledge. She never finds out. The footage is never shared. He enjoys it.
• Well-being verdict. Her well-being is untouched (she never knows), his rises. On a strict well-being view, nothing wrong happened, or something good did.
• Why it fails. She has been wronged even though she never feels it. Some wrongs are violations of what a person is owed, not of how she feels. The well-being standard cannot see wrongs that are never experienced.

8. The painless killing of a friendless man

• Setup. A man with no family, no friends, and no one depending on him is killed painlessly in his sleep. No one grieves. No one notices. His life, while pleasant, was not especially full.
• Well-being verdict. Little lost. If his property goes to someone who will enjoy it more, the act may even be positive.
• Why it fails. Murder is not wrong because of the grief it causes others or because the victim suffers. It is wrong because of who the victim is. Genesis 9:6 gives the reason: "in the image of God made he man." The friendless man bears the image as fully as anyone.

9. The broken deathbed promise

• Setup. A dying friend asks you to promise that his savings will go to his estranged son. You promise. He dies. No one else knows. You could give the money to a charity that would produce far more well-being.
• Well-being verdict. Break the promise. The charity does more good.
• Why it fails. A promise creates an obligation that does not dissolve because breaking it would pay better. Scripture praises the one who "sweareth to his own hurt, and changeth not" (Psalm 15:4, ASV). Faithfulness is a good the well-being standard cannot register.

10. Eliminating low well-being lives

• Setup. If average or total well-being is the measure, then lives that score low (the severely disabled, the chronically depressed, the very old) drag the numbers down. Prevent them, or end them, and the figures improve.
• Well-being verdict. In principle, approve it. This is the logic of eugenics.
• Why it fails. The twentieth century showed where this road goes. The worth of a person does not rise and fall with their score. The Christian claim that every human, at every stage and ability, bears the image of God is what stands between the vulnerable and the calculator. See Imago Dei.

What the ten cases show

• In each case, the act that maximizes well-being is wrong.
• In each case, what makes it wrong is something the well-being standard cannot see: the individual's claim not to be used, justice to the innocent, contact with reality, truth, the special weight of the weak, the difference between good and evil pleasures, being wronged without being harmed, intrinsic worth, faithfulness, and dignity independent of score.
• Therefore well-being is not the ground of morality. At best it is one good among several, and the others need a ground of their own.

P4, The root problems

The ten cases are symptoms. These are the causes.

1. Whose well-being, and measured how?

• Harris speaks of "conscious creatures." Do animals count equally with humans? Insects? Future people who do not exist yet? Each answer is a moral judgment that the standard needs before it can be applied, so the standard cannot be the source of it.
• Well-being itself has rival definitions: pleasure, desire-satisfaction, or an objective list of goods. Choosing among them is a moral choice. The Brave New World case shows how much depends on it.
• There is no common unit for comparing one person's well-being with another's. The structural treatment, including Arrow's theorem, is in Harm-Reduction Cannot Ground Morality (Defeater).

2. Aggregation sacrifices individuals

• Any standard that maximizes a sum will trade a person for a larger sum. That is not a bug in the cases above; it is the definition at work.
• The objection, in a line: maximizing views do not take seriously the distinction between persons. Each person has one life, and it is not a resource for the total.
• Paul refuses the logic outright: "Let us do evil, that good may come? whose condemnation is just" (Romans 3:8, ASV).

3. The is-ought gap

• "Act X increases well-being" is a description of what is. "You ought to do X" is a prescription. No set of descriptions entails a prescription without a moral premise added (David Hume; see Is-Ought Fallacy).
• Harris's needed premise is "we ought to maximize well-being." Science cannot supply it, because science measures what is.
• Frank Turek puts it plainly in his response to Harris:

science can't tell you whether or not you ought to hurt someone. (Frank Turek, Stealing from God, PDF p. 110)

4. Definition is not justification

• Harris sometimes answers the is-ought objection by saying that "good" simply means what promotes well-being, so asking why we should promote well-being is confused.
• But defining a word does not make anything true about the world. One can always ask, "this maximizes well-being, but is it right?" and the ten cases show the question is not empty. That is G. E. Moore's open-question test.
• Knowing what helps people is epistemology. What makes helping them obligatory is ontology. Turek states the distinction:

The question is not what method should we use to discover what is moral, but what actually makes something moral? (Frank Turek, Stealing from God, PDF p. 110)

5. It ignores truth, faithfulness, and justice

• The experience machine and Brave New World show truth matters for its own sake.
• The deathbed promise shows faithfulness matters for its own sake.
• The scapegoat and the organ harvest show justice matters for its own sake.
• "Thou shalt not bear false witness" (Exodus 20:16) is not a well-being rule. It holds when a lie would make everyone happier.

P5, The pushback: "Just add rights and justice"

The most common reply from a thoughtful opponent: "Fine, raw well-being has problems. So add rights. People have a right not to be killed for organs, a right to a fair trial, a right to privacy. Well-being operates inside those limits."

Why this concedes the point

1. If rights can override well-being, well-being is not the standard. Whatever overrides it is the higher standard. The opponent has moved to a different theory.
2. The rights now need a ground. Why does the friendless man have a right to life that outweighs the gain from his death? Not because of well-being, since by hypothesis well-being favors killing him. So what grounds the right? That is the original question, now asked about rights.
3. "Rights maximize well-being in the long run" collapses back. If rights are justified only because they usually pay off, then in the case where breaking one pays off (secretly), the right should give way. That returns all ten counterexamples.
4. The Christian answer grounds rights directly. A person has a claim not to be killed because he bears the image of God (Genesis 9:6). That claim is not a forecast about totals. It holds in the dark, with no one watching.

Failure mode named: a hybrid in which the load-bearing part is not well-being, presented as a defense of well-being.

Harris the determinist

Harris also holds, in Free Will (2012), that free will is an illusion: every choice is the product of prior causes the chooser did not choose, so no one could ever have done otherwise. See Hard Determinism and Free Will and Determinism.

This creates a tension worth raising.

1. "Ought" implies "can." It makes no sense to say a man ought to have done what he could not have done. If no one could ever have done otherwise, then no one ever ought to have done otherwise.
2. Moral blame loses its object. Harris accepts that retributive blame makes no sense on his view. But the realism of The Moral Landscape says some acts are objectively wrong. Wrong for whom, if the agent had no alternative? A landslide that kills a village is a disaster, not a wrongdoer.
3. Moral reasoning loses its point. If my conclusions are fixed by prior causes, then my judgment that Harris is right and my judgment that he is wrong are equally the output of causes. Neither was reached by weighing reasons freely.
4. Harris's reply is that choices still matter as causes, and we can still restrain dangerous people and shape behavior. Grant it. That is a theory of management, not of moral responsibility. It treats people the way we treat weather.

Tactical use: do not lead with this. Raise it after the counterexamples, as a second front: "You say some things are objectively wrong. You also say no one could ever have done otherwise. How can a person do wrong if he could not have done right?" See Libertarian Free Will for the Christian alternative.

The Christian alternative

1. Goodness is grounded in God's character. God is not one more being who happens to be good; He is the standard of goodness. What is good is what accords with His nature. This is the conclusion of the Grounding Argument from Cognitivism, and it avoids the Euthyphro dilemma (see Euthyphro Dilemma Objection Defeater).
2. Human dignity is grounded in the image of God. Genesis 9:6 (quoted below) gives the reason murder is a capital wrong: not the grief it causes, not the loss to the total, but what the victim is. That one sentence answers cases 1, 2, 8, and 10.
3. The individual cannot be spent. Because worth is grounded in the image, not in a score, it does not shrink when someone's well-being is low or grow when it is high. The shepherd leaves the ninety-nine for the one.
4. Well-being still matters, for the right reason. Christians care about health, food, and joy because the people who have them matter. Well-being is downstream of dignity, not the source of it. "Thou shalt love thy neighbor as thyself" (Leviticus 19:18, ASV) commands care for his good, but it never licenses using him for someone else's.
5. Truth, faithfulness, and justice are goods in their own right, because they reflect the character of a God who is true, faithful, and just. That is why they hold even when breaking them would pay.

Master objections

1. "Your counterexamples are artificial thought experiments." Reply: thought experiments test whether a definition is correct. If "right" means "maximizes well-being," the definition must hold in every possible case, including unusual ones. And several are not artificial: eugenics programs, show trials, and the sacrifice of the few for the many have all happened.
2. "In the real world those acts would lower well-being." Reply: stipulate secrecy and permanence. The wrong remains. So the wrongness never depended on the side effects.
3. "Harris is talking about well-being in a rich sense that includes justice, truth, and love." Reply: then well-being has been redefined to include moral goods, and the theory is circular. "Right is what promotes the good, including justice" assumes justice is already known to be good. The question was what grounds it.
4. "Christianity also cares about consequences." Reply: yes. Christian ethics weighs consequences within limits set by commands and by the dignity of persons. That is exactly the structure this page defends: consequences matter, but they are not the foundation.
5. "Without God you still have empathy, and that is enough." Reply: empathy is a feeling, and feelings vary and fail. The question is whether we ought to act on empathy when we do not feel it. See Atheism Cannot Justify Compassion.

Tactical opening / closing

Opening line: "I want to start by agreeing with you. You think some things are objectively wrong, not just disliked. So do I. My question is what makes them wrong. You say well-being. Can I test that with one case?"

Force-commit move: "A surgeon can save five patients by killing one healthy visitor for his organs. No one will ever know. Does that maximize well-being? Is it right?"

Closing landing strip: "Every time well-being and right came apart, you sided with right. That tells us which one is the real standard. The man on the operating table has a claim no total can override, and Genesis tells us why: he is made in the image of God."

Live-cite kit

• Scripture: Genesis 9:6 (image of God as the reason murder is wrong); Genesis 1:27 (every human made in the image); Romans 3:8 (do not do evil that good may come); John 11:50 (Caiaphas's scapegoat reasoning); Matthew 18:6 and Matthew 18:12-14 (the little ones, the one sheep); Exodus 20:16 (truthfulness); Psalm 15:4 (keeping an oath to one's own hurt); Leviticus 19:15 (no partiality in judgment); Proverbs 17:15 (condemning the righteous is an abomination).
• Scholarly: Sam Harris, The Moral Landscape (2010) and Free Will (2012); Robert Nozick, Anarchy, State, and Utopia (1974); Judith Jarvis Thomson on the transplant case; Aldous Huxley, Brave New World (1932); Ursula K. Le Guin, "The Ones Who Walk Away from Omelas" (1973); David Hume, Treatise III.1.1; G. E. Moore, Principia Ethica (1903); Frank Turek, Stealing from God (2014); William Lane Craig vs Sam Harris (Notre Dame, 2011).
• Aphorisms:
  • "Science can tell you what hurts. It cannot tell you that you must not hurt."
  • "A total will always be willing to spend you."
  • "Every time well-being and right came apart, you chose right. That is your real standard."

Tactical notes

• One case, not ten. In live debate, pick the organ harvest or the scapegoat and stay on it. Ask, wait, and let the answer land. Keep the others in reserve for when the opponent patches the first.
• Always stipulate secrecy. The reflex reply is "it would cause fear and distrust." Close it before it opens: "No one will ever know."
• Credit the realism. Harris is closer to the Christian than the relativist is. Say so. It narrows the dispute to grounding.
• Do not deny that well-being matters. That hands the opponent an easy point. Say it matters because people matter.
• Watch for redefinition. If "well-being" starts to include justice, truth, and faithfulness, point out that the theory now assumes the moral goods it was meant to explain.

Connection to Scripture

Whoso sheddeth man's blood, by man shall his blood be shed: For in the image of God made he man. (Genesis 9:6, ASV)

• Genesis 9.6, the image of God as the ground of the prohibition on murder, given to all humanity through Noah
• Genesis 1.27, every human made in God's image
• Romans 3.8, Paul rejects doing evil that good may come
• John 11:50, the high priest's expedient sacrifice of one for the many
• Matthew 18.6, the weight of harming one little one
• Exodus 20.16, truthfulness as a command, not a calculation
• Proverbs 17.15, condemning the righteous is an abomination, whatever it achieves
• Leviticus 19.15, justice without partiality

See also

• Grounding Argument from Cognitivism, the positive argument this defeater supports; well-being is option (b) on its menu
• Harm-Reduction Cannot Ground Morality (Defeater), sibling defeater on the structural failures of harm-based ethics
• Atheist Moral Realism Defeater, sibling defeater against brute moral facts
• Atheism Cannot Justify Compassion, why empathy without a ground does not bind
• Moral Argument, the classic positive case
• Euthyphro Dilemma Objection Defeater, the reply to the usual counterattack
• Sam Harris, the author of the view this page answers
• David Hume, the is-ought gap
• Is-Ought Fallacy, the gap stated as a fallacy
• Hard Determinism, Harris's view of free will
• Free Will and Determinism, the wider debate
• Imago Dei, the ground of human dignity
• Subjective Morality Defeater, against the relativist alternative
• Arguments, master index