Comments
50
J.S. (2026). Resolution Ethics: Structural Foundations for Moral Reasoning. https://doi.org/10.17605/OSF.IO/XTPFS
Moral philosophy has a set of problems that no existing framework handles cleanly. Not because people have not tried. They have. For decades. Some for centuries.
I have been working on a framework called Resolution Ethics, or RE (J.S., 2026). Rather than explain it abstractly, I figured the most useful thing would be to just run it against three of the hardest problems on the books and let you see what happens.
Think of this as a stress test.
Derek Parfit is widely considered one of the most important moral philosophers of the twentieth century. In Reasons and Persons (1984) he dropped a bomb that the field still has not recovered from.
The setup is simple. Imagine a world of ten billion people living excellent lives. Call it World A. Now imagine a vastly larger world where every person's life is barely worth living. Muzak and potatoes, as Parfit put it. Call it World Z.
Total utilitarianism says World Z is better. More people times even a tiny positive value equals a bigger number. The math checks out. The conclusion is repugnant.
You would think someone could patch this. They have tried. Every patch breaks something else. Average utilitarianism avoids Z but says one ecstatic person outweighs billions of happy ones. Critical-level theories avoid Z but say adding happy people can make the world worse. Person-affecting views avoid Z but collapse under Parfit's Non-Identity Problem.
Then Gustaf Arrhenius (2000) proved impossibility theorems showing that no population axiology can satisfy a small set of reasonable conditions simultaneously. This is not "we have not found the answer yet." This is "the answer does not exist" within the framework space everyone has been searching.
Parfit spent his entire career looking for what he called Theory X. A satisfactory population ethics. He died without finding it.
Here is the thing though. The impossibility proof constrains systems that rank total world-states. It constrains frameworks that ask, "which world is better?"
RE does not ask that question.
RE operates on a dependency ordering I call PTF: Protection, then Trust, then Free Agency. These are not the everyday definitions of those words. They have structural precision within the framework. RE also identifies a specific failure mechanism called Self-Deception (SD). This is not lying to yourself in your head. SD is not a mental state at all. In RE, SD is action that is incoherent with your own existence and navigation of reality. It is detected through what you do, not what you think or feel. A torturer who says "I know this is wrong, I just don't care" has not escaped SD through self-awareness. The torture IS the SD. The act is what you evaluate, not the internal monologue.
So when someone proposes a policy that deliberately engineers billions of lives at bare subsistence, RE does not need to compare World A and World Z on some cosmic scoreboard. It looks at the action itself.
A policy whose design goal is maximal population at near-minimum welfare, when higher-welfare alternatives are feasible within the same resource envelope, is a full PTF domain violation. Protection is violated because the policy deliberately harms the conditions of life when better options exist. Trust is violated because the people in World Z are counting on policy makers to act in their genuine interest, not to sacrifice their wellbeing for a numerical abstraction. Free Agency is violated because subsistence-level existence, "muzak and potatoes," constrains people's capacity to navigate reality in any meaningful self-directed way.
The people pushing that policy know it degrades those conditions. They are deceiving themselves about what the PTF ordering actually requires, in order to chase an aggregate number.
RE flags the SD in the policy. It does not tell you which world is "better." It verifies that the reasoning behind the policy is incoherent with PTF structure. The impossibility proof does not apply because RE is not playing the game the proof constrains.
Parfit was looking for Theory X inside a room that was mathematically locked. RE is not in that room. If it had existed in his time, I think it would have at least caught his attention.
In 1976, Bernard Williams and Thomas Nagel independently published papers that exposed something uncomfortable at the foundation of moral judgment (Nagel, 1979; Williams, 1981).
The setup is dead simple. Two people get drunk at a bar. Both drive home. Both run a red light. One gets lucky and the road is empty. The other hits and kills a pedestrian.
Same choice. Same recklessness. Same moral failure. The only difference is whether a person happened to be in the crosswalk. Something entirely outside either driver's control.
We treat them radically differently. One gets a manslaughter charge and a destroyed life. The other gets a hangover and a close call story.
Every major framework has a core principle that says you should only be judged for things within your control. Kant was explicit about this. Your moral worth comes from your will, not from dice rolls. Yet our actual practices violate this constantly.
Nagel showed how deep the problem goes. It is not just outcomes. Your circumstances are luck. Whether you were ever tested is luck. Your character traits are luck, shaped by genetics and upbringing you did not choose. If you follow the control principle consistently, you keep peeling layers away until there is nothing left to judge.
Fifty years later, philosophers are still stuck. Accept moral luck and morality becomes partly arbitrary. Reject it and you are saying something almost nobody actually believes.
This is where SD earns its keep as a concept.
Both drivers performed the same act of SD. They knew alcohol impairs driving. They knew people use crosswalks. They convinced themselves the risk was acceptable. Both corrupted their own justification for getting behind the wheel: "I can handle it," "it will be fine," "the odds are low." That corruption is identical in both cases regardless of what happens at the intersection.
And both drivers, by choosing to drive drunk, already violated two PTF domains through that single choice. Trust was violated because every other person on the road reasonably expects other drivers to operate responsibly and not drive intoxicated. Free Agency was violated because imposing your impaired driving on others compromises their capacity to navigate safely, a choice they never consented to.
That is what drunk driving IS in RE. A two-domain violation. Both drivers share it equally.
The difference is what happens at the intersection. The driver who killed the pedestrian manifested a third domain violation. Protection was breached through fatal harm. A moral entity's existence was terminated. That is the condition that generates more Moral Entropy, not a different moral failure, but the same SD cascading into a full three-domain violation when it collides with reality.
That Moral Entropy is not abstract cosmic damage. In RE, it is stored in actual relationships. The destroyed relationship with the victim. The shattered trust in the victim's family. The degraded confidence in the community that roads are safe. The other driver's SD generated the same two-domain violation, but the third domain never manifested. No relationship was destroyed. No life was terminated.
This is where RE does something no other framework manages. It validates the moral intuition people already have, that the driver who killed someone carries more weight, while simultaneously showing that the SD is identical in both cases. The moral failure is the same. The Moral Entropy is different. RE explains why your gut was right and gives you the structural reason behind it.
The pedestrian's presence changes the legal outcome, and it changes the Moral Entropy. It does not change the SD. Both drivers made the same moral failure. One got unlucky enough to see it manifest fully.
The reason moral luck has been a stalemate is that consequentialism evaluates outcomes, deontology evaluates intentions, and virtue ethics evaluates character. Outcomes are contaminated by luck. Intentions are hard to verify. Character is shaped by forces you did not choose.
SD is none of those things. It is not outcome-dependent. It is not hidden behind intention. It is not genetically determined. It tracks moral failure without being infected by luck.
That is what RE is looking at when it processes this problem.
Peter Singer's 1972 paper "Famine, Affluence, and Morality" is one of the most cited in modern philosophy, and probably the most disobeyed.
Here is the argument. You walk past a shallow pond and see a child drowning. You can save them by wading in. It will ruin your expensive clothes. Should you do it?
Everyone says yes. Obviously.
Singer then makes the move. Distance is morally irrelevant. If you can prevent suffering without sacrificing something of comparable moral importance, you are obligated to do it. The same principle that says save the drowning child says donate most of your income to prevent equivalent suffering overseas.
The logic is valid. The premises are hard to reject. The conclusion follows. And almost nobody does it.
Fifty years of objections have not found a clean flaw. Scheffler (1982) proposed that people are allowed to weigh their own interests more heavily. Murphy (2000) proposed each person should only do their fair share. Wolf (1982) pointed out that a fully devoted moral saint would live an impoverished life, never enjoying anything for its own sake. These responses feel reasonable but none of them actually identifies where Singer's logic breaks.
Here is what happens when you run it through RE.
The child in front of you is a moral entity in danger of fatal harm that you can directly influence. You are aware of the immediate danger. The most coherent permutation available to you in that scenario is saving them. Doing anything else while knowing what you know is SD. Protection is clear and unambiguous.
Now for distant suffering. Singer says the same obligation scales up to everyone, everywhere. RE says something more precise. Your Trust-level obligation operates in concentric circles of moral responsibility. You have an inner circle, the people and communities where your moral responsibility is direct and immediate. And you have outer circles that extend outward from there.
But these circles have a dual structure. Moral responsibility tracks what you owe inward, stronger obligations to those closer. Accountability tracks what your circles owe outward. You cannot excuse harm to outsiders simply because it benefits your inner circle. The concentric circles establish graduated obligation, not moral license for in-group favoritism.
Resources are finite. You must allocate them to your inner circle before extending to the outer ones. This is not moral indifference toward distant suffering. It is the structural reality of being a finite agent in a world that demands prioritization.
Here is the move that matters. If your resources were infinite, RE says you absolutely should help everyone in that Trust circle, including the entire human race. The obligation is real. The constraint is not moral, it is material. Singer treats the constraint as a moral failure. RE identifies it as a resource allocation problem within a real structural ordering.
This does not let anyone off the hook. You still have genuine obligations at the Trust level, and those obligations grow as your capacity grows. But the drowning child in front of you and suffering you can only reach through institutional coordination are different layers of the hierarchy, not the same obligation scaled up.
Singer's argument works by flattening everything into a single utilitarian calculus where all suffering is fungible and all obligations are identical. RE maintains a structural ordering that Singer's framing does not account for.
The reason Singer's argument has survived fifty years without a clean refutation is that every counter-argument tries to fight within his utilitarian frame. RE operates outside that frame.
Each of these problems has been treated as unsolvable for decades. The Repugnant Conclusion has a formal impossibility proof behind it. Moral luck has had Nagel himself say it has no solution. Singer's drowning child has resisted refutation for half a century.
What I notice is that they all share a pattern. They break frameworks that evaluate outcomes, or intentions, or character, or aggregate world-states. RE evaluates something different: whether an action is coherent with the PTF structural ordering, whether SD is present in the reasoning, and what Moral Entropy actually manifested into the relationships involved.
There is a deeper pattern here too. Parfit believed that Kantians, consequentialists, and contractualists were climbing the same mountain from different sides. He spent 1,400 pages in On What Matters trying to demonstrate convergence. RE suggests he was right. Kantians were tracking Free Agency. Consequentialists were tracking Protection. Contractualists were tracking Trust. They were all seeing the same structure. They differed on which domain to prioritize when conflicts arose. P > T > F is the dependency ordering they were missing.
RE does not claim to have solved these problems. It does not prescribe. It verifies. It catches incoherent reasoning. Among coherent options, you still must choose. But when a framework can engage three of the hardest problems in moral philosophy without breaking, that is at least interesting enough to warrant a closer look.
Arrhenius, G. (2000). An Impossibility Theorem for Welfarist Axiologies. Economics and Philosophy, 16(2), 247-266.
J.S. (2026). Resolution Ethics: Structural Foundations for Moral Reasoning. https://doi.org/10.17605/OSF.IO/XTPFS
Murphy, L.B. (2000). Moral Demands in Nonideal Theory. Oxford University Press.
Nagel, T. (1979). Moral Luck. In Mortal Questions (pp. 24-38). Cambridge University Press.
Parfit, D. (1984). Reasons and Persons. Oxford University Press.
Parfit, D. (2011). On What Matters. Oxford University Press.
Scheffler, S. (1982). The Rejection of Consequentialism. Oxford: Clarendon Press.
Singer, P. (1972). Famine, Affluence, and Morality. Philosophy and Public Affairs, 1(3), 229-243.
Williams, B. (1981). Moral Luck. In Moral Luck: Philosophical Papers 1973-1980 (pp. 20-39). Cambridge University Press.
Wolf, S. (1982). Moral Saints. The Journal of Philosophy, 79(8), 419-439.
I don't really understand what you mean by protection, but there seems to be something fundamentally wrong in the way you both equate and differentiate it with consequences. You claim that consequentialism tracks this “protection,” but it is clearly something different. If it weren't, your framework would lead to the same conclusions as consequentialism on all three of these problems. This is because you put the highest priority on protection out of all your proposed principles.
Resolution ethics doesn't really have any more explanatory power than consequentialism for our moral intuitions in the case of moral luck. Any sensible version of consequences clearly recognizes that the actual results are more important than potential consequences while also recognizing that the act itself is equally wrong or right in both cases. Also, these violations of the principle of protection still seem to be based completely on luck regardless of terminology.
What bothers me the most is that you seem to completely dismiss the possibility that the so-called “repugnant conclusion” is actually acceptable. This also applies to the case of the drowning child. I don't find it self-evident that these conclusions you are so determined to resist are actually bad.
I think we are looking at this through genuinely different lenses, and I appreciate you engaging.
Protection in RE is not a standalone principle and it is not a synonym for welfare maximization in consequentialism. It is part of an inseparable structural ordering: P > T > F. You cannot pull Protection out of that structure and compare it to consequentialism in isolation, because it only functions as part of the whole. A forum post is a thin slice of the framework, so I completely understand the blurriness. I would genuinely love to know whether you have had a chance to read the RE paper itself, because I think several of your concerns dissolve once the definitions are on their own terms rather than mapped onto familiar ones.
On moral luck, you are actually closer to RE's position than you might think. You said the act is equally wrong in both cases while the actual results carry more weight. RE agrees, and it gives you the structural reason why both things are true simultaneously: the Self-Deception is identical in both drivers, and the Moral Entropy is different. That is not a repackaging, it is a mechanism.
On the Repugnant Conclusion, you are absolutely right that some serious philosophers accept it. That is a legitimate position. My argument is not that it feels bad. It is that the reasoning process that generates it contains a specific kind of incoherence when run through PTF structure. If you want to defend accepting it, I would be genuinely curious what you make of that argument specifically.
No final word claimed here. Just something I think is worth a closer look.