TL;DR
“You can’t derive an ought from an is” is the last move left once every positive case for the state has been answered. It claims that libertarian ethics is only a preference. This is Hume’s guillotine: the is-ought gap, used to cut away someone’s ethics. It gets there through one premise: an ought has standing only if facts alone entail it. Stated that way, the premise is itself an ought no fact entails, so it fails its own test. Softened to “if you want true beliefs, hold nothing underived,” it no longer charges any ought with being underived, and the dispute moves to which ends and means. Either way it condemns the statist’s oughts as much as the libertarian’s, and held consistently it isn’t a position at all, it’s leaving the conversation.
- The test fails itself. The skeptic needs an ought, “hold nothing underived,” that his own rule says no fact can produce.
- One standard, one side. The rule hits every ought equally, so swung at only one side it’s a different logic for each.
- The line in the sand. Held all the way down, it’s a refusal to argue, and a refusal isn’t answered, only declined.
What it claims
Follow any case for intervention downward, one question at a time (The Descent). Nobody should starve, so we need a safety net. Who pays? Everyone, by tax. And if someone declines? He’s made to. On whose say-so? The majority’s. Each answer is a place where the other side can say “fine, you’re right” and stop. Refuse every one and the conversation reaches the bottom of the descent, and the last move left is the guillotine. It never opens the conversation. It appears only after the positive case has been answered.
At full strength, it says three things:
The gap. No set of purely descriptive premises entails a moral conclusion. Facts about self-ownership, scarcity and human action, however well established, never entail “aggression is wrong.”
The unproved ought. So the libertarian’s ethics rests on a premise that is a commitment, not a finding. His claim to have proven the NAP goes further than logic allows.
Only a preference. So “taxation is theft” and “taxation is justified” are the same kind of statement: two commitments with no fact between them. The careful version adds: I’m not telling you what to believe. I’m noting that you can’t call the statist wrong in a way that’s more than a difference in values. He wants the libertarian to drop the verdict “wrong” and treat the dispute as a clash of tastes.
The one premise under all three
The gap is real, and the first claim is true. The second and third don’t follow from it. They follow from a bridge:
An ought that facts alone don’t entail has no standing.
This is the categorical bridge, the version the essay refutes. A fact about validity says what follows from what. It doesn’t say what you should hold. “Unjustified,” “just a preference” and “drop it” all come from the bridge, not from the gap.
The bridge fails in three places. The rest takes them in turn.
1. The test fails itself
Grant the gap. Hume’s own point was modest: writers slide from “is” to “ought” without noticing, and the step should be observed and explained. The modern version, Hume’s Law, is sharper: no argument from purely descriptive premises validly reaches a moral conclusion. The mistake libertarians make is to attack it with “but that’s a norm too!” The Humean has an easy reply: I didn’t say you ought not. I said you cannot, validly. And he’s right. “Cannot” here doesn’t mean physically unable. It means the result won’t be valid, and “that inference is invalid” is a fact, just like “no triangle has four sides.”
That concession is where the argument turns. Take the skeptic’s claim as what it is, a report of a fact, and follow it:
- The gap is a descriptive claim. It says how inference works.
- By its own rule, a descriptive claim alone can’t entail a moral conclusion.
- So the gap alone can’t entail “your ethics is unjustified, so abandon it.”
- To get there he needs a second premise, the bridge named above: one ought not hold what isn’t validly derived.
- That premise is an ought. By his own rule, no set of facts produces it.
So the objection to underived oughts rests on an underived ought. Now ask whether that norm binds:
If it binds, then some oughts bind everyone without any fact producing them. “Your ethics rests on an ought that wasn’t derived” stops being an objection, because his rests on one too. The question becomes which oughts bind, and arguments settle that, not pointing at the gap.
If it doesn’t bind, then invalidity has no force against anyone. He can’t say you’re wrong to believe anything: not the NAP, not a flat earth. “You can’t derive an ought from an is” shrinks to “I don’t feel like accepting your conclusion,” a statement about his mood.
Either way the guillotine can’t be swung at libertarian ethics. Holding it up takes an ought he can’t account for. Dropping it disarms him.
“But the norm is only conditional: if you want true beliefs, hold nothing underived. Hume allowed oughts that follow from a fact plus an end.” Granted, and that is a different premise. In the instrumental form an ought follows from a fact plus an end, so the charge that an ought is “underived” is dropped, and what remains is a dispute about which ends and which means. The norm then binds whoever has the end of reaching truth by argument, and that is the skeptic when he objects and the statist when he answers. It does not bind a man who doesn’t care, and that man isn’t objecting to anything. The same route is open to the libertarian: anyone who argues has the end of settling a disagreement by reasons rather than force. If Argumentation Ethics succeeds, non-aggression is presupposed by that end too. That is the positive case, which this essay doesn’t rest on.
“I never said underived oughts can’t bind. I said you can’t claim to know yours.” That’s a standard of knowing: claim only what’s validly derived. It is itself a claim about what one should hold, and by its own terms he doesn’t know it. It’s the same bridge, and it fails the same way.
“So you’ve proven the NAP?” No. This clears the ground; it doesn’t build on it. The positive case comes from Argumentation Ethics, and it works the same way the skeptic’s own norm does: it takes what is already in force the moment two people settle a disagreement by reasons instead of fists, and shows that non-aggression is part of it. The epistemic norm and the NAP are siblings. He can’t keep one and throw away the other. He accepted both when he opened his mouth.
Bottoms out at: Logic, where the test, applied to itself, defeats itself.
2. One standard, one side
A man says taxation is theft is just your preference. The same man says billionaires ought to pay their fair share, children ought to have healthcare, nobody ought to go hungry. He doesn’t wave off “you shouldn’t torture children” as a mere preference with no rational standing. The test he used on one ought has no subject matter built into it. It condemns every ought exactly the same way, or none.
So a man who uses it only on the oughts of the side he’s arguing against, and only once his own have been answered on the merits, doesn’t hold it. He’s using it. He is a moral realist for the length of the conversation and a nihilist for the thirty seconds it takes to avoid conceding. That isn’t skepticism. It’s Polylogism: one logic for the domains where he likes the results, another for the domains where he doesn’t. He wouldn’t let an engineer mix metric and imperial because consistency felt inconvenient that day. The contradiction was never between the libertarian and Hume. It’s between the statist and himself.
And the standard doesn’t favour him even applied evenly, but here the word has to change, from underived to supported. Underived, the two sides are level: no one’s ought follows from facts alone, his included. The difference is what holds each one up. The libertarian’s ought survives every question put to it, and each answer lands on the descent‘s next step down until it bottoms out in logic. The statist’s has nothing under it. Take earn over $500k, owe the state 35%. No fact fixes that number, and no libertarian claims a fact fixes his rules either. The question is what the rate stands on. Economics can’t supply a calculation for it, a vote can’t supply a right nobody holds, and “taking is fine when the state does it” is Mixed Law, one rule for rulers and another for you. What’s left is a bare assertion with a tax collector attached. Yet the guillotine is only ever pointed at the side of the table that has foundations.
Bottoms out at: Equality → Logic
3. The Line in the Sand
Suppose you meet the rare consistent one, who takes the “it doesn’t bind” answer all the way down and says: no oughts bind anyone, including the norms of reasoning. Contradictions don’t bother me.
You can’t argue him out of that, and you shouldn’t try. Every argument assumes the listener already cares about the Law of Non-Contradiction, so no argument can be what makes someone start caring. Aristotle made this point about the man who denies non-contradiction: if he says anything meaningful, he’s using the law, and if he says nothing, he’s “no better than a vegetable” (in Ross’s translation). The consistent nihilist hasn’t lost the argument. He has left it. What’s left to him isn’t discourse but the only other way humans settle conflicts: force. That’s the Primacy of Consciousness in its purest form: reality, logic, and your property answer to what I want.
Be precise about who stands on the other side of the line. It isn’t everyone who has ever contradicted himself. A man who lied for his own benefit, noticed the contradiction, and changed his conduct is inside the line. Correction is the norm working. The man outside is the one who adopts contradictions as a standing policy and says “so what?” when they’re pointed out. Nothing physically stops a man from saying that, but it isn’t a counter-argument. It’s a refusal to make one, and the honest thing for someone who has made it is to stop talking, because he has nothing left to say. One is a fallible reasoner. The other has resigned from reasoning and would like to keep the benefits of membership, starting with other people’s property.
That’s the fundamental divide. It isn’t left versus right, or compassion versus greed. It’s between people who accept that consistency binds them, even when it’s inconvenient and costs them something they want, and people who accept it only when it pays. With the first, every disagreement is resolvable in principle, because both sides answer to the same standard. With the second there’s nothing to resolve. There’s only a decision about how to live alongside them.
Libertarians already have the answer, without needing to convert anyone. You don’t debate the man who has renounced reasons. You decline to share a covenant with him, you defend your property if he comes for it, and you let Law, the part of ethics that handles Conflicts and nothing more, do the only job it was ever for. The line isn’t a failure to persuade. Persuasion was only ever on offer to people who accept what persuasion requires. (See The Woke Precedent for why screening at the gate is ordinary, not exotic.)
Bottoms out at: Logic → Below logic
Where they go next
“I’m not telling you to do anything. I’m just noting the gap. Believe what you like.” Then he’s said nothing against you. A man who notes that your argument has a gap, while insisting the remark carries no implication for what you should believe, has made a remark about the weather. He can’t then call a libertarian’s conviction unjustified, irrational, dogmatic or wrong. Each of those words is normative, and he gave them up when he stepped back. All that’s left is “you haven’t derived it from facts alone,” which is true of every ought, his included.
“Then ethics is sentiment, and our shared sentiment favours the safety net.” That’s the vote again. A feeling, however widely held, isn’t a right to force someone who doesn’t share it. A majority can’t hand over a right none of its voters have (Democracy), and the question of what settles a clash between two sentiments is exactly what Law exists to answer.
“Fine. No oughts bind anyone.” Then he’s the man in section 3. Nothing obliges him to anything, and nothing obliges you to let him take your property.
Step back and see how far down this went. It started with don’t take other people’s things, a rule every child grasps and every thief expects to see applied when someone takes his things. To get around it, the statist had to go below economics, below ethics, below the norms of his own reasoning, to the claim that nothing obliges anyone to anything. That’s the hill he chose to die on, and the prize at the top is someone else’s property. The libertarian doesn’t need the whole descent to know stealing is wrong. He needs it to show that the man denying it had to saw through the ground under his own feet to do so.
Sources
- Is–ought problem — Hume’s passage at the end of Treatise 3.1.1 and the modern “Hume’s Law”
- Aristotle, Metaphysics IV.4, Ross translation — the man who denies non-contradiction