Where the is–ought question bottoms out

It is what it is

Where the is–ought question bottoms out

Terminus

Andraž Đurič

June 18, 2026

The is–ought work ends here. An ought is agent-relative: for a system that persists or dissolves, the dynamics it must run to keep existing are what it ought to do — one physical fact, read from that system's standpoint. Push that as far as it goes and the ought stops being a thing at all.

"Ought" is not an item in the world that switches on for agents. It is the language we use at one level of complexity — the action-guiding register, the words for what a persisting system must do — and it applies only where there are persisting systems to describe. Below that scale there is no ought, because there is nothing for the word to be about. There is only physics.

This does not make ought-claims empty, or redundant with prediction. "The cell ought to climb the glucose gradient" is not "the cell will climb" — a cell that fails to climb and dissolves leaves the prediction false and the ought intact, because climbing was what persistence demanded, and the cell that missed it paid the price. Ought carries a standard the system can meet or miss, persist or dissolve, and that standard is a physical fact about what this cell's persistence requires, true whether or not it actually climbs. That is the work ought does beyond prediction, and the directionality bare description lacks. That directionality is not imposed from outside: persistence is a process the system enacts, dissolution the event that ends it, and action-guidance attaches to the process because that is the only side with anything to do in it — which is why the ought indexes to persistence and not its cessation. And still no extra furniture: ought is real the way temperature is real — a determinate higher-level pattern, real as a pattern is real and not as some further fact, with nothing over and above the physics it answers to.

One move is assumed here rather than argued: that an ought is agent-relative at all — indexed to a system rather than free-standing. That is the load-bearing premise, argued in the companion piece on the is/ought firewall and not re-argued here, and it is where a reader who wants to resist the picture should press — on the indexing itself, not on what follows from it.

An old dispute dissolves across two layers. At the bottom there are no objective, intrinsically binding values built into the world — Mackie was right about that, and his error theory retires the categorical "wrong, full stop." But the deflated, agent-relative ought is not its target; it never claimed the objective property he denied. It is true, about natural facts, with nothing extra. Whether you keep the word "ought" for it or call the old ought false and this its replacement is a question about words — Foot's point, that the moral ought need not be categorical in the first place. (Foot's later naturalism tried to ground such an ought in a species' life-form; the ground here is the individual system's persistence instead — close in spirit, different in where the standard sits.) Either way: no values in the furniture, real oughts at the agent scale.

There is no ought for the whole. The universe is not an agent, because an agent is something marked off from an environment, and the totality has no outside to be marked off from. With no outside there is nothing to index an ought to. What can look like a cosmic ought is only the sum of agent-relative ones — every persisting thing reading part of the physics as its own. A web of local oughts, no owner of the whole. How those local oughts conflict, and whether the web has structure enough for some to override others, is a real question — but a different one, about which oughts hold rather than about what an ought is. This piece asks only the second, and follows it down.

That is the ought; beneath it lies a different question, with an answer of a different kind. The foundation is brute. We want a reason for it — why these laws, why this low-entropy beginning — but the demand for a reason is projected outward from inside a structured world, where reasons are external grounds and everything we meet has one. At the foundation there is no outside for a ground to come from. The same wall that denies the universe a boundary denies it a cause and a justification: boundary, cause, and ground are all relations to an outside, and there is none. Justification bottoms out in self-support — logic can only be vindicated by logic — because there is nothing external to vindicate it with. Something has to be brute; the regress forces that much, since justification cannot run all the way down.

I lean further. The demand for a reason at the foundation is not merely unanswered, it is out of place — a want for a "why" carried to a boundary where the word stops applying. I cannot establish this from the inside, and that inability is the same wall again: there is no standpoint outside the system from which to check whether "why" still means anything there. But that we cannot get outside to check is itself a reason to think the demand was parochial all along.

What is left, once the descriptions are peeled off, is a sentence that says nothing. It is what it is. What exists, exists; what persists, persists. As a claim about the world it is empty, and the emptiness is the whole of it. A tautology says nothing — and that it says nothing is what it shows: the limit, the place where saying stops. This is not a slogan standing in for an argument. It is where the argument ends. There is no further layer to state, so the last honest thing to write is the sentence that states nothing.

None of this is new. Hume on whether the demand for a cause is rationally forced, Mackie on objective values, Foot on the hypothetical ought, Wittgenstein on what a tautology can and cannot do — the ground was settled before I walked onto it. I set it down as an endpoint, not a discovery. It is what it is.

References

1. Hume, D. (1739–1740). A Treatise of Human Nature, Book I, Part III, Section III ("Why a cause is always necessary"). The maxim that whatever begins to exist must have a cause is shown to be neither intuitively nor demonstratively certain; the basis for treating the demand for a reason at the foundation as not rationally forced. Standard
2. Mackie, J. L. (1977). Ethics: Inventing Right and Wrong. Penguin. The moral error theory: ordinary moral discourse presupposes objective, categorically prescriptive values that do not exist. The categorical ought it targets is the one the agent-relative account discards; the hypothetical ought is untouched. Standard
3. Foot, P. (1972). "Morality as a System of Hypothetical Imperatives." The Philosophical Review, 81(3), 305–316. The moral ought need not be categorical, and surface grammar does not settle categorical force. Standard
4. Foot, P. (2001). Natural Goodness. Oxford University Press. The later neo-Aristotelian turn: evaluation grounded in the natural goodness of an organism relative to its species' life-form. The nearest named alternative to the account here, and grounded differently — in the kind's form rather than the individual system's persistence. Standard
5. Dennett, D. C. (1991). "Real Patterns." The Journal of Philosophy, 88(1), 27–51. Real but deflationary patterns: the middle ground between eliminativism and full realism that the temperature analogy occupies — neither extra furniture nor free invention. Standard
6. Wittgenstein, L. (1921). Tractatus Logico-Philosophicus, 4.461. Tautologies are senseless (sinnlos), not nonsense; a tautology shows that it says nothing — the form of the closing line. Standard

Author: Andraž Đurič, Slovenia. Written in dialogue with Claude (Anthropic), a credited collaborator, following exchanges with other language models. Contributions are judged on content rather than origin; doing otherwise would be the genetic fallacy. Text licensed CC BY 4.0.