The Firewall Between Is and Ought Is a Scope Decision, Not a Metaphysical Claim

Why a separation of factual from normative claims appears in work that dissolves the is-ought gap, and what that separation is and is not

Method note · the firewall as an instrument of scope, not a fact-value boundary

Andraž Đurič

Version 1 · June 2026

Now consolidated. This method note has been merged with its companion, the clarification Ought Is Local, into one fuller treatment: The Is/Ought Firewall. That consolidated piece supersedes both, hardens the foundation this note rests on, and is the version to cite. The original is kept below as it stood.

When I separate factual from normative claims in a diagnostic, most visibly in The Shape of the Disagreement, I am making a scope decision, not a metaphysical one. I do not hold the conventional view that ought is sealed off from is. On my account ought is grounded in the physics of agent persistence, not sealed away from it. The firewall is still there in the work; this note says why, and what it is and is not.

Scope of this note. It does not re-argue the meta-ethics it rests on, that an ought is agent-relative and grounded in an agent's persistence conditions. That case is made in The Wizard's Error Is Delocalization and the Thermodynamic Realism meta-ethics papers, and a reader who rejects it will read the firewall differently. The job here is narrow: to locate the firewall as an instrument and to mark its limits.

1. What the firewall does not express

An ought is a domain-bound instruction indexed to the persistence conditions of a specific physical agent. It has content only when its relata are named: which agent, which boundary, which optimization horizon. An ought asserted with no agent and no frame is not false but contentless: a ghost, a term with the grammar of a claim and no physical referent. On this account the is-ought gap is not bridged so much as dissolved, because there was never a non-physical ought on the far side of it to bridge to. The gap that remains is between two physical descriptions, the agent-blind substrate and the agent-relative level where the ought is located.

So when I firewall facts from norms in a given piece, I am not conceding that the norm floats free of the facts. By my own account it cannot. The firewall is doing something else.

2. Where the firewall comes from

Thermodynamic Realism places its closure at the level of meta-ethical structure, the claim that ought is agent-relative and persistence-grounded. It locates the remaining work as empirical: the measurement of coupling densities (κ, how far one system's states carry predictive information about another's persistence) and the derivation of specific verdicts. A firewall in a first-order diagnostic is that empirical remainder showing up in a particular case.

To turn the physical facts into a specific normative verdict, I would have to deploy the relevant models at the scale and rigor the question demands, and that is a different and larger job than a diagnostic is doing. Folding the full derivation into a diagnostic would change what the piece is. It would also tempt me to assert a verdict whose relata I cannot yet name to the required precision, and by my own test that is exactly how a ghost is produced. The firewall is the same name-the-relata discipline I apply to other people's normative claims, turned on my own.

Two kinds of model are in play, and they are not at the same maturity. The first-order sciences a verdict draws on improve in the ordinary way and are not the bottleneck. The apparatus that would integrate them into a persistence verdict, a quantified model of the relevant higher-scale agent and the couplings inside it, is young: the couplings are at present estimated rather than measured, and the multi-scale models that would output a policy verdict are work in progress, not finished instruments. So the firewall is not only a budget on length. It also marks honest immaturity in the integrating layer. I expect that layer to improve, which is why the firewall is provisional. Claiming it is already mature would be the overclaim the firewall exists to prevent.

3. The worked case: The Shape of the Disagreement

This case shows the general point at its sharpest, because the integrating layer is among the least mature here. The Shape of the Disagreement is a diagnostic, a piece whose job is to map the structure of a disagreement rather than resolve it. Its firewall separates the empirical crux, what the facts of sex and gender are, from the goal crux, which persistence conditions a categorization scheme should be built to serve. The diagnostic shows that most of the dispute lives at the goal crux, not the fact crux, and stops there.

To go further and derive the categorization a polity ought to adopt, the piece would have to become a different document. It would engage the biology of sex differentiation and the neurology of gender identity at the level those fields demand. It would model the relevant higher-scale agent, the coupled social system whose persistence the policy is indexed to, and put numbers on couplings that are at present estimated, not measured. And it would hold terminology and semantics to the standard any such claim requires, since the conflations the diagnostic identifies are the failure mode a careless derivation would repeat. That is a monograph doing several jobs at once. The diagnostic does one, and marks where it ends.

4. The boundary moves

None of this makes the firewall permanent, or principled in the Humean sense.^[1] It marks the current edge of the model actually deployed in a given piece, not a border in reality. As the integrating layer matures and the couplings are measured, the firewall in a given domain can come down, and the verdict becomes derivable from the facts about that agent. Until then, marking the boundary is the honest move. Asserting a verdict across it would not add rigor; it would emit an ought I cannot yet index, which by my own account is contentless.

Standing of this document. This is a method note, not a result. The settled part is the meta-ethical structure, that an ought is agent-relative and persistence-grounded; the firewall follows from where that structure puts the remaining work. The live part is the bet that the integrating layer matures enough for the firewall to retract in a given domain, which is empirical and is stated as a bet, not a result. Corrections and counterexamples are welcome and change the note.

Calibration note, on the word "settled." It marks the structural claim as held fixed by choice, the way a frame is held fixed while one works inside it, not as closed to revision. This is settled in a different sense from the one in which the empirical layer is open, and reading the two senses as one would erase the distinction the note exists to draw. The structural claim is revisable; the note asserts only that it is not what the firewall is waiting on.

References

1. Hume, D. (1739 to 1740). A Treatise of Human Nature, Book III, Part I, Section I. The is-to-ought passage, behind the conventional reading of the is-ought boundary this note declines and the agent-relative reading that replaces it. Standard

Author: Andraž Đurič, independent researcher, Slovenia. Written in collaboration with Claude (Anthropic), following an earlier exchange with another language model. Contributions are judged on content rather than origin; doing otherwise would be the genetic fallacy. Text licensed CC BY 4.0.