Discovery · the ethic · ← all discoveries
The is–ought gap remains.
We name where we cross.
Descriptive premises alone do not supply a single ought without a normative bridge. Emergentism does not pretend otherwise. It separates a conditional model of extraction from a cooperative commitment, then declares that commitment in the open as a revisable vow—not a theorem.
The discovery in one breath
A conditional model suggests an asymmetry: in sufficiently closed, host-dependent systems over a long enough horizon, unrestricted extraction can erase its own conditions. The model does not command a response. We choose a cooperative side, say we chose it, and never claim the universe chose it for us.
First, the gap — honestly
Many systems blur the gap.
Keep the bridge visible.
The usual trick is smuggling. A framework describes the world in loaded adjectives — healthy, natural, optimal, fit — and the ought rides in wearing the costume of an is. "Science says you must" is always this trick, and it is always a trick: the must was packed in the luggage before the argument left the station.
We refuse the trick. Our selected equations model balance and collapse; they do not command. A command requires an explicit normative premise or commitment. The point is not that every ethical theory fails, but that its bridge premises must remain visible.
Then, what the selected model proposes
The conditional asymmetry:
does extraction erase its host?
Here is the proposed is-side. Consider a strategy that grows by draining a resource or host on which it remains dependent. Under closure, insufficient regeneration, no fresh hosts, and a long enough horizon, generalizing that strategy depletes its own precondition. Change those assumptions—open the system, add substitution, regeneration, or a short horizon—and the result may change.
The asymmetry is therefore a testable conditional, not a universal law. Even if the model survives in a domain, it remains an is: knowing that one strategy preserves its conditions better does not force anyone to prefer it.
Credit where it is due
The universalizability test is Kant's, two centuries old. What is ours is a proposed comparison with a capacity-and-dependency model. The comparison must earn its keep on declared systems; resemblance between vocabularies is not independent evidence.
The last step
The ought is a vow. We sign it. We do not prove it.
After the model, a step remains, and no equation takes it for you: prefer moves that expand rather than secretly consume another bearer's capacity. This is the framework's selected ethical commitment. We call it a vow because that is what it is: a commitment made by persons, in the open, on the record, renewable and refusable.
Why this matters to you: it inoculates in both directions. Against the moralist who says the facts oblige you — they don't, and anyone who claims otherwise is smuggling. And against the nihilist who says nothing follows, so nothing matters — almost everything follows. The facts carry you to the very lip of the gap and show you that the two sides are not equal. Only the last bit is yours. Which means you know exactly where your signature goes, and exactly how small the thing is that you are being asked to sign.
Everything modeled, tested. Everything chosen, declared.
Held against adversarial audit — see the trial record.
[C] the extraction asymmetry beyond stated assumptions · [I] the cooperative vow — chosen, revisable, not derived