The Architecture

An ethical system constrains the behaviour of agents within a shared substrate. You already know why. You have lived with other people.

Without agreements about who does what, when, and at what cost, shared spaces degrade. Cooperation requires rules. Rules require a source.

The question is not whether rules are needed. The question is where the rules come from.

There are two possible answers. Only two. Not because other answers have been excluded by fiat, but because the question itself is binary.

The rules either derive from the invariant structure of reality — from what is measurably, testably, falsifiably true about how coupled systems work — or they derive from something else. A claimed authority.

A declared source. A text, a tradition, a revelation.

There is no third option. To see why, consider: an ethical system either reads the structure it governs, or it imports a claim about that structure from outside.

Any system whose authority depends on interpretation belongs to the second category, no matter how sophisticated the interpretation.

Architecture A — authority-based ethics. The constraint is derived from an authority external to the substrate. A god declares. A prophet transcribes. A text preserves. An institution interprets.

The authority is not the structure of reality. The authority is a claimed source — revelation, tradition, divine will — that stands outside the system it governs. The rules are not derived from the substrate.

They are imposed on it.

Architecture B — first-principles ethics. The constraint is derived from the invariant structure of the substrate itself. The rules are not imposed. They are read. The speed of light is not commanded.

The terminal ethic is not commanded. Both are consequences of the same axioms acting on the same substrate. The constraint is the structure, described honestly.

Now consider the traditions that seem to fall between these two.

Virtue ethics, developed by Aristotle, derives its norms from a conception of human flourishing. But “flourishing” is defined by the philosopher, the culture, the era.

A Greek philosopher’s flourishing is not a medieval monk’s flourishing is not a Silicon Valley founder’s flourishing. The definition is not invariant. It is interpreted. And what can be interpreted can be reinterpreted.

Contractualism, developed by Scanlon and others, derives its norms from agreements that rational agents would make. But “rational” is defined by the theorist. The agreements are hypothetical constructs, not structural measurements.

Different theorists produce different agreements.

Pragmatist ethics, developed by Dewey, derives its norms from what works. But “works” is defined relative to goals that are themselves chosen, not derived. Change the goals and you change what works.

Care ethics, developed by Held and others, derives its norms from relational responsiveness — which is closer to the coupling geometry of the axioms than any other tradition.

Care ethics understands that relationships are structurally primary, not secondary. But it still grounds its authority in human relational experience rather than in the invariant structure from which that experience emerges.

Under sufficient pressure, the definition of “care” can be reinterpreted to serve the interpreter.

Each of these traditions contains genuine insight. Several map closely to consequences that the axioms derive independently. But none of them derive their normative authority from a structure that is invariant, testable, and falsifiable.

Each ultimately rests on a claim — about human nature, about rational agreement, about relational primacy — that can be interpreted differently by different agents.

That interpretability is the structural property that makes them susceptible to the same forcing chain as Architecture A, even if the forcing operates more slowly and with less violence than in the religious case.

They are variants of Architecture A operating at lower pressure. Under sufficient pressure, the same chain activates.

Architecture B is not a third option alongside these.

Architecture B is the structural category to which any ethics belongs if and only if its normative authority derives from the invariant structure of the substrate — structure that is testable, falsifiable, and carries explicit kill switches.

To date, the 420 Code is the only complete implementation of Architecture B. The binary is not asserted.

It is derived from the question of where normative authority originates, and the answer is exhaustive: either from the invariant structure, or from something else.

Now here is the thing that matters. Please take this slowly, because everything that follows depends on it.

Architecture A is unstable. Not contingently — not because bad people use it, not because specific religions are flawed, not because the original revelations were impure — but necessarily.

The instability is a consequence of the architecture itself.

A bridge whose resonant frequency matches the wind will oscillate to destruction regardless of the quality of the steel. The steel is not the problem. The frequency is the problem. The architecture is the frequency.

The instability unfolds in five stages. Each follows from the one before it.

Together they form a forcing chain — a chain of the same shape as the ones that force the speed of light, the proton mass, and the terminal ethic, though, as stated below, at the strength of a reliable structural tendency rather than a mathematical forcing.

Except this chain forces collapse.

Stage 1: Declaration. An authority is declared. God spoke. The prophet received. The text was revealed.

The declaration places the source of ethics outside the structure of reality and inside a claimed event — a revelation, a vision, a covenant, a burning bush, a cave, a mountaintop.

The event is historical, singular, and unrepeatable. It cannot be re-run. It cannot be verified. It cannot be tested. It cannot be falsified. It can only be believed.

This is the first structural flaw: an ethical foundation that cannot be tested is an ethical foundation that cannot be corrected.

Stage 2: Transcription. The authority’s output is recorded. Tablets. Scrolls. Books.

The recording is performed by human agents — Moses, Paul, Muhammad, the disciples, the compilers, the translators, the councils that decided which books to include and which to burn. Every human agent introduces noise.

Not dishonesty — noise. The signal passes through a channel with finite bandwidth. The transcription is an approximation of the declaration. The approximation is all that survives. The original signal is gone.

What remains is a human product — written in human language, shaped by human context, carrying human limitation — that claims divine origin.

The claim cannot be verified because the original signal is unavailable for comparison.

Stage 3: Interpretation. The transcription requires interpretation because language is ambiguous and situations are infinite. The text says “thou shalt not kill.” A thousand years of commentary asks: kill whom? When? Enemies in war? Heretics?

Apostates? The unborn? The terminally ill? The text does not answer because the text is finite and the situations are not. Interpretation fills the gap.

Interpretation is performed by human agents — priests, rabbis, imams, scholars, councils, popes, ayatollahs, televangelists. Each interpreter brings their own window. Each window has a different view. Each interpreter claims fidelity to the original declaration.

The interpretations diverge. They must diverge. A finite text applied to infinite situations by different minds in different centuries will always produce divergent readings. The divergence is not a failure of the interpreters.

It is a mathematical certainty produced by the architecture.

Stage 4: Divergence. The divergent interpretations produce competing claims to absolute truth. Sunni and Shia. Catholic and Protestant. Orthodox and Reform. Theravada and Mahayana. Each claims fidelity to the original declaration.

Each accuses the other of distortion.

The claims are irreconcilable because each is derived from an absolute — a god who does not negotiate, a text that does not update, a revelation that does not repeat.

When two parties each hold an absolute that contradicts the other, no structural mechanism exists for resolution. Negotiation requires that both parties hold their position provisionally. An absolute is not provisional. An absolute is absolute.

Compromise is apostasy. Concession is betrayal of God.

The architecture has produced two groups, each certain they are right, each certain the other is wrong, each certain that their certainty comes from the highest possible authority, and has given them no mechanism for resolving the disagreement short of one group ceasing to exist.

Stage 5: Collapse. Competing absolutes in a shared substrate with finite resources produce violence. Not as aberration. As consequence. The Sunni does not kill the Shia because the Sunni is evil.

The Sunni kills the Shia because the Sunni’s architecture tells the Sunni that the Shia’s reading of God’s word is a corruption, and the Shia’s architecture tells the Shia the same about the Sunni, and neither architecture contains a mechanism for admitting that it might be wrong.

Kill switches do not exist in Architecture A. The absence of kill switches is the structural flaw. A system that cannot admit error will, given sufficient time and sufficient pressure, convert the error into violence.

This is a strong structural tendency, evidenced across the historical record — a reliable consequence of interpretive divergence under pressure, not a mathematical forcing on par with the physics derivations. The ball-in-valley below is an analogy for that reliability, not an identity with it.

The same way a ball at the top of a hill is forced into a valley by gravity, competing absolutes are forced into violence by the geometry of irreconcilable claims on a finite substrate.

This is not a moral judgement. It has the shape of thermodynamics: the authority provides the asymmetry — these rules, not those; this god, not that one; this reading, not that one.

The architecture does not provide the structural support — the rules are declared, not derived; believed, not tested; interpreted, not proved. Entropy does the rest. The timeline varies — centuries, decades, sometimes years.

The outcome does not vary. The scaffold falls. It has always fallen. It is falling now.

And when it falls, it falls on the people underneath it.

Architecture B does not have this property. The axioms cannot be interpreted because they are not ambiguous. The speed of light does not require a commentary tradition.

The proton mass formula does not require a council of scholars. The terminal ethic does not require a pope. The axioms are tested, not believed.

They carry kill switches — over five hundred and fifty explicit, stated, falsifiable conditions under which they die. A system that can admit error can correct error. A system that cannot admit error can only escalate.

That is the structural difference. That is the only structural difference that matters.