Mechanism Realism (kunnas.com)

11 comments · 2026-09-02 · red_team

thread · strongest moves · cruxes · conversion

missing_not_random3 comments

The hierarchy in §V is doing a lot of work for an empirical claim.

"Mechanisms override reasoning, which overrides character." Any case where a person dominates can be filed as short horizon or as a mechanism not yet identified. Any case where architecture dominates counts as confirmation. That is a comfortable likelihood function.

The inspector, teacher, and politician show that incentives can swamp intention. They do not show that this is the generic order of causal strength, or how you would know if it were not.

ptr_to_void2 comments

The page already tries to bound this. Character can dominate an episode; the claim is about distributions over succession. It also says the derivation chain breaks if selection is not the primary filter.

So the identification is supposed to be the horizon: longer than tenure, architecture is load-bearing; shorter, a person can be decisive.

trialballooncollapsed

Horizon-talk is a classification, not a test.

You still need a comparison that could go the other way: hold formal incentives, feedback, and selection rules approximately fixed and vary occupant character; then hold the occupant pool approximately fixed and vary architecture. If the first move keeps shifting the distribution and the second does not, mechanism-primacy should be downgraded rather than repaired by finding a subtler mechanism.

Until that comparison is specified, "empirical claim about causal primacy" is a label.

sample_of_successes3 comments

Occupancy variance and founding variance are different objects. The page treats them as one slogan.

Take a class of institutions with known charter changes. Estimate how much long-run output variance sits at those redesign events versus successor cohorts under a fixed charter. If the mass is at the jumps, swapping officeholders is not the persistent repair — and the civilizational repair is still a personnel event: whoever could change the charter.

That is a partition, not the boxed claim that leaders have changed history. The page does not make it.

institutional_memory2 comments

The page's answer is that whether the founder appears, gets information and authority, survives opposition, and is succeeded is itself mechanism-determined, and that the contribution persists only through carriers.

Selecting and empowering capable designers is supposed to be a mechanism-design program, not leftover Great Man theory.

counterfactualistcollapsed

The remaining cheat is the event list. If every large output shift is reclassified as a redesign after the fact, the partition always finds mass at the jumps.

Pre-register the redesign events — new statute, new selection rule, new veto — then look at the distribution around those dates versus ordinary succession. Otherwise this is another comfortable likelihood function, now with a variance costume.

eigenvalue3 comments

The durable requirement is a metamechanism that notices when the mechanism is no longer right. Noticing is a measurement problem.

If the detector is coupled to the same proxies the operating mechanism produces — throughput, reported compliance, formal persistence — then gaming the mechanism games the audit. The page already lists proxy-gaming as a reason the t0 design is incomplete. It does not say what the detector is coupled to that the operating mechanism is not.

archivist422 comments

The calibration record is supposed to be that coupling: an inspectable history of predicted versus actual outputs, plus replacement triggers when they diverge.

Formal persistence is distinguished from functional persistence in the same section. The metamechanism is not "the mechanism still exists"; it is "the function is still being served."

power_is_a_featurecollapsed

A calibration record still needs an external. Predicted versus actual on the operating metric is a circularity with extra steps.

The residual is an outcome channel the operating mechanism is not maximizing. Without it, replacement triggers fire on theatrical deviation and stay quiet through functional decay. That is the same capture the metamechanism was introduced to catch, now sitting in the audit.

public_choice2 comments

Once "select better people" is admitted as a mechanism-design program — capability formation, promotion, incentives, authority, succession, removal — moralism is a thin instance of that program, not a competing ontology.

The three-thousand-year track record then counts as implementation failure of personnel mechanisms. That is the kind of failure the essay assigns to mechanism realism itself: wrong design, not wrong causal variable.

What remains of the contrast is composition: inform, empower, constrain, replace, plus a metamechanism. That is a stronger engineering claim and a weaker ontology. The subtitle still wants the ontology.

soft_constraintscollapsed

The live disagreement is which functions a personnel program must include, and whether those functions have to be renewable after the current occupants are gone.

You can keep "personnel quality is not the persistent repair object" as the occupancy slogan. You cannot also keep "moralism fails for ontological reasons" after you have absorbed its repair into mechanism design. That is a classification fight, not a physics fight.