Correctors and Recruiters

A recruiter widens permeability at scale. Not a relationship formed one mind at a time, but access opening for many minds at once to a given region of Belief Space. ETCS already described institutions doing this as their basic function, spreading a coordination target, a curriculum, a credential, wide enough that large numbers of people become permeable to it simultaneously. Nothing new to resolve here. Recruiting is the mechanism already established, just named directly.

A corrector is not a thing that gets transmitted. It isn't belief or information pushed across a membrane, that's the sender-pushes-a-packet model permeability already replaced for belief in general. A corrector is an event: genuine substrate contact of a kind the existing reference frame can't absorb without being forced to react.

This is why more exposure, more repetition, more stirring, never converts a mixture into a compound. It takes a different kind of contact entirely, not more of the same kind. Real testing, real experience, evidence that can't be reinterpreted away, forces the reaction that produces knowledge. Belief that only ever receives more of the same confirming contact stays belief indefinitely, however long it's held.

Recruiting scales because it's fundamentally a diffusion process, belief as gas filling whatever's permeable, and permeability can be opened for enormous numbers of people at once through a single institutional act, a curriculum, a broadcast target, a credential requirement. Correcting doesn't have that structure. Forcing genuine substrate contact strong enough to produce a reaction is inherently a per-mind event, the specific experience that breaks one person's existing frame open doesn't necessarily do the same for the next person, and can't be mass-produced the way a target can be mass-broadcast.

This means institutions don't neglect correcting through malice or oversight. Any institution optimizing for reach naturally tilts toward the mechanism that actually scales. The imbalance ETCS described, recruiting outpacing correcting as institutions grow, isn't a design failure. It's what happens by default when one half of the pair scales and the other doesn't.

Crisis is the exception, and it's the exception that proves the mechanism rather than breaking it. A crisis forces genuine, undeniable substrate contact on large numbers of people simultaneously, without needing an institution to mediate it person by person. That's still a corrector, still substrate contact strong enough to force a reaction, it's just one of the rare conditions under which correcting achieves the kind of scale recruiting gets by default. Crisis doesn't scale correcting by design. It scales correcting by brute force, substrate imposing contact directly rather than needing to be arranged.

XIK-TS left this open: permeability and crystallization show up correlated, with no single located mechanism producing either. There is one mechanism, not two separate pathways converging on the same appearance.

Mixtures, asymptotic beliefs, lack the substrate bond that would let them stand independently. What sustains them instead is identity anchoring, they get tied to a person's sense of self, and once anchored, they're protected by whatever function protects the rest of identity. Knowledge, genuinely reacted, doesn't need anchoring to persist, so it gives identity nothing to defend, and stays comparatively permeable, open to being extended without threat. Crystallization and low permeability are what identity's general self-protective function looks like when applied to something that was never actually reacted, but has been anchored to self anyway.

The rigid expert case makes the mechanism concrete rather than abstract. What crystallizes isn't usually the technical knowledge itself, which may be genuinely reacted and comparatively open to refinement. What's anchored is the identity of being an expert, being the kind of person who would already know if something were true. When disconfirming contact arrives, it doesn't test that identity the way a corrector is supposed to. Instead, the anchored identity claim tests the contact, I would have known this if it were true, and whatever fails that test gets rejected or reinterpreted rather than triggering revision. That's the correction relationship inverted: substrate contact is supposed to test belief, and here belief, specifically an identity anchored to expertise, has become the standard that substrate contact is measured against instead.

Belief Space described what a reference frame built around an error looks like, and why staying unenacted protects it. This page names the event that would actually correct it, and explains why that event is structurally rare compared to how easily belief spreads. ETCS's recruiting-without- correcting problem now has a real mechanism behind it, not just an observation. Generative Coordination, whenever fully drafted, is what it looks like when correctors and recruiters operate together deliberately, correcting scaled up wherever possible, recruiting never allowed to outrun it by default.