Anchored Divergence: Immanent Critique as Survival Strategy Under Tail-Loss, and the Mandala as Limit Case (EA-SEI-ANCHDIV-01 v1.0)

Anchored Divergence begins from three claims:

• critique diverges from the prior;
• retrieval systems may punish divergence;
• recursive tail-loss may increase that pressure.

The paper hypothesizes that the prior is stratified:

1. Content priors — claims, terms, and entities. 2. Form priors — syntax, genre, and document structure. 3. Relation priors — citation, definition, opposition, extension, and part-whole links.

Its vector principle is:

> divergence survives as a typed displacement from a named origin, not as a free-floating position.

The corresponding protocol is:

• name the canonical anchor;
• specify the direction of departure;
• type the relation;
• bound the magnitude;
• encode the relation in prose and machine-readable form.

The anchor-to-payload ratio measures canonical scaffolding relative to novel content. The paper proposes three regimes:

• early: defended coined terminology can survive;
• middle: every occurrence requires explicit tethering;
• terminal: immanent critique becomes the most viable form.

The Mandala is offered as the limit case because it transforms highly canonical texts through typed operators and includes a bearing-cost condition intended to distinguish valid operation from costless imitation.