V.2 — Agency Conservation and the Sacrifice Pattern

Paper Summary

Formalizes a recurrent failure mode in governance systems: systematic destruction of individual agency as instrumental means of achieving system-level objectives. Shows structural isomorphism between ancient ritual sacrifice and modern “collateral damage.”

Key Concepts

Agency Operationalized

Agency = the set of future trajectories an agent can still steer toward from a given vantage state.

Quantifying Agency Loss

Hard Loss: Futures that become unreachable \(H_i(\pi_1 \rightarrow \pi_2) = P_i^{\pi_1}\left(\Omega_i \setminus \operatorname{supp}\left(P_i^{\pi_2}\right)\right)\)

Soft Deformation: Futures that remain reachable but become statistically implausible \(K_i(\pi_1 \rightarrow \pi_2) = D_{\mathrm{KL}}\left(P_i^{\pi_1}(\cdot \mid s)\,\|\,P_i^{\pi_2}(\cdot \mid s)\right)\)

Individual Agency Loss: \(\Delta A_i(\pi_1 \rightarrow \pi_2) = \alpha\,H_i(\pi_1 \rightarrow \pi_2) + \beta\,K_i(\pi_1 \rightarrow \pi_2)\)

Standing Sensitivity

\(S_i = \left|\frac{\partial C}{\partial \Delta A_i}\right|\)

Standing Asymmetry: When expected sensitivity for victim-candidate class V is much less than for decision-bearing class D.

The Sacrifice Pattern

Definition (All must hold)

  1. Instrumentality: ∃i ∈ V such that ∂G/∂ΔAᵢ > 0 (reducing agency improves goal)
  2. Non-consent: Members of V lack effective exit or veto
  3. Standing Asymmetry: Harm to V weakly affects institutional cost
  4. Epistemic Avoidability: No documented agency-conserving exploration of lower-agency-loss alternatives proportional to stakes

“Instrumentality, not intent, is the bright line.”

The Sacrifice Attractor

Under standing asymmetry and responsibility diffusion, policies concentrating harm on V while improving G are locally stable outcomes of institutional selection.

“Sacrifice is an attractor.”

Structural Persistence Mechanisms

The pattern persists not by malice but by:

  • Standing Asymmetry: Who counts
  • Responsibility Diffusion: Who’s accountable
  • Cosmological Abstraction: Framing that obscures individual harm

Gradient Suppression (Modern Theology)

Modern systems stabilize sacrificial regimes by suppressing moral gradients through:

  • Aggregation
  • Abstraction
  • Responsibility diffusion
  • “Cosmology tokens” (greater good, progress, necessity)
\[\frac{\partial C}{\partial \Delta A_i} \downarrow \quad \text{while} \quad \Delta A_i \not\downarrow\]

“This is functionally identical to theological sanctification in ancient sacrifice.”

Admissibility Conditions

A policy π is admissible only if:

  1. Targeting: Directed at specific threat/constraint
  2. Minimality: Least harmful option
  3. Non-instrumentality: ∂G/∂ΔAᵢ ≈ 0 for non-coercive agents
  4. Responsibility Localization: Clear accountability
  5. Robustness: Admissibility invariant across reasonable α/β range

Anti-Ethics-Washing Constraint

Exploration process E is admissible only if:

  • Standing preservation
  • Model diversity
  • Gradient visibility
  • Auditability

“Simulated exploration that suppresses victim gradients constitutes a second-order sacrificial violation.”

Implications for Algorithmic Governance

“Any governance system—human or artificial—that aggregates utility while hiding per-agent marginal agency loss is structurally sacrificial.”

Safe governance must expose ∂C/∂ΔAᵢ as a first-class metric for all i.

Key Insight

“Ritual sacrifice and modern collateral damage differ in presentation, not structure.”

Prevention requires:

  • Restoring standing
  • Enforcing exit and veto
  • Localizing responsibility
  • Auditing exploration
  • Rejecting abstractions that override individual agency

Connection to Axionic Framework

  • Operationalizes “harm” from Axionic Injunction (III.5)
  • Provides diagnostic criteria for detecting phase-destruction
  • Extends non-consent analysis from ARC (IV.5)