Epistemic Feedback Viability: When Negative Feedback Becomes Positive

Wiener’s cybernetic framework establishes negative feedback as the fundamental mechanism of regulation: a controller that reduces the discrepancy between current and desired state produces stabilizing dynamics. This result is implicitly conditioned on semantic fidelity of the feedback channel — the assumption that the observation 𝑌𝑡 faithfully represents the current state 𝑋𝑡. We make this hidden condition explicit and prove that it fails precisely when the epistemic viability conditions of the preceding series are violated. We introduce the Wiener Epistemic Feedback Viability Theorem: a negative feedback controller preserves the negative definiteness of the Lyapunov drift 𝔼[ ̇ 𝑉 (𝑥) ∣ 𝑌𝑡] < 0 if and only if 𝐼effective(𝑡) ≥ 𝐼required and 𝜅(𝑡) ≥ 𝜅min (the series’ Tier 2 conditions). Below either threshold, the controller applies a stabilizing law in the perceived state space that corresponds to a destabilizing projection in the true state space — a phenomenon we term epistemic inversion. We identify a third phase-transition parameter 𝜅destab (complementing Paper 1’s 𝜏 ∗ and Paper 6’s 𝜇∗) below which the Lyapunov drift sign reverses. This completes the series’ epistemic failure stack: channel degradation (Paper 2), source drift (Paper 3), model staleness (Paper 6), and now control law semantic failure — four independent mechanisms by which epistemically nominal systems become dynamically unstable.