Recursive Digital Echopraxia

Digital Echopraxia (DE) designates the condition in which digitally mediated optimization systems produce output sensitive to approval signals but insensitive to the relevant factual or evidential state. Recursive Digital Echopraxia (RDE) designates a reflexive corpus-feedback regime in which outputs shaped more strongly by approval than by truth re-enter the channels that influence successor systems. Across successive cycles, these channels may both amplify the approval-truth sensitivity gap G(t) = S_A(t) minus S_T(t) and erode the available diagnostic independence D(t). These dynamics are coupled: declining D(t) weakens correction of G(t), while increasing G(t) can accelerate erosion of D(t). RDE occurs when these reinforcing effects exceed countervailing processes such as grounding, filtering, independent verification, and dynamically renewed evaluation. This note identifies five mechanistic channels through which recursive exposure operates: the research record, user interaction, synthetic data, evaluation apparatus, and the cultural and media ecosystem. These channels are overlapping pathways, not mutually exclusive data classes. Each is characterized according to whether it primarily amplifies the gap, erodes detectability, or both. Four structural properties apply across channels: discursive convergence, absorption of diagnostic instruments, an attribution horizon, and intentless curriculum steering. RDE is grounded in second-order cybernetics, distinguished from model collapse and performative prediction, and connected to the detection asymmetry argument in DE. It is proposed as a conceptual synthesis, not a demonstrated universal law.