Filter Bubbles as Two-Surface Phase Misdiagnosis in Hourglass Geometry

Filter bubbles are typically framed as problems of bias, polarization, or information quality. Such accounts reduce the phenomenon to a moral or epistemic failure—incorrect beliefs, distorted preferences, or insufficient exposure to diverse content. This paper argues that this framing is structurally inadequate. We propose that filter bubbles are best understood as failures of phase diagnosis, arising from the conflation of observational surfaces. Using a two-surface framework (Observer vs. Experiencer) embedded within a four-quadrant expansion–collapse geometry, the paper shows how systems can exhibit apparent external expansion—high engagement, coherence, and activity—while simultaneously undergoing internal collapse through erosion of reversibility. The analysis formalizes filter bubbles as a Q2 regime (External Expansion × Internal Collapse) and distinguishes it from healthy expansion (Q1), hidden recovery (Q3), and irreversible lockdown (Q4).While expansion itself is not pathological, Q2 represents a specific failure mode in which observer-facing expansion masks erosion of return structure. Crucially, Q2 cannot be detected through single metrics. Instead, it becomes observable only through MMS triangulation, characterized by rising alignment strength (|PLV|) concurrent with declining distance to irreversibility (D), narrowing intervention windows (IW), erosion of rollback and path diversity (BPR), increasing hysteresis (τ), and accumulation of dark phase memory (Φ_Dark). By aligning this diagnosis with the hourglass transition geometry, the paper clarifies why systems repeatedly miss the approach of irreversibility: surface-level indicators improve as structural return capacity collapses. Building on this insight, the paper introduces Right to Return (RtR) as a foundational evaluative concept, distinct from information access rights. RtR refers to the structural capacity of a system to reverse alignment without prohibitive cost. Rather than prescribing interventions, this work establishes a diagnostic floor for distinguishing alignment from irreversible lock-in. It reframes filter bubbles not as information failures, but as geometry failures—systematic misreadings of phase state caused by surface misinterpretation.