A Foundational Framework and Benchmarking Methodology for Observer-Dependent Entropy Retrieval in Linguistic Computation
Evlondo Cooper · Preprints.org · 2025
Comprehension is not merely passive decoding, but rather an observer-relative process of entropy retrieval. We formalize how individual cognitive differences modulate linguistic uncertainty via Observer-Dependent Entropy Retrieval (ODER). ODER makes two novel predictions that no current model jointly captures: (1) ∇C spikes during garden-path resolution will correlate with P600 amplitude only in low-working-memory observers, and (2) coherence terms in the observer’s density matrix (μ) will predict priming interference patterns. ODER’s quantum-inspired framework explains divergent processing costs across observer types without resorting to arbitrary parameter tuning. Using a constructed language (Aurian) as a controlled testbed, we provide implementable tools for measuring how processing difficulty is shaped by observer-specific attention, memory capacity, and prior knowledge. We further compare ODER’s quantum formalism to classical alternatives—such as Bayesian mixture models and fuzzy logic—to clarify its purpose: modeling ambiguity and interference without implying that the brain itself performs quantum computation.