This preregistration describes a secondary analysis of existing public lexical, affective, sensorimotor, corpus, and bodily-sensation norm datasets. The project examines whether the association between interoceptive grounding and estimated age of acquisition is specific to emotion words, beyond general abstract-word effects. The primary confirmatory analysis tests whether Lancaster interoception ratings show a stronger negative association with age of acquisition for emotion words than for matched non-emotion abstract words. A secondary confirmatory analysis tests competing predictions about interoceptive differentiation among emotion words, using a prototype-margin index derived from corpus-based body-term co-occurrence vectors and Nummenmaa/world_emBODY bodily-sensation prototypes. All analytic sample sizes, feasibility checks, index-construction decisions, and design simulations were fixed before running any main age-of-acquisition outcome models. The registration package includes the preregistered analysis plan, pre-outcome scripts, configuration files, provenance records, feasibility outputs, design-analysis outputs, and checksum manifest. No new data are collected, and no main RQ-A or RQ-B AoA–predictor outcome model has been run before registration.