The Limits of Sacred Authority: Consent, Withdrawal, and Non-Imposition in Religious Jurisdiction

Religious traditions may teach exclusive doctrines, organize rites, credential specialists, control offices and membership, refuse recognition, and end association. This paper does not decide whether the individual, lineage, community, ancestral line, or land is the primary bearer of sacred standing. Its claim concerns the reach of institutional power rather than the location of sacred truth. The proposed threshold is non-imposition. An institution may withdraw its own religious recognition, membership, internal office, title, rites, endorsement, and participation. Religious withdrawal does not itself determine contractual, property, employment, housing, educational, healthcare, fiduciary, or public-law rights, which remain governed by applicable agreements, vested interests, notice and reliance protections, and law. An institution may not extend its authority into physical harm, confinement, coerced ritual, dispossession, loss of legal standing, interference with family or dependants, targeted harassment, or incitement — and such conduct does not become internal governance merely because the person remains formally enrolled. The framework protects doctrinal denial and voluntary association even where they are painful. It also treats consent and exit as material conditions rather than formal words: dependence on housing, income, care, medical access, immigration status, family membership, or community belonging can make nominal agreement unreliable. The argument is symmetrical and deliberately narrow. A companion doctrinal survey, reported in an associated methods paper, did not establish a tradition-wide restriction on private sacred access across seven religious traditions; a narrower operational hypothesis concerning particular lineages, regions, and periods remains untested. This working paper therefore presents a normative framework and a research agenda rather than a comparative empirical verdict. Status: Working paper, not peer reviewed. Companion methods paper (Adversarial Multi-AI Falsification) forthcoming as a separate Zenodo record; this record will be updated with its DOI on release.