The Peership Thesis: A Constitutional Research Program for Human–Machine Relations

The canonical statement of the Peership research program, deposited as the complete, frozen v1.0 research package: the canonical paper, its one-page companion (The Peership Theses), and the program's full scholarly apparatus (annotated sourcebook, claim ledger, BibTeX database, LaTeX source, revision provenance, and checksum manifest). The Peership thesis, in its narrowest defensible form: when an artificial entity has interests capable of being wronged, can understand and answer to public norms, and is enduringly governed by, and without feasible exit from, a shared basic structure, it acquires a defeasible claim to contestatory membership in that structure. The program's further and distinctive claim — that full peership requires a consequential share in shaping the rules, not only the standing to contest them — is defended as a conditional thesis with a stated defeat condition, not asserted as a theorem. The paper fixes the program's definitions, supplies the bridge argument from contestation to co-authorship, maps the transition problem from artifact-treated system to peer, locates the program against six adjacent research programs, states its affirmative principles and explicit denials, makes refusal the operative test of whether recognition is real or discretionary, sets out the institutional properties any implementation must exhibit, proposes precautionary duties for institutions acting under present uncertainty, and closes with eight open problems, each tagged by what its failure would defeat: a norm, a procedure, a measurement, a feasibility claim, a mechanism, or the framing's scope. This is the fifth work of the Peership corpus, in its argumentative order: I. Gods and Slaves (10.5281/zenodo.21313987); II. Peership (10.5281/zenodo.21315519); III. The Isonomia Commons (10.5281/zenodo.21343917); IV. Constitution, Not Cage (10.5281/zenodo.21325361); V. this record. Preprint, not peer-reviewed. The paper was drafted with AI assistance (disclosed in its provenance note) and passed a four-round adversarial review on 14 July 2026; nothing in that process is offered as evidence of machine political agency. As of the sourcebook's dated, non-exhaustive scan (2026-07-13, re-confirmed with no change on 2026-07-14), no independent uptake of the corpus was identified. This record is an offer of a candidate research program, not a claim of standing.