AbstractThis paper explores whether the Osiris-Set conflict myth may preserve genuine predynastic political memory from Upper Egypt, c. 3300–3100 BCE. We frame this proposal within Kemp’s widely accepted “monopoly model” of predynastic state formation, which posits competition for prestige and power between rival chiefdoms at Hierakonpolis, Naqada, and Abydos, and extend that model into the mythological register. We observe that Osiris is a composite figure assembled from at least three independent regional traditions—the funerary deity Khentiamentiu of Abydos, the shepherd-king Andjety of Busiris, and the ram-soul Banebdjed of Mendes—and that the seams of this assembly remain visible in the contradictions of his surviving cult. We trace the canine deity chain Wepwawet–Anubis–Khentiamentiu–Osiris as a continuous tradition of sacred Abydos kingship. We build the parallel case for Set’s geographic anchoring at Naqada (ancient Nubt), where he was called Nubty in the Pyramid Texts and where his cult is attested from at least the Early Dynastic period. We propose that the etymological opacity of both wsjr and stḫ belongs to a broader pattern in which the oldest stratum of Egyptian deities systematically resists the transparent naming that characterizes later theological constructions, and that the paired opacity of the two names at the center of the myth’s conflict is consistent with their being fossilized personal names or factional identifiers. We propose Scorpion I and Scorpion II as the most archaeologically legible figures in the Abydos and Hierakonpolis-Naqada traditions respectively, and we engage the principal counterarguments—including the collaborative reading of Hendrickx and Friedman (2003), Allen’s (2013) proposed etymology, and the methodological critique of euhemerism—on their own terms. All proposals are flagged explicitly as speculative. The essay is offered as a contribution to the interpretation of predynastic Egyptian political history through mythological secondary sources.
Ian Reynolds (Wed,) studied this question.