What Is Light? Why a Finite Public World Must Have a Frontier

This preprint asks a question normally hidden by the success of Maxwell theory, relativity and quantum electrodynamics: why must an intelligible world possess any public frontier of influence at all? The argument begins before spacetime, fields and particles. A consequential history is represented through non-identical restitution: a controlled difference remains recognisable while changing the profile of possible continuations. In a genuinely single and nonfragmented public world, a world-level difference cannot remain permanently quarantined. If its publication is finite rather than instantaneously total, public reach must proceed through proper compositional stages. The Necessary Public Frontier Theorem proves that every proper stage of public reach has a nonempty next frontier. The construction requires no prior metric, spatial distance, clock duration, invariant speed or null cone. It identifies a premetric light role: the boundary through which a difference first becomes available to further public continuation. Independent lock packets then select the realised physical branch. Source, support, nuisance and no-bypass locks identify a common physical signal frontier. Smooth single-cone locks reconstruct a Lorentzian conformal structure and its null boundary. Independent U(1), conserved-current and common-image locks identify the electromagnetic sector. Locality, weak-field quadraticity and derivative-order locks select Maxwell electrodynamics. Positive-energy quantum closure, a massless spin-one pole, helicities +1 and -1, two transverse modes and coupling to the same electric current select the photon as the unique terminal occupant of the complete electromagnetic lock packet. The paper therefore distinguishes three levels that are usually collapsed: the necessary public frontier, its electromagnetic realisation and its photonic quantum closure. Light is not first a thing travelling through an already public spacetime. Light is the finite frontier through which the world becomes public. The paper answers fourteen foundational questions, supplies formal proofs and countermodels, distinguishes front velocity from phase and group velocity, and states explicit empirical failure conditions.