The Generative Relational Account of Time: Calendrical Practice, the Stratification of Cosmological Content, and the Authority over Time in the Chinese Case

A temporal system does several things with an occasion: it labels the occasion, it fixes the occasion's position by tracking and reconciling recurrences, and in many traditions it qualifies the occasion as fit or unfit for classes of action. This paper distinguishes those operations as three strata of a temporal system, unit naming, cycle construction, and situational qualification, and proposes that cosmological content may be carried at any of them, so that its absence from one leaves its presence at another undetermined. A criterion follows: a comparison of two temporal systems in respect of the cosmological content they carry is well formed only when it compares corresponding strata, and a comparison across strata yields an artifact, an inference from a stratum to a whole community's conception of time. The motivating case is the familiar judgment that reads planetary day-names against numbered and cyclical Chinese day-labels; the paper argues that the judgment compares non-corresponding parts, and that the same error is available from either tradition's starting point. Conducted at matched strata, the comparison finds each system carrying content at each of its three strata and differing in the distribution of that content, in the unit at which occasions are qualified, and in whether the qualifying apparatus stands apart from the dating apparatus. Reading the historical material through a generative relational account of temporal emergence required five constructs the paper introduces and marks as its own: the externalization of registration into an artifact holding the joint configuration of several cycles; the passage by which a computed position that answers to no physical body comes to be named, characterized, and addressed; the transposition of a qualifying apparatus across scales of addressee, with its cycles unchanged and its map to those it qualifies rebuilt; the location, at the naming stratum, of a registration running backward over occasions already passed; and a distinction between carrying a warrant and owning one, which turns on whether the terms in which a warrant could be disconfirmed are supplied by the party the warrant supports. The last of these yields the paper's most compact result: a temporal system may be closed against disconfirmation at one stratum while remaining answerable to observation at another, so that a single locus owns part of its warrant and carries the rest, and the part it carries is the part that admits of public failure. The paper contributes no sinological finding. Every historical statement in it is conceded to and cited from the specialist who established it, and the account is offered as a coupling device for findings that several complete literatures have made and that compose poorly across the boundaries between them. It asks to be judged by what it generates: questions are returned to each of those literatures in its own idiom, and refutation conditions are stated for every load-bearing claim.