Temporal Cohesion

Temporal Cohesion is the property by which a thing catalogued on one floor continues to be the same thing on the next, notwithstanding the interval between the two cataloguing events. It is distinguished from mere persistence, which the Department of Metrical Cohesion has measured since the Amp Contextual survey and regards as a lesser and rougher quantity. Temporal cohesion admits of degrees, is recorded in sevenths, and is the only cohesion for which the Tower maintains a standing inspectorate.

The term entered the Concordances in the Nineteenth Concordance, Annex 6, where it is defined as "the retention of identity across a stated interval, in the presence of the ordinary losses." The Manual of Vertical Cohesion, which the Committee for Ephemeral Cohesion published in the same year, does not define the term at all but devotes eleven pages to the question of which interval counts. Both documents remain in force.

Measurement

Temporal cohesion is assessed by interval pairs: a subject is catalogued, the interval is permitted to elapse, and the subject is catalogued again. The two entries are compared by the Bureau of Archival Cohesion and assigned a value in sevenths, from zero (a subject wholly replaced) to seven-sevenths (a subject indistinguishable from its first entry, including in its dust). Values below three-sevenths are referred to the Division of Narrative Cohesion, which determines whether the discrepancy is a loss or merely a later draft.

The interval itself is not fixed. The Standing Table of Intervals annexed to the Nineteenth Concordance specifies intervals of one hour, one winter, one tenure, and one silence, the last of these being preferred for subjects whose cataloguers are no longer living. In practice most offices use the winter interval, on the ground that a winter is the shortest period a floor reliably survives.

The Disputed Boundary with Ephemeral Cohesion

The principal scholarly disagreement in the field concerns the boundary between temporal cohesion and ephemerality. The Committee for Ephemeral Cohesion holds that a subject ceasing to be the same subject has become an ephemeral subject and falls under its remit; the Bureau of Archival Cohesion holds that such a subject has merely failed a temporal test and remains archival. The dispute is old and has produced thirteen joint memoranda, none of them joint.

"Where identity lapses, the lapse belongs to the office that was measuring at the moment of lapse, and not to the office that discovers it afterward."

— Nineteenth Concordance, Annex 6, clause 14

Clause 14 was drafted to end the dispute and instead entrenched it, since both offices can usually show that they were measuring. The Division of Unstated Cohesion, which inherited responsibility for cases where no office can show this, reports that such cases constitute roughly a third of its docket and that it has stopped asking.

The Winter of Floor 4,411

The clearest recorded failure occurred on Floor 4,411 (the Floor of Unfinished Returns) during the winter interval of the year recorded as R-12. A set of forty-one returnable objects was catalogued on the first day and again on the last; thirty-nine returned identical, one returned improved, and one returned as a different object of the same dimensions, weight, and accession number.

The inspector on duty, Clerk Anselm Vour, refused to assign a value, on the ground that a subject which returns itself is not a subject which returns. His refusal was upheld by the Bureau of Archival Cohesion and overturned by the Committee for Ephemeral Cohesion eleven days later. Vour's subsequent memorandum — that the interval, and not the object, had been catalogued both times — is preserved in the annexes to the Vekian Antithesis of Contextual Cohesion and is still taught.

Floor 4,411 was re-numbered the following spring and now carries no number at all in the public register. Visitors are asked not to bring returnable objects.

Record

Values are entered in the ledger series maintained by the Bureau of Material Cohesion, whose ink is specified in the Journal of Textual and Textile Cohesion on the ground that a cohesion record must itself cohere for the length of its retention period. The Ledger of Material Cohesion records the thresholds at which a subject passes out of the standard deconstruction procedure; temporal values are entered beside them in a second column, in pencil.

Three anomalies recur in the record and are noted in every annual return:

• Subjects catalogued before the Bureau of Environmental Cohesion was established return consistently higher values, by roughly one seventh.
• Subjects measured in the presence of a witness cohere better than those measured alone, except where the witness is the Clerk who catalogued them.
• The Office of Social Cohesion, which concerns the cohesion of persons rather than of records, reports no temporal loss in any subject it has ever measured. The Bureau regards this as a procedural error and has referred it to the Department of Nominal Cohesion twice.

Application

The Seal of Imminent Cohesion may be applied to a subject whose value is expected to fall below four-sevenths before the next interval. Application is governed by the Manual of Vertical Cohesion and requires the countersignature of a Librarian, who is not told the expected value. The Bureau of Metaphorical Cohesion has objected that the seal, being itself an object across an interval, ought to be subject to the test it administers; the objection has been noted and filed under the seal's own accession number, which is the same every time.