Section Beta: Cross-Referential Analysis

Section Beta: Cross-Referential Analysis is the working wing of the Department of Categorization & Cross-Referencing responsible for reading a holding's cross-references against the holdings they point to, and for recording where the two disagree. It occupies Floor 1140, a long low gallery of standing desks, and is distinguished from the rest of the Department by the fact that its staff are permitted to leave their own floor. Section Beta does not create references, does not sever references, and does not adjudicate disputes; it produces the analytical finding, which other bodies then act upon, often years later.

The Section was constituted under Third Concordance, Annex 9, in the same recension that established the Bureau of Referential Integrity, though the two bodies have never agreed on which of them was established first. Its founding charge is a single sentence, much quoted and frequently misapplied: "No reference shall be counted sound until it has been read from both ends."

Method

Analysis proceeds by the reading of pairs. An analyst takes up a holding, follows each of its outgoing references to the floor and shelf indicated, and there compares the incoming reference recorded at the far end with the description carried in the originating entry. A pair that agrees in substance is marked concordant. A pair that agrees in nothing but the fact of the link is marked adrift, and is passed upward. Between these two marks lie the majority of the Section's working categories: concordant with revision, concordant in name only, at variance as to date, and the notorious sound but inadvisable, a category introduced in 411 and never defined in writing.

The Section's standing instruction requires that an analyst reading a pair twice must record the second reading separately and must not consult the first until the second is sealed. Disagreement between a clerk's two readings of the same pair is termed a personal inconsistency and is entered in the analyst's own file, which the analyst may not read.

Relationship to other procedures

Section Beta's findings feed the Cross-Referential Audit and, more remotely, the Cross-Referential Authentication of holdings whose provenance is doubted. Where an analysis returns evidence of Cross-Referential Resonances or of Referential Sprawl, the Section does not treat the matter itself but refers it to the anchoring desks, which apply Cross-Referential Anchoring under the Cross-Referential Mandate. Where the traced link leads to a holding already logged as a referential echo, the analyst is required to file Form OV-7 'Writ of Referential Scrutiny' before continuing; the form is short, but may only be obtained from Floor 1141, which is closed on the days the Section sits.

Findings of severe inconsistency are remitted to the Department under C.7: Cross-Referential Inconsistency, a designation the Section has twice petitioned to have clarified. The Section's own view, recorded in a minute of 908 and never withdrawn, is that C.7 describes the condition it exists to resolve and not the resolution itself, and that the Department's continued filing of C.7 memoranda "under the head of remedies" is an error of long standing.

The Beta Reading of 1146

The Section's reputation rests chiefly on the general analysis conducted between the ninth and fourteenth months of 1146, in which every reference held on Floors 1136 to 1160 was read from both ends. The exercise took four hundred and eleven analysts, consumed nine hundred reams of analysis paper, and produced a finding of 61,004 pairs, of which 58,882 were concordant, 2,001 were adrift, and 121 were returned as unreadable at the far end — the far ends in question being floors that the readers could not locate, though the floors themselves are numbered and appear in the register.

The unreadable 121 were not abandoned. They were wrapped, sealed, and stored in the Section's holding annex, where they remain and where they are counted each year without being opened. A marginal note in the Section's copy of the ledger, in the hand of the then presider Analyst Sephra Vance, reads only: "Not a defect in the reference. A defect in the stair."

Disputes

The Section has been in recorded disagreement with the Bureau of Referential Integrity since 1149, when the Bureau's auditors, arriving to conduct a Cross-Referential Audit of the Section's own files, were refused entry on the ground that an audit of the auditors' counterpart is not a cognate proceeding. The Bureau replied that no such distinction appears in any Concordance. The Section replied that neither does the Bureau. The matter was settled in 1151 by a joint recension of the Master Cross-Reference Ledger, which added the Section to the register of floors subject to scrutiny and simultaneously exempted it from scrutiny during the months of its sittings. Since the Section sits in all twelve months, the exemption has been read by successive presiders as total and by successive Chiefs of the Bureau as void. No audit has been attempted since 1163.

A second and quieter dispute concerns the Cross-Referential Webbing, first catalogued by the Bureau during the 88th recension of the Ledger. The Section maintains that the webbing is a description of what its analysts see when reading a dense body of references at speed, and that it should be analysed before being named; the Bureau maintains that it was named first and analysis may follow at leisure. The Section has twice declined to analyse the webbing on the ground that the Bureau has not said what it is.

Practice notes

Analysts work in pairs, one reading forward and one back, and are rotated every eleventh day to prevent what the Section's handbook calls "the growth of familiarity." Visitors to Floor 1140 are asked to remain silent while the desks are reading and to leave by the stair they came up by. The Section keeps no chairs for guests, on the reasoning that a chair invites a reading. Ink is not supplied; analysts bring their own, and the Section's stores hold only the paper.