The standard and the method

The history of a work is read from the documents that record it: the registers of swords kept by the great houses, sale catalogues, journals, the designation records of the NBTHK, and the inscriptions, sayagaki and papers that accompany the work itself.

The full-text collections of the National Diet Library of Japan are searched for every name, valuation and description the work carries, and every decisive passage is read on the image of the page. Each source is cited so that a reader can open it: a book by its title and page, a page of the National Diet Library by its identifier and frame.

A record publishes findings at two standings only, given in the margin of each finding.

Leads that fall short of these standings stay in the museum’s research file, where the work on them continues; when one reaches a published standing, it enters the record as a new version.

A reading the museum has published or circulated and later found to be wrong is not deleted. It stays in the record, struck through, with the reason it fell.

Each source carries a grade for the kind of evidence it gives:

A record states what the evidence establishes, and stops there. Searches that found nothing are kept in the research file, not in the record. The one exception is the identification, where a record names the other works the sword is distinguished from.

A record names the people who held a work when they acted in a public role: the houses and their lords, princes, appraisers, officials, and dealers named in print. Private individuals of the post-war years are not named unless they or their heirs consent.

From 1980 on, a record includes only documents about the work itself: designations, sayagaki, publications and exhibitions. It never names or points to an owner, a dealer or a sale in those years, and it never states or implies who owns the work today.

A document that has been lost is cited only through a published report of it, and only at the standing that report supports.

When a record establishes that a house held the work, its title page carries the crest of the house and one concluding line. The line is derived from the established findings by a fixed rule; it is never written for the occasion.

“Documented Heirloom of …” (伝来) requires the house’s holding to be established at two or more dates at least fifty years apart, under at least two heads of the house. Otherwise the line reads “Formerly in the Collection of …” (旧蔵). Without an established holding there is no concluding line.

In the war and the occupation many swords left the hands that held them, through requisition and the surrenders after 1945. For these years, as for any other, a record states what the documents establish about the work, and the documents of the occupation (licences and registers) are read as primary sources like any other.

A record is not an authentication, an appraisal or a valuation, and it is not a paper of the NBTHK or of any other body. Questions of authenticity, quality, condition and value are left to those whose office they are.

A record makes no statement about legal title, about export, or about whether a work is recorded as stolen. For those questions an owner or a buyer should take the advice of those whose office they are.

Each version of a record is kept unchanged at its own address. Its fingerprint (the SHA-256 of the record’s data, which can be downloaded and checked) is printed in its colophon, with the fingerprint of the version before it, so that the history of a record cannot be rewritten unseen.

New evidence, a finding strengthened or weakened, a typographical correction: each makes a new version, and the record’s revisions say in one line what changed. If a record is ever withdrawn from public view, its number stays in the register with the date and the fingerprints of its versions.

Every record is examined by the Curators of the Meihin Museum before it is issued. No finding is reviewed only by the person who researched it: each Established finding is checked against its decisive source by a curator who did not research it. Until that review a record is shown as a draft.

Research undertaken on commission follows the same standard. Its fees are fixed in advance and never depend on the findings or on any sale, and a commission never brings a work into the museum.