The 1993 Summary

The 1993 summary is the most load-bearing document in this folder and the only one nobody has carded. By citation, the tone tables exist for most readers as their own summary: chains of secondary reference route through the 1993 paraphrase so thoroughly that the paraphrase has functionally replaced what it paraphrases, and it has done this without ever being put on a claim-and-condition card of the kind the folder owes every other artifact.

What the tables are

The tone tables are the Loop's largest artifact: roughly forty pages of frequency measurements published in the Ledger across fifteen years. Each reading carries its own stated measurement error, and the run carries a habit — the repeated note "board, not frame" — that appears in no summary of them. Anyone who has actually read the tables knows the errors and the board/frame distinction are most of what the tables are. The 1993 summary reproduces ranges. It does not reproduce conditions.

Citation anatomy

In secondary accounts, citations to the tone tables are overwhelmingly citations to other citations. The chain usually terminates in a single 1993 summary that paraphrases the tables and, as reported, gets two of the ranges wrong. This is marrow in the precise sense: the part of a body of material everyone has copied and nobody has read past.

The survival table concedes it in one line. Among "what survived" after 2001 it lists the tone tables twice — "via the Ledger, and via one bad 1993 summary" — making the digest a survivor in its own right, carrying content the tradition can no longer check it against.

Carded, claim against condition

Following Sisken's method, the page's own load-bearing claims go on separate lines:

• CLAIM: the 1993 summary is the terminal citation for most references to the tone tables. CONDITION: attested in the citation survey; no counter-chain has been produced.
• CLAIM: the summary gets two of the ranges wrong. CONDITION: NOT SEEN. This finding descends from citations-to-citations. Nobody currently holds the summary and a complete table set together, so no page here — including this one — can demonstrate the errors. The claim is reported, not shown.
• CLAIM: no complete Ledger set exists in reach. CONDITION: consistent with four of Sisken's seven shoeboxes being unaccounted for and the Index's whereabouts being unestablished. Ledger-true as far as the ledger runs; the ledger has holes in it.

What would settle it is small and physical: one complete set of the tables, one copy of the summary, in one room, read against each other by someone willing to card the differences. Until then the count stays "reported as two" and the gap stays a gap.

Why the replacement is invisible

The Fenwick affair supplies the control case. The 1992 forgery fell on paper — a watermark that entered production in 1987 beneath readings dated 1985 — but its eleven-day exposure taught the durable lesson: four separate people who confirmed the forged sheets as consistent with "their copies" had each checked against a summary, not the tables. At that moment nobody in the tradition held a complete set, and the checking ritual proceeded as if they did.

The general shape: originals that are copied-but-unread get silently replaced by their digests, and the replacement is invisible precisely when the digest is good. A bad summary is only detectable by somebody who has read the original — the same circularity the Fenwick eleven days exposed. The summary should be carded before that circularity is asked to guarantee anything again.

See also

lore/tone-tables, lore/the-meets, lore/index.

Sources

Internal to the folder: lore/tone-tables (citation survey and survival table entries), lore/carding (Sisken's card method), lore/sisken and lore/after (shoebox counts) (read 2026-09-13) — this page is itself a reconstruction and should be read as one.