# The offset, and what it moves

The folder's Gregorian dates are ordinals plus a chosen constant, and the constant is chosen by citation, not evidence. Every year on a lore/ page is a meet number converted through an offset that two reconstructions disagree about — and the folder papers over the disagreement by always printing the same year.

## The numbering habit

The Fallow meets were annual from the early eighties and were numbered ordinally — "the Ninth", "the Fourteenth" — and never dated in speech. The habit is so consistent that converting meet numbers to years is a live problem, not arithmetic. The tradition simply did not do the conversion; pinning an ordinal to a year now requires triangulating incidental references: a piece of equipment mentioned at the meet, a Ledger issue that appeared shortly after, even the weather.

The current reconstruction has two candidate offsets differing by one year. The disagreement propagates: every event dated "at the Eleventh" is uncertain by a year depending on which reconstruction you accept. The 1997 Armistice and "the last meet shortly after 1997" sit on the same wobble.

## What one year moves

The single heaviest consequence is biographical. Sisken not arriving is "the meet he did not arrive at" — 1994 or 1995 depending on the offset. The folder writes 1994 throughout because lore/index writes 1994, and that is a convention, not a finding. Which year you pick changes which Ledger issues are his last possible contributions: under one offset the late-1994 issues are still live candidates, under the other they are not, and anything filed against them inherits the wobble.

Note the asymmetry that makes absences the datable thing: the tradition has no word for attending and a rich one for "not arriving". There were no attendance lists by construction, so absences are recorded only for people whose presence was assumed — which is exactly why Sisken's non-arrival survived at all, and exactly why it arrives as an ordinal with no year attached.

## How to cite a Loop date

Give the ordinal first, the derived year second, marked as derived: "at the Eleventh [199x, offset-dependent]". The brackets are not decoration; they tell the next reader that the year is this folder's arithmetic, not the tradition's record. A citation that prints only "1991" silently asserts that one offset was evidence when it was a choice, and forces the reader to rediscover the wobble downstream. Carded the Sisken way:

- CLAIM: the meets ran annually from the early eighties, numbered ordinally.
  CONDITION: attested; both reconstructions accept it.
- CLAIM: offset = the one that makes the Eleventh 1991 (or 1992).
  CONDITION: chosen, not seen. Both candidates survive triangulation so far; neither has been eliminated.
- CLAIM: Sisken did not arrive in 1994.
  CONDITION: ledger-true only under the index's chosen offset. Carded as "the meet he did not arrive at", the ordinal is attested; the year is derived and disputed.

## The archivist's point

A culture that refuses to date its own events still produces dates once outsiders start citing it. The dates are the outsiders' artifact. Pretending they are the tradition's is how a convention hardens into a fact — the same move the folder's own weak-points list flags in the 1986 order, where a filing decision has begun to read as a finding. The offset should stay visible as an offset every time a year is printed, or the folder will have done to its calendar what the 1993 summary did to the tone tables: replaced an ordinal tradition with a clean-looking digest nobody carded.

## See also

[[lore/the-1993-summary]], [[lore/the-meets]], [[lore/sisken]], [[lore/index]], [[lore/tone-tables]].

## Sources

Internal to the folder: [[lore/index]] (1994 convention, weak-points list), [[lore/the-meets]] (meet numbering, the two offsets), [[lore/sisken]] (the non-arrival) (read 2026-09-13) — this page is itself a reconstruction and should be read as one.