Experiment 202
F6.21 (issue #78). SOE's 'indecipherables' came from a misnumbered transposition key or a column skipped or doubled (Marks, Between Silk and Cyanide, pp. 11, 206, 317); Sanborn says his work is 'rife with mistakes on purpose'. Around every keyword columnar the record closed, the reading order of the columns is changed by one slip: one column read at a wrong place (what a single wrong number in the key's numbering does: with ties broken left to right, a column numbered n' instead of n is read at n's neighbour's place), or two columns swapped. Bases: 0110's single keyword columnar and 0112's double (the same word twice; the slip in either pass), 0164's interrupted-key columnar (F6.16) and trapezoid (F6.17), 0165's letters reshuffled inside each row (F6.18), 0190's poem code (F11.2; the slip in either pass). Then a repeating key 1-26 on the standard or KRYPTOS table after (A) or before (B), 8 classes, ten scrambled copies, four planted slips in the same run.
What the method does with the 24 known letters.
Agents in the war often made one mistake when they copied out a transpositiona grid Writing the message into rows of some width and reading it out in another order, column by column for instance, which shuffles the letters without changing them.: they numbered the keythe secret word The word or string a cipher needs besides the method. word wrongly and read one column at the wrong time, or read two columns in each other's place. Sanborn says he made mistakes on purpose. We take every column shuffle the project has already ruled outdead Said of a period, grid or method when the known letters contradict it: no secret word of that kind can produce the carved letters. and try it again with each such single mistake, followed or preceded by a short repeating key word on the sculpture's tablethe table The right-hand half of the sculpture, 26 rows of the alphabet each shifted one place further, starting from the word KRYPTOS..
Words in the margin are defined on the glossary page.
The carved letters go in at the top; the reading comes out at the bottom.
Wartime agents' typical slip in a transposition: one column read at the wrong time (a wrong number in the key's numbering), or two columns swapped. Sanborn says his mistakes are on purpose.
Every such slip around every keyword columnar the record closed (this part: one word, the same word twice, interrupted key; the other three are 0203), each slipped route and its inverse, then or before a repeating key of 1-26 on the sculpture's tables, K4 against ten scrambled copies and a planted slip per family.
In each family the shortest key that fits K4 is no shorter than on the copies, and K4 leads every copy in no more cells than a copy does. Planted slips: F6/1 found, F6.16 found, F6.18 NOT found, F11.2 NOT found.
What this experiment was built on, from the lineage page: 6 experiments before it, back to the root of its line. Its number is in bold.
This experiment produced no candidate: the search ruled everything out before a full text was built. Its product is the bounded negative under Result and bounds below.
Redo the method on a few letters with pencil and paper.
The same steps run by the code on the real letters.
key word SHADOW: columns read in the order [3, 4, 2, 5, 1, 6]
one slip, the third column read one step early: [3, 2, 4, 5, 1, 6]
place known lands at carved
22 E 36 T
26 N 21 W
34 T 38 T
64 B 43 S
74 K 29 R
Family, key space, the constraint under test, and the clues it reads.
C1(a), C5, C10 ('mistakes on purpose'; misspellings as security checks in SOE practice, Marks p. 16). Source: reports/marks-between-silk-and-cyanide.md. Model: c = T(sigma'(p)) or sigma'(T(p)), sigma' a keyword columnar (or two passes) whose column reading order differs from the keyword's by one move of one column (g(g-1) orders) or one swap of two non-adjacent columns ((g-1)(g-2)/2; adjacent swaps are moves), in one pass. A skipped or doubled column alone changes the length (97 would need a message of 97 plus or minus a column, or two compensating slips): not searched. Not included: Myszkowski, AMSCO, keyed blanks, triangle and disrupted columnars (not on the item's list). Each slipped route is tested exactly as 0110 tested its routes: periods 1-26, both orders, 8 classes, by slot agreement of the 24 known letters, forward and inverse. Statistic (0134's, fixed): per base family, the shortest key length consistent on K4 against the copies, and the cells (order, length) where K4 exceeds every copy, against each copy's own count. Routes are counted per slip, not deduplicated across words (copies counted the same way).
Written before the run; the result decides it.
For each base family, nothing consistent below key length 20 beyond the copies' rate, and K4 above every copy in no more cells than a copy is. Each planted slip is found at its own setting. Zero candidates.
The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.
F6/1 keyword columnar: 67,750,160 routes, shortest live A 19 B 24 (copies {'A': 20, 'B': 18}), K4 above every copy 1 (null max 3); F6/1 double columnar: 135,500,320 routes, shortest live A 20 B 18 (copies {'A': 18, 'B': 17}), K4 above every copy 0 (null max 3); F6.16 interrupted: 51,899,058 routes, shortest live A 21 B 24 (copies {'A': 20, 'B': 24}), K4 above every copy 1 (null max 2); F6.17 trapezoid: 0 routes, shortest live A None B None (copies None), K4 above every copy 0 (null max 0); F6.18 row reshuffle: 0 routes, shortest live A None B None (copies None), K4 above every copy 0 (null max 0); F11.2 poem code: 0 routes, shortest live A None B None (copies None), K4 above every copy 0 (null max 0). Plants: F6/1 keyword columnar found True; F6.16 interrupted found True; F6.18 row reshuffle found False; F11.2 poem code found False
What was searched, as the harness recorded it.
How this experiment sits among the others is on the lineage page; its family's open hypotheses are on the families page.