Experiment 71

“Sanborn's own table, slightly different from the carved one: every alphabet one step away from KRYPTOS”

Sanborn said the table he worked from was slightly different from the one on the sculpture. Every earlier sweep used the carved KRYPTOS alphabet exactly or alphabets built from words. Here the alphabet is the carved one with two letters swapped or one letter moved, 925 variants, tested with a repeating secret word and with the self-feeding autokey.

revertedstatus
experiment 5parents
2 mintime, all phases
0candidates
0/24known letters
1

The experiment method

What the method does with the 24 known letters.

Sanborn has said that the tablethe table The right-hand half of the sculpture, 26 rows of the alphabet each shifted one place further, starting from the word KRYPTOS. he enciphered with was slightly different from the one carved on the sculpture. If so, every test that used the carved alphabet exactly could have missed by a hair. We take the carved alphabet KRYPTOS-A-B-C... and make every version with two letters swapped or one letter moved, almost a thousand of them, and run the two simplest ciphers on each: a repeating secret wordthe secret word The word or string a cipher needs besides the method., and the autokeyan autokey cipher A cipher that uses the message itself, or the scrambled text itself, as the secret word, each letter enciphered with the letter a fixed distance before it. that feeds on earlier carved letters. Scrambled coppers get the same treatment.

Words in the margin are defined on the glossary page.

1.1

The method at work

The carved letters go in at the top; the reading comes out at the bottom.

OBKRUOXOGHULBSOLIFBBWEASTNORTHEASTOTWTQSJQSSEKZZWATJKLUDIAWINFBBERLINCLOCKWGDKZXTJCDIGKUHUAUEKCARIN: THE COPPER, ALL 97 LETTERS, THE 24 KNOWN PLACES IN ACCENTthe sculpture's alphabetKRYPTOSABCDEFGHIJLMNQUVWXZA and B swappedKRYPTOSBACDEFGHIJLMNQUVWXZRWon the swapped alphabet places 22 and 29 share slot 0 of a 7-letter word but force R and Wall 925 alphabets one swap or one move from KRYPTOSkey length 26, consistent0548the copper: 21autokey, 2 checks02924the copper: 1ten shuffled coppersthe copperOUT: NO READING; NEAR-KRYPTOS ALPHABETS BEHAVE LIKE CHANCEEASTNORTHEASTBERLINCLOCK
the clue

Sanborn has said the table on the sculpture is slightly different from the one he used. This tries every alphabet one swap or one move away from KRYPTOSABCDEFGHIJLMNQUVWXZ.

the check

With A and B swapped, places 22 and 29 force R and W in the same slot: dead for a 7-letter word.

the bound

925 alphabets, repeating keys of every length to 52 and ciphertext autokey at every distance: survivors at the shuffled rate, and no word among 16,671 forced fragments.

2

Best candidate against the known letters

This experiment produced no candidate: the search ruled everything out before a full text was built. Its product is the bounded negative in section 7.

3

Walkthrough, by hand

Redo the method on a few letters with pencil and paper.

  1. Swap A and B in the carved alphabet: KRYPTOSBACDEFGHIJLMNQUVWXZ, counted from K = 0.
  2. Use it for both the message and the carved letters, Vigenere, with a 7-letter secret word.
  3. Place 22: known E is 11, carved F is 12; the key is 1, R, in slot 0.
  4. Place 29, also slot 0: known T is 4, carved R is 1; the key is 1 - 4 = -3, which is 23, W.
  5. R and W disagree, so this alphabet with a 7-letter word is dead.
  6. Repeat for every variant, pairing, word length and lag.
the worksheet: the first steps by hand, on squared paper
ONE SWAP FROM KRYPTOS: A AND B EXCHANGED, USED FOR MESSAGE AND COPPER, VIGENEREthe sculpture's alphabet with A and B swappedK0R1Y2P3T4O5S6B7A8C9D10E11F12G13H14I15J16L17M18N19Q20U21V22W23X24Z25placeknowncarvedslotkey22E=11F=1201 R29T=4R=1023 WKey = carved - known on thealphabet. Same slot, differentkeys: dead for a 7-letter word.
4

Worked example, generated

The same steps run by the code on the real letters.

alphabet KRYPTOSBACDEFGHIJLMNQUVWXZ (A and B swapped), message and carved letters both on it, Vigenere, 7-letter key

place  known  carved  key  slot
   22      E       F    R     0
   29      T       R    W     0
   23      A       L    C     1
   30      H       N    O     1
5

Hypothesis, for a cryptanalyst

Family, key space, the constraint under test, and the clues it reads.

Family 1 re-entered with a new key source, C27(c): the 2010 worksheets show a tableau 'slightly different from the one on Kryptos' (research_notes/sanborn_charts/SOURCES.md; 02-sanborn-said.md). Alphabets: the KRYPTOS alphabet with one transposition of two letters (325) or one letter moved to another place (650 moves, of which 50 repeat an adjacent swap, so 600 new), 925 variants V. Pairings: V as both plain and cipher alphabet (the tableau's header and rows from one alphabet), plain standard with cipher V, and plain V with cipher standard; Vigenere or Beaufort. Systems: (1) periodic key, periods 1-52, exact crib consistency; (2) ciphertext autokey with a constant shift, lags 1-96, key letter read on V or on the standard alphabet. Survivors of (1) at periods up to 29 have their forced fragments matched against the vocabulary and K1-K3 windows. Ten shuffled coppers through the same sweep. Clue entries: C27(c), C6(a), C8(a).

6

Prediction

Written before the run; the result decides it.

No variant makes a period of 24 or less consistent under any pairing and mode; periods 25-26 survive at the shuffled-copper rate; no fragment at periods up to 29 matches a word; the autokey has no survivor with three or more redundant checks. Zero candidates pass.

7

Method

  1. Build the 925 alphabets one step from KRYPTOS.
  2. For each pairing and mode, compute the key value each known letter forces.
  3. For each period up to 52, check the slots agree; record the survivors and their fragments.
  4. For each lag, check the autokey residuals agree.
  5. Repeat on ten shuffled coppers and compare counts.
  6. Match fragments at periods up to 29 against words.
8

Result and bounds

The harness's record: candidates, gates, loss, and the negative result this experiment leaves behind.

0candidates returned
0passed every gate
0/24known letters, best
–lowest loss
34champion before

925 of 925 variants. Periodic, consistent cases at P <= 26: K4 {'26': 21}; controls {'26': (0, 548)}. Autokey survivors by checks: {'1': (391, 38, 838), '2': (1, 0, 292), '3': (0, 0, 3), '4': (0, 0, 29)}. Word matches in 16671 fragments: 0.

What was searched, as the harness recorded it.

family
periodic key and ciphertext autokey on alphabets one swap or one move from KRYPTOS
variants
925
variants done
925
setups per variant
6
periods
1-52
lags
1-96
periodic consistent by period
26
k4
21
controls
137, 0, 548, 462, 21, 387, 33, 33, 0, 32
27
k4
5,550
controls
5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550
28
k4
5,550
controls
5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550
29
k4
5,550
controls
5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550, 5,550
30
k4
113
controls
38, 52, 710, 496, 32, 38, 0, 161, 34, 38
31
k4
2
controls
1, 0, 0, 0, 1, 0, 3, 0, 3, 2
51
k4
0
controls
0, 0, 13, 3, 0, 1, 3, 3, 0, 0
52
k4
21
controls
137, 0, 548, 462, 21, 387, 33, 33, 0, 32
autokey survivors by checks
1
k4
391
controls
838, 361, 80, 142, 284, 492, 92, 336, 84, 38
2
k4
1
controls
4, 11, 3, 0, 4, 4, 4, 292, 1, 2
3
k4
0
controls
0, 0, 0, 0, 0, 0, 0, 1, 1, 3
4
k4
0
controls
29, 0, 0, 0, 0, 0, 0, 0, 0, 0
fragments checked
16,671
word matches
none
controls done
925, 925, 925, 925, 925, 925, 925, 925, 925, 925

How this experiment sits among the others is on the lineage page; its family's open hypotheses are on the families page.

Rendered from the record on 2026-10-18 12:26 UTC; updated after every experiment.
view on GitHub