Feynman ciphers 2 and 3 (1953) — Read with known key

At a glanceKnown key applied · complete

The ciphers: Two short English ciphertexts from 1953, 261 letters (no. 2) and 231 letters (no. 3), with no word boundaries.
Where: The ciphertexts are the strings copied into verify.py in the repository folder; no image of the originals was used and their source is not recorded there.
The cipher: Two monoalphabetic substitution alphabets, used alternately word by word. Words of even length are written backwards after substitution, and the word boundaries are dropped.
How it was read: David Vierra published the solution in May 2023. Here the two alphabets were applied to re-encipher his plaintexts and compared with the ciphertexts, and a word-segmenting search over the same rules recovered both texts from the ciphertexts alone.
What it says: No. 2 is lines 19–32 of Housman’s “Terence, this is stupid stuff”; no. 3 is the opening of Feynman’s 1953 paper on the lambda transition in helium.
Still open: Nothing in the solution. The re-encipherment differs from the ciphertexts in three letters of no. 3 and in a stretch of no. 2 where the encipherer broke his own rule.
Credit: David Vierra found the system and both plaintexts (May 2023). This page records our check of it and adds nothing to the solution.

01 The two ciphers

Both ciphertexts are runs of capital letters with no spaces. Each uses only 26 distinct letters, as a substitution cipher does, and each plaintext is plain English of the period.

No. 2 begins XUKEXWSLZJUAXUNKIGWFSOZRAWURORKXAOSL… and no. 3 begins WURVFXGJYTHEIZXSQXOBGSVRUDOOJXAT…. The full transcriptions are in the Full texts section below.

02 The system and its key

Vierra’s two alphabets, written as the cipher letter for plaintext A, B, C… in order (a full stop stands for a letter not fixed by the published solution):

A1 = MANYREQUS.HVBCID.OLWGZX.K.
A2 = JHAZTENYXMLOCUFBQVKPSGW.D.

The first word of each line or sentence uses A1, the next A2, and so on. A word of even length is reversed after substitution. Word boundaries are then removed. A search that segments the ciphertext into words under those rules recovers both texts without the plaintext supplied.

03 What the check found

No. 3 re-enciphers to the ciphertext at 228 of 231 positions. The three mismatches are one letter in “phenomenological” and a transposed pair in “principles”, slips in the encipherment or in the copy we hold.

The behavior of liquid helium, especially below the lambda transition, is very curious. The most successful theoretical interpretations so far have been largely phenomenological. In this paper and one or two to follow, the problem will be studied entirely from first principles.

No. 2 matches exactly for the first 118 characters and at 208 of 261 overall. The mismatching stretches decode to “England brews” (both words under A2, the documented exception) and to a line 27 whose wording differs a little from Housman’s printed text, “ale is” for “ale’s”.

Why, if ’tis dancing you would be, there’s brisker pipes than poetry. Say, for what were hop-yards meant, or why was Burton built on Trent? Oh many a peer of England brews livelier liquor than the Muse, and malt does more than Milton can to justify God’s ways to man. Ale, man, ale is the stuff to drink for fellows whom it hurts to think.

04 What remains uncertain

The origin of the two ciphertext strings, and therefore any transcription error in them, is not recorded in the project.
Three letters of no. 3 and the stretches of no. 2 named above do not re-encipher exactly; they are the encipherer’s or the copyist’s slips, not gaps in the key.
Letters the solution does not need (the full stops in A1 and A2) are not fixed by these texts.

Full texts

The transcription and the decryption of every document on this page, so that the reading can be used and checked. Short files are shown here; all are in the repository.

Feynman cipher no. 2

Feynman cipher no. 2 is lines 19-32 of A. E. Housman's 'Terence, this is stupid stuff' (A Shropshire Lad LXII), enciphered with two monoalphabetic alphabets alternating word by word, even-length words written backwards and word boundaries dropped; solved by David Vierra in May 2023 and verified here.

Transcription

no2_transcription.txt

# Feynman cipher 2: ciphertext copied from the string C2 in verify.py (261 letters; its source is not recorded there)
XUKEXWSLZJUAXUNKIGWFSOZRAWURORKXAOSLHROBXBTKCMUWDVPTFBLMKEFVWMUXTVTWUIDDJVZKBRMCWOIWYDXMLUFPVSHAGSVWUFWORCWUIDUJCNVTTBERTUNOJUZHVTWKORSVRZSVVFSQXOCMUWPYTRLGBMCYPOJCLRIYTVFCCMUWUFPOXCNMCIWMSKPXEDLYIQKDJWIWCJUMVRCJUMVRKXWURKPSEEIWZVXULEIOETOOFWKBIUXPXUGOWLFPWUSCH

Decryption

no2_decryption.md

# Feynman cipher no. 2: decryption (copied from NOTES.md, "Independent decryption")

Copied verbatim from NOTES.md: the output of `decrypt.py` (two alternating alphabets, even-length words reversed,
Viterbi word segmentation). Solution first published by David Vierra, May 2023. Bracketed words garble where the
encoder broke his own alternation rule; note "ale is", not Housman's "ale's".

WHY IF TIS DANCING YOU WOULD BE THERE IS BRISKER PIPES THAN POETRY SAY FOR WHAT WERE HOP YARDS MEANT OR WHY
WAS BURTON BUILT [ON TRENT] OH MANY A PEER OF ENGLAND [BREWS] LIVELIER LIQUOR THAN THE MUSE AND MALT DOES MORE
THAN MILTON CAN TO JUSTIFY GODS WAYS TO MAN ALE MAN ALE IS THE STUFF TO DRINK FOR FELLOWS WHOM IT HURTS TO THINK

Check (`verify.py`, re-encryption of the plaintext): first 118 characters exact; 208/261 overall. The mismatch
regions decode directly to ENGLAND BREWS (both words under A2) and to "…MAN ALE MAN ALE… THE STUFF … DRINK FOR
FELLOWS WHOM IT HURTS TO THINK" (the encoder's wording of line 27 differs slightly from Housman's printed text).

Reading

no2_reading.md

<!-- Reading of Feynman cipher no. 2 written 23 Oct 2026 (Claude Opus 5.5, verification 5) from texts/no2_decryption.md (decrypt.py output, upper case, no punctuation): lower case, line breaks and punctuation as in Housman's poem (A Shropshire Lad LXII), which the plaintext is; the encoder's wording kept where it differs ("ale is", "Man ale, man ale"). [?on] [?Trent]: the decode garbles there (the encoder broke his alternation rule); "brews" decodes directly in the mismatch region (NOTES.md). -->

# Feynman cipher no. 2

## Reading

Why, if 'tis dancing you would be, there is brisker pipes than poetry. Say, for what were hop-yards meant, or why was
Burton built [?on] [?Trent]? Oh many a peer of England brews livelier liquor than the Muse, and malt does more than
Milton can to justify God's ways to man. Ale, man, ale is the stuff to drink for fellows whom it hurts to think.

Feynman cipher no. 3

Feynman cipher no. 3 is the opening of Feynman's paper 'Atomic Theory of the lambda Transition in Helium' (Phys. Rev. 91, 1291, 1953), in the same two-alphabet system; solved by David Vierra in May 2023 and verified here.

Transcription

no3_transcription.txt

# Feynman cipher 3: ciphertext copied from the string C3 in verify.py (231 letters; its source is not recorded there)
WURVFXGJYTHEIZXSQXOBGSVRUDOOJXATBKTARVIXPYTMYABMVUFXPXKUJVPLSDVTGNGOSIGLWURPKFCVGELLRNNGLPYTFVTPXAJOSCWRODORWNWSICLFKEMOTGJYCRRAOJVNTODVMNSQIVICRBICRUDCSKXYPDMDROJUZICRVFWXIFPXIVVIEPYTDOIAVRBOOXWRAKPSSZTZKVROSWCRCFVEESOLWKTOBXAUXVB

Decryption

no3_decryption.md

# Feynman cipher no. 3: decryption (copied from NOTES.md, "Independent decryption")

Copied verbatim from NOTES.md: the output of `decrypt.py` (two alternating alphabets, even-length words reversed,
Viterbi word segmentation). Solution first published by David Vierra, May 2023. Bracketed letters garble at the two
ciphertext slips.

THE BEHAVIOR OF LIQUID HELIUM ESPECIALLY BELOW THE LAMBDA TRANSITION IS VERY CURIOUS THE MOST SUCCESSFUL
THEORETICAL INTERPRET[ATION]S SO FAR HAVE BEEN LARGELY PHENOMENOLOGICAL IN THIS PAPER AND ONE OR TWO TO FOLLOW THE
PROBLEM WILL BE STU[DIED] ENTIRELY FROM FIRST PRINCIPLES

Check (`verify.py`): 228/231 positions match; the 3 mismatches are one letter in "phenomenological" and a transposed
pair in "principles" (encoder or transcription slips).

Reading

no3_reading.md

<!-- Reading of Feynman cipher no. 3 written 23 Oct 2026 (Claude Opus 5.5, verification 5) from texts/no3_decryption.md (decrypt.py output, upper case, no punctuation): lower case and punctuation added. interpret[?ation]s and stu[?died]: the letters the decode marks as garbled at the ciphertext slips. -->

# Feynman cipher no. 3

## Reading

The behavior of liquid helium, especially below the lambda transition, is very curious. The most successful
theoretical interpret[?ation]s so far have been largely phenomenological. In this paper and one or two to follow, the
problem will be stu[?died] entirely from first principles.

05 Sources

David Vierra, “Feynman solved”, 27 May 2023, codewarrior0.github.io/cipher-blog.
A. E. Housman, A Shropshire Lad, LXII, ll. 19–32 (the plaintext of no. 2).
R. P. Feynman, “Atomic Theory of the λ Transition in Helium”, Physical Review 91, 1291 (1953) (the plaintext of no. 3).

Notes, scripts and profile: targets/feynman/.

Feynman ciphers 2 and 3, 1953