Ancient Cryptography

December 08, 2022, 04:51:50 AM
Welcome, Guest. Please login or register.


Ancient Cryptography » General » Other Mysteries » Feynman ciphers

Author Topic: Feynman ciphers  (Read 13043 times)

0 Members and 1 Guest are viewing this topic.

tonybaloney

  • Full Member
  • Posts: 173
Feynman ciphers
« on: June 26, 2008, 02:03:27 PM »
Quote
Richard Feynman, physicist and pioneer in the field of nanotechnology, received three encoded messages from a scientist at Los Alamos and shared them with his graduate students when he couldn't decipher them himself. Currently, they are posted on a puzzle site. Cryptanalysts have only managed to decipher the first message, which turned out to be the opening lines of Chaucer's "Canterbury Tales" written in Middle English.

When I was a graduate student at Caltech, Professor Feynman showed me three samples of code that he had been challenged with by a fellow scientist at Los Alamos and which he had not been able to crack.  I also was unable to crack them.  I posted them to Usenet and Jack C. Morrison of JPL cracked the first one.  It is a simple transposition cipher: split the text into 5-column pieces, then read from lower right upward.  What results are the opening lines of Chaucer's Canterbury Tales in Middle English.

1. Easier
MEOTAIHSIBRTEWDGLGKNLANEA
INOEEPEYSTNPEUOOEHRONLTIR
OSDHEOTNPHGAAETOHSZOTTENT
KEPADLYPHEODOWCFORRRNLCUE
EEEOPGMRLHNNDFTOENEALKEHH
EATTHNMESCNSHIRAETDAHLHEM
TETRFSWEDOEOENEGFHETAEDGH
RLNNGOAAEOCMTURRSLTDIDORE
HNHEHNAYVTIERHEENECTRNVIO
UOEHOTRNWSAYIFSNSHOEMRTRR
EUAUUHOHOOHCDCHTEEISEVRLS
KLIHIIAPCHRHSIHPSNWTOIISI
SHHNWEMTIEYAFELNRENLEERYI
PHBEROTEVPHNTYATIERTIHEEA
WTWVHTASETHHSDNGEIEAYNHHH
NNHTW

2. Harder
XUKEXWSLZJUAXUNKIGWFSOZRAWURO
RKXAOSLHROBXBTKCMUWDVPTFBLMKE
FVWMUXTVTWUIDDJVZKBRMCWOIWYDX
MLUFPVSHAGSVWUFWORCWUIDUJCNVT
TBERTUNOJUZHVTWKORSVRZSVVFSQX
OCMUWPYTRLGBMCYPOJCLRIYTVFCCM
UWUFPOXCNMCIWMSKPXEDLYIQKDJWI
WCJUMVRCJUMVRKXWURKPSEEIWZVXU
LEIOETOOFWKBIUXPXUGOWLFPWUSCH

3. New Message
WURVFXGJYTHEIZXSQXOBGSV
RUDOOJXATBKTARVIXPYTMYA
BMVUFXPXKUJVPLSDVTGNGOS
IGLWURPKFCVGELLRNNGLPYT
FVTPXAJOSCWRODORWNWSICL
FKEMOTGJYCRRAOJVNTODVMN
SQIVICRBICRUDCSKXYPDMDR
OJUZICRVFWXIFPXIVVIEPYT
DOIAVRBOOXWRAKPSZXTZKVR
OSWCRCFVEESOLWKTOBXAUXV
B

Chris Cole

(I've pasted the above straight from another web page - I'm sure they won't mind, especially if we solve it for them?!)

This is the solution to No.1

Whan that Aprill, with his shoures soote
The droghte of March hath perced to the roote
And bathed every veyne in swich licour,
Of which vertu engendred is the flour;
Whan Zephirus eek with his sweete breeth
Inspired hath in every holt and heeth
The tendre croppes, and the yonge sonne
Hath in the Ram his halfe cours yronne,
And smale foweles maken melodye,
That slepen al the nyght with open eye-
(So priketh hem Nature in hir corages);
Thanne longen folk to goon on pilgrimag

« Last Edit: March 14, 2009, 02:54:10 PM by Aaron »

tonybaloney

  • Full Member
  • Posts: 173
Re: Feynman ciphers
« Reply #1 on: July 19, 2008, 11:22:32 AM »
Anyone looked at this cipher?

2. Harder
XUKEXWSLZJUAXUNKIGWFSOZRAWURORKXAOSLHROBXBTKCMUWDVPTFBLMKE
FVWMUXTVTWUIDDJVZKBRMCWOIWYDXMLUFPVSHAGSVWUFWORCWUIDUJCNVT
TBERTUNOJUZHVTWKORSVRZSVVFSQXOCMUWPYTRLGBMCYPOJCLRIYTVFCCM
UW
UFPOXCNMCIWMSKPXEDLYIQKDJWIWCJUMVRCJUMVRKXWURKPSEEIWZVXU
LEIOETOOFWKBIUXPXUGOWLFPWUSCH

3. New Message
WURVFXGJYTHEIZXSQXOBGSVRUDOOJXATBKTARVIXPYTMYABMVUFXPXKUJV
PLSDVTGNGOSIGLWURPKFCVGELLRNNGLPYTFVTPXAJOSCWRODORWNWSICLF
KEMOTGJYCRRAOJVNTODVMNSQIVICRBICRUDCSKXYPDMDROJUZICRVFWXIF
PXIVVIEPYTDOIAVRBOOXWRAKPSZXTZKVROSWCRCFVEESOLWKTOBXAUXVB


Of the two unsolved messages numbered 2 and 3 -
CJUMVR appears twice consecutively in message 2
WUID x 2 in message 2
CMUW x 3 in message 2
OJUZ x 2 once in each message
SQXO x 2 once in each message
WUR x 4 twice in each message
PYT x 4 once in message 2 & three times in message 3
ICR x 3 all in message 3


RKX; OBX; VTW; UFP; GSV; WUF; SVR; KPS; XPX; RVF; GJY; RUD are the other recurring trigrams each appearing only twice

WU x 9; YT x 6; VR x 6; VT x 5; RO x 5; PX x 5; CR x 5; OJ x 5; JU x 5; XA x 4; IC x 4; UF x 4; FP x 4; UR x 4; OR x 4; OS x 4; XU x 4; MU x 4; SV x 4; IW x 4; PY x 4; are the most frequent digrams

The quantity & variety of these ngrams corresponds to what I would expect to find if each word of the message was enciphered in the same manner but not each ngram ie.

THE WEATHER IS MISTY THE ...
123 1234567 12 12345 123
UJH XGDXMKY JU NKVXD UJH

or with a keyword starting again with each new word –
 
THE WEATHER IS MISTY THE ...
KEY KEYWORD KE KEYWO KEY
DLC GIYPVVU SW WMQPM DLC


note ‘THE’ in ‘weaTHEr’ does not encipher the same as ‘THE’ nor does the ‘IS’ in ‘mISty’ repeat the encipherment of ‘IS’

hence the most frequent (but by no means all) of these repeated trigrams & tetragrams will represent whole words

in message 2 we find the sequence ICRBICR if ICR is a word the B trapped betwixt is probably ‘A’ or ‘I’
also in message 2 we find the sequence SQXOBGSV now SQXO & GSV are both repeating ngrams, the B is again trapped

If B represents the encipherment of the letter A are all initial letters of words enciphered by simply advancing one letter in the alphabet?

WUR is our most frequent trigram – retreating one letter gives us a 3 letter word beginning with V (not very probable even if it’s in middle English)

but if WUR represents ‘the’ the initial letter has advanced by 3 places –

is it possible the initial letter of each word is enciphered by advancing the length of the word?

in the first of the above messages we also find the sequence CJUMVRKXWUR where the KX is trapped (very probably a 2 letter word), reversing 2 letters it will begin with an I - KX appears 3 times in the messages - 

the most common tetragram CMUW, if a 4 letter word will begin with a Y

the sequence SQXOCMUWPYT becomes O???/Y???/M?? -

Thats about as far as i've got at the moment - it looks a promising line of attack - how the 2nd, 3rd etc letters of each word are enciphered is possibly some combination of the word length and the previous letter or its encipherment.

Tony


tonybaloney

  • Full Member
  • Posts: 173
Re: Feynman ciphers
« Reply #2 on: August 17, 2008, 06:39:23 AM »
CJUMVRCJUMVR

Considering what the above repeating sequence could represent I thought it more likely to be a combination of 2 words – using the previous possibility that the initial letter of the word is advanced by the number of letters in the word we get -

W ? ? ? ? ?/W ? ? ? ? ?
X ? ? ? ?/Q/X ? ? ? ?/Q
Y ? ? ?/T ?/Y ? ? ?/T ?
Z ? ?/J ? ?/Z ? ?/J ? ?
A ?/R ? ? ?/A ?/R ? ? ?
B/? ? ? ? ?/B/? ? ? ? ?
C J U M V R/C J U M V R


looking at the above possibilities the most likely would seem to be as highlighted and stand for ‘YEAR TO YEAR’ – assuming ‘FROM’ to be the most likely word to precede this I was pleased to find the F falls above the J –

...... FROM YEAR TO YEAR TO
------ G--- Z---    Z---       T--
------ H--- A--- T- A--- T- I- U--
------ I--- B--- U- B    U- J- V--
...QKD JWIW CJUM VR CJUM VR KX WUR KPSEE...

there are 4 repeating tetragrams in the 2 ciphers –

SQXO, CMUW, OJUZ & WUID

treating these the same as ‘FROM’ [advancing the word length on the first letter and using the caesar alphabets +5, +20, +10 (all multiples of 5!! – coincidence!?) on the remainder] they become –

OLDE, YHAM, KEAP & SPOT

looking up YHAM & KEAP in google they appear to be Middle English variations of  ‘them’ & ‘keep’

going through every letter of the cipher in turn assuming it to be the start of a 4 letter word the only other recognizable/pronouncable words thrown up are – ASCI, SAYE, WRAH, SHAN, ROCK, PRAY, ERET, BRUH, NOAD, ROCA, NUYL, RAIS, YISS, ERSI, SHYA, AYRO, TRAH, KACA, WOHA, CIME, HRAH, SMUT, KNOW, GNIL, CEES, RIZE, LYST, TOFA, GOUR (I may have missed a few) – the other 400+ tetragrams producing gibberish.

The fact that the 4 repeating tetragrams all produce words (3 of them common words) looks very promising.

unfortunately CJMU becomes YEAC not YEAR so maybe CJMUVR is a 6 letter word after all –
eg. ‘…. received a letter from Walter. Walter is ….’ or something vaguely along those lines.


All the above may look too simplistic for an unsolved cipher – but the first solved one was a simple transposition it’s only difficulty being to recognize the original plaintext (Chaucer’s middle English) – the composer would see this simple method as a more difficult system – but I think the only really difficult bit is not knowing if its middle English, old English or what dialect and recognizing the words when you find them.

There could be some buried treasure here if X marks the SPOT.

Help - anyone tell me what the following says – the capital P stands for ‘thorn/th’, the 3 for ‘gh’ (I think) – I just want to know what SPOT means in this sentence –

‘Pe spot of hor is Pe couaytise of Pe wordle Pet ne ssel na3t by ine his herte PeP wyle queme god.’

Tony

tonybaloney

  • Full Member
  • Posts: 173
Re: Feynman ciphers
« Reply #3 on: August 20, 2008, 10:10:51 AM »
Comparing WUR(4 off), PYT (4) & ICR(3) advancing word length on first and finding possible vowel/consonant combinations the only possibility seems to be TAL, MEN & FIL respectively – MEN & FIL appear in Chaucers works, TAL I can’t find anywhere (is it a variation of TIL??)
If correct 3 letter words are +3 +20 +6 (there’s that +20 again!)

1   2   3   4
O   L   D   E
+4 +5  +20 +10
S   Q   X   O

1   2   3
M   E   N
+3 +20 +6
P   Y   T

looking for some method in all this madness – the final letter of each word could be enciphered by summing (1+2+3+4=10), (1+2+3=6)
so a 2 letter word would be +2 +3

1  2
T  O
+2  +3
V  R


(VR occurs 6 times in the 2 ciphers)

other 2 letter words produced by this method – UP, IF, MO, AT, PI, IM, RE, AS, OF, IS but these could occur by chance

returning to CJUMVRCJUMVR which I guessed was YEAR/TO/YEAR./… but rejected because it came out YEAC/TO/YEAC/… I noticed looking through Chaucers works a lot of words beginning with Y (ycome, ytaught, yfalle etc.) – I can’t find YEAC anywhere but EAC (and, also) seems to be quite common in some old texts – I found Y EAC (by googling it) in some old English text – is it possible to run these 2 words together??

That’s it so far and it’s beginning to feel as if I’m building a castle on quicksand!!!

Tony

Aaron

  • Administrator
  • Sr. Member
  • Posts: 326
  • Gender: Male
Re: Feynman ciphers
« Reply #4 on: September 07, 2008, 07:44:48 PM »
Good luck! It's hard enough deciphering clean english, never mind olde english. At least it sounds like you have a foothold. :)

tonybaloney

  • Full Member
  • Posts: 173
Re: Feynman ciphers
« Reply #5 on: September 08, 2008, 04:55:35 AM »
Perhaps a 'toehold' would be more accurate?!

tonybaloney

  • Full Member
  • Posts: 173
Re: Feynman ciphers
« Reply #6 on: March 14, 2009, 10:08:32 AM »
Anyone got any more info on this cipher - such as who sent it, when, where etc -

every piece of 'olde English I look at seems to be different' - different dialects I suppose - what are common words in one are not in another -
Think it would be easy to solve if we could narrow to a specific time frame and place.

Aaron

  • Administrator
  • Sr. Member
  • Posts: 326
  • Gender: Male
Re: Feynman ciphers
« Reply #7 on: March 14, 2009, 02:53:22 PM »
The only thing I can figure out is that by "Los Alamos scientist", Feynman meant one of the people he worked with on the Manhattan Project. I wish I could find the original Usenet post, but at least I confirmed that Jack C Morrison is a real person (Jack.C.Morrison@jpl.nasa.gov).