A coded telegram from 1937. Deciphering help needed.
Louis Jaffe
In 1937 my father, Louis I. Jaffe, an American newspaper editor, toured
central Europe. From Vienna he sent a telegram outlining his travel
plans to Lenoir Chambers, a trusted associate in the US. The telegram
is apparently in code.
Jaffe's trip may have been more than a vacation, and he may have been
concerned for his safety in that age of rising fascism, since he was
from a Jewish family background and had written anti-fascist
editorials.
I have a photocopy of the telegram and Chambers' handwritten
translation. A biography of my father is in progress, and the writer,
Prof. Alexander Leidholdt at Purdue University, has shown the telegram
to language experts, but no one has been able to determine what code or
language the telegram might be written in.
If anyone in this newsgroup can help, it will be greatly appreciated,
and I'm sure Dr. Leidholdt will duly credit you in the published work.
Here is the brief text of the cable:
AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
And here is how Chambers interpreted it:
Shall remain here until 14 and then go to Burda. Hope you and your
family are well. Inform friends of your having heard.
Thanks for any help--
Louis I. Jaffe, Jr.
Ken Shirriff
In article <32592D...@well.com>, Louis Jaffe <lo...@well.com> wrote:
>Here is the brief text of the cable:
>
> AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
>
>And here is how Chambers interpreted it:
>
> Shall remain here until 14 and then go to Burda. Hope you and your
>family are well. Inform friends of your having heard.
My guess would be that they just set up a code beforehand:
ABOZO = remain until
AMACO = will go to
DOEKA = Hope you and your family is well.
NUXIL = Inform friends of your having heard.
LOUIS = his name
The code would presumably have other code words that they didn't use.
It looks like he's using 5-letter code words for everything. This could
be some commercial code and both people had a code book. Or they could
have just made up their own code with a small number of phrases. One
reason to suspect this sort of code is that the telegram uses pronouncible
nonsense words, and each word corresponds to a fairly long phrase.
It seems unlikely that they're using some language for several reasons.
First, all 5-letter words is unlikely. Second, one of your language
experts should have recognized it. Finally, the message is much longer
than the telegram; it's unlikely that the language would have one word
meaning "inform friends of your having heard".
(Although DOEKA is apparently a word in Czechoslovakian, so this might be
a clue.)
Another possibility is that they used some simple encryption algorithm for the
message. This is unlikely because the telegram used pronouncible nonsense
words, rather than scrambled letters. Also, the letter patterns don't
look right; you wouldn't expect a K, X, and Z if you just scrambled it.
Finally, the telegram is too short compared to the message length.
Without more telegrams, or finding a code book in their possessions, I
can't see any way of proving that they used a code that assigned 5-letter
words to phrases, but I highly suspect this is what they used.
Ken Shirriff shir...@eng.sun.com http://www.cs.berkeley.edu/~shirriff
John Flanagan
They may have been using one of the commercial code books that were
available for commercial use at that time - in the dim and distant past
when I was in the Export business I remember many companies used to quote
on their letterheads which code system they used - the books we available
at our local reference library here in the UK - you may find copies
through one of your main libraries if they keep old books - they were as
big as the bible and very heavy
Maybe this is the answer to your problem
BUD
Internet address = b...@cix.compulink.co.uk
Mark McCutcheon
In article <53c8a5$p...@engnews1.Eng.Sun.COM> shir...@jaywalks.eng.sun.com
(Ken Shirriff) wrote:
| In article <32592D...@well.com|, Louis Jaffe <lo...@well.com| wrote:
|| Here is the brief text of the cable:
||
|| AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
|
| ABOZO = remain until
| AMACO = will go to
| DOEKA = Hope you and your family is well.
| NUXIL = Inform friends of your having heard.
| LOUIS = his name
|
| The code would presumably have other code words that they didn't use.
| It looks like he's using 5-letter code words for everything. This
| could be some commercial code and both people had a code book.
I'd almost bet real money on it ;-) It seems to me that early-day
telegraphy did use 5-letter groups from a codebook to represent common
phrases. It may even have been the Western Union Telegraph Code.
Unfortunately I can't find a reference at the moment to confirm it.
There's an old sci-fi story based on this: "BEROM". I think it's the
one written by John Berryman, first published in Astounding in 1951.
After long study, an alien visits earth with what it believes is a
fluent command of the language, but no-one here understands. BEROM is
apparently the code group for "greetings".
Mark
Will Ware
Mark McCutcheon (mjm...@cs.ubc.ca) wrote:
: It seems to me that early-day
: telegraphy did use 5-letter groups from a codebook to represent common
: phrases. It may even have been the Western Union Telegraph Code.
In the use of Morse code by amateur radio operators, there are at least
three groupings of abbreviations, and I think some of them date back to
telegraph days. There are numerical codes (73 = best wishes, 88 = hugs
and kisses), Q codes (QTH = where are you? or my location is, QRZ = who
is trying to talk to me, QLF = try tapping the telegraph key with your
*left* foot, your right foot is too clumsy), and another group that I
forget. A few of the codes clearly post-date telegraph use (QSY = change
carrier frequency) but many of them would have been serviceable for
telegraph use and probably date from then.
--
-------------------------------------------------------------
Will Ware <ww...@world.std.com> web <http://world.std.com/~wware/>
PGP fingerprint 45A8 722C D149 10CC F0CF 48FB 93BF 7289
da...@iinet.net.au
Louis Jaffe <lo...@well.com> wrote:
>In 1937 my father, Louis I. Jaffe, an American newspaper editor, toured
>central Europe. From Vienna he sent a telegram outlining his travel
>plans to Lenoir Chambers, a trusted associate in the US. The telegram
>is apparently in code.
[snip]
>If anyone in this newsgroup can help, it will be greatly appreciated,
>and I'm sure Dr. Leidholdt will duly credit you in the published work.
>Here is the brief text of the cable:
> AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
>And here is how Chambers interpreted it:
> Shall remain here until 14 and then go to Burda. Hope you and your
>family are well. Inform friends of your having heard.
If AB OZO are run together as ABOZO, all the code-groups are of 5
letters. This strongly suggests that a commercial book-code was used.
These were not considered secret: their purpose was to save telegram
costs (in effect, a primitive data compression technique). Some museum
may have copies of the old commercial code-books, and with a lot of
desk-work you could presumably match up the code-words and plain text.
-- Dave Brooks <http://www.iinet.net.au/~daveb>
PGP public key: finger da...@opera.iinet.net.au
servers da...@iinet.net.au
fingerprint 20 8F 95 22 96 D6 1C 0B 3D 4D C3 D4 50 A1 C4 34
What's all this? see http://www.iinet.net.au/~daveb/crypto.html
Wim Van Berlo
da...@iinet.net.au says...
>
>Louis Jaffe <lo...@well.com> wrote:
>
>>In 1937 my father, Louis I. Jaffe, an American newspaper editor, toured
>>central Europe. From Vienna he sent a telegram outlining his travel
>>plans to Lenoir Chambers, a trusted associate in the US. The telegram
Lenoir Chambers (f: The black chambers)? What a coincidence!
>>is apparently in code.
[...]
>>Here is the brief text of the cable:
>
>> AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
>
[...]
> If AB OZO are run together as ABOZO, all the code-groups are of 5
>letters. This strongly suggests that a commercial book-code was used.
Except for BUDA, unless there is a typo in the code here. Name of
place is given as BURDA in the interpretation.
The message is too short for anything other than guesses.
(One more guess would be AB=from (Latin), which would give:
from OZO 14 BU(R)DA. OZO meaning 'here'?)
On the other hand, if it were a code to conceal, rather than to shorten
the message, why aren't numbers and names encoded as well?
>These were not considered secret: their purpose was to save telegram
>costs (in effect, a primitive data compression technique). Some museum
>may have copies of the old commercial code-books, and with a lot of
>desk-work you could presumably match up the code-words and plain text.
>
Considering the name of the recipient, the type of message (2 typos
if it is to be a 5-letter code) and the lack of encoding the
number, I would like to consider one other possibility: a joke from
Louis Jaffe jr. (no offense to mr Jaffe, I like a joke occasionally).
Wim van Berlo
Stanley Chow
In article <53fvbr$j...@erinews.ericsson.se>,
Wim Van Berlo <eka...@eka.ericsson.se> wrote:
>>> AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
>>
>[...]
>> If AB OZO are run together as ABOZO, all the code-groups are of 5
>>letters. This strongly suggests that a commercial book-code was used.
>
>Except for BUDA, unless there is a typo in the code here. Name of
>place is given as BURDA in the interpretation.
I assume that is to avoid ambiguous decoding. Since there is no
code word for BURDA, the actual word is to be used, but then it
(being 5 letters and appearing in the code word list) would be
incorrectly decoded. So the solution is to shorten the word.
>The message is too short for anything other than guesses.
>(One more guess would be AB=from (Latin), which would give:
>from OZO 14 BU(R)DA. OZO meaning 'here'?)
>On the other hand, if it were a code to conceal, rather than to shorten
>the message, why aren't numbers and names encoded as well?
I tend to think the book-code theory is almost certainly right (but then,
I don't think I have actually such a book).
--
Stanley Chow; sc...@bnr.ca, stanley....@nt.com; (613) 763-2831
Bell Northern Research Ltd., PO Box 3511 Station C, Ottawa, Ontario
Me? Represent other people? Don't make them laugh so hard.
Christopher Biow
eka...@eka.ericsson.se (Wim Van Berlo) wrote:
>Except for BUDA, unless there is a typo in the code here.
Buda is half of the city of Budapest. Each half is on one side of the
river. I don't know how much formal basis this distinction has, but it was
prominently mentioned in the book I read on the the 1956 uprising, which
was primarily confined to Buda, if I recall correctly.
Allen J. Baum
> If AB OZO are run together as ABOZO, all the code-groups are of 5
> letters. This strongly suggests that a commercial book-code was used.
> These were not considered secret: their purpose was to save telegram
> costs (in effect, a primitive data compression technique). Some museum
> may have copies of the old commercial code-books, and with a lot of
> desk-work you could presumably match up the code-words and plain text.
I just bought one of these at a book sale. It's from (roughly) 1919, and
has a few caveats printed on it in red letters about how you're allowed to
use this code during the war...
It's a big thick book, and had many error-correcting properties ( like all
code words differed by at least 2 letters, and a list of codewords in
letter-reversed order used to recover from transcription mistakes). It
also has several cost saving features, in that many of the codewords could
be concatenated to form a single 10 letter word, and that all the words
were pronounceable (often consonant-vowel-consonant-vowel-consonant. Both
of these features save telegraph costs, since they were charged by the
word, (even if 10 letters) and words had to be pronouncable
The book has a set of standard codewords, and pages of 'custom' words that
individual companies can use. The company that owned this book (an
international bank) had all sorts of interesting banking phrases, the
names of lots of banks, etc,. and a set of 'escape' words that changed the
codebook!
Fun to look through.
--
***********************************************
* Allen J. Baum *
* Digital Semiconductor *
* 181 Lytton Ave. *
* Palo Alto, CA 94306 * ***********************************************
Vaughn Herbert Seward
In <32592D...@well.com> Louis Jaffe <lo...@well.com> writes, in
part:
>Here is the brief text of the cable:
>
> AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
>
>And here is how Chambers interpreted it:
>
> Shall remain here until 14 and then go to Burda. Hope you and
your
>family are well. Inform friends of your having heard.
I can add no new information to the most helpful of the earlier
replies, but as further replies are now coming that are of limited
usefulness, and may lead to confusion, (illustrating the limits of
USENET as an information source!) I would like to summarize what is
already known:
AB OZO - possible typo for ABOZO, "Shall remain here until ~"
FOURTEEN - 14 (meaning the 14th of the month?)
BUDA - Burda ( in clear, misspelled, _possibly_ to avoid confusion
with codegroup )
AMACO - "~ is my next destination", "Will go to ~", "~ is next on
the itinerary" (word order a problem here)
DOEKA - "Extending wishes of good health to you and family"
NUXIL - "Give notice of receipt of message"
LOUIS - Louis I. Jaffe, Sr., signature, in clear
The wordings are changed here from the previous helpful post, from my
acquantance with some commercial code books in my collection. (I only
have 2, and both are before the changes in cable regulations...instead
of 5-letter pronounceable groups, they have words under 10 letters
taken from several European languages.)
At this stage, only someone with access to copies of commercial code
books can help, either to identify or eliminate a suspect. Note that
the code book may be a European one, rather than one of the common U.S.
ones (ABC, ACME, Western Union...).
The fact that FOURTEEN is given in clear, however, is unusual. Most
commercial code books included a long section of equivalents for
numbers, both cardinal and ordinal. (That is, 14 had a codeword, and
14th also had a different one.)
While military codes often used four or six digits, or five-letter
groups without regard for pronounceability, some diplomatic codes also
used five-letter pronounceable groups. One example is the U.S. State
Department's GRAY code, which included ZOXIL as the group for "quote"
(as we know from Kahn[The Codebreakers]).
John Savard
Louis Jaffe
Wim Van Berlo wrote:
> Lenoir Chambers (f: The black chambers)? What a coincidence!
>
> Considering the name of the recipient, the type of message (2 typos
> if it is to be a 5-letter code) and the lack of encoding the
> number, I would like to consider one other possibility: a joke from
> Louis Jaffe jr. (no offense to mr Jaffe, I like a joke occasionally).
An interesting and amusing theory, Mr. Van Berlo. I can see why you
thought the name Lenoir Chambers was manufactured. But he was actual.
In the 1960's he won the Pulitzer Prize for editorial writing, American
journalism's highest honor, as in fact my father had done in 1927.
Both men were leading progressives in the American South of their era.
A biography of Chambers, by the same Prof. Leidholdt who's now
researching Louis Jaffe Sr., is being published by University of Alabama
Press.
The question of the telegram thus has real historical interest. Thanks
to everyone for their responses.
-LJ
Phil Herring
In article <53fvbr$j...@erinews.ericsson.se> Wim Van Berlo, eka...@eka.ericsson.se writes:
>Except for BUDA, unless there is a typo in the code here. Name of
>place is given as BURDA in the interpretation.
'Buda' == 'Budapest'. I.e., not a code word, just a common abbreviation.
---------------------------------------------------------------------------
Copyright 1996 Phil Herring. This article may not be reproduced for profit.
---------------------------------------------------------------------------
WHorton877
It is not Bentleys or Marconi's code. Checked both, have others that I
will check though.
GYEJY WIOHR (Best wishes, Bentleys code of 1934)
Wesley Horton
Aquiles Luna-Rodriguez
Louis Jaffe (lo...@well.com) wrote:
: Here is the brief text of the cable:
: AB OZO FOURTEEN BUDA AMACO DOEKA NUXIL LOUIS.
: And here is how Chambers interpreted it:
: Shall remain here until 14 and then go to Burda. Hope you and your
: family are well. Inform friends of your having heard.
Maybe it's a hoax, designed to confuse someone who spyed on your father.
The sytem appears indeed to be a 5-char code, but assuming that AB OZO
is a typo for ABOZO, there are 3 exceptions: FOURTEEN, BUDA and LOUIS.
FOURTEEN is very unusual, should be replaced in a normal code for a
5-letter codeword.
BUDA seeems to refer to Budapest, since Burda is a surname, not a place
(at least in german, a fashion magazin s named after it); Maybe it's
a typo on Mr. Chambers interpretation or in your post.
If your father was keeping a secret, it seems incredible he would
send in the clear a) date of departure, b) destination, and c) his name.
If it's a commercial code used for compression, FOURTEEN makes no sense;
the only alternative I can think of, would be your father using a very
common commertial code, and the operator decoding it on the fly. In this
case the codebook would be the standard one of the company carrying the
telegram.
My guts tell me that your father was trying to fool somebody. The trick
is quite obvious, so either Mr. Jaffe had no experience with codes (why
should he use one then?) or had a low opinion about the intelligence of
his followers. The test of this idea is easy: check if Mr. Jaffe really
departed on 14th, direction Budapest. If he didn't, there may be still
a message hiddenen the others words, wich nobody has decoded until now.
*********************************************************************
* Aquiles Luna-Rodriguez //there is a bg in this line *
* Universitaet Hamburg, Germany //Nobody expects... *
* pz4...@rrz.uni-hamburg.de //..the Spanish Inquisition! *
*********************************************************************
Jim Reeds
In re ABOZO, .. AMACO DOEKA NUXIL.
I have copies of 3 old commercial code books: "5th edition improved
ABC", "Bentley's", and "United Telegraph Code". The first 2 date
from before WWI and the last after. None of them contains the
code word "NUXIL". So none of those codes were it.
The National Cryptologic Museum is said to have a very complete
collection of commercial code books, in case you are looking for
library work.
--
Jim Reeds, AT&T Research, Room 2C-357
600 Mountain Avenue, Murray Hill, NJ 07974, USA
email: re...@research.att.com, phone: +1 908 582 7066, fax: +1 908 582 2379
home page: http://www.research.att.com/~reeds/index.html
Jim Reeds
In article <53gse2$g...@tor-nn1-hb0.netcom.ca>, sew...@netcom.ca(Vaughn Herbert Seward)
(or is it John Savard?) cogently writes:
|> I can add no new information to the most helpful of the earlier
|> replies, but as further replies are now coming that are of limited
|> usefulness, and may lead to confusion, (illustrating the limits of
|> USENET as an information source!) I would like to summarize what is
|> already known:
|>
|> AB OZO - possible typo for ABOZO, "Shall remain here until ~"
|> FOURTEEN - 14 (meaning the 14th of the month?)
|> BUDA - Burda ( in clear, misspelled, _possibly_ to avoid confusion
|> with codegroup )
|> AMACO - "~ is my next destination", "Will go to ~", "~ is next on
|> the itinerary" (word order a problem here)
|> DOEKA - "Extending wishes of good health to you and family"
|> NUXIL - "Give notice of receipt of message"
|> LOUIS - Louis I. Jaffe, Sr., signature, in clear
|>
|> The wordings are changed here from the previous helpful post, from my
|> acquantance with some commercial code books in my collection. (I only
|> have 2, and both are before the changes in cable regulations...instead
|> of 5-letter pronounceable groups, they have words under 10 letters
|> taken from several European languages.)
This is very useful. Commercial code books were one-part codes, that is,
the plain text phrases were listed in the same alphabetical order as the
code equivalents. Some were in strict dictionary order, & some were arranged
under main words of phrases (so "Give notice of receipt..." might be found
under "Notice", for example.) We should be be able to see traces of this
in the given crib.
|>
|> At this stage, only someone with access to copies of commercial code
|> books can help, either to identify or eliminate a suspect. Note that
|> the code book may be a European one, rather than one of the common U.S.
|> ones (ABC, ACME, Western Union...).
Agree 100%, except that I thought ABC was English. BTW, I just checked ABOZO
and AMACO against the ACME Code's "mutilation table" and they are not valid
ACME code words. DOEKA is, NUXIL is not. Conclusion: if its ACME, its horribly
garbled.
|>
|> The fact that FOURTEEN is given in clear, however, is unusual. Most
|> commercial code books included a long section of equivalents for
|> numbers, both cardinal and ordinal. (That is, 14 had a codeword, and
|> 14th also had a different one.)
|>
A code book user was always tempted to not encode a given word if he did not
know where to look it up in the book.
|> John Savard
--
Jim Reeds, AT&T Labs -- Research, Room 2C-357
John Flanagan
Anyone know if these old telegraphic codes (Bentley Marconi etc.) have
ever been put into disk form - or are they still subject to copyright??
I would like to add them to my collection
Vaughn Herbert Seward
In <Dz501...@cix.compulink.co.uk> b...@cix.compulink.co.uk ("John
Flanagan") writes, in part:
>Anyone know if these old telegraphic codes (Bentley Marconi etc.) have
>ever been put into disk form - or are they still subject to
copyright??
Many of the well-known commercial codes were compiled in the 1920s,
and, as they are 'institutional works', they would remain copyrighted
for 100 years under present copyright law.
However, it is entirely possible some of them have not had copyright
renewed, and thus are in the public domain in the U.S.; but this would
be somewhat difficult to research. In any case, many are probably still
copyrighted.
The codes that are 100 years old won't use five letter groups, since
the cable regulations were different back then.
John Savard
John Flanagan
Thanks for the information John :-)
Jim Reeds
Here are some more speculations about the Jaffe code message. They are
similar to Ken Shirriff's and John Savard's, but I reach slightly
different conclusions.
I would parse the message as follows
AB OZO = typo for ABOZO, meaning: "Shall remain here until ___
and then go to ___".
FOURTEEN = fill-in for first blank in ABOZO
BUDA = fill-in for second blank in ABOZO. (Buda is a city
near Pesht.)
AMACO = "Hope you and your family well."
DOEKA = "Inform friends of"
NUXIL = "your having heard"
LOUIS = Jaffe's signature.
This is based on general knowledge of commercial codes, & on particular,
on inspection of the "United Telegraphic Code" of 1933, a code with a
very large vocabulary (with more than twice the size of general vocab.
as 5th ABC). The United code has a group SUJEN which means exactly
"Shall remain here until ___ and then go to ___", and has phrases close
to the others I have listed above. I would guess that Jaffe used a
large (and hence late) code, possibly the "Western Union Code", which
was even larger than the United code. A small code, such as Bentley's,
which had comparatively few very short phrases, would have needed more
code groups to encode the message.
Now we come to the question of alphabetic order. I suppose ABOZO would be
found under "remain", AMACO under "hope", DOEKA under "inform", and NUXIL
under "your" or "heard." This is hard to reconcile with alphabetic order.
I believe, however, that it was common for users of commercial codes to
use a simple form of superencipherment, such as simple substitution mapping
vowels to vowels and consonants to consonants, or transposition, as in
sending the code words reversed. If we find a code book in the library
with the phrase "Shall remain here until ___ and then go to ___",
and with 3 phrases whose concatenation equals "Hope you and your family well.
Inform friends of your having heard," we can then try to correlate the
corresponding 4 code groups with ABOZO AMACO DOEKA NUXIL and see if we
can figure out the superencipherment method. I would start with recent
(that is, 1930's) code books first, especially fat ones.
Where to look for these code books? Well, some readers of this groups have
old commercial code books. Large libraries have a few. I have heard that
the Library of Congress has a good collection of old commercial code books,
as well as the National Cryptologic Museum. Another place would be in
whatever library received Jaffe's own collection of books.
So, if you have access to some commercial code books, see if you can
encode the message as the Jaffe message seems to have been: g1 number
place g2 g3 g4. If you can, post the results!
--
Jim Reeds, AT&T Labs - Research, Room 2C-357
600 Mountain Avenue, Murray Hill, NJ 07974, USA
re...@research.att.com, phone: +1 908 582 7066, fax: +1 908 582 2379