Jump to content


Anatomy of an Encrypted Communication

  • Please log in to reply
No replies to this topic

#1 Herbert Blenner

Herbert Blenner

    Advanced Member

  • Members
  • PipPipPip
  • 52 posts

Posted 04 January 2015 - 09:09 AM

A detailed examination of anomalies in an article on Lee Harvey Oswald and an adjacent advertisement show Nat Sherman intentionally included the irregularities in his ad.

The United Press International article, "American Awaits Soviet Word," originated in Moscow on November 2, 1959 and appeared the next morning in the New York Times. This article on Lee Harvey Oswald contained fifteen lines of text and about seventy words. They published two incorrectly spaced words, on his and nothing to, which appeared as "o nhis" and "nothin gto."


Around 1960 the publication of any United Press International article would start with a written story being sent by teletype-link to subscribing newspapers whose teletype would print a draft quality article and simultaneously punch a paper tape. If the story were selected for publication then the punched tape would be feed into a Justo-writer that produced letter quality print justified to one column width. Next they would cut and paste up the justified article.

Newspapers used mechanical Justo-writers because figuring out where to insert spacers to produced a justified line of text was too complicated for typists. The machine would count character by character the accumulated width of a line until it reached the width of the column. Then the machine would backtrack to the first space while decreasing the accumulated line width. Next the machine would take the difference between the column width and the decreased line width as the padding space. The final algorithm would place an additional space between words until the line width reached the column width. That is right, Justo-writers were mechanical computers.

I should emphasize a Justo-writer inserts padding spaces between words by recognizing the space character as word separators. Where they were no spaces, no additional padding would be added. At most, a Justo-writer would increase the number of spaces between incorrectly spaced words. The incorrectly spaced words, "o nhis" and "nothin gto," came out of the Justo-writer because they were feed in with incorrect spacing.

Suppose United Press International in Moscow transmitted the words "on his" and "nothing to" with correct spacing. If The New York Times received "o nhis" and "nothin gto" then the communications link caused four character recognition errors resulting from sixteen individual bit errors.

Two successive character recognition errors could have caused the first pair of incorrectly spaced words. The New York Times would have received the transmitted bit sequence for lowercase n, 1101110, as the 0100000 bit sequence for a space. Next, they would have received the following transmitted bit sequence, 0100000, for a space as the 1101110 bit sequence for a lowercase n.

Bit sequences are numbers and the transmitter and receiver keeps a tally, known as a checksum, of all numbers sent and received during a communication. Following the message the transmitter sends its checksum. If the checksum calculated by the receiver did not equal the checksum sent by the transmitter then the receiver detected an error. This error detection scheme would work when two errors occur with just one exception.

The second error was the singularly unique error out of 127 possible errors that frustrated the error detection scheme. United Press International sent 1101110 and 0100000 whose partial checksum is 10001110 and The New York Times received 0100000 and 01101110 whose partial checksum is 10001110.

Similar remarks apply to the second pair of incorrectly spaced words. The New York Times would have received the transmitted bit sequence for lowercase g, 1100111, as the 0100000 bit sequence for a space. Next, they would have received the following transmitted bit sequence, 0100000, for a space as the 1100111 bit sequence for a lowercase g. They would have caught the third communication error if the fourth singularly unique communication error did not occur.

Every one of the sixteen individual bit errors affected only data bits. Not once did an error occur on the start or stop bits sent with every bit sequence. The communications link missed sixteen opportunities to flag a framing or overrun error. If the optional parity bit were employed then this test would have failed because each of the four erroneous bit sequences had an even number of bit errors.

Individually each character recognition error would have been an unlikely event. Frustration of the checksum as an error detection scheme by a singularly unique error would have been a rare event. The failure to detect any error in the sixteen individual bit errors and the double frustration of the checksum test is not a tenable explanation of the origin of the two incorrectly spaced words. Either United Press International in Moscow sent two incorrectly spaced words or The New York Times engineered the two incorrectly spaced words.

The explanation that UPI in Moscow sent two incorrectly spaced words ignores the role of Soviet censorship. Censors had a keen interest in Western press reports on the "defectors." They delayed the first article on Robert E. Webster for about three days and held part of the initial report of the "defection" of Lee Harvey Oswald in censorship. Without question, Soviet censors reviewed the second article on Oswald before they permitted UPI to send their story. I doubt the censors approved a story with incorrectly spaced words or someone at UPI risked prosecution for espionage by embedding the incorrect spacing after receiving approval.

The alternate explanation that The New York Times engineered the two incorrectly spaced words avoids the problems caused by Soviet censorship and would explain why the spacing errors got by the proofreaders.

The Nat Sherman' advertisement contains an anomaly. In place of the opening quote in "It's a Boy" a large comma and a period appear as superscripts.


However, in our language we print commas and periods as subscripts and use the superscript position for apostrophes and quotation marks. Our keyboards, typewriters, replaceable font wheels, and typesets do not include the comma or the period as superscripts.

Distracted, incompetent or drunk typesetters may only mistakenly set a character belonging to their typeset. Physically the span of any typesetting error is limited to characters in the typeset. I dismiss the possibility of error by a typesetter because two types from alien typesets of two different font sizes would have to have been selected to replace a single type.

I doubt that Nat Sherman produced advertisements by the typeset method. He combined six different styles of text to give his advertisements their distinctive appearance. Elaborate advertisements would have been too complicated to typeset and too expensive to publish six days a week.

Advertisers who published the same ads repeatedly affected a significant cost reduction by using the photographic method. Nat Sherman alternated between two advertisements between October 1959 and December 1959. He published about forty advertisements from two pieces of artwork.

Nineteen advertisements were the same type as the advertisement published on November 3, 1959 but did not contain the anomalous punctuation. The print published on November 3, 1959 was photographically distinct from nineteen other similar advertisements. An examination of the photographic technique will illuminate this discrepancy.

People would use an electric typewriting with a replaceable font wheel to print lines of text of common font and size then cut and paste the individual lines to master artwork. In this manner, they generated advertisements containing various fonts and font sizes. When the artwork was finished, they photographed it. After they retouched the negative to remove any streaks or alignment guidelines, they made non glossy prints for inclusion in the newspapers.

The production of a line of text containing a large superscript comma and a superscript period from an ordinary keyboard requires the five-step sequence of half-down, comma, change font wheel, period, and half-up. If anyone were to assert that the anomaly in the November 3, 1959-Nat Sherman' advertisement was an accident then they would incur liability of showing how someone could mistakenly perform five steps in place of one.

The Nat Sherman' advertisement published on November 3, 1959 was the product of deliberate photographic manipulation. Amateur photographers had the tools and techniques to superimpose two images on one sheet of photographic paper. They required a photographic enlarger, removable red filter to cover the projection lens, and a red mask to define the size and position the altered section. First they would place the unaltered negative in the enlarger and project the image through the red filter onto non glossy photographic paper. Next, they placed the red mask over the section they wanted to change. They removed the red filter for a couple of seconds to make their first exposure. Then they replaced the original negative by a negative containing a large comma and a period.

They positioned the negative so that the new image fell on the red mask. Finally they would lift the red mask and remove the red filter for a couple of seconds to produce the second exposure. The doubly exposed print would then be processed in the normal manner.

Nat Sherman wrote his own advertisements and had the opportunity and knowledge to do his own photography. His business was walking distance from Willoughby's, a primary source of photographic equipment and darkroom supplies, where they knew Nat as a regular.

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

Web Work by: XmasZen.com