Research Notes:
November 2024

However, there were tradeoffs for the speed and convenience offered by telegrams—namely, privacy and price. Snail mail letters are low tech, but the simple envelope, which employs tamper-evident flaps and thick or patterned paper, protects its contents from prying eyes. In Dr. Barnes’s time, as today, it was expected that a letter’s contents would be known only by the sender and the recipient. Telegrams, however, required the parties involved in the transport and transmission of the message to be apprised of its contents. For birthday messages or holiday greetings, this was no issue. For more sensitive information, the security risks are obvious.

Cost was another factor. Because human labor was involved in transmitting telegrams, customers were charged by the word, which led to adaptation. For example, a 1926 telegram from Dr. Barnes to Guillaume simply reads, “ARRIVED NEWDOG.”¹ Based on other correspondence around the same time, we know that Dr. Barnes had just returned to his home in Merion from Cherbourg, France, and had bought a new dog. However, “ARRIVED NEWDOG” is a much more economical message.

Luckily for Dr. Barnes, many before him had also faced the challenge of making telegrams more secure and affordable. This led to the development of telegraph codes, which served to encrypt messages while keeping word counts low. The idea of encoding messages to aid delivery and security had been around for centuries with systems like smoke signals, semaphores, and naval signal flags.² Eventually, the latter evolved into numerical systems, where flags would stand for numbers, which in turn would correlate to specific words, phrases, and sentences. Telegraphic codes, which arose shortly after the invention of the telegraph in the mid-19th century, resemble this system.

The simplest codes used existing words to stand in for longer phrases. Known as “commercial codes,” they were effective at lowering the cost of a message but offered little security since decoding keys were available in published codebooks.³ Effectively, commercial codes brought telegrams the same level of security as an envelope—protecting a message from wandering eyes and casual ne’er-do-wells.

However, situations like labor strikes and military and government operations called for greater security. This led to the creation of “superenciphered” codes, which combined surrogate words with numerical codes, keywords, and numerals known only to the sender and receiver that could be used to unlock the telegram with the aid of a cipher.⁴ Superenciphered codes were even deemed secure enough to be used by striking railway workers on telegraph lines owned and operated by the railways they were striking against.⁵

Communicating in Code

The first coded telegram between Dr. Barnes and Guillaume came in December 1925. In it, Guillaume used code words to provide measurements of paintings to Dr. Barnes. The utility of this is multifaceted: numbers would often be transcribed incorrectly or otherwise obfuscated by damage to a telegram, leading to inaccuracies upon delivery. For example, it is much more reliable to send the word “postulant” than the numeral 54½: a smudged 5 could look like a 6 or an 8, a 4 could be misinterpreted as a 9, the ½ that follows could be misread as 12, and so on.

By using a distinct word, the chance for misinterpretation is greatly diminished, provided the recipient has the correct codebook on hand. In fact, clarity was likely Guillaume’s primary goal in sending this first encoded message. Secrecy would not have been a concern, as Barnes had already purchased the paintings in question.

Shown here is a sampling of coded telegrams between Barnes and Guillaume during the 1920s.

In one, Barnes sends, “SMOKER POCKET PREDICTION LANDSCAPE PLANTADOR POODLE GAUGUIN PRESSGANG PULSE ABRUPTLY REBUKABLE PRECINCT.”⁶ Though a string of nonsense at first glance, the message is a mixture of plain, unencoded text, referencing specific paintings, and encoded measurements and phrases. The message can be decoded as follows: “SMOKER 46 ½ 62 LANDSCAPE 39 ½ 51 GAUGUIN 65 ½ 90 ½ BUSINESS IS GOOD, AND EVERYTHING AND EVERYBODY ALL WELL 6,660 ¼.”

“SMOKER,” “LANDSCAPE,” and “GAUGUIN” refer to paintings, and the numbers that follow each are their dimensions. The message also contains “ABRUPTLY,” which translates to the “Business is good . . . ” phrase above, and ends with two words that together make the number 6,660.25, which may refer to a sum of money. Similar coded figures appear elsewhere in Barnes and Guillaume’s telegrams. Researchers have speculated that the numbers refer to account balances, either for Barnes’s accounts with dealers or his own cash accounts, rather than prices for individual works.

A telegram sent from Guillaume to Barnes on January 13, 1927, provides some clues that point to their chosen code. The telegram reads, “BANK ASUNDER INDUSIAL SHIELDING.”⁷ The message uses a mixture of encoded and unencoded text to convey the message “BANK HAVE NOT RECEIVED DRAFT $3,000 28TH DECEMBER.” Using code to discuss transfers of funds related to artwork purchases was typical of Barnes and Guillaume.

We can see that the decoded message was written on the telegram in pencil by Dr. Barnes or one of his staff members. Clues like this helped Barnes volunteer Roger Hutchin begin to understand the codex, or code system, used by Barnes and Guillaume and track down its source: the code must use complete, pronounceable words as surrogates for the words, numbers, and phrases they replace, and it must contain the words in this telegram, which must correspond to the meanings scribed in pencil.⁸

This ruled out a large swath of code systems, as many used random letter or number sequences to stand in for words and phrases. This also indicated to Hutchin that Barnes and Guillaume were using a commercial codebook, as opposed to one meant for enhanced security. Hutchin was also able to rule out hyper-specific examples like The Liverpool Cotton Telegraph Code (1882) and Cinema Code; A Practical Telegraph-cable Code Specially Compiled for the Motion Picture and Allied Industries and for the Personal Use of Individuals Connected Therewith (1923), assuming that Barnes and Guillaume would have a greater need for the phrases contained in a general-use codebook than for any such domain-specific phrases.

After searching archives and special collections for codebooks that were once in Barnes’s possession (unfortunately, there are none), Hutchin turned to the internet, relying on digitized versions of codebooks from around the same period that matched his constructed criteria.⁹ Luckily, he came across The Adams Cable Codex, made digitally available by the HathiTrust Digital Library. The words in the Adams codex matched the January 13, 1927, telegram and others, which all made sense based on the context of surrounding letters.

Barnes and Guillaume exchanged over two dozen coded telegrams between 1925 and 1928. Though there are other examples of Barnes communicating in code, their correspondence was the most consistently coded. Naturally, we wonder why Barnes and Guillaume chose to encode their messages. First, it should be noted that not all of their telegrams were in code. In fact, there are entire years of their correspondence that contain no coded messages. This suggests that whatever the reason for corresponding in code, it was not a consistent concern. Although Barnes had reason to worry about privacy—his high-profile purchases and ill-tempered antics often caught the attention of the press—it likely was not the primary motivator. Second, their use of a commercial code indicates that any interest in security was at a superficial level.

While this removes some of the intrigue from the whole situation (it’s safe to say Barnes wasn’t using secret codes to scoop up masterpieces), the mystery of why codes were in use remains. Clarity and convenience seem the most likely culprits. As discussed, telegram codes are helpful when transmitting multiple sets of numbers, preventing them from becoming jumbled and misinterpreted. It makes sense, then, that many of Barnes’s coded telegrams focus on numbers—measurements, prices, account balances, requests for wire transfers, and dates. On numerous occasions, Barnes lashed out at art dealers after receiving incorrectly sized frames for new purchases. He no doubt saw the value of codes in preventing such mix-ups.

Convenience, on the other hand, also explains why Barnes and Guillaume used the codes, and why they did not use them consistently. Several code words for common phrases appear repeatedly in their telegrams. “ABRUPTLY” and “AWKWARD” are two examples, meaning “Business is good, and everything and everybody all well,” and “Send telegraphic money order,” respectively.¹⁰ Given the regular use of these words, we can assume that, once the codes were in their vocabulary, Dr. Barnes embraced the ease of communicating routine updates and orders with a single word, rather than explaining things anew each time.

Conversely, a lack of convenience may explain why Barnes and Guillaume never stuck to consistently encoding their telegrams. While it may be easy to remember that “AWKWARD” means “Send telegraphic money order,” other less-used phrases would necessitate flipping through the codebook each time one needed to send a message. Barnes and Guillaume both traveled frequently and likely did not always have their preferred codebook at hand, in which case, they would be out of luck.

Overall, Barnes’s use of telegram codes may have more to do with the collector’s personality than any desire for secrecy or cost savings, or even convenience. Dr. Barnes’s interest in telegram codes follows a familiar pattern for the collector: a period of initial, passionately focused interest, followed by a precipitous, unexplained drop-off. Initially, Barnes invents his own codes, writing to Guillaume in letters which words he will send to indicate different courses of action for the dealer to take.¹¹ Things escalate from there, with Barnes requesting that Guillaume devise an entire bespoke code that the pair can use to communicate.¹² As seen, they ultimately settle on a readily available commercial code. After a time, even their use of this code wanes, until they abandon it completely without another mention.

Dr. Barnes repeats this pattern in countless areas of his life: in his interest in the ideas of authors and philosophers, in different artists and schools of artistic thought, and in professional relationships and collaborations. This pattern even appears in Barnes’s relationship with Guillaume—the two get off to a hot start, with Guillaume snatching up works on Barnes’s behalf and Barnes singing his praises, even naming Guillaume the Foreign Secretary of the Barnes Foundation. Their relationship peaks when Barnes renames a gallery the Paul Guillaume Room, hanging Guillaume’s portrait opposite his own. But as quickly as it blossomed, the relationship wilted, as Barnes picked fights with the dealer seemingly out of thin air until they ceased doing business completely, despite Guillaume’s interest in mending a relationship that had no clear reason for fracturing.

As with other topics, artists, and people, Dr. Barnes’s interest in telegram codes was a passing fancy. However, like those pursuits, it added even more variety and interest to a man who was full of whims and had no lack of means to act on them. After all, it was Dr. Barnes’s mercurial nature that led him to create the eclectic collections of the Barnes Foundation, with an abundance of equally eclectic stories surrounding it to match.

Closing Thoughts

A heartfelt thanks is owed to Roger Hutchin, who volunteers his time to the Archives, Library, and Special Collections department, transcribing and translating from French Dr. Barnes’s correspondence with Paul Guillaume. Without Roger’s attention to the coded telegrams and his efforts in finding the codex used, this Research Note would not have been written.

Readers are encouraged to read the Paul Guillaume correspondence for themselves, as well as Barnes’s exchanges with other important figures in art history, at the Albert C. Barnes Correspondence Spotlights.

Endnotes