Bob Schneider is a Partner in 
Wilson So is the Director of Hitachi America, Ltd. XBRL and the Telegraph Code Book
Written by Bob Schneider Posted February 9, 2007
The laying of transatlantic cable in the second half of the 19th century revolutionized communications between the US and Europe. But the resulting telegraphy could hardly be described as instant messaging. It was not until the 20th century that speeds reached even 120 words a minute. Moreover, cabling was extremely expensive. It cost a quarter to send a single word from New York to London in 1900, a lot of money back then.
To pack more information into a message, senders and receivers employed code dictionaries of phrases and other expressions. The sender would find a phrase with his intended meaning and include the accompanying code in his message. (For example, in one code book, caninero stood for “have ceased selling” one word standing for three.) The code message would be transmitted by the telegraph or cable company. The recipient would look up the code words in another copy of the same book or follow other procedures to find their meaning.
Thousands of code books were created. The most popular was the all-purpose ABC Code (which went through seven editions between 1873 and 1936). There were also books for specific industries and segments, with terms useful for a particular trade. Some books were sold to the general public; others were developed by individual firms for their own use and remained proprietary. The codes particularly the specialized codes arranged their phrases by topic, sometimes in tables, to save space and facilitate message coding. (If you’d like to know more about the code books, take a look at Jim Reeds’s fine introduction, or at the material and extracts assembled by my friend John McVey at his site.)
I’m hesitant to compare 21st century XBRL with a 19th century technology. Still, I do see direct parallels. The code books are analogous to XBRL taxonomies, the vocabularies or dictionaries that provide the foundation for information exchange. The telegraph transmission can be compared to the instance document, a file that contains content structured in the XBRL language. Whoever deciphers the cable’s meaning does the work of the XBRL reader, which can access, view, and analyze an instance document.
The levels of specificity reflected in code books from all-purpose ABC Code to, say, Acme Inc. Code also yield rough equivalents in the taxonomy hierarchy. If we consider XBRL for financial reporting alone, ABC Code corresponds to the Commercial and Industrial taxonomy for US GAAP. Industry code books are aligned with industry taxonomies, eg, that for oil and gas. The 19th century firm had its own company code book; the 21st century firm has its own company XBRL taxonomy.
Then as now, competition for “standards” existed. As John writes: “The codes competed on various mixes of convenience, depth and/or breadth, error-resistance, reputation of compiler, novel system, even naturalness of their expressions (for the so-called “verbatim” codes).” Old books and systems were discarded as better ones were created, while others remained to complement the new works. Today we see XBRL competing with other XML-based standards (RIXML, FIX, MDDL, etc.) for broad-based recognition and acceptance, but also at times working in collaboration.
I don’t pretend that there are any great lessons in comparing highly elaborate XBRL with the relatively straightforward telegraph code book, which in the end are dissimilar communication tools, introduced in business and regulatory environments separated by 150 years. But I do think it helps to put the challenges XBRL faces in perspective. It informs us, once again, that technology adoption is a process, where competition and collaboration work in tandem to respond to user needs.