On 15th April 1912 the passenger ship Titanic hit an iceberg and sank with the loss of nearly 1500 lives.  Some claim that communication failures were as significant as anything else in scale of the tragedy.

According to Neal McEwen writing in the Telegraph Office Magazine…

“When the Titanic sank, 5 ships were next to the Titanic, but did not have their radios on. Some ships only turned their radios on once or twice a day, and receive their messages from a land station and turn off their radios. The Carpathia had steamed 4 hours past the Titanic and had to turn around. The SOS they heard was not from the Titanic. A 14 year old operator, David Sarnoff, of American Marconi in New York was at the key and heard the Titanic's SOS, and he kept repeating long after the Titanic had gone down, waiting for the ships to turn on their Radios.”

Tragically, the Californian, a large ship located only five to ten miles away from the Titanic, could have rescued many of the ship’s passengers but for “an entirely avoidable and extremely unfortunate lack of communication”[1].

How can it be so difficult to communicate such a life critical message?  What does this tell us about modern business communications?

From the late 1800’s ships had started using wireless signals for communications. But while Trans-Atlantic ships encoded their messages using International Morse, US coastal ships used American Morse. The dots and dashes meant different things.  There were also two dominant technology providers, Marconi and Telefunken[2].  Each provider used differing protocols for scheduling signals and as signalmen were employees of the service provider not the shipping line, ships would listen for signals at different times. 

Another problem was understanding the meanings of the messages.  Telegraphic messages used various codes to shorten messages and reduce errors.  In effect these codes created their own languages for messages.  Unfortunately there was no consistent or common agreement on the use and meaning of these codes (the vocabulary of the language). This was especially significant with respect to distress signaling in emergencies.

At the first international congress of wireless telegraphy in 1903, the Italians recommended the use of "SSSDDD" to signal an emergency.  "D" had previously been used internationally as the signal for an urgent message. Though discussed, it was not adopted. Deciding on a distress signal was put on the agenda for the next meeting in 1906.  And so in 1904, Marconi started using the code "CQD" for distress signals.  In Britain "CQ" was widely used to precede time signals and special notices and meant "for all stations".  

Then, in 1908 the international congress ratified the German distress signal "SOS" as the official code for ships in distress.  “SOS” was adopted simply on account of its easy radiation and its unmistakable character. There is no special significance in the letters used.

The use of "CQD" lingered for several more years, especially in British services where it originated. Which is why the radio operators on the Titanic first used "CQD" to call for help.  Later, they interspersed their calls with "SOS."  In fact they had a choice of also using “HELP” or “NC”.   All were in use at that time.

The ineffectiveness of the Titanic’s distress calls emphasizes that there are many different types of agreement required for effective communications. Signals, frequencies, protocols, identities and codes all have to be agreed.  Anyone familiar with agreements and standards for electronic business will recognize this situation.

One of the effects of the Titanic disaster was the recommendation that the International Telecommunication Union (ITU) assume governance of all maritime radiography standards. The ITU was at that time the meeting place for governments that owned and regulated all things related to postal, telephone, telegraph, and radio.

In the world of electronic business we are confronted with several frameworks that identify various “signals, frequencies, protocols, identities and codes” required, like layers in a cake.

It appears there are quite a few layers in the cake required for effective communications, such as...

Within electronic business services there are relatively precise specifications for the infrastructural parts of the communication cake (such as the TCP/IP, SMTP, HTTP, etc.).  Which is why the Internet, Word Wide Web and emails work.  The nature of the technology is that the options for these protocols are finite and can be unambiguously defined. 

It is also interesting that at the environmental levels there are also relatively formalized trade agreements, laws, and contracts supported by common business practices.

The challenge for electronic business seems to lie in the transactional layers.  This may be because they link between the technology and policy – the “how?” and the “why?”. 

Taking this one step further, the pivotal layer for effective communication appears to be the semantics – reaching agreement on the meaning of things.  It is the way we can assure ourselves that the answer the question “Do I understand you correctly?” will be “Yes”.

In the U.S. Senate hearings following the Titanic disaster Harold Bride, the surviving wireless operator was asked,  "Is CQD in itself composed of the first letter of three words, or merely a code?"  Bride responded, "Merely a code call sir."  As with “SOS” there is no special significance in the letters used. 

This answer is significant because it reminds us that these codes (the vocabulary of the language) are nothing more than meaningless labels that need to be associated with a meaning. It is the same with human languages – we use various terms to mean the same things.  And if the terms “SOS” and “CDQ” have no meaning in isolation, then it can only be by common agreement that they are associated with a maritime distress call.

On board the Titanic the question was which code should be sent?  Which would be most understood by the greatest audience?  In the end messages went out using the two commonly used codes, either of which had a good chance of being understood by the receiving stations nearby.   This was because those receiving stations understood the semantics of at least one of those codes.

The lesson is that while it is convenient to use the same standardized coded references (or language), this is not critical to communication.  What is critical is that codes or terms reference the same meanings.  This is because the vocabulary of different languages can be associated with the same meaning.  Sometimes this is incorrectly paraphrased,  “semantics are more important than syntax”. What is meant is “semantics are more important than language”.

Humans do this mapping between language and semantics naturally.  We can translate between different languages, read musical notations and appreciate works of art.  We can extract the meaning by applying our intellect. 

When we attempt to exchange business information we rely on computer applications that do not have the advantage of intellect. Systems have to be instructed using unambiguous rules.  Understanding the meaning of business data requires understanding aspects such as context of use, business rules, structures, associations and cardinalities. Controlled vocabularies, dictionaries, pattern libraries, data models, markup and data definition languages are all aids to formalizing these rules, but this is an evolving craft.  Formal ontologies are moving us towards better semantic definitions. But we aren’t there yet.

Today, the best we can do to define the semantics of electronic business is to use a standardized language based on agreed semantics.  As with human languages we need a language that can convey the required semantics.  Some languages are more suitable than others.

To address this the OASIS Universal Business Language (UBL) offers supporting semantic artifacts such as conceptual process and data models, a controlled vocabulary of terms used consistently for names and definitions as well as an international data dictionary (to explain these definitions in other languages).

As was the case with the Titanic’s distress calls these semantic artifacts can aid in the translation between other languages.  But, as with all languages, these translations are not perfect. Semantic interoperability is greatly assisted if a common language is used.  This is because semantic alignment is only part of the overall communication framework required for the effective exchange of information. It may be true that “semantics are more important than language” but that does not mean agreement on a common language is unimportant.

It took a major tragedy to realize the importance of formal agreements on an interoperability framework for telegraphy.  What will it take for electronic business interoperability?

[1] Patrick S. Ryan - The ITU and the Internet's Titanic Moment, Stanford Technology Law Review, 2012: 2012 STAN. TECH. L. REV. 8

http://stlr.stanford.edu/pdf/ryan-theituandtheinternetstitanicmoment.pdf

[2] The Titanic was outfitted with Marconi radios and had a subscription for Marconi personnel services, whereas the Californian used the radios and personnel services of Telefunken, one of Marconi’s competitors. In spite of a non-enforceable gentlemen’s agreement to work together it is possible that the two companies’ rivalry took an ominous turn on that fateful night. [Patrick S. Ryan - The ITU and the Internet's Titanic Moment, Stanford Technology Law Review, 2012: 2012 STAN. TECH. L. REV. 8

http://stlr.stanford.edu/pdf/ryan-theituandtheinternetstitanicmoment.pdf]