Information Theory part 7: Morse Code
In 1832, mathematician, Carl Gauss, and physics professor, Wilhelm Weber, designed a system which allowed them to communicate at a distance while they worked on their experiments – connecting the observatory with the physical laboratory. They solved a really important problem, which was more of a puzzle: how to send all letters of the alphabet using one circuit – or a line. And their system used a galvanometer, since it was known that electric current, passing through a coil, creates a magnetic field pointing through the center of the loop, which could deflect a needle. But instead of merely moving a needle at a distance, their system used a switch which could reverse the direction of current instantly. This would cause the magnetic field around the coil to reverse, and the needle would deflect either to the right or left, depending on the direction of current – thus, giving them two different signaling events – or ‘symbols’ – right or left deflection. Most importantly, he assigned shorter symbols for the most common letters. [For instance,] ‘A’ was a single right deflection. And ‘E’ was a single left deflection. And he used the longer codes for less common letters, such as ‘K,’ which was three right deflections. And at the time, the speed of transmission was around nine letters per minute. All the needle telegraphs which followed suffered from a similar limitation – and it was an engineering problem. The ‘signaling rate’ was slow. Now the the signaling rate was the number of deflections per minute which could be accurately transmitted and received. And if you squished signaling events together, the receiver would get confused, due to jitter – resulting in errors – similar to how sustained notes on a piano will bleed together and become less recognizable – if you play rapidly. And over time, the signaling rate was incrementally improved. One improvement was to add a small permanent magnet to the outside of the coil. This helped pull the needle back to neutral position, after each deflection. And these designs led to a wide range of needle telegraphs, which were deployed across Europe. The Electric Telegraph Company was the first public telegraph company. It was formed, in 1846, after its owners purchased the key needle-telegraph patents at the time. But the speed of these various needle telegraphs never surpassed around 60 letters per minute – as each needle couldn’t signal much faster than one deflection per second. And initially, the company billed customers based on single messages – which could hold up to twenty words – which is about the length of a ‘tweet.’ And by 1848, the cost of sending a single message from London to Edinburgh was sixteen shillings. And this was around one week’s salary for, say, a shop owner at the time. So this technology was initially out of the [reach] of common people. In the United States, the commercialization of the telegraph was led by a portrait painter, named Samuel Morse, who had followed development of the needle telegraphs in Europe. Morse is important. because he focused on speeding up the rate at which letters could travel. He did away with needles. And in , he initially submitted a patent based on the idea that electric current could either flow or be interrupted – and interruptions could be organized to create meaning. Though his designs on how to produce these interruptions were complicated – involving a convoluted system of gears, levers and electromagnets. However, his system was greatly simplified after his collaborations with Albert Vail. This led to an iconic piece of user interface – the simple spring-loaded lever – or ‘key’ – which could be controlled with the tap of a finger. And on the receiving end was a spring-loaded lever that could be pulled and released by a strong electromagnet. To create a difference akin to the left-right deflection, he varied the length of a key press, or the pulse width. The closure of a switch for a very short time was called a ‘dot.’ And the dot can be thought of as the basic unit of time in Morse code. And the closure of the switch for three units of time represented a ‘dash.’ [SOUND OF LETTERS BEING SENT BY MORSE CODE.] Spacing exactly right. Very small, tight spaces between the dits and dahs in a character. Didah dit. [LETTER BEING SENT BY MORSE CODE.] Didah dit dit. [LETTER BEING SENT BY MORSE CODE.] And this was the source of difference in their coding strategy. Starting with an initial dot or dash – left or right branch – which then leads to another dot or dash, and so on. And the scheme assigned shorter symbol sequences to more probable letters – based on the letter frequencies – which could be tabulated from books. So nodes high up in the tree – such as a single dot – represented ‘E.’ A single dash represented ‘T.’ And as we move down the tree, we place less common letters. And after a letter, this system inserts a three-unit pause. Spacing between the characters in a word or group is uniform too – but longer. [LETTERS BEING SENT BY MORSE CODE.] It’s important to realize that the meaning of these messages was intertwined with the timing [used when sending] them. Are you wondering if proper spacing is really so important? Or is it no more than an extra refinement – a nice thing to do – like neat handwriting? If you think so, you’re wrong. And I’ll show you why. [LETTERS BEING SENT BY MORSE CODE] Dit for dit, and dah for dah, they match. Only the spacing makes the difference between one word and the other. So to send the word ‘Paris,’ we would first need to think of it as ‘P [space] A [space] R [space] I [space] S.’ The signaling rate of this system was directly related to the tempo of the signal. And music analogies were used inside training videos. What he was sending was standard test word: ‘PARIS.’ And there you are. Each peak is a dit – or a dah. Each valley, a space. This is excellent sending. Uniform and rhythmic. This is an example of poor hand sending. Same word: ‘PARIS.’ But look at the difference. Irregular dits and dahs. Haphazard spacing. No uniformity. No rhythm. Amazingly, it was the simplicity of this keying system which made it much faster than any of the buttons and cranks employed by the needle telegraphs in Europe. The letter rate jumped to 135 letters per minute – or more, with trained operators. And on May 24th, 1844, the first successful transmission was the message, “What hath God wrought?” And the next day, it was reported by the New York Tribune that, “The miracle of annihilation of space is at length performed.” Consider that, at the time, 90% of messages were still transported by horseback. Immediately, this technology was becoming critical to the success of [the] military, newspapers, financial traders, crime-fighting [organizations, etc.]. Any business that relied on information now relied on the telegraph – and Morse code. By 1900, the prices had dropped to 30 cents per message – as traffic surged to over 63.2 million messages sent that year. As people begin using this system, they naturally thought of ways to save money. This led to popular code books that mapped words to common sentences. For example, ‘Blade’ would actually mean ‘Please name and reserve, for myself and family, the following accommodations.’ The telegraph companies frowned upon this, as they were happily charging people to be verbose. More letters equals more profit. It was now clear that information was an elastic term. A specific meaning was needed. An obvious question remained unanswered. If you are [charging to transmit] information – no matter the system – how should you [charge] to be fair to everyone? Number of letters – as a measure of information – would no longer suffice.