Morse code

Washington, the Capitol, 24 May 1844. Samuel Morse, a portrait painter by trade, sits at his apparatus and starts pressing a lever. Some forty miles away, in a railway station in Baltimore, his partner Alfred Vail watches a stylus press short and long marks into a moving paper tape. A few minutes later Vail sends the words back: “What hath God wrought”. The line from the Book of Numbers was chosen by young Annie Ellsworth, the daughter of the Commissioner of Patents. From that moment news could travel at the speed of electricity.

A telegraph wire carries exactly one thing: current on or current off. To turn that into text you need an agreement about which patterns of on and off stand for which letters. That agreement is Morse code.

A code is not a cipher

First, the words. A cipher hides meaning: without the key the text can’t be read. A code simply writes the text in other signs, and the table is open to everyone. Morse code has no key; it is printed in handbooks and on posters. It exists not to hide a message but to carry it: down a wire, over the radio, as light or as knocks.

So why is it in a course about ciphers? Because everything went over the air, plain telegrams and secret ones alike. A wartime radio operator copied columns of meaningless letters in Morse, and a code clerk then turned them into an order. The code is the envelope; the cipher is the locked box inside.

Long and short

The code has just two signs: the dot, a short signal, and the dash, a long one. A letter is a few of these in a row. A is ·−, B is −···, E is a single dot, T a single dash.

Why was E so lucky? Because it is the most common letter in English, the language the code was built for. Every extra mark on a frequent letter costs time on almost every word, and over thousands of telegrams that adds up to hours. So frequent letters got short codes and rare ones long codes: Q and Z have four marks each. A century later mathematicians would call this idea data compression: the more often a symbol occurs, the shorter it should be written.

All the codes fit neatly into a tree. Start at the root. A dot is a step left, a dash a step right. Wherever you stop, that is your letter. One step left is E, one step right is T. Two steps give I, A, N, M. The deeper you go, the longer the code and the rarer the letter.

Notice the gaps in the bottom row: English doesn’t need all sixteen four-mark codes, so some slots stay empty. Switch to Russian and every one of the sixteen is taken. Russian has more letters, so its Morse fills the tree almost completely.

It’s all about timing

A dot and a dash differ only in length. But the silences between them carry meaning too: they tell you where a letter ends and where a word ends. Without them the code turns to mush. ···−−−··· could be SOS, but it could also be SMB, VMS, IJS and several more.

So everything in Morse is measured in one unit, the length of a dot:

• a dot: 1 unit of sound;
• a dash: 3 units of sound;
• the gap between marks inside a letter: 1 unit of silence;
• the gap between letters: 3 units;
• the gap between words: 7 units.

The 1–3–1–3–7 rule is easy to remember, and you hear it as rhythm. Skilled operators don’t count dots; they hear a whole letter as a little tune. That is why learners are told to learn the sound of each letter, not a picture of dots and dashes.

Speed is measured in words per minute. The standard word is PARIS: with the gap after it, it takes exactly 50 units. Twelve words a minute means 600 units in 60 seconds, a dot one tenth of a second long. Try it yourself.

If letters come out wrong, your dashes are probably a bit short or the gaps inside letters a bit long. That is why operators spent months training their hand: a steady rhythm matters more than speed.

Sound, light and SOS

A wire isn’t the only way. When radio arrived at the end of the nineteenth century, Morse code moved onto the air: the transmitter switched its carrier wave on and off, and the operator heard it as a beep in his headphones. Navies flashed Morse with signal lamps from ship to ship, at night, with no wires and no extra ears. The code fits any signal you can switch on and off.

The most famous signal of all is SOS: ··· −−− ···. It was adopted by the international radiotelegraph convention in Berlin in 1906 and came into force in 1908. Despite the legend, SOS doesn’t stand for anything, neither “Save Our Souls” nor “Save Our Ship”. It was chosen for its rhythm: three short, three long, three short can’t be mistaken for anything else, even through crackling static. It is sent as one run, without the usual gaps between letters.

Morse was taken down by ear, but it can be seen too. Below is an intercept made right in your browser. Under the button is a trace of the signal, the way a pen recorder would draw it: black bars are sound, the spaces are silence.

Russian Morse

Morse code soon spread beyond English. Russian letters were given the codes of similar-sounding Latin ones: А as A, Б as B, В as W, Г as G, Д as D, К as K, Т as T. Sounds that Latin lacks got the spare combinations: Ч is −−−·, Ш is −−−−, Щ is −−·− (the same as Q), Я is ·−·−.

As a result the same code reads differently in each language. ·−− is W in English and В in Russian; −··− is X in English and the soft sign Ь in Russian. An operator copying an intercept has to know what language is being sent. Digits and punctuation, on the other hand, are shared by everyone.

What to take with you

• Morse code is a code: it doesn’t hide the text, it lets you send it by wire, radio or light.
• Common letters got short codes: E and T are a single mark. All the codes form a tree: dot to the left, dash to the right.
• Everything is measured in dot lengths: a dash is 3, the gaps are 1, 3 and 7. Without the gaps the code can’t be read unambiguously.
• SOS isn’t an abbreviation but an unmistakable rhythm. Russian Morse borrows from the Latin table, so the same code can mean different letters in different languages.

A dot and a dash are really two states: signal or no signal. Computers live exactly the same way, only they call them one and zero. In the next lesson we see how letters become bits and bytes, and why Base64 encrypts nothing. Next: binary, hex and Base64 →