Why Are Some Letters Quick to Send in Morse Code and Others Slow?

Two Signals and a Rule About Time

Morse code writes every letter with just two kinds of signal, a short one called a dot and a long one called a dash. It has been in use in one version or another since 1844 [1]. It works with sound, light or an electric switch, because all the receiver has to tell apart is short from long.

The international rules, written down by the International Telecommunication Union, are really rules about time. A dash lasts as long as three dots. Between the signals inside one letter there is a pause one dot long. Between two letters the pause is three dots long, and between two words it is seven [1].

So the code is not only made of sounds. It is also made of silences of different lengths, and we will come back to why. First, the question in the title: why are the letters different lengths?

The Most Common Letters Got the Shortest Codes

Count the signals in the official table and a pattern appears. Only E (one dot) and T (one dash) have a single signal. Four letters have two: A, I, M and N. Eight have three: D, G, K, O, R, S, U and W. The remaining twelve letters, such as B, J, Q and Z, take four [1].

The shortest codes went mostly to letters that English uses a lot. Wikipedia, for example, says the most common letter, E, gets a single dot [3]. Anyone who sends a lot of English then spends less time on average, because the busiest letters are the quickest.

You can see the saving in the timing rules. Counting a dot as one beat, a dash as three, and the one-beat pause between signals, E costs 1 beat. T costs 3. A costs 5. Q costs 13, and Z costs 11 (my own count, from the table and rules in [1]). Sending Q takes thirteen times as long as sending E.

How Alfred Vail Worked Out Which Letters Are Common

Someone had to decide which letters were common, and in the 1830s there was no computer to count them. According to Wikipedia, Alfred Vail, who worked with Samuel Morse on the telegraph, estimated the frequency of English letters by counting the movable type he found in the type cases of a local newspaper in Morristown, New Jersey [3]. A type case is the tray of small metal letters a printer sets into words, and a printer keeps more of the letters that are used more often. A popular history of communications describes the same trip, saying he went to the local newspaper office and found what he needed in the compositors’ type cases [5].

The idea was clever because the printers had already done the counting for him: a drawer holding many E’s and few Z’s tells you something about the language. I could not find the exact numbers in a source I could read in full, so I leave them out. The story comes from a few secondary accounts, and I describe it as reported rather than certain.

It also was not the first plan. Morse’s earliest design sent only numbers, with a codebook to look up each word by its number. Wikipedia describes it as a cipher of three- or four-digit numbers for words, like the one used on existing semaphore telegraph lines [4]. The letter-by-letter code replaced that. A letter Vail wrote in February 1838, as quoted there, says Morse had “thrown aside the Dictionaries” [4]. With letters, nobody needs to carry a codebook.

Who deserves the credit is still argued over. The Linda Hall Library says that who worked out the dot-and-dash letters is “still up for debate”, with many thinking it was Vail [2], and Wikipedia notes several scholars favor Vail while Morse’s supporters point out that Vail never claimed the code in his writings [4]. This note does not settle that. It only says what the code was designed to do.

Why the Silent Gaps Are Part of the Code

Short codes for common letters create a problem. Look at A (dot, dash) and J (dot, dash, dash, dash) in the table [1]. The code for A is the beginning of the code for J. It is also the beginning of R, L, P and W. If you only hear dot, dash, how do you know whether the letter is finished?

Wikipedia puts it this way: Morse code is not a “prefix code”, one where no letter’s code is the start of another’s, and A and J are the example it gives [3]. The answer is the pause. A pause of three dots tells the listener that one letter has ended, while pauses of one dot are inside letters [1].

Here is a small example you can try. E followed by T sends one dot, a three-dot pause, then one dash. The letter A sends a dot, a one-dot pause, then a dash. The signals are the same. Only the timing of the silence tells them apart. That is why the rules fix those gaps so carefully, and why the Morse operators’ skill lies partly in rhythm.

Test the Plan With Paper and Pencil

You can check whether the shortest letters really are the common ones. You need a pencil, paper, and any book or magazine you have at home, and nothing else.

Copy out about a hundred letters from a page of ordinary English, skipping spaces and punctuation. Make a tally mark for each letter. Then ask: are E and T near the top of your list? Where do Q and Z land? Next, give each letter its beat cost (E 1, T 3, A 5, Q 13, Z 11) and add up the beats for your hundred letters. Compare that with a pretend code in which every letter costs 8 beats, which would be 800 beats for a hundred letters.

You can also tap it out on a table with a pencil: a short tap for a dot, a tap held three times as long for a dash, and a steady count in your head for the gaps. Try tapping E, then T, then A, and have a friend guess without seeing.

A single page is a small sample, so your tally may not match the printers’ counts exactly. Part of the fun is seeing how close a page of your own can get. If you want more detail, the sources below, especially the official ITU table, go well beyond this note.

Sources

Update history