Morse Code: The Encoding System That Made the Telegraph Possible

Morse code is the encoding system that made the single-wire telegraph practical. By assigning a unique sequence of short and long electrical pulses — dots and dashes — to each letter, it allowed the alphabet to be transmitted through a single circuit of wire by the simplest possible arrangement of contacts. Its descendants are still in use today in amateur radio and in some aviation and maritime applications.

The Problem of Encoding

The earliest electrical telegraphs — Cooke and Wheatstone’s 1837 needle telegraph, for example — used multiple wires and multiple needles, with each needle deflected left or right to indicate a letter. This worked but required bulky, expensive multi-wire cables. A telegraph using a single wire and a single on–off pulse required an encoding step: how to map 26 letters (and digits, and punctuation) into a stream of two states.

Two approaches were available. The first was to assign each letter a fixed number of pulses, as in the early Cooke–Wheatstone 5-needle system. The second was to use a variable-length code in which the most common letters get the shortest codes. The second approach, by a rough application of what we now call entropy coding, minimizes the average transmission time of a message. It is the approach Vail chose.

The Development of Morse Code

The encoding system that became “Morse code” was developed jointly by Samuel Morse and Alfred Vail between 1837 and 1840, with Vail doing the bulk of the design work on the codes themselves. Vail’s notebooks — preserved in the Smithsonian — show that he assigned shorter codes to the most frequent letters by counting letter frequencies in a New York Morning Herald type-case.

A concrete example: in the original 1838 American code, the letter E (the most common in English) was a single dot, T a single dash, A ·−, I ··, N −·, S ···, O −−−. The distress sequence SOS — ··· −−− ··· — was an obvious choice precisely because the three letters it spells are among the simplest to recognise by sound.

The original 1838 American code was different in several respects from the code in use today. The American code was substantially revised by the German-Austrian standardisation at the 1865 International Telegraph Conference in Berlin, and the International Morse Code adopted there is what is in use today. The most familiar modern example: S is ···, O is −−−, and the distress call SOS is therefore ··· −−− ···.

The Elements of Morse Code

In International Morse Code, a dot is one unit of time long, a dash is three units long, the gap between elements of a letter is one unit, the gap between letters is three units, and the gap between words is seven units. This timing convention — implemented as the absolute length of the “dit” — is the most important practical detail of the system. An operator transmitting too fast or too slow is unintelligible.

The most common letters have the shortest codes, and the most common words have their own standard abbreviations: CQ (−·−· −−·−, “calling any station”), DE (−·· ·, “from”), K (−·−, “over to you”), R (·−·, “received”), 73 (“best regards”), 88 (“love and kisses”). The most important prosign — a sequence that functions as a single symbol — is the distress call SOS, adopted as the international maritime distress signal at the second International Radiotelegraph Convention in Berlin, 3 November 1906. Mayday for voice distress was adopted in 1923.

The Reception of Morse Code

A skilled telegraph operator in the 1850s could read Morse code by ear at 25–40 words per minute; top operators reached 50–60 wpm. The two principal instruments for reception were:

The register, a clockwork-driven mechanism that pulled a strip of paper tape under a stylus and produced a permanent record of the dots and dashes. The register is what Morse originally designed and is the instrument used for the famous 1844 message.
The sounder, a small electromagnetic device that clicks once for a dot and twice as long for a dash. The sounder, developed by several inventors in the mid-1850s (the credit is usually given to Charles Sumner in the United States and William Fothergill Cooke in Britain), allowed operators to read messages by sound rather than by reading paper tape. By 1860 most U.S. telegraph offices used sounders; the register was kept as a backup.

Tapping out Morse code (continuous-wave keying) was the standard operator skill, but the first commercially successful automatic transmitter — a punched-tape device that could feed a perfect stream of pulses at high speed — was introduced in the 1860s, and high-volume commercial lines were routinely operated at 100–200 wpm by punched-tape machines.

The Decline and Legacy of Morse Code

Morse code began to decline in the early twentieth century as the telephone, then continuous-wave radio, displaced the telegraph as the dominant means of long-distance communication. The end-points are well documented:

Maritime Morse — the international requirement for ships to be able to communicate in 500 kHz Morse — ended on 31 January 1999.
Aviation Morse — the international requirement that aviation stations be able to communicate in Morse code on the VHF band — was withdrawn by the ICAO Council on 1 January 1991 (rather than later, as is sometimes asserted).
U.S. commercial Morse — the last commercial Morse service in the United States ended on 31 December 1999, when the last commercial Morse operator in North America (at KPH, the Marconi/RCA coastal station at Bolinas, California) was withdrawn.

Morse code is still alive in three communities: amateur radio operators, who treat it as a craft skill; the U.S. military, which still trains some personnel in it; and the aviation, maritime, and railroad industries, where older equipment continues to use it. Its principles — variable-length coding, agreed international standards, a single-channel on–off signal — underlie every digital modulation scheme used since. The Huffman coding used in modern data compression is the same idea as Vail’s letter-frequency counts, in rigorous form.

The Continuing Question

The interesting design question about Morse code is whether variable-length encoding is still the right model. The original Morse system was a 1838 frequency-optimized code: the most common English letter (E) got a single dot, and the rarer letters (Q, J, Z, X) got the longer combinations. In information-theoretic terms, this is the same logic that Claude Shannon later formalized and that David Huffman generalized in 1952 to produce the optimal binary tree for any given frequency distribution. The interesting unresolved question is why Morse — and not, say, a fixed-length code of 5 bits per letter like the later Baudot code — won the international competition of 1865. The standard answer is that variable-length coding was a real efficiency gain for English text, and the on-off sounder could be read by ear at 25-40 wpm. The revisionist answer, developed in the work of the information historian Thomas Petzold and in Andrew Odlyzko’s quantitative analysis, is that fixed-length codes of the same period (Baudot’s 1874 5-bit code, the Murray code of 1901) were actually faster to transmit and easier to automate, and the persistence of Morse into the 1990s reflects sunk costs in operator training and equipment rather than any inherent technical advantage. The Morse-vs-Baudot question is still relevant: the ASCII character set was a fixed-length 7-bit code, and the variable-length encoding has only returned to dominance with the modern data-compression formats (gzip, MP3, JPEG) where the trade-off between code length and complexity can be offloaded to silicon.

Selected Sources

Kenneth Silverman, Lightning Man: The Accursed Life of Samuel F. B. Morse (Knopf, 2003)
Tom Standage, The Victorian Internet (Walker, 1998)
Ira Rhodes, “Vail, Alfred Lewis,” American National Biography (Oxford University Press, 1999)
International Telecommunication Union, Radiotelegraph Code (ITU, 1990)

For the related article on the technology, see the telegraph and Samuel Morse.