Zum Inhalt springen

Data Communication Prehistory: Telegraph, Morse, and Teletype

Zusammenfassung

Before computers ever exchanged a bit, humanity spent a century learning to send symbols over wires at the speed of electricity. The electric telegraph of the 1830s and 1840s established the radical idea that information could be separated from any physical object and transmitted across continents in moments. Morse code introduced a binary-like encoding of letters into dots and dashes; the Baudot code turned characters into fixed-length 5-bit patterns; and the teletype machine automated the whole process into a keyboard-to-printer link that prefigured the computer terminal. These nineteenth- and early-twentieth-century technologies invented, in embryonic form, nearly every concept that digital data communication would later formalize: encoding, multiplexing, error, flow control, character sets, store-and-forward switching, and the very notion of a network. This article traces that lineage — the roots from which ARPANET and the internet eventually grew.

The Telegraph: Information Leaves Its Body

For most of history, a message could travel no faster than the fastest horse, ship, or runner that carried the physical letter. Optical telegraphs — most famously Claude Chappe’s semaphore network in revolutionary France (from the 1790s), a chain of towers relaying coded arm positions — broke the speed limit for line-of-sight distances but depended on daylight and clear weather.

The electric telegraph severed information from matter entirely. Building on the discovery that electric current could be detected at a distance, several inventors converged on practical systems in the late 1830s. In Britain, William Fothergill Cooke and Charles Wheatstone patented a needle telegraph in 1837 and deployed it along railway lines. In the United States, Samuel Morse, together with Alfred Vail, developed a simpler single-wire system. On May 24, 1844, Morse sent the famous message “What hath God wrought” over a line from Washington, D.C., to Baltimore — and a communications revolution began.

The telegraph’s consequences were immense and immediate. Within decades, wires linked cities, then nations; the first lasting transatlantic telegraph cable went into service in 1866 after several failed attempts, collapsing the communication time between Europe and America from weeks to minutes. The telegraph created the first information networks, the first network operators, the first standardized rate structures, and — as historian Tom Standage argued — a “Victorian Internet” complete with its own slang, online romances, and frauds.

Morse Code: Encoding as Compression

Morse code is, in a real sense, the ancestor of digital encoding. It represents each letter and numeral as a sequence of short signals (dots) and long signals (dashes), separated by gaps. Crucially, the code was designed with efficiency in mind: Morse and Vail assigned the shortest sequences to the most frequent letters of English (a single dot for E, a single dash for T), reportedly by counting the type in a printer’s font case to estimate letter frequencies. This is the same insight — assign short codes to common symbols — that Huffman coding would formalize mathematically a century later.

Morse code is sometimes called binary, but strictly it is more than two symbols: the durations of dots, dashes, and the various gaps form a variable-length code. It was designed to be sent and received by humans by ear or by hand, which made it robust but fundamentally limited in speed by human reflexes.

Baudot and the 5-Bit Character

The drive to mechanize telegraphy — to remove the skilled human operator — produced one of the most important conceptual leaps in the prehistory of computing. In 1874, the French engineer Émile Baudot devised a 5-bit, fixed-length code in which every character occupied exactly five units. Five bits yield 32 combinations, which Baudot extended to cover letters, figures, and control functions by using “shift” characters to switch between a letter set and a figure set — a trick that effectively doubled the available symbols.

The Baudot code mattered for two reasons. First, fixed-length encoding made automatic machinery far easier to build than Morse’s variable-length symbols, because every character took the same time and space. Second, Baudot pioneered time-division multiplexing: his system let several operators share a single telegraph line by interleaving their characters in timed slots — the same principle that would later let many phone calls share one cable, and that underlies TDMA in digital cellular networks. The unit of signaling speed, the baud, is named in his honor. A later refinement by Donald Murray adapted the code for typewriter-style keyboards and rearranged it for mechanical efficiency; the result, the Baudot–Murray code, became international standard ITA2 and remained in use for decades.

The Teletype: Keyboard to Printer Across the World

The teleprinter or teletype machine, maturing in the 1910s and 1920s through companies such as the Teletype Corporation in the U.S. and Creed in Britain, fused a keyboard, an encoder, a transmission line, and an automatic printer into a single end-to-end text system. An operator typed on a keyboard; each keystroke was encoded (in ITA2) and sent over the wire; at the far end, a machine decoded the signal and printed the characters automatically. No Morse skill was required.

Teletypes turned telegraphy into something that looked remarkably like later computer communication. They introduced practical mechanisms that data communication would inherit wholesale:

  • Start and stop bits framing each character — asynchronous serial transmission, still the basis of serial ports (RS-232) and the structure of UART communication today.
  • Control characters for carriage return, line feed, and “bell” — survivals of which still lurk in text files and terminal codes.
  • Paper tape for storing and replaying messages, enabling store-and-forward operation: a message could be punched, then transmitted automatically, then relayed onward — exactly the logic later used in Usenet’s store-and-forward news and in message-switching networks.

The global teletype network, Telex, connected subscribers worldwide with their own addressing and switching system — in effect a planet-spanning machine-to-machine text network operating decades before the internet. News agencies, businesses, and governments ran on it.

From Teletype to Terminal to Network

The bridge to computing was direct and physical. The earliest interactive computer terminals were teletype machines: the Teletype Model 33 ASR, introduced in the early 1960s, became the standard input/output device for minicomputers and early time-sharing systems. The device file name /dev/tty in Unix — still present on every Linux and macOS system — is a fossil abbreviation of “teletype.” The ASCII character set (1963) was designed in part to serve teleprinters as well as computers, replacing the 5-bit codes with a richer 7-bit standard, and its control characters (carriage return, line feed, bell, escape) inherited their meanings directly from teletype hardware.

When ARPANET researchers in the late 1960s built the first computer network, the conceptual toolkit they drew on — encoding characters into bits, framing them for transmission, switching messages between nodes, multiplexing many conversations over shared lines, and handling errors — had all been pioneered, in mechanical and electrical form, by the telegraph and teletype engineers of the preceding 130 years. Packet switching was a genuinely new idea, but it was an answer to problems the telegraph network had already taught the world to recognize.

Dead End: The Pneumatic Tube and the Fax Detour

Two parallel branches of message technology mostly led nowhere for data communication. Pneumatic tube networks — pressurized tubes shooting physical canisters between buildings — carried telegrams, mail, and documents in cities like Paris, London, and New York from the 1850s into the twentieth century. They were fast for short distances but could never scale to long-haul electrical speeds, and they moved physical objects rather than information, the very limitation the telegraph had transcended.

The facsimile machine offers a subtler lesson. Alexander Bain patented an electric fax concept as early as 1843 — before Morse’s famous message — yet fax remained a commercial curiosity for over a century. It transmitted images as analog scan lines rather than discrete symbols, so it never developed the encoding, error-handling, and switching abstractions that made the telegraph–teletype line the true ancestor of digital networking. Fax finally boomed in the 1980s as a digital technology riding the telephone network, but by then it was a side branch, soon overtaken by email. The lineage that led to the internet ran through the symbolic, switched, multiplexed telegraph network — not through systems that merely moved pictures or parcels.

📚 Sources