The problem a switchboard solves
If every telephone had to be wired directly to every other one, the number of wires would grow absurdly. Ten telephones that could all call each other would need 45 separate lines, since each of the ten has a wire to the other nine and each wire is shared by two telephones, giving 10 times 9 divided by 2. Fifty telephones would need 1,225. A city of thousands could not be wired that way at all.
The solution is a central point. Each subscriber has one line running to an exchange, and at the exchange a switch can join any two lines together for the length of a call. The first exchanges were manual: a person did the joining with a plug and a cord.
How a manual exchange worked
The details varied by period and country, but the main steps were similar in the early manual exchanges.
- A subscriber turned a crank on the telephone or lifted the receiver, which sent a signal to the exchange. In the earliest boards, the signal dropped a small shutter or lit a lamp next to the subscriber’s line.
- An operator saw the signal, plugged one end of a cord into that subscriber’s jack, and asked who was wanted.
- The operator then plugged the other end of the cord into the jack of the person called, and rang that line.
- When the call ended, a signal told the operator, who pulled the cords out so the lines could be used again.
Calls to a subscriber on another exchange added a step: the operator would contact a colleague at the other exchange over a dedicated line, and that operator would make the final connection.
The key limit was physical. An operator could only reach so far along a panel, so each person could handle only the lines within arm’s length. Larger exchanges used “multiple” boards, a design associated with Charles Scribner of Western Electric in the late 1870s and 1880s, in which every line appeared in front of every operator. Operators sat side by side, each with a share of the calling lines but access to all of the lines that could be called.
Why manual exchanges lasted
They did not simply last because nobody had thought of something else. They worked well for small and medium numbers of subscribers. A human can understand an unclear request, remember that Dr. Smith is out, or know that the grocer is usually at the other number. Operators provided service that no early automatic system could match, such as reporting emergencies, finding someone, or giving the time.
They also had the advantage of low capital cost per subscriber when the number of lines was small. Many small towns kept manual exchanges for decades. The same hub-and-spoke logic reappeared later in early computer networks, which also needed a way to join many machines without wiring each pair.
The costs
The costs were mainly the people and the growth problem. A busy exchange employed dozens or hundreds of operators, working in shifts. Doubling the number of subscribers more than doubled the traffic, because each new subscriber could call all the others, and so operators and cords multiplied. There were also delays, misconnections and privacy worries, since an operator could in principle listen in.
Large cities strained the model first. By the early twentieth century, planners were trying to reduce the load by building bigger multiple boards, introducing lamp signals instead of drop shutters, and moving to a “common battery” design in which the exchange supplied power, so subscribers no longer had to turn a crank.
Automatic switching
Automatic switching was developed in the late nineteenth century, long before electronic computers or integrated circuits could be used for the job. The most famous early device is associated with Almon Strowger, an undertaker in Kansas City, Missouri, who patented an automatic switch in the early 1890s. A popular story says he was worried that calls for his business were being diverted to a competitor by the local operator; whether or not that explanation is accurate in detail, it appears often in histories of the technology. The first Strowger exchange in regular service opened in La Porte, Indiana, in 1892.
Early automatic systems used pushbuttons, then a dial, which was developed in the following years. The caller’s dialing pulses controlled selector switches at the exchange that moved step by step to find the called line. The mechanics of those systems, including the later crossbar type and electronic switching, are a technical story in their own right.
A slow transition
Automation did not happen all at once. Many exchanges, especially large ones, used semi-automatic arrangements for years, with operators handling long-distance and special calls. Some cities switched in a single night, with operators watching the equipment they had worked beside for years take over. Rural and small-town manual exchanges lingered into the middle of the twentieth century in many places, and in some, even later.
Long-distance calls kept operators relevant longer, since connecting through several exchanges needed coordination. Direct customer dialing for long-distance calls began in the United States in the early 1950s, and operator assistance gradually became a service for special cases, such as collect calls.
What the switchboard left behind
The manual switchboard established ideas that still shape the network: a central point connecting many subscribers, numbered or named lines, and a human-friendly service model. Its successors, from step-by-step and crossbar switches to the electronic and digital systems of later decades, did the same job faster and with far less labor. Behind every modern call, a series of automatic decisions still plays the role the operator once played, and in networks built on packet switching the decisions are made for every small piece of a message.