Three kinds of calculating tool
It is useful to separate three stages of calculation. First come aids, such as the abacus, where a person does the thinking and the tool stores numbers. Second come mechanical calculators, where gears and levers carry out the arithmetic. Third come electronic calculators, where circuits do the work and a display shows the answer. The calculator as most people know it belongs to the third stage, but it grew out of the first two.
The abacus and early aids
The abacus, in its various forms, was used for thousands of years across many cultures. Beads on rods or stones in grooves stand for values by position, and an experienced user can calculate at impressive speed. In parts of East Asia, abacus use remained widespread well into the twentieth century and is still taught in some places. For centuries, it was the practical standard against which any calculating machine had to compete.
Pascal and Leibniz
In the seventeenth century, the first geared machines appeared. Blaise Pascal built an adding machine in the 1640s, with a mechanism that carried digits from one wheel to the next. Gottfried Wilhelm Leibniz designed a machine in the 1670s that aimed to multiply and divide as well, using a stepped drum. Their devices were remarkable but not reliable enough for general use, and the details of their construction and the earlier work of Wilhelm Schickard are told in the story of the earliest calculating machines. What matters here is that the key mechanical ideas were already in place by about 1700.
Making calculators practical
It took until the nineteenth century to produce calculators that businesses could buy and use daily. In 1820, the French inventor Charles Xavier Thomas de Colmar patented the Arithmometer, a machine based on Leibniz’s stepped drum principle. It was produced for decades and sold in significant numbers, mainly to insurance companies, government offices and similar customers who handled large amounts of arithmetic.
Later in the century, other designs spread. A pinwheel mechanism, used in machines developed by Willgodt Odhner in Russia in the 1870s, became a common alternative to the stepped drum and was manufactured in several countries for decades. In the United States, adding machines with keyboards and printed tapes, such as those of William Seward Burroughs, were sold to banks and offices from the 1880s onward. Key-driven machines meant that an operator could enter numbers quickly, without turning knobs or cranks.
In the twentieth century, electric motors replaced hand cranks, and the best desk machines could do multiplication and division automatically. Offices employed rows of operators with machines for payroll, accounting and engineering work. A separate and interesting design, the Curta, was a small hand-cranked cylinder invented by Curt Herzstark and produced from around 1948. It was used by engineers, surveyors and rally navigators who needed a portable device before electronic calculators existed.
Electronic calculators
Electronic desk calculators began to appear in the early 1960s. An early commercial example is the British ANITA, launched in 1961. It used vacuum tubes for its calculating circuits and was large, but it was silent and fast compared with mechanical rivals. Soon afterward, transistor-based machines replaced tubes, and the machines grew smaller and cheaper. Several Japanese companies competed strongly in this market through the 1960s.
The next step relied on integrated circuits. Putting many transistors on a single piece of silicon made it possible to reduce a calculator’s electronics to just a few silicon chips, a change covered in the story of integrated circuits. A prototype handheld calculator developed at Texas Instruments in the late 1960s, involving Jack Kilby and colleagues, is commonly cited as an important early example, though it was a research project first and a product later.
By 1971, commercial pocket calculators were reaching the market. Among the best known is a Japanese Busicom model that used a set of custom chips; the work on that design led to Intel’s 4004 microprocessor, one of the first processors on a single chip. Other companies, including Bowmar in the United States, sold small handheld calculators in the early 1970s. Early pocket models were expensive and sold first to engineers and business people.
In 1972, Hewlett-Packard introduced a scientific pocket calculator, the HP-35, which could perform trigonometry and logarithms and began to replace the slide rule among engineers and students. Prices then fell quickly, as competition grew and chips became cheaper. Within a few years, basic calculators cost a small fraction of their early price, and by the end of the 1970s they were common in schools and offices. Displays of light emitting diodes were gradually replaced by liquid crystal displays, which used far less power and allowed for solar cells.
Calculator to computer
Pocket calculators changed several habits. Slide rules and printed tables of logarithms faded from everyday use. Teachers argued about whether students should be allowed calculators in math class, a debate that did not fully settle. Businesses spent less on arithmetic work done by hand.
The calculator also led toward more general devices. Programmable calculators let users store sequences of steps, an early form of programming for many people. The microprocessor that came out of calculator design went on to power personal computers, and the display and keyboard conventions developed for calculators carried over into other products. Today, a phone includes a calculator as a minor feature, and it is the same ancient task of calculating that is being done, but with a very different device behind it.