As you’ll recall from the first section of this series, some of the first types of data were simple tallies. These tallies became more complex and the need to count larger numbers of things grew. Simple counting and equations were no longer adequate. Early Sumerians found a solution to this by inventing the abacus around 2700 BCE, which allowed them to perform more advanced mathematical equations and tally larger numbers. Interestingly enough, the user of an abacus was called an abacist, but later in history, a person that performed mathematical calculations was called a computer, you no doubt know that word from a more modern convenience.
As you may know, Big Data only seems to get bigger and bigger. As humanity developed, they eventually outgrew the abacus. This simple tool for counting and computing could no longer keep track of the growing masses of data.
Contained within these masses of data are massive ideas that help shape society, assuming they can be shared with enough people. The best known ideas are those that are recorded and spread to as many people as possible and we’ve established that all important ideas must be written down and backed up on a piece of paper or even cave wall if they are to be preserved for any substantial period of time. As populations grew and territories were discovered and claimed, new ideas took root. Ideas associated with religions, or even political ideals began to proliferate and people even started to migrate to places where their ideas wouldn’t be oppressed.
One particular group of freedom-seekers helped found a land built around one of the most important ideas ever recorded: the idea that all men are created equal. But the document in which that iconic statement is found also has another important decree that would come to affect the technology world forever.
When the United States Constitution was ratified in 1788, the framers recorded the philosophies and ideals on which the entire nation was founded. Knowing that the nation would continue to grow, the framers decided they would need to gather information about the population. In order to ensure this information was gathered, they mandated that the government would take a census after each decennial. But this nation was growing rapidly, and it became difficult to gather information on such a large scale. This meant that the government would need a new method of gathering data, and a new way to store it.
Punched card technology was the first method used to track and store data on a large scale and made gathering census information much faster. Let’s again set the foundations of this technology before we examine its use in the census and computers.
Punched cards were first used to control textile looms in France around the year 1801. Basically, the punched sheets of paper contained pattern instructions for the looms that allowed the loom to produce the various textile patterns automatically—this is often recognized as the world’s first computer program as it acts as a sort of primitive binary code.
On a textile loom, punched cards correspond to a set of pins that are connected to a string. As the pins pass over the card, the pin can only move the string when there is a hole in the punched card for the pin to enter. In this way, the punched holes determine which string goes where in the pattern. From here, the technology was improved by Semen Korsakov who used them in informatics and information storing and searching around 1832.
Using similar concepts, Herman Hollerith invented the first method of recording data on a medium that could be read by a machine. Prior to this, punched cards had been used to control machines like automatic pianos rather than to record or read data. Hollerith used the technology to assist the 1890 U.S. census mandated by the U.S. Constitution, in which his device revolutionized large information handling. Hollerith’s method was a much more organized, faster, and accurate way to handle large quantities of data than counting and marking by hand.
Interestingly enough, the original census data taken in 1890 was destroyed in 1921 by—you guessed it—a fire and people learned the need for data protection once more. Following the destruction caused by a fire in the basement of the Commerce Building in Washington, DC, there was a large outcry or a permanent National Archive.
Hollerith’s method worked by using large data sheets that were created with standardized perforations, each representing specific individual traits like gender, nationality, occupation, and so forth. Then, in what could be considered one of the first data processing centers, operators looked at the data sheet and placed a card in the card puncher, which is essentially an early version of a keypunch. If there was a mark on the data sheet the puncher would punch the information into the corresponding position on a thick card.
Once punched, the card would be placed in a card reader. The card reader had 80 electrically connected pins that hung above the card. The pins were connected to a lever that, when lowered, moved the pins against the card. The pins are either stopped where there were no holes, or continued through the hole allowing the pin to complete an electrical circuit connected to another machine that marked all of the data on a panel of gauges. This method allowed all of the data from the card to be recorded instantly without counting by hand. Each of the punched cards was then stored in a box for safe keeping. This method of counting large quantities of people was also
Following his work with the census, Hollerith founded the Tabulating Machine Company, one of four companies that merged to form Computing Tabulating Recording Company (CTR) which later became the corporation we’re all familiar with, International Business Machines Corporation or IBM. IBM also allegedly sold their punched card technology to the German army during WWII where it proved to be a critical tool for the Nazi party, who used it to count and track prisoners—a sordid reminder that any technology can be misused.
Punched card technology continued to advance from the early 1900s into the 1950s, where punched cards became the most widely used medium for data storage, data entry, and processing in institutional computing. It also proved useful for things like voting, punch clocks, and yes, computer programming. Punched cards were essential to computer programmers because they were used to store binary processing instructions for computers—everything from executables to source codes were read from punched cards. Having little or no memory, computers had to read data from punched cards. In fact, NASA used punched cards and computers to read them to perform calculations as part of the first manned space-flight to the moon. Luckily, backups could be fairly simple. Punching an exact copy or punching two cards at once could produce instant backups.
As punch cards grew in proliferation, storing them became a hassle; eventually requiring large storage facilities to house cartons upon cartons of punched cards. In the 1950s, UNITYPER introduced magnetic tape, which promised faster speed and required less storage space. Along with the tape, UNITYPER introduced keystroke machines that enabled data entry directly to magnetic tape for UNIVAC systems. These were somewhat successful, though tape wasn’t popular until the 1970s, when tape steadily began to replace punched cards and keypunch technologies. Even so, punched card programs were still in use until the mid-1980s.
Stay tuned for History of Backup Part 5: Magnetic Tape