The story of his calculating and computing ‘engines’ are well known. What’s less widely recognised is that they include the world’s first computer printers, which are in fact the whole point of the designs. They guaranteed error-free output all the way to the printing press. Today we call this maintaining data integrity.
Babbage was a mathematician with a genius for invention. He was Lucasian Professor of Mathematics from 1828 to 1839, a prestigious post held by Sir Isaac Newton before him and Sir Stephen Hawking long afterwards. He was a skilled networker, a popular raconteur and threw great parties for the cream of London society at his London home. He’d probably have his own TV series today. However, he also had a genius for shooting his mouth off and upsetting powerful people. His fallings out with his chief engineer and the British government over money, and the Royal Society and the Astronomer Royal over everything, meant that none of his calculating machines were completed in his lifetime. All that were built were small test devices that could be used as demonstrators in Babbage’s ceaseless quest for finance.
In the early 1800s, computers were human and fallible. They compiled reference books of tables of astronomical data, vital for the star and sun positions needed for ships’ navigation, and the logarithmic tables that simplified long division and multiplication. Unfortunately, errors were common and could in extreme circumstances lead to shipwrecks and lost lives.
Printing of books was by letterpress, with metal type that was set by hand from individual letters and numbers. Errors could creep in at several stages: the distribution of the type into type cases; the picking of type by the compositor, the missing of errors by proof readers; and loose letters being pulled out on the press and replaced in the wrong places.
If the calculating and typesetting stages could be automated to remove human error, accuracy could be guaranteed and lives would be saved. The only losers would be the compositors and their employers.
The story goes that in 1812 Babbage was examining astronomical tables in Cambridge with his astronomer friend John Herschel. Despairing of the errors they found, Babbage exclaimed: “I would that this were done by steam!” The quote may be dubious historically, but by 1822 Babbage had conceived the idea of an automatic mechanical calculator to derive logarithms and other complex numbers. He called it the Difference Engine because the method of calculation it used is called ‘divided difference’.
The first design was very complex and was abandoned in 1834 following a dispute over money with Babbage’s chief parts manufacturer Joseph Clement. A small demonstration section is in the Science Museum.
By then Babbage had refined his ideas to create a more versatile general-purpose computer, which was programmable through punched cards. He called that the Analytical Engine and likewise managed to get a demo section built, with printer. This is also on display in the Science Museum. It could only be built with funding from the UK government, but a lack of progress coupled with Babbage irritating powerful people, led to a withdrawal of government finance and the project stalled, though he never quite abandoned it.
Using what he had learned along the way, Babbage designed a simplified Difference Engine No2 (shortened to DE2) in 1846, finally giving up on it in 1849 and continuing to complain about his funding problems until his death.
In the 1980s an Australian, Dr Alan Bromley from the Basser Department of Computer Science in Sydney, decided that Babbage’s engines could be built in the 20th century. In 1985 he met up with Doron Swade, who was newly appointed as curator of computers at the Science Museum in London. To cut a long story short the pair worked hard to get approval and funding to build a working Difference Engine No.2 based on Babbage’s extensive original plans and copious notes. The actual engineering and assembly was supervised by Reg Crick.
Originally only the calculating section was constructed, being completed by December 1991 to become the star of the Science Museum’s display to mark Babbage’s 200th anniversary. The publicity this got worldwide led to US sponsorship to build a second machine, plus printing sections for both. Construction of the printer was completed in 2000. The device is 3.4m long by 0.5m wide and 2.1m tall and entirely powered by the operator who tuns a crank handle at the opposite end from the printer. Four turns are needed for each set of numbers in the calculation.
An important part of building the DE2 was documenting how it was done. Swade was responsible for this, discussing it with Crick. He produced copious academic notes and a thesis, but also wrote a popular book about Babbage’s life and inventions. This book is called The Cogwheel Brain, and was published in 1999. It’s still in print. However it doesn’t go into a lot of detail about the printer.
Swade did write about the printer in 1996, which can be downloaded and read from the following link: bit.ly/pw-babbage-de2. He says that “the printing and stereotyping section is integral to the concept of the machine and forms an essential part of the control system. The apparatus is at least as complex as the calculating section and requires an additional 4,000 parts – about the same number as the calculating section already built, but with substantially less repetition of similar parts.”
How the DE2 printer works
The drive for the printer/stereotyper is taken from a shaft linked to the main hand-cranked cam shaft via a bevel gear. There are 31 numbering wheels taking positions from the calculating section, with raised metal wrong-reading numbers from 0-9 on each.
There are two output media: paper and two sizes of stereotype pages. The paper printer has an inking system for the numbering wheels which are pressed onto the roll of paper after each calculation, then the paper is advanced. This could be used if only one copy was needed, to check the machine was working and for a permanent hard copy.
The other output method is the main purpose of the whole Difference Engine. It creates stereotypes, meaning moulds from which one-piece metal letterpress plates could be cast, with no chance of errors creeping in by transcribing or loose type.
Stereotypes were relatively new in Babbage’s day, but enjoyed a long life in the newspaper industry until well into the 1980s. In the DE2 there is a page-sized tray that contains a flat layer of soft material. Plaster of Paris was planned, or alternatively wet papier-mâché.
In the DE2 the tray was held on a carriage below the metal numbering wheels and was advanced a row at a time as the numbers changed. There was provision for basic page layouts. Interchangeable wheels could give either a single page-wide column, or two varieties of twin columns: side-by-side numbers or automatically wrapped columns. Inter-line spacing could also be varied with different wheels. When a tray was full a mechanism disengaged the cranking handle so the machine stopped until the tray was replaced by one with fresh plaster/paper.
Completed trays would have been taken to a casting room. Molten lead alloy would be poured on top. When it cooled and set, the plaster would be removed (probably being chipped and brushed off), leaving a full-page relief printing plate. After shaving the back to make it level, it could be mounted on a wooden block and loaded into the press.
If papier-mâché had been used, it could be peeled off the solidified plate and re-used. There was apparently also a plan to impress the numbers into a sheet of soft metal, such as copper.
Babbage did manage to have one printer built in his lifetime. This was a part of the small demo/test section of the Analytical Engine. This can be seen in the Science Museum, not far from the DE2. The printer is smaller than that on the DE2, with 26 rows of numbers. Babbage also had plans for a curve plotter, which would have been the first computer graphics printer.
Babbage’s efforts can be counted as wasted in some ways: they had little influence on the development of mechanical adding machines, or indeed the Linotype and Monotype mechanical hot metal typesetters later in the 19th century. It took until the 1940s for recognisable computers to be built and they used electronics not moving metal parts.
The Difference and Analytical Engines might have changed the world, but didn’t. Since the 1980s this has inspired a genre of science fiction literature called cyberpunk, that imagines alternative realities where Babbage and his brilliant collaborator Lady Ada Lovelace succeeded and triggered a Victorian computer age. A good example is the 2015 graphic novel The Thrilling Adventures of Lovelace and Babbage by Sydney Padua. Her alternative Victorian London has comic takes on real characters such as Queen Victoria, Isambard Kingdom Brunel, Mary Ann Evans (aka George Eliot), the Duke of Wellington (and his horse), woven in with clever explanations of how the Analytical Engine was supposed to work.
Grateful thanks to Doron Swade, who provided reference material, some unpublished, to help with our simplified explanation of the very complex printers. If you need entertainment over Christmas, visit the Science Museum to view Babbage’s visions, or buy the books.