Jacquards' Web (29 page)

By the early
1880
s, however, the continued pace of industrialization was prompting inventors to look again at the problem of finding a mechanical calculation aid. They knew that anyone who wanted to build a better calculator than the arithmometer would have to solve three crucial challenges. Firstly, there was 206

Howard Aiken dreams of a computer

the urgent need to speed up the rate at which numbers could be entered. Secondly, the speed of calculation had to be dramatically improved. Thirdly, if possible the machine needed to supply a permanent record of the result of the calculation.

This last point was particularly important for banks, which were obliged by law to keep a permanent record of customer transactions.

For some inexplicable reason, many pioneers in automated calculation have curious names. Babbage was one perhaps.

Another was Dorr E. Felt, a Chicago-based engineer who had the idea of building a key-driven calculating machine. Felt’s innovation was probably inspired by the typewriter. The idea appears obvious now, but at the time it seemed an extremely ingenious and innovative development.

Felt used keys to enter specific numbers instead of laboriously entering them by turning a wheel or by pushing against a geared rack with a stylus. This approach offered two great advantages: speed and convenience. The keys of his machine—which he christened the ‘Comptometer’—were arranged in columns, labelled from
0
to
9
in each column. Mechanically the Comptometer was relatively primitive. It necessitated ten keys (i.e.

0
to
9
) for each column of integers. So, for example, a calculator designed to handle numbers up to
99 999 999
would need eight columns of ten keys.

Still, this machine was a big advance on any previous calculator. Yet its usefulness was limited. Its main function was to add and subtract; multiplications or divisions still had to be done manually by breaking down each component of the calculation into additions or subtractions. In
1886
a new company, Felt and Tarrant, started manufacturing the Comptometer on a production line. As the nineteenth century turned into the twentieth, Felt and Tarrant were selling
1000
Comptometers every year.

Gradually the pace of invention increased. By around
1910
, the best key-driven calculators had become relatively sophisticated 207

Jacquard’s Web

machines that used a combination of keys and gear-wheels to complete the calculation and in some cases even employed bells to indicate to the user when the calculation was complete.

Generally the user was obliged to read off the result from a little

‘window’ in the machine and transcribe it. However, some calculators were fitted with paper rolls so that the result could be printed using a printing mechanism.

Despite advances in calculator manufacturing technology, calculators still used the cumbersome, slow method of repeated addition in order to carry out the all-important function of multiplication. The first calculation machine that could undertake multiplication and division properly instead of by repeated addition was known, optimistically perhaps, as the ‘Millionaire’.

The Millionaire was about the size and weight of a small suitcase that had been filled to the brim with cogwheels. This, incidentally, was basically what it was. Its users, however, still belonged to a generation that was not especially mobile. For them, portability was not a major concern. The Millionaire sold well; by
1912
more than
2000
were in use. Two years later a motor-driven version became available. This model was also a success.

Mechanical calculators were put to use compiling a new generation of mathematical tables used to engineer all kinds of large-scale projects, including the great ocean liners of the first two decades of the twentieth century. For example, the marine engineers who built
Titanic—
at the time of its maiden voyage in
1912
the largest man-made object on the planet—relied on a variety of mechanical calculation aids at every stage of the work.

Naturally, the great ship’s unfortunate fate does not detract from the remarkable engineering achievement it represented.

Nor were mechanical calculators the only aid to calculation available in the early twentieth century. Hollerith tabulation machines, too, could be used for calculation purposes. Tabulators 208

Howard Aiken dreams of a computer

were by their very nature adding machines, and IBM eventually devoted considerable effort to modifying tabulators to maximize their usefulness for calculation purposes. By the
1930
s it was routine for automatic tabulators to handle a wide range of calculations.

An additional resource that had become available during the late nineteenth century was the slide rule. This was still essentially a mechanical way of calculating, usually by moving two calibrated scales against each other. Slide rules were first invented in the
1860
s, becoming popular about twenty years later. The best ones were able to carry out sophisticated arithmetical functions including multiplication, division, extraction of square roots, and even calculation of trigonometric functions and logarithms. But the slide rule’s accuracy was restricted due to the problem of parallax (meaning that a precise reading varies according to the eye’s position in relation to the calibration) and also because the accuracy of the result depended on the precision of the slide rule’s calibration. There was, and is, always a limit to how precisely slide rules can be calibrated.

Despite the progress that had been made, Charles Babbage’s ambition still remained well out of reach even during the
1930
s.

The problem was that if Babbage’s dream of
automatic
calculation were to come true, a new kind of technology would have to come into play. In a similar way, for example, the invention of hot-air balloons did not make it possible for people to travel to the Moon; that had to wait until the rocket engine was perfected.

Even by
1940
, a device had yet to be invented that could undertake any type of mathematical calculation a user might reasonably want to compute.

In particular, there was no machine available to carry out tasks needing extensive and prolonged calculation. Some calculations could, in theory, have been enacted by clerks using state-of-the-art arithmetical calculating aids, but it would have taken many weeks for some of the more complex calculations to be completed, and many calculations could never have been 209

Jacquard’s Web

realistically handled manually at all, even if years could have been devoted to the task, because the calculations were simply too demanding.

During the
1920
s and
1930
s there was increasing and justified concern among scientists, mathematicians, engineers, and many others that the lack of reliable automatic calculation machines presented a serious obstacle to continued progress in all the sciences, and of course also in mathematics itself. The concerns expressed in the
1820
s about the inaccuracy of mathematical tables had returned, and with renewed force. Yet still there was no machine available that could placate them.

The first person in the twentieth century to try to solve the problem was a man called Howard Aiken. Aiken is as important a character in the story of Jacquard’s Web as Joseph-Marie Jacquard himself, Charles Babbage, or Herman Hollerith. Aiken’s work provides the final connection between the Jacquard loom and the modern computer.

Howard Hathaway Aiken was born in Hoboken, New Jersey on
8
March
1900
. His father, Daniel H. Aiken, came from a wealthy and well-established Indiana family. His mother, Margaret, was a child of German immigrants. Howard was their only child. When he was still a boy, he moved with his parents and maternal grandparents to Indianapolis, the capital of Indiana and about
150
miles south of Chicago.

Life was difficult for the family. Daniel Aiken was an alco-holic and would often beat his wife. During one such episode, young Howard—already big and strong at the age of twelve (he would eventually reach six foot four)—grabbed a poker from the fireplace and drove his father out of the house. The family never saw him again.

Unfortunately, yet perhaps not surprisingly, Daniel’s wealthy relatives responded to what Howard had done by declaring that they would have nothing more to do with either Howard or his 210

Howard Aiken dreams of a computer

mother. For much of his adult life, Howard worked to support his mother and his maternal grandparents. None of the relatives he supported seemed to think of actually working to support
him
.

The possibility of Margaret Aiken engaging in remunerative work did not arise. In those days a middle-class ‘lady’ was usually unskilled and could not simply go out and find work. More to the point, she would have considered it an appalling social disgrace to have done so.

Aiken left school when he was still a teenager so that he could continue to provide for his mother and her parents. He found a job installing telephones. Later in life he enjoyed telling friends that he had installed all the telephones in the red-light district of Indianapolis. Despite the full-time commitment of his job, he did not abandon his education, but started taking correspondence courses in the subjects he enjoyed. One of his former school teachers went to see Margaret Aiken and pleaded with her to let him return to school full-time, but the family’s financial circumstances meant that this was impossible. The teacher sought another solution. He found Aiken a job as an electrician’s assistant for the Indianapolis Light and Heat Company.

In his new job Aiken worked a night shift so he would be able to attend school during the day. Probably few ordinary students would have been able to cope with such pressure, but Howard was very far from being ordinary in any respect. He not only graduated from high school, but continued his correspondence courses, too.

Throughout his career Aiken had a particular gift for winning the help and support of teachers and officials who got to know him. By all accounts he was an inspiring student, and others wanted to do their best for him. This talent, which stayed with him throughout his student days and beyond, played an important part in his success.

One of the officials who came to have the highest opinion of Aiken was the local superintendent of public instruction, one Milo Stewart. Because Aiken’s need to work in paid employment 211

Jacquard’s Web

had left him short of some of the school credits he needed in order to go to university, Stewart created a special examination so that Aiken could get these credits.

Stewart also wrote to every Midwestern public utility in a university town asking it to employ the young man. Aiken was eventually offered a job as night-time telephone operator by the Madison Gas and Electric Company, based in Madison, Wisconsin. It was because of this job offer that he chose to go to the University of Wisconsin. Aiken was twenty when he moved to Madison with his mother. He enrolled in an electrical engineering course at the University of Wisconsin, working at his new job from four o’clock in the afternoon until midnight and attending university during the day.

Edward Bennett, the chairman of the University’s electrical engineering department, had a huge influence on the young Aiken. A quarter of a century later, in a letter Aiken wrote to Bennett inviting him to the ceremony marking the dedication of the world’s first electromechanical computer, Aiken wrote, ‘I sincerely hope it will be possible for you to be with us, for in a large part, successful completion of this machine was due to the careful preparation which I had as your student.’ He signed the letter ‘Respectfully yours’.

In a lecture he was to deliver in
1955
both in Sweden and in Germany, Aiken referred to the importance of Bennett’s teach-ings in the shaping of his own intellect. Aiken recalled that Bennett had taught him how the development of any new body of knowledge passed through four stages.

The first stage, Aiken said, was
observation
, when the investi-gator knows almost nothing about the subject and so can do little else than make observations from nature and observe new facts.

To take the study of electricity and magnetism as an example to illustrate Aiken’s point, one might say that the study of these subjects had been in the observation stage throughout most of history until Michael Faraday started his important and hugely influential work in the early nineteenth century.

212

Howard Aiken dreams of a computer

The second stage was
classification
, when the observer has sufficient facts to be able to separate them, place them under major headings, and rank the facts in the order of their importance. The study of electricity and magnetism engineering had reached the classification stage by about
1880
, following major discoveries made by Faraday and other pioneers.

The third stage,
deduction
, was when a new science is born.

By the start of the twentieth century the study of electricity and magnetism had become a science of its own:
electrical engineering.

Electrical engineering was a thrilling science to be studying at the start of the twentieth century, as much an exciting field of exploration as physical exploration by sea was for mariners in the fifteenth and sixteenth centuries. Electricity had yielded up many of its secrets. The more research there was, the more it became clear that the potential for applying electricity in everyday life was fabulously exciting and possibly close to unlimited.

Furthermore, the speed of electrical processing was creating the possibility that new, powerful types of machines and communication devices might be made which had the potential to revolutionize technology in every aspect of life. In Bennett’s terminology, the deduction stage was when it became possible to deduce new facts from old. Knowledge would have attained what was, in effect, a critical mass and this fertile bed of knowledge would have created a new discipline in which all sorts of new ideas and initiatives could flourish.