It Began with Babbage (6 page)

There was more to the Analytical Engine than this philosophical advance. There was technological promise. A new language was at hand to manage the “truths” of mathematical analysis in a far speedier, more accurate, yet practical manner than had ever existed before. The Analytical Engine promised not only to bridge the material and the mental, but also the theoretical and the practical.
¹⁷

The use of punched cards to communicate the sequence and nature of the operations to the machine led to other possibilities that added to the machine's universality. For one, the initial values of the variables (numbers) on which to be operated could be set either manually in the store or by way of additional cards that Babbage called “variable cards”. More interestingly, it was possible in Babbage's scheme to
iterate
through a sequence of operations. For this purpose, he proposed a combination of what he called “combinatorial cards” and “index cards”, the roles of which were to direct, at specific intervals,
the return of variable cards to particular “places” in the computation, and to direct the number and nature of the repetitions to be performed.
¹⁸
From a present-centered perspective, Babbage's index cards were the forerunners of what later came to be called index registers.
¹⁹

Babbage did not, of course, just conceive the architectural principles for his machine. In the long years he devoted to the project—from 1833 until 1849, and then again from about 1856 until his death in 1871
²⁰
—he obsessed incessantly over its actual physical design and implementation. He kept extensive notes, now in the collections of the Science Museum in London. The massiveness of the documentation is impressive: 300 sheets of engineering drawings; 600 to 700 representations of the mechanisms in the form of (what we would now call) timing diagrams, flowcharts, logic diagrams, and walk-throughs; and 6000 to 7000 pages of notes.
²¹

Of course, the Analytical Engine, like the Difference Engine, was a mechanical machine, a complex of gears, wheels, chains, levers, rack-and-pinions, weights, barrels, pulleys, and the like.
²²
Babbage's was a mechanical world
²³
; yet, despite this, his work illustrates vividly that there are logical, conceptual, organizational, and behavioral (in other words, what in present-centered language we would call architectural) principles that can be abstracted out of the purely physical, purely technological. To put this in another way, there was something that computer scientists of the late 20th century would recognize: what makes a digital computer a digital computer is not the material or the kind of physical technology out of which it is built, it is its conceptual, organizational, and architectural principles.

One person who studied Babbage's Analytical Engine documents in great detail was Australian computer scientist and Babbage scholar Allan Bromley. Through his eyes we get a detailed sense of Babbage's mechanical world. Bromley, however, belongs to what is called the age of electronic computers. To describe the machinations of the Analytical Engine, he
had
to translate Babbage's ideas into present-centered language. He could do this, fortunately, because the concepts Babbage invented had direct modern correspondences, which is why Bromley could say (using modern-day jargon), that the Analytical Engine was a “decimal machine” and that it used the “sign-and-magnitude” scheme to represent numbers.
²⁴

A particularly vivid instance of the modernity of Babbage's ideas was discussed by the British electronic computer pioneer (later, Sir) Maurice Wilkes (on whom much will be said in later chapters), who was probably the first to examine Babbage's notebooks on the Analytical Engine (in 1971, the centenary of Babbage's death).
²⁵
One of Wilkes's discoveries was that Babbage's design of what we now call the computer's
control unit—
as the name suggests, that part of the computer that directs the very basic step-by-step actions to carry
out the operations encoded in the punched cards (the program in present-centered language)—bears a remarkable resemblance to the very technique Wilkes invented in 1951. Babbage had anticipated, a good 100 years earlier, the principle of microprogramming.
²⁶

A significant portion of the Difference Engine was built by Babbage, and there was the Scheutz implementation.
²⁷
Alas, the Analytical Engine was never built, not even as a prototype, not even some substantial portion of it, although Babbage made a number of experimental components.
²⁸

We have seen that government funding of the Difference Engine (totaling about £17,000
²⁹
) evaporated completely by 1833. Thereafter, Babbage could not interest the government in financing his work on the Analytical Engine. Without such support, there was never a possibility of the machine being built. As it was, he had spent some £20,000 of his own money on the Difference Engine.
³⁰
Yet, during the first 16 years (1833–1849) that he devoted to the project and then again during the later period (1856 until his death), he was at it constantly, ideating, thinking, refining, elaborating, conceptualizing, obsessing. There is something quite poignant in our imagined vision of this man, working in complete isolation, obsessed with an idea, the archetypal lonely inventor, the solitary scientist of Romantic lore,
³¹
churning out drawing after drawing, in their hundreds, note after note, in the thousands, conceiving new ideas, designing new mechanisms, for a machine he must have realized would never get built in his lifetime. There lies, perhaps, the poignancy of his venture. Babbage was leaving something
for posterity
, a time capsule to be retrieved in the future—as, indeed, it was, by such Babbage scholars as Martin Campbell-Kelly, Bromley, and Wilkes.

Babbage's work on the Analytical Engine is an example
par excellence
of what it means to be a designer–inventor rather than a craftsman–inventor. His vast outpouring of documents on the project not only represents a design activity of almost epic proportions, but also demonstrates the separation of designing from making, of the role of designs as blueprints, of the separation between designer and builder.

This separation of designing from making marks the distinction between someone like Babbage and the traditional craftsman (for example, even in Babbage's time in Victorian England, the wheelwright
³²
). For the craftsman, designing and making are inextricably intertwined: one designed while making and one made while designing, and, indeed, one made while using and one used while making.
³³
In separating designing—and the creation of drawings and explanatory and descriptive notes
are
acts of design—from building, Babbage left a blueprint for posterity. In fact, more excitingly, he left behind a
theory
for, and of, the Analytical Engine. We find in Babbage's work an idea that I have already mentioned (see Prologue, section VI), to which I return to later, that has a vital place in the history of computer science: the idea of designs-as-theories.
³⁴

We cannot end Babbage's story without further mention of Lovelace, whom we have, so far, encountered only cursorily as the translator of Menabrea's 1842 memoir on the Analytical Engine, and one who added a lengthy gloss on the translated article. The exclusion of women from intellectual establishments such as universities and scientific organizations is a well-known feature of the Western intellectual tradition
³⁵
—thus, their absence in most histories of science. It would not be until the 19th century that women began to have a presence—although still a devastatingly minor one—in science and mathematics, and their minority status continued to prevail into and through the 20th century. The history of science, including mathematics, is a long, sad history of misogyny.
³⁶

There were, of course, stellar exceptions, especially in the mathematical sciences, some of whom were noblewomen who were privileged with “a room of their own” (to borrow from Virginia Woolf). In 18th-century France, for example, there was the grandiosely named Gabrielle-Emilie Le Tonnier de Breteuil Marquise du Châtelet (1706–1749), mistress to philosopher and writer Voltaire (1694–1778), who discoursed on Newtonian physics for the French Enlightenment
philosophes
, espoused Leibnizian ideas, and even embroiled herself in intellectual disputes with the likes of the mighty Swiss mathematician Leonhard Euler (1707–1783).
³⁷
Across the English Channel, born the year after du Châtelet died, the German–born observational astronomer Caroline Herschel (1750–1848)—aunt of Babbage's friend Herschel—when not assisting older brother, Sir William Herschel (1738–1822) in his discovery of the planet Uranus and unimagined cosmic universes, would go on to find many comets on her very own.
³⁸
In 19th-century Britain there was the remarkable Scotswoman Mary Fairfax Somerville (1780–1872), mathematician and natural philosopher (as physicists were then called), who wrote the first paper (on magnetism) by a woman to be read before the Royal Society and published in its venerable
Philosophical Transactions
³⁹
; who wrote
The Mechanism of the Heavens
(1831), an exposition on the mechanical philosophy of the French
savant
Pierre Simon, Comte de Laplace (1748–1827); who wrote
On the Connection of the Physical Sciences
(1834), a scientific “bestseller” in her lifetime; who became associated with Babbage's friend Herschel; and yet who would be denied fellowship in the Royal Society because she was a woman.
⁴⁰

And then there was Augustus Ada, the Countess of Lovelace (1815–1852), born Ada Byron, a daughter of Lord Byron (1788–1824). The poet and his daughter had no relationship, although they apparently pined for each other, and Byron wrote her into his long narrative poem, “Childe Harold's Pilgrimage”(1812–1818). They were buried next to each other in a church in Nottinghamshire; both died at age 36.

Married at 20 to the First Earl of Lovelace (hence her title), she showed mathematical ability at an early age. While in her teens, she came to know Somerville,
⁴¹
and was tutored by the mathematician Augustus de Morgan (1806-1871)—a pleasing coincidence
as de Morgan's celebrated theorems in Boolean algebra play a role in the later design of computer circuits (as we will see).

Babbage was a regular visitor at the Lovelace's country seat.
⁴²
When he learned that Lovelace had translated Menabrea's French article on the Analytical Engine into English, he suggested that she add notes to the English memoir,
⁴³
which she did. Her notes, appended to the Menabrea memoir, were three times as long as the original article. The most celebrated of these was Note G, the last one, for two very different reasons.

The first was because of her remark that the Analytical Engine could not initiate any task. It had no such “pretensions”; it could only do what it was ordered to do by humans.
⁴⁴
Here, Lovelace anticipated an issue that was much debated more than a century after she wrote these words (as we will see): whether computers can exhibit “intelligence” or “original thought.”

The second point worthy of mentioning in Note G is that it laid out in great detail the sequence of operations (an
algorithm
) for the computation of a mathematical entity called Bernoulli's numbers; this computation would be performed on the Analytical Engine.

Whether this algorithm was constructed by Lovelace or whether she was communicating Babbage's work remains unresolved. Babbage himself remembered that he and Lovelace discussed possible problems to use as illustrative examples. He suggested several, but ultimately the ones selected were chosen by Lovelace. As for the algorithm to compute Bernoulli's numbers, this was originally written by Babbage but there was a “grave mistake” in it that Lovelace detected and returned to him for amendment.
⁴⁵

If we accept this account, the algorithm for computing Bernoulli's numbers was really Babbage's work, not Lovelace's. Claims made about her of being “the world's first programmer” in this light are thus exaggerated. Lovelace's place in the history of computing lay in her translation of the Menabrea article, and in her lengthy gloss on how mathematical functions could be reduced to operations on Babbage's engine. And even had she been the originator of this particular algorithm, she could scarcely be called its programmer. At the very least, she would have had to write a program that
executes correctly
on an actual Analytical Engine, but such a machine did not exist physically.

Still, if not the world's first programmer, Lovelace must be regarded, along with Menabrea, as the one who demonstrated how an analytical problem in mathematics could be executed by a machine.
⁴⁶
We must also appreciate her sharp insight into the very nature of a computing machine, its scope and limits, as her comment cited earlier illustrates, and which is amplified further by her remark that a machine such as the Analytical Engine would have
consequences
for the course of science itself, for many problems in science would become resolvable and “more profoundly” tractable with the help of the Analytical Engine.
⁴⁷

She was anticipating the role of this machine as an instrument of science itself. We must also not forget that she was a poet's daughter with a flair for an elegant turn of phrase—that she could speak of the engine as a machine for “weaving algebraic patterns.”