Who Was Ada Lovelace? The First Programmer, and the Argument About It

Augusta Ada Byron was born in 1815, the only legitimate child of the poet Lord Byron, who left the country permanently five weeks later and never saw her again. Her mother, herself mathematically educated and called by Byron the princess of parallelograms, had her taught mathematics and science intensively, in a deliberate effort to suppress any inherited poetic tendency. She was tutored by Mary Somerville, one of the leading scientific writers of the period, and through her met Charles Babbage in 1833 when she was seventeen, at a party where he demonstrated a working section of his Difference Engine. She married William King, who became Earl of Lovelace, had three children, corresponded with Babbage and with the logician Augustus De Morgan, who taught her advanced mathematics by post, and died of uterine cancer in 1852 at thirty-six, the same age as her father.

Babbage designed two engines and completed neither. The Difference Engine was a mechanical calculator for producing mathematical tables by the method of finite differences, intended to eliminate the human errors that made navigational and actuarial tables unreliable; a section was built, and a full one constructed to his drawings by the Science Museum in London in 1991 works exactly as specified. The Analytical Engine, designed from 1834, was something else entirely: a general-purpose machine with a separate store for numbers and a mill for performing operations, controlled by punched cards borrowed from the Jacquard loom, capable of conditional branching and of looping. That architecture is recognisably a computer more than a century before one was built, and Babbage never obtained the funding or the manufacturing precision to construct it.

An Italian engineer, Luigi Menabrea, published a French account of a lecture Babbage gave in Turin, and Lovelace was asked to translate it. She produced the translation with seven notes of her own, labelled A to G, running to around twice the length of the source and containing the substance she is remembered for:

• •Note G contains a detailed table of operations for computing Bernoulli numbers on the engine, including a loop structure and the reuse of intermediate variables, which is the document usually described as the first published algorithm intended for a machine
• •Note A argues that the Analytical Engine differs in kind from a calculator, since the Jacquard cards mean it weaves algebraic patterns as the loom weaves flowers and leaves
• •She proposes that if the relations of pitched sounds could be expressed and adapted to the engine, it might compose elaborate pieces of music of any complexity, which is the earliest suggestion that a computer could handle non-numerical material
• •She distinguishes clearly between the machine's mechanical operation and the analysis required to set it up, which is the distinction between hardware and programming
• •Note G also states what became known as her objection, that the engine has no pretensions to originate anything and can do only what we know how to order it to perform, a line Alan Turing quoted and answered directly in his 1950 paper on machine intelligence

Her status is genuinely contested and the dispute is worth understanding rather than resolving glibly. The sceptical case, argued most forcefully by the historian Bruce Collier and by others since, is that Babbage had already written programs for the engine, unpublished, that his correspondence shows him supplying material and corrections, and that the Bernoulli table was substantially his work with her doing the exposition. The supportive case notes that the surviving correspondence shows her finding a mistake in Babbage's own working, that Babbage himself described the notes as her production and praised them, that the conceptual claims in note A are absent from anything he wrote, and that the question of who first wrote a program is less interesting than who first understood what the machine was. The reasonable position is that Babbage built the architecture and wrote earlier unpublished sequences of operations, and that Lovelace published the first substantial program with a full explanation and was the first person to articulate what general-purpose computation would mean.

The notes were largely forgotten for a century. Alan Turing read them while writing his 1950 paper and engaged with her objection seriously, which brought the work back into view, and B. V. Bowden reprinted them in 1953. Her reputation grew from there and accelerated considerably from the 1980s, when the American military named its new programming language Ada in 1979 and when the shortage of women in computing made a nineteenth-century female figure useful to point at. That usefulness has cut both ways, since the reaction against what some historians see as inflation of her role has occasionally been as unmeasured as the claims it answers. What is not in dispute is the content of note A, which describes a machine that manipulates symbols rather than quantities, at a moment when the machine did not exist and nobody else had said it.

Taught mathematics intensively by a mother determined to suppress any inheritance from Byron, she met Babbage at seventeen and in 1843 translated an Italian account of his unbuilt Analytical Engine, appending notes three times longer than the original. They contain a detailed operation table for computing Bernoulli numbers with a loop, and the claim that such a machine could work on any symbols, including composing music. Whether Babbage wrote earlier programs is disputed, and the conceptual argument about general-purpose computation is hers.