Shadowy vision: spanners in the mechanization of mathematics (Q1776886)

This paper explores the claims made by observers of the time on behalf of Babbage's Analytic Engine that it could handle general symbolic operations. The authors find that the claims are essentially unsubstantiated and that some later writers have simply assumed, on the basis of those claims, that the Analytic Engine was closer to the modern computer than it really was. The authors find Babbage, Lady Lovelace and Luigi Menabrea convincing when, in the early 19th century, they argued that the Analytic Engine could solve two simultaneous equations in two unknowns and could compute the Bernoulli numbers. However, their descriptions were detailed, step by step, and, in any case, based on numerical calculation. But the authors are rightly very skeptical of the less substantiated claims of Lady Lovelace to the effect that the Analytic Engine could perform symbolic manipulation. The machine has never been built to this day (unlike Babbage's Difference Engine, which presented enough of a challenge) and no one has shown how the inventor's designs could allow such a usage. As the authors hint, this early example in the history of computers of overly grand expectations was certainly not the last.