The Mechanical Turk
Innumerable pamphlets followed the Turk's progress across Europe to London in the mid-1780s, where, having already been used in Vienna to bemuse aristocratic visitors to the Imperial court, and in Paris to contest the mastery of the chess wizards at the Café de la Régence, it now rivalled the shows of Merlin and Maillardet in Mayfair. The Turk's arrival in western Europe coincided precisely with that of another Viennese guru, Franz Mesmer - while von Kempelen had the ability to build an automaton which displayeé human intelligence, mesmeric s/ances seemed able to reduce the most rational humans to the condition of automata. Indeed, von Kempelen seems first to have built the Turk to distract attention from Viennese interest in the phenomena of animal magnetism. Enlightened philosophers drew the appropriately arrogant moral: 'these days', claimed a gossipy German periodical in 1783, 'physics, chemistry and mechanics have produced more miracles than those believed through fanaticism and superstition in the ages of ignorance and barbarism'. Others simply reckoned that von Kempelen must be dealing with the devil. Like Mesmer, however, the Turk was also the target of committed exposés. 'The machine cannot produce such a multitude of different movements, whose direction couldn't be foreseen in advance, without being subject to the continual influence of an intelligent being'. Some London commentators immediately alleged the automaton contained a child, or a dwarf, inside the box, without ever quite managing to explain where the diminutive prodigy lay hidden, nor did the hostile stories which appeared in such profusion yet damage the enthusiasm of the automaton's public.
On von Kempelen's death in 1804, the Turk was soon bought by a brilliant Viennese musical engineer, Johann Maelzel, court mechanician for the Habsburgs and a close ally of one of their favoured composers, Beethoven. Maelzel swiftly saw the patronage he could win by trading on von Kempelen's automaton, and the Turk became a temporary habitué of the new Napoleonic courts in Germany. E. T. A. Hoffmann, a fellow musician, found the figure of a mechanical Turk a suitably exotic subject for his pen, and in 1814 sent a Leipzig musical magazine a story entitled The Automata, in which he 'took the opportunity to express myself on everything that is called an automaton', teasingly hinting that the Turk might work by setting up a musical harmony with the mind of its audience. Hoffman used his story to debate the most up-to-date views of occultist German philosophies of nature, much devoted to the inner rhythms of human mental life, then turned his attention to the equally modish attempts to mechanize musical composition. Meanwhile, Maelzel threw himself into another lucrative mechanical project for regulating musical performance with a device he baptized the 'metronome'. The metronome was, of course, a rather more potent means of mechanizing and standardizing artistic creativity than any mere chess-player would ever be: 'an universal standard measure for musical time is thus obtained', chorused the musical journalists, 'and its correctness may be proved at all times by comparison with a stopwatch'. After furious patent suits with rival inventors, and complex negotiations with Beethoven, Maelzel established himself as the monopoly distributor of these newfangled musical timekeepers, re-purchased the Turk from the Bavarian court, and then, in 1818, set off on a marketing and publicity tour of Paris and London.
Maelzel's London show was very carefully staged. In autumn 1818 at Spring Gardens, in a pair of candlelit drawing rooms equipped with sloping benches, he displayed the Turk alongside a fine mechanical trumpeter, and when he soon moved round the corner to a larger chamber in St James' Street he added a moving diorama of the burning of Moscow, 'in which Mr Maelzel has endeavoured to combine the Arts of design, mechanism and music so as to produce by a novel imitation of Nature a perfect facsimile of the real scene', and a set of automatic rope dancers, 'scarcely to be distinguished from those of a living performer', moving 'with the utmost and correctness without any apparent Machinery'. Maelzel also helped realize his metronomic dream of a completely automatic orchestra, the Panharmonicon of 42 mechanized musicians, for which Beethoven had specially composed his ghastly Battle Symphony, a composition rather likely to appeal to jingoistic British audiences. One London paper praised such an orchestra which 'displayed none of the airs of inflated genius, but readily submitted to being wound up'.
The wind-up Turk, of course, occupied pride of place amidst these other wonders. Maelzel faithfully followed von Kempelen's recipe, imitating precisely the ritual opening of the cabinet front then back, moving a candle round the interior, and winding up the mechanism at regular intervals. And his metropolitan audience faithfully reproduced their earlier enthusiasm for the show. Ever considerate to this public, Maelzel even announced that the Automaton would purposely make bad moves so as deliberately to lose if the company seemed bored with over-lengthy games, while the Turk's opponents were ordered to move as fast as possible to alleviate the tedium. A pamphlet authored by a pseudonymous Oxford graduate alleged the whole trick relied on a hidden piece of wire or catgut: 'it seems to be a thing absolutely impossible', the Oxonian alleged, 'that any piece of mechanism should be invented which possessing perfect mechanical motion should appear to exert the intelligence of a reasoning agent'. Unmoved by this futile revelation, during the summers of 1819 and 1820, when the London 'season' ended, Maelzel took his show to the provinces and to Scotland, with apparently equal success. Back in St James' at the end of 1820, however, his nemesis was ready at last.
Robert Willis was an ingenious young Londoner, heir to a distinguished medical family - his father famously attended George III during the monarch's madness. In later life Willis himself became, like Babbage, a pre-eminent Cambridge mathematician, then distinguished professor of applied mechanics and untiring surveyor of ecclesiastical architecture. He also produced one of the best mechanical analyses of the principles of speaking machines, including those von Kempelen had tried to build. During 1819, still a teenager, he patented a new mechanism for harp pedals and toured all the major instrument-shops, including high-class machinists such as Holtzapffel and Bramah, and John Newman's works, supplier of equipment to many of the best scientific lecturers in the city. Early in the year he noticed a telling advert placed in the London papers by one 'Monsieur Novoski' of Knightsbridge Post Office, who offered to sell the secret of the chess-player for 2000 guineas. Willis took up the challenge at once. He bought copies of the games the Turk had played and won. He made sure to visit Maelzel's show while it occupied the cramped space of Spring Gardens, 'more favourable to examination as I was enabled at different times to press close up to the figure while it was playing'. Then he smuggled an umbrella into the room so as to measure 'with great accuracy' all the dimensions of the Turk's celebrated cabinet. He went back to St James' for the 1820 season, and by the autumn had completed an obsessively detailed analysis of just how the 'automaton' really worked.
Willis' accurate umbrella and his command of wheel-work made all the difference. The chest, he demonstrated, was much larger than it seemed, giving more than enough room for a fully grown (and doubtless experienced) human chess player to fit inside. 'Instead of referring to little dwarfs, semi-transparent chess boards, magnetism, or supposing the possibility of the exhibitor's guiding the automaton by means of a wire or piece of catgut so small as not to be perceived by the spectators', Willis' Attempt to Analyze the Automaton Chess Player, finished in December 1820, proffered the simplest possible scheme of the Turk's hidden intelligence. The noisy gear-wheels were there simply so that their sound would conceal any noise made by the concealed player. Even more influentially, he ponderously laid down the law of mechanism's limits: 'the movements which spring from it are necessarily limited and uniform, it cannot usurp and exercise the faculties of the human mind, it cannot be made to vary its operations so as to meet the ever-varying circumstances of a game of chess. This is the province of intellect alone'. Despite the care with which he drew the dramatic plates for his pamphlet, there was still some continuing debate about Willis' story: for example, while he reckoned the hidden chessplayer put his arm inside that of the Turk, more reflective analysts guessed that the machine must use the kind of pentagraph which Babbage described in the case of drawing automata. These debates received their fullest publicity in the best-selling Letters on Natural Magic (1832) by the Edinburgh optical expert David Brewster. Maelzel himself left London, tried to sell the Turk in Paris, then went to the United States in 1825. Willis' pamphlet was frequently reproduced there by newspapers eager to exploit public interest in the automaton. After Maelzel's show had stunned Richmond and Baltimore, the local writer Edgar Allan Poe lifted Willis's report directly from Brewster, brazenly passed it off as his own brilliant detective job in the Southern Literary Messenger (1836), then used it as a precedent for a whole series of rather more original, and certainly better-known, stories about the application of analysis to cunning mysteries, whether purloined letters or concealed bodies. Poe never forgot von Kempelen, nor his deceit, for in the midst of the California goldrush Poe teasingly announced that a New Yorker of that name, doubtless connected with the Slovak engineer, had discovered how to transmute lead into precious metal. If the Turk's progenitor had this shadowy afterlife, its promoter Maelzel never came back to Europe - he died on board ship off Cuba in 1838 - and nor did the Turk, who went up in flames in Philadelphia scarcely fifteen years later.
Moralists found the story of the chess automaton irresistible. The pious drew the obvious implication that if a genius like von Kempelen had not been able to build a rational machine, what must be the skill of the divine Creator who had pulled off this trick ? Others sang the praises of the Turk's hidden managers, now revealed as William Lewis and Jacques Mouret, who'd played chess so well and under such apparently overwhelming disadvantages in Britain between 1818 and 1820. George Walker of the Westminster Chess Club, writing in Fraser's Magazine in 1839, mourned Mouret's 'beautiful emanations of genius' when the Frenchman 'burnt out his brain with brandy' and died in Paris 'reduced to the extremest stage of misery and degradation'. It was a commonplace that such machines belonged at court, whether in the Orient of the Arabian Nights or the grandiose palaces of the Tzars. 'To the half-bred savages of the north', Walker sneered, 'the exhibition could not fail to be striking'. Novels and reviews told how the Turk had conned the powerful and humbled the great: 'even Bonaparte, who made automata of Kings and Princes at his will, was foiled in an encounter with the automaton chess player'. Later in the century, a successful French play put the Turk on stage in a victorious contest with Catherine the Great. It was rather the point, so populist writers explained, that self-styled experts, the politically powerful, and the superstitious mob, could all be deceived by a mechanism effortlessly unmasked in public prints directed at a new and confidently rational readership. 'Had the gulled mob reasoned on the matter earlier', Walker noted in dangerously republican terms, 'King Automaton would have been speedily deposed from his high places'.
But the most telling lesson of the Turkish chess-player was the relationship between machine intelligence, technological progress and the puzzles of concealment. This was a moment when, as the Automaton's admirers never hesitated to remark, 'political economists amuse themselves and the public with the nicely-balanced powers of man as a propagating and eating animal and philosophers and divines often assure us that he is, in other and higher respects, but a machine of a superior description'. Correspondents from the London chess clubs predicted that 'a man inside will most assuredly never again work the charm, but, advanced as science is during the present generation, a Brunel or a Stephenson could easily and successfully vary the deception'. And in the 1830s one economic journalist, describing the rapid growth and progress of automation in the Lancashire cotton industry, told the apocryphal story of the invention of the power loom, half-a-century earlier, by Edmund Cartwright. Cartwright had allegedly seen the Turk in London, trusted its purely mechanical origin, and thus been thoroughly convinced that a weaving machine could scarcely be harder to make than one which could play chess so well. Rather similar stories appeared in Poe's favourite source, Brewster's Natural Magic, which was dedicated to Brewster's close friend Walter Scott and entirely devoted to teaching his fellow-citizens the inner secrets on which all apparently miraculous and surprising mechanical devices really depended. Part of the point was characteristically Presbyterian: gaudy tricks conned the ignorant into idolatry. Part, however, was economic. In his chapter on automata, Brewster devoted pages to von Kempelen's chess player, summarised the other notorious stage-shows of the age, then moved straight to Babbage's calculating engines themselves. 'Those mechanical wonders which in one century enriched only the conjurer who used them, contributed in another to augment the wealth of the nation. Those automatic toys', he concluded, 'which once amused the vulgar, are now employed in extending the power and promoting the civilization of our species'. Apparently theatrical automata really had inspired the industrial revolution.
Surrounded by stories which made the intimate link between the plausibility of mechanizing intelligence and the reality of automatic manufacture, it was scarcely surprising that Babbage's new calculating engines, first seriously proposed to the new Astronomical Society in 1820 and first publicized in a pamphlet of summer 1822, should raise in such a lurid way the puzzles of mechanical intelligence which the Turk had just dramatised. It was also rather predictable that his reflexions on the intelligence of calculating machines should eventually culminate in a detailed analysis of whether chess could be reduced to a program and an engine whose inner workings would be completely hidden from view behind a gaudy exterior. In his public letter to the President of the Royal Society, Humphry Davy, written barely eighteen months after Willis' pamphlet on the chess automaton, Babbage conceded that his own plans for a machine to 'substitute for one of the lowest operations of human intellect' might 'perhaps be viewed as something more than Utopian, and that the philosophers of Laputa may be called up to dispute my claim to originality'. He was, as usual, absolutely right.