The Chess-Playing Mechanical Turk

Today we're headed back in time, all the way back to the Vienna court of Archduchess Maria Theresa of Austria, in the year 1770. There the scientific polymath Wolfgang von Kempelen, then thirty six years old, brought forth a mechanical automaton: the figure of a man seated at a large wooden chess table, the cabinet below filled with clockwork. A volunteer from the audience stepped forward. Kempelen wound up the machine, and it reached out and made the first move, the clockwork whirring and ticking. The astonished volunteer was quickly defeated. Delighted with the mechanical marvel, Maria Theresa ordered many more performances. In fact, the Turk, as it was nicknamed for its Turkish clothing, toured the world for the next 80 years, defeating the world's top chess players plus luminaries such as Napoleon Bonaparte and Benjamin Franklin, without anyone ever discovering its secrets. Although many skeptics published fine articles purporting that the Turk actually contained a child, dwarf, or legless adult chess player, or that it must have been secretly controlled by its exhibitor, the workings of the Turk remained one of history's best kept secrets.

But all secrets are fleeting, and shortly before the Turk's destruction in an 1854 fire, its last owner's son, Silas Mitchell, published the revelation, proving that no skeptic had ever correctly guessed how it worked. In fact, no one had even come close. Over the years, three authors in particular had put forth the best known hypotheses, and Kempelen had fooled them all.

But the most intriguing mystery about the Turk would not turn out to be how it worked, but rather why a man like Kempelen would have built it. Kempelen was no Barnum. He was neither a showman nor a magician; he was an inventor and engineer of the highest caliber and held a series of important public works appointments in Maria Theresa's government. The last thing he'd do would be to construct some sort of sideshow trick. The first of the three most notable proposed explanations came in 1789, by Joseph Friedrich, Freiherr zu Racknitz. He wrote a book based on his many viewings of the Turk and his friendship with Kempelen. Racknitz noted that the Turk's exhibitor would always first open and close the cabinet's various doors for the audience's inspection. He concluded that a very small human operator was inside the cabinet, lying flat during the opening of the doors; and then, during game play, sat up, played the game on a small secondary chessboard, and watched magnetized needles on the bottom of the tabletop to learn what move the opponent had made. By Racknitz's measurements, the hidden human would have had to be less than five feet tall, and less than seven inches high when lying flat. Kempelen refused to offer any assessment of Racknitz's proposed solution.

In 1821, Robert Willis, an engineer of musical instruments, published a pamphlet with his own explanation of the Turk. Willis noted in particular that the order in which the doors were opened for inspection never varied. This, he proposed, was to allow a hidden human operator to move from one part of the cabinet to another, allowing the various cabinets to be shown empty in sequence. Then, to play, the operator would sit up, place his own hand inside the Turk's arm, and watch the board through the thin fabric shirt covering the Turk's chest.

The best known analysis was that of Edgar Allan Poe, published in 1835, which ultimately found in favor of Willis' explanation but differed in that it offered far deeper reasoned analysis of why it must be so. For example, Poe noted that William Schlumberger, secretary of the Turk's current owner and exhibitor Johann Maelzel, was never present during the performances and that when he was ill, the Turk did not perform. Poe also reasoned that although the Turk won most of its games it did sometimes lose, whereas a truly mechanical solution would be incapable of error. Like Kempelen before him, Maelzel refused to respond to any such guesses.

It wasn't until Maelzel himself died that the Poe family's doctor, John Mitchell, purchased the Turk at an estate auction for $400 in 1840. To finance it, the founded a club, each member of which paid $5 or $10 and was let in on the secret. At last the Turk's true workings were published in 1857 by Dr. Mitchell's son Silas. The solution? (Drum roll, please.)

The Turk was indeed controlled by a human operator hidden inside, but due to Kempelen's ingenuity in designing its interior and moving panels, there was no need for the operator to be small; any average adult would fit. This allowed Kempelen and Maelzel to hire the finest chess players available, including (as Poe had surmised) Schlumberger. No known record survives of any of the players engaged by Kempelen, but author Tom Standage managed to track down most of Maelzel's operators for his excellent book As many had guessed, the operator did move around during the display of the cabinet's interior, though on a sliding seat gliding soundlessly on greased rails. The operator did play on a small secondary chessboard inside, and did watch the game via magnetic indicators on the underside of the board above. To move the Turk's arm, Kempelen had designed what was essentially a pantograph. When the Turk moved, the operator would move a small arm over the chessman on his secondary board, twist its end to open and close the Turk's fingers, to lift and move the piece; all of the operator's movements smoothly reproduced via the armature extending through the Turk's hand. The operator also had controls to shake the Turk's head, roll its eyes, strike its fist upon the table, or make some especially noisy clockwork run to cover any noises like coughs. There was also a means of simple communication between the exhibitor and the operator, a small knob with a pointer that could be set by either person to the numbers 0 through 9, the meaning of which could be whatever they planned. This might be used by the operator to inform the exhibitor that his candle had gone out, that he wished to retire from the current game, or anything else.

The Turk was, in fact, not an automaton at all, but merely a magician's prop. It was clever and well built, but was not revolutionary and was unworthy of Kempelen's extraordinary talents and predilection for hard science. Why, then, had such a man gone to the trouble of building it, and spent so much of the latter half of his life traveling around Europe as a showman? It was completely out of character for him.

Somewhat paradoxically, the key lies in Kempelen's position as a scientist and engineer for his government. Most of his European tours were by official command. Once the word spread across Europe that Maria Theresa had an extraordinary chess playing automaton that could challenge all comers, official visitors to Austria nearly always requested demonstrations. Kempelen had dismantled the mechanism when he was ordered to bring it back for a party of Russian dignitaries. Kempelen was most distraught when he was told to take it on a two-year tour of Europe, including France, England, and Germany, all under the closest scientific scrutiny. Imagine Kempelen's state of mind: the engineer responsible for Austria's great hydraulic waterworks forced to perform an illusion for his fellow scientists worldwide, all the while having to smile and claim the device was a true automaton. His solace was his ongoing work in artificial speech machines, one of which Maelzel later adapted to make the Turk say "Check" in the desired language. He must have asked himself a thousand times how he ever could have gotten into this awkward position.

The answer to that was found in the character of his patron, the Archduchess Maria Theresa. She loved science, physics, and engineering, and adored Kempelen's work. She was also uncharacteristically skeptical for her day, once pardoning a condemned witch with the comment "Witches can only be found where there is ignorance" and was a pioneer in advocating for inoculation against smallpox. It was in pursuit of her skeptical interests that she first begged Kempelen to accompany her to a performance by a French conjurer named Pelletier, hoping he might be able to explain to her how the tricks were accomplished. Kempelen's response was to offer a challenge: that his scientific ingenuity could produce an automaton capable of giving a performance more impressive and mystifying than any she had yet seen. And, neither of them realizing the impact it would have on the rest of Kempelen's life, Maria Theresa accepted, excusing him from his official duties for six months in order that he might deliver on his promise. The result of this challenge was the Turk.

And this was where Kempelen's genius shone through. Besides its decoy clockwork and its crucial pantograph controls, there was nothing especially marvelous about the Turk from an engineering perspective. Instead, Kempelen had had to create a masterpiece in a field that was far outside his expertise: the stagecraft of a magician. Kempelen learned how to fool an audience. He taught himself the order and manner in which to reveal the contents of the cabinets, even lighting them inside to apparently prove the absence of a human operator. The Turk's largest cabinet was empty by necessity to give the operator room to move and work, so Kempelen turned this liability into an advantage. He crafted a small wooden casket stored inside this empty space, thus giving the empty space an apparent function; then for the performance, Kempelen removed the casket and set it on the stage a short distance away, occasionally touching or moving it, thus creating a diversion that led the audience to suspect he was using it to somehow control the Turk's moves. The Turk was not a mechanical wonder; it was a magnificent achievement in the art of magician's stagecraft.

By the time of Kempelen's death in 1804 at the age of 70, he had managed to somewhat distance himself from what had turned out to be his most notorious accomplishment. His patron Maria Theresa was long dead and Kempelen had kept his post through several of her successors, continuing his civil engineering to the degree that his health allowed, the Turk successfully disassembled and packed away into forgotten crates, its secret safe and all but forgotten. Kempelen lived amid circles of philosophers, scientists, engineers, writers, and artists, in the wealth and comfort bestowed upon him by his government. After he passed, the inscription on his monument, non omnis moriar, was ultimately to prove more than a little prophetic: "Not entirely do I die."

Corrections: An earlier version of this incorrectly identified Maria Theresa as Empress of Austria-Hungary. She was actually the Archduchess of Austria (Austria-Hungary did not yet exist). She also held many other titles, including Holy Roman Empress, German Queen, and Queen of Hungary and Croatia.

It also incorrectly had Von Kempelen's first name as Werner instead of Wolfgang, which is correct.

I also originally had the epitaph wrong; I copied it down as Non orbis moriar, or "Not as a circle do I die", suggesting that I probably had James Bond's family motto Orbis non sufficit on the brain for some reason as I was typing.

By Brian Dunning
Follow @BrianDunning