February 2, 2000

THE FABULOUS AUTOMATON CHESS-PLAYER

PART TWO. HOW IT WORKED AND WHAT HAPPENED TO IT.

(If you missed last week's Part One, click here to read it first!)

When we left you last week, you were wondering -- I hope -- just how the illusion of the man-in-the-box could have been accomplished before such learned and numerous audiences who were captivated and deceived by the demonstration. Read on.

Please study the illustrations just above. The shaded outline indicates where the concealed operator was located at various stages of the "examination" presentation of the Chess Automaton. The operator was always toward the back of the box, away from the audience, as shown in the first drawing. The wide drawer at the bottom contained the chess-pieces as well as several boxes containing other minor "props." I consider this to be one of the most ingenious bits of misdirection that Kempelen came up with -- from the point of view of a magician. Had the drawer been empty, the spectators could have begun to wonder why it was there the first place.

At the point in the examination routine where we left you, all the doors were re-closed and re-locked, the drawer being the last thing to be replaced, after the contents had been removed. The spectators had every reason to believe that they had been shown the entire inside of the box, and indeed they had -- but not all at one time. The illustrations will clearly show how the figure inside was able to move and maneuver so as not to be glimpsed during the carefully thought-out procedure described above.

The occupant, at the beginning of the examination routine, was seriously uncomfortable. He had to be squeezed into the volume between the level of the top of the drawer (there was doubtless a partial shelf there upon which he perched for the brief time at the beginning) and the actual table-top. He could not occupy the drawer space because it was filled with the props. However, when the front drawer had been opened -- and left open -- he could lower himself into that now-empty space. Remember, that was the first move the operator made, which we now see was done to relieve the discomfort of the occupant. Simply leaning forward at that point, he was entirely concealed within the center and right portions of the chest, and the space left by the drawer. The first drawing, above, shows this situation, and how it appeared when the front left door was opened.

When both the front and back doors of the left-hand compartment were opened, the spectators became satisfied that that portion was empty except for the mechanism. The candle-light proved that. The mechanism that they saw did not go very far back, occupying only about one-third of the front volume of the left-hand compartment. Then, the back door having been closed, the occupant could place an opaque partition (a dark cloth, perhaps) behind the cogs and wheels and now have plenty of room for the upper part of his body. He would lean back into that volume as the center and right hand doors -- the front ones first, then the back doors -- would now be thrown open.

I'm sure it's becoming plain to my readers how the rest of it went. With the five doors and the drawer opened, the spectators would see completely through the center and right hand compartments, even if the chest were turned around, as it would be at that point. The robes covering the "Turk" figure would be raised to reveal nothing there but more mechanism. Turned to face front again, the machine only had to have its doors closed, and the occupant would be able to raise himself up so that he occupied all three compartments -- though only in the back two-thirds of the volume -- and the drawer would be closed. Once that was done, he would lower himself again, and would have the additional volume of the now-empty drawer to use for his lower body.

Though the Turk figure had much mechanism inside it, there was room for the concealed occupant to then insert his head and shoulders up into the lower part of the figure. There, perhaps through slits or some sort of see-through fabric, he would be able to view the chess board. The left arm of the figure was an ingenious construction whereby the fingers could be opened over a chosen chess-piece, then closed and the arm raised, after which the chosen piece would be placed where needed. Et le voilą!

Kempelen prove himself braver than we might think when he actually sold his invention to Frederick the Great, King of Prussia! Of course he had to reveal the secret to the purchaser, who was not at all amused at how he'd been duped. Frederick considered the machine useless -- which is not difficult to understand -- and threw it into a basement room somewhere in Berlin, where it languished for some years.

The inventor died in 1804, and a man named Maelzel eventually purchased and took over the operation of the Chess Automaton, making several improvements among which were that the occupant of the box had a magnetic sensing apparatus which enabled him to tell what pieces had been moved on the board without his directly viewing them, as his predecessors had done. Maelzel had inside the chest a miniature chess board with pegged pieces, so that the player could duplicate the game seen outside, and work out the appropriate moves. A small bellows and vibrating reed, a sort of artificial speaking mechanism, was incorporated whereby the operator could signal "check!" by forcing air through a tube. The approximation of the word "check" was said to lack clarity.... This new owner of the Automaton was an inventor himself, as Kempelen had been. He had to his credit a clever music metronome and several other machines. He created for himself a number of other "performing" devices that worked along with the famous Automaton, and on tour he did very well financially for himself.

While Kempelen had first produced his invention only as a curiosity, and he often described it as "a mere trifle," he had not intended to use it as anything but an amusement. He delighted in showing it to interested persons, particularly royalty and other influential individuals who could help his career. But, following certain financial reversals, it soon became obvious to him that charging admission to see the device in operation could be very lucrative, and he eventually made a great deal of money touring throughout Europe with his show. Though he offered frequent mild disclaimers, he never actually stated that the machine was a hoax, and to responsible persons he would often hint broadly that things were not as they seemed to be.

Maelzel, however, commenced his own association with the Automaton on a fully commercial basis, and at one point he, too, actually sold the device -- and its secret -- to the King of Bavaria, Eugene Beauharnois, in 1806. The price was 30,000 francs, a very sizable sum. This member of royalty was also miffed, as he had every right to be, though he ended up laughing at his own naivete, and at one point actually occupied the interior of the device to play chess games with his friends. Though we might suspect that these friends were well aware not only of how the machine worked, but who was working it, it appears that they never "blew the whistle" on His Majesty.

Maelzel held a special command demonstration of the Automaton for Napoleon in 1806 in Berlin -- a city which Napoleon was occupying for the moment. The general tried to upset the machine by performing illegal moves -- for which the protocol laws well-prepared, since the understanding was that the figure would nod three times if such a thing were to happen, and when Napoleon persisted in making the wrong moves again and again, the Automaton finally swept all the pieces to the floor, and the game was over. Later, when the general was behaving himself and obeying the rules, he lost his game -- and was reportedly not happy.

Along the way, throughout its illustrious career, the Chess Automaton came to the attention of many prominent persons who essentially solved -- at least partly -- how it was being operated. Booklets and articles appeared detailing the secret, and more importantly pointing out that there were obvious discrepancies in the performance that quite gave it away as a hoax. Why, asked several of them, was it necessary to close the doors during the period in which the figure was operating? And many observed that when the owner "wound up" the machine -- which was done after each dozen or 15 moves had been made -- he often turned the key a different number of times. This, to technically-minded persons of the day, would have been a dead giveaway, since wind-up mechanisms were very familiar to them, and would be expected to perform in accordance with their experience. That was a serious defect of the machine.

Of course, from our position of very highly developed technology, we can see that building a real artificial chess-player is now quite possible. In 1956, a massive assembly of primitive computers first played a chess game. Even at that period of time, it was felt that any chess-playing machine would be incapable of strategy. That was when it was not possible to handle more than 100,000 "operations" a second in electronics. I hesitate to say how many are possible today, because by the time you read this, that figure will have grown, I'm sure. Suffice it to say that home computers are presently working at 1,000,000,000 operations a second, and commercial machines at a thousand times that rate, and more. A chess book published in 1960 suggests that we should "abandon all ambitions of constructing an ideal electronic chess player." The author wrote, ". . . even the best electronic computer cannot possibly be programmed with an objective unambiguous pattern of operation as would guide it towards success in scanning the astronomical number of various individual combinations."

In 1961, one Mikhail Botvinnik said, "we shall continue to fail as long as we attempt constructing a mechanical champion." Well, Mikhail was wrong. Today, with such vastly improved means at hand, we have been able to make artificial players that have beat world champions, and that is humbling indeed.

When the Chess Automaton went abroad, it was never to return to the Continent. It showed up in New York City in 1826, where Maelzel exhibited it with great success until his death in 1837. It passed through several hands until it was eventually completely reconstructed and refurbished by a Professor J. Mitchell, and it finally ended up at the Chinese Museum in Philadelphia. In June of 1854, fire swept the museum and the Marvelous Chess Automaton was reduced to ashes.

You'll see in our first illustration, in Part One, that there is a small box standing on the floor next to the Automaton. That was another stroke of genius invented by Kempelen. Note the keyhole. The operator would carry this box about with him as he circled in the area near the Automaton. From time to time, he would unlock the box, consult the interior, and then mysteriously re-lock it. That was a marvelous piece of misdirection on his part, since the spectators would have their attention diverted by this action, and any pause in the proceedings -- which might be found necessary in order to give the actual chess-master time to plan a move -- was thereby excused. There were several similar "gimmicks" used by the inventor, and I deeply suspect that one time he had come under the influence of a conjuror. I can only wonder.

So, we have told you the strange tale of a talented inventor/hoaxer, and also of the man in the box, a person who would never be credited with his ability to play chess on a very high level, and probably suffered from backache of a magnitude that we cannot imagine. It is also a tale of the naivete of a public all too willing to accept a miracle simply because they could not solve a trick. Have we really changed very much in that respect?