Seoul Hero

Please see the important updates at the end of this post. I will continue to update as information comes in.
–
Some time ago, I came across a fascinating Chessbase story by Ramón Jiménez about the world’s first chess computer, “Torres y Quevedo’s rook endgame automaton.” This “chess automaton” dated to 1911 or 1912 (see below), and was invented by the famous Spanish inventor Torres y Quevedo. The world’s first analogue chess computer, invented prior to the advent of digital technology, was capable of solving King + Rook vs. King endgames beginning in any position. The machine did not always mate in the minimum number of moves, but it did always finish the job. The machine had electrical sensors under the board, and used a kind of arm to move its pieces. Unfortunately, this first chess computer is not well-known, and appears to be represented by only one, or possibly–see below–two or three pictures on the internet.

The purpose of this post is to present a new picture of the first machine; this follows a brief discussion regarding some discrepancies in the available information on this machine. This first chess computer is shown below; the actual new photograph appears last in the original post. It is the small colored square thumbnail–click on it for a larger view. Meanwhile, my thanks go to my wonderful friend Brian in Montreal for helping me get this larger version of the Chessbase picture out of the PDF file it was in.

There is some conflicting information on the internet as to this machine. A scholarly PDF file–which was a source for the Chessbase article–entitled, “From Analytical Engine to Electronic Computer: The Contributions of Ludgate, Torres, and Bush,” attributes the first machine to 1911. The Chessbase article places the creation of this machine in 1912, as does this Spanish webpage, which was another source of the Chessbase article. The English Wikipedia article seems indebted to the Chessbase article and its sources, and does not appear to use much independent information, although there are a few interesting tidbits in it.

The PDF file, a scholarly paper authored by Brian Randell, speaks of a second “automaton,” created “some years later”; the Chessbase article places the creation of this machine to the year 1920, which seems reasonable. On the other hand, there seems to be some confusion within the Chessbase article as to which machine the second picture refers to. The second picture of the machine in the Chessbase article runs this caption: “Gonzalo Torres y Quevedo showing his rook endgame machine to Norbert Weiner”; the PDF file, in contrast, says

Some years later Torres made a second chess automaton, which used magnets underneath the board to move the pieces. Like a number of his other inventions, this one still exists and is still operational. Figure 11 shows Torres’s son Gonzalo demonstrating the automaton to Norbert Wiener.

The text seems to imply that it is the second machine that is shown in the picture. Possibly, the writer of the Chessbase article, who is from the Dominican Republic, may well have seen the machine in a museum, or had access to better information in Spanish, although it is possible that the caption is based on a misinterpretation or an overly fast reading of the PDF file. The figure 11 picture from the article, that is, the second Chessbase picture, is shown below.

Another picture, which can be found in a link from the Spanish page linked to above, provides another hint as to which machine the above picture shows. The Spanish page calls this “el segundo,” i.e. the second chess machine, which it dates to 1920. It appears to be identical with the picture shown above, although it possible that I am mistaken.

This means that there is essentially only one photograph of the original first machine on the internet at this time: the picture from the Randell PDF file used above, and in the Chessbase article. This is a great pity, for “el jugador ajedrecista” is a landmark not only in the field of chess, but also in the field of computing studies. It is unfortunate that the machine, and its inventor, Torres y Quevedo, remain largely unknown to the English-speaking world.

To remedy this lack of material, I emailed some time ago an inquiry to the staff at the Spanish embassy in Toronto, Canada, asking for pictures, movie clips, contacts, or information relating to this first chess computer. They were very kind, and were able to provide me with some email addresses. One contact kindly supplied two other contacts, one of whom I heard from. Although I was not able to hear from anyone from the museum where el ajedrecista is housed, I did get one email contact from another museum who kindly provided me the picture shown below (click for a larger view). In his email to me, he wrote that this was the only “remnant” left of the photographs in his collection for the “ajedrecista.” The following picture, then, comes courtesy of Esteban Moreno, Master Museo Virtual de la Ciencia, and I remain much in his debt for sending me this picture.

The picture itself is very detailed, in color, and shows the workings of the machine much better than the black and white photograph that began this post. I hope that those interested in either chess or in the history of computing will enjoy the picture, and find it useful. Please click on the thumbnail, below, for the full-sized picture. Depending on your browser settings, the picture may be shown smaller than its actual size.

The Picture (click to enlarge)

I had hoped that further material would be forthcoming from other contacts, but none has been so at this time. However, I have sent out inquiries to other places, and if further material turns up, I will update this post, which I shall put on my sidebar for ease of reference. From now on, I hope that I will be joined by other information gatherers who will turn up additional visual material on this important piece of human and computing–and chess–history. It is my hope that, just as Ramón Jiménez inspired me to find a new picture of this machine, one or more readers will be inspired who will actually visit the museum that houses this piece of human history, the Colegio de Ingenieros de Caminos, Canales y Puertos in Madrid, and will succeed in taking, and then posting on the internet, pictures or short video clips of this amazing electronic chess computer.

–
Nathan Bauman, Seoul, South Korea
Email: nathan [a t ] nathanbauman.com with “Chess Automaton” in the subject heading.
–
UPDATE 1: Edward Winter, the author of A Chess Omnibus and other fascinating books of chess lore and history, has just posted a picture of Tartakower being mated by the second machine, which is associated with the next generation of the Torres family. He also wrote some additional information about the first chess computer, which may have been invented as early as 1890. Readers interested in these machines are strongly encouraged to click on the hyperlink above, and read Chess Notes 4470 for July 18th, 2006. In the mean time, I would like to thank Mr. Winter for responding to my query so kindly, and I remain much in his debt. Readers may see a slightly larger version of the photographs that he provided, below, by clicking on the above hyperlink.

UPDATE 2: In Chess Notes 4482 (click and scroll down), Mr. Winter posts some fascinating information, including a link sent by reader Christian Sánchez to the webpage of the museum that houses the first chess machines. That page has three small, but very good pictures; the one below shows the first two chess computers side by side.

Unfortunately, I have yet to see a picture of the first machine that shows the chessboard it used. It is my intention to write another post on this subject, when I have more time, in which I shall try to take advantage of the more correct information that I have been learning thanks to Edward Winter’s columns.

UPDATE 3: In Chess Notes 4495, Mr. Winter posts another photograph, with more information, sent by reader Leonard Barden.

UPDATE 4: In Chess Notes 4525, Mr. Winter touches on the question of the date of the first machine, noting two pieces of evidence which, if accurate, would place the prototypical chess machine as early as 1890.

UPDATE 5: In Chess Notes 4547 (click and scroll down), Edward Winter returns to the issue of the 1890 date of the first chess computer by Torres y Quevedo.

UPDATE 6: Lawrence Totaro of Las Vegas sends me this newspaper article, taken from the “Daily Gleaner,” of 1970. The Reuters article on the top left of the page makes reference to a 50 year old machine in Madrid made by “Torres Quevedo,” whom the article also mentions as having invented a cable car that crossed Niagra Falls in North America. That would be the senior Torres y Quevedo, and this machine, then, might be the first machine he invented. Curiously, the article claims that the machine cries out “Checkmate” in Spanish when it wins.

UPDATE 7: Lawrence Totaro sends me this picture from a German book by Herbert R. Grätz: “Schach – mehr als ein Spiel.” I’m indebted to him for his kindness in sending material two times! The picture itself is quite a good one.

UPDATE 8 – Sept. 25th, 2009

I was recently contacted by a Japanese publishing company, which must remain unnamed for the moment, about permission to use the new photograph I presented in a book on robots.   The company has now received permission to use the photograph from the relevant people in Spain.  I will update this post further when the book comes out.

UPDATE 9 – Sept. 25th, 2009

An important explanation of how the first chess automaton worked can be found here.