## Programming a Computor [sic – Computer] for Playing Chess. Offprint from: The London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science, Ser. 7, Vol. 41, No. 314, March 1950.

[London: Taylor & Francis], 1950.

First edition, **Shannon’s own copy ****inscribed by him to his sister**, of the very rare offprint of “the first technical paper on computer chess” (OOC) and one of the earliest significant works of artificial intelligence. “The very first paper [on computer chess], Shannon’s seminal work dating back to 1949, was first presented as a lecture on March 9th of that year to the National Convention of the Institute of Radio Engineers in New York. Shannon pioneered computer chess as we know it today, and his ideas have been employed in almost every chess program ever written” (Levy, Computer Chess Compendium*). “*His work on computerized chess would, in time, be recognized as another instance of Shannon dropping into a field and, in one stroke, defining its limits and unearthing many of its central possibilities. Decades after the publication of his paper ‘Programming a Computer for Playing Chess,’ *Byte *magazine would put it succinctly: ‘There have been few new ideas in computer chess since Claude Shannon’ … Shannon imagined some future applications of a chess-playing artificial intelligence: routing phone calls, translating text, composing melodies. As he reminded his readers, these machines were right around the technological corner, and no one doubted their economic utility. As diverse as these applications were, they had an important quality in common: they didn’t operate according to a ‘strict, unalterable computing process.’ Rather, ‘solutions of these problems are not merely right or wrong but have a continuous range of ‘quality.’’ In this way, chess was a valuable test case for the emerging generation of artificial intelligence. Nearly a half-century before Deep Blue defeated the world’s human champion, Shannon anticipated the value of chess as a sort of training ground for intelligent machines and their makers. The chess machine is an ideal one to start with, since: (1) the problem is sharply defined both in allowed operations (the moves) and in the ultimate goal (checkmate); (2) it is neither so simple as to be trivial nor too difficult for satisfactory solution; (3) chess is generally considered to require ‘thinking’ for skillful play; a solution of this problem will force us either to admit the possibility of a mechanized thinking or to further restrict our concept of ‘thinking’; (4) the discrete structure of chess fits well into the digital nature of modern computers … What followed in the paper, and what Shannon would later popularize in a less technical article for *Scientific American, *was the range of strategies that could be programmed into a computer: a blueprint for turning a machine into a good, if not a great, player. It is an admittedly broad survey: he studied each move’s possible outcomes, considered game-theoretic approaches, outlined how a machine might go about evaluating moves, and concluded that a computer could be programmed to play a perfect game of chess, but that such an outcome would be wildly impractical. This was, in a way, a limitation of the technology of the time: if a contemporary computer’s goal were to calculate all possible moves for itself and its opponent, it would not move its first pawn, Shannon calculated, for 10^{90} years” (Soni & Goodman). This paper was first presented at the National IRE Convention, March 9, 1949, in New York, but was first published in the Philosophical Magazine, as here. OCLC lists one copy (in Germany); no copies of the offprint in auction records.

*Provenance**: *The personal files of Claude E. Shannon. Inscribed ‘Kay’ by Shannon on front wrapper (i.e., Catherine Kay, Claude Shannon’s sister).

“The relevant history [of chess-playing programs] begins with a paper by Claude Shannon in 1949. He did not present a particular chess program but discussed many of the basic problems involved. The framework he introduced has guided most of the subsequent analysis of the problem. As Shannon observed, chess is a finite game. There is only a finite number of positions, each of which admits a finite number of alternative moves. The rules of chess assure that any play will terminate: that eventually a position will be reached that is a win, loss, or draw. Thus chess can be completely described as a branching tree, the nodes corresponding to positions and the branches corresponding to the alternative moves from each position. It is intuitively clear, and easily proved, that for a player who can view the entire tree and see all the ultimate consequences of each alternative, chess becomes a simple game. Starting with the terminal positions which have determinate payoffs, he can work backwards determining at each node which branch is best for him or his opponent as the case may be, until he arrives at the alternative for his next move. This inferential procedure – call minimaxing in the theory of games – is basic to all the attempts so far to program computers for chess …

“Now minimaxing might have been the ‘wheel’ of chess if the tree were not so large that even current computers can only discover the minutest fraction of it in years of computing. Shannon’s estimate, for instance, is that there are something like 10^{120 }continuations to be explored, with less than 10^{14 }microseconds in a century to explore them.

“Shannon then suggested the following framework. Playing chess consists of considering the alternative moves, obtaining some effective evaluation of them by means of analysis, and choosing the preferred alternative on the basis of the evaluation. The analysis – which is the hard part – could be factored into three parts. First, one would explore the continuations to a certain depth. Second, since it is clear that the explorations cannot be deep enough to reach terminal positions, one would evaluate the positions reached at the end of each exploration in terms of the pattern of men on the chessboard. These static evaluations would then be combined by means of the minimaxing procedure to form the effective value of the alternatives. One would then choose the move with the highest effective value. The rationale behind this factorization was the reasonableness that, for a given evaluation function, the greater the depth of analysis the better the chess that would be played. In the limit, of course, such a process would play perfect chess by finding terminal positions for all continuations. Thus a metric was provided that measured all programs along the single dimension of the depth of analysis.

“To complete this scheme, a procedure was needed to evaluate positions statically – that is, without making further moves. Shannon proposed a numerical measure formed by summing, with weights, a number of factors or scores that could be computed for any position. These scores would correspond to the various features that chess experts assert are important. This approach gains plausibility from the existence of a few natural quantities in chess, such as the values of pieces, and the mobility of men …

“To summarize: the basic framework introduced by Shannon for thinking about chess problems consists of a series of questions:

- Alternatives. Which alternative moves are to be considered?
- Analysis.
- Which continuations are to be explored and to what depth?
- How are positions to be evaluated strategically – in terms of their patterns?
- How are the static evaluations to be integrated into a single value for an alternative?
- Final choice procedure. What procedure is to be used to select the final preferred move?

“We would hazard that Shannon’s paper is chiefly remembered for the specific answers he proposed to these questions: consider all alternatives; search all continuations to a fixed depth, n; evaluate with a numerical sum; minimax to get the effective value for an alternative; and then pick the best one. His article goes beyond these specifics, however, and discusses the possibility of selecting only a small number of alternatives and continuations. It also discusses the possibility of analysis in terms of the functions that chessmen perform – blocking, attacking, defending” (Newell, Shaw & Simon, pp. 42-44).

“Claude Elwood Shannon was born in Petoskey, Michigan, on Sunday, April 30, 1916 … The first sixteen years of Shannon’s life were spent in Gaylord, where he attended the Public School, graduating from Gaylord High School in 1932. As a boy, Shannon showed an inclination toward things mechanical. His best subjects in school were science and mathematics, and at home he constructed such devices as model planes, a radio-controlled model boat and a telegraph system to a friend’s house half a mile away. The telegraph made opportunistic use of two barbed wires around a nearby pasture. He earned spending money from a paper route and delivering telegrams, as well as repairing radios for a local department store. His childhood hero was Edison, who he later learned was a distant cousin … In 1932 he entered the University of Michigan, following his sister Catherine, who had just received a master’s degree in mathematics there. While a senior, he was elected a member of Phi Kappa Phi and an associate member of Sigma Xi. In 1936 he obtained the degrees of Bachelor of Science in Electrical Engineering and Bachelor of Science in Mathematics. This dual interest in mathematics and engineering continued throughout his career” (*Collected Papers*, p. xi).

“After graduating from the University of Michigan in 1936 with bachelor’s degrees in mathematics and electrical engineering, Shannon obtained a research assistant’s position at the Massachusetts Institute of Technology (MIT). There, among other duties, he worked with the noted researcher Vannevar Bush, helping to set up differential equations on Bush’s differential analyzer. A summer internship at American Telephone and Telegraph’s Bell Laboratories in New York City in 1937 inspired much of Shannon’s subsequent research interests. In 1940 he earned both a master’s degree in electrical engineering and a Ph.D. in mathematics from MIT. He joined the mathematics department at Bell Labs in 1941, where he first contributed to work on antiaircraft missile control systems. He remained affiliated with Bell Labs until 1972. Shannon became a visiting professor at MIT in 1956, a permanent member of the faculty in 1958, and professor emeritus in 1978” (Britannica).

Hook and Norman, *Origins of Cyberspace, 882* (complete journal issue); Sloane & Wyner, *Claude Elwood Shannon Collected Papers* 54*. *Levy*, Computer Chess Compendium**, *1988*.* Newell, Shaw & Simon, ‘Chess-playing programs and the problem of complexity,’ pp. 39-70 in:*Computers and Thought**, *Feigenbaum & Feldman (eds.), 1963.

8vo (252 x 172 mm), pp. [1], 256-275, [1, blank]. Self-wrappers, stapled and glued. The first leaf (wrapper title) is tipped onto a gathering formed by pages 257-272 (signed ‘U’); pp. 273-276 (276 is blank) are formed by folding a single sheet once and tipping that unsigned gathering to the verso of the ‘U’ gathering. Near fine.

Item #5548

**
Price:
$25,000.00
**