All forms of chess are thought to have a common ancestor, but the dating and placing of the prototypical game are contentious. Following the lead of the chess historian H.J.R. Murray (whose scholarship was perhaps wider than it was deep), it has frequently been asserted that chess originated in India as chaturanga around the middle of the first millenium CE. Others, citing the lack of direct literary or archaeological evidence for chess in India at that time, point to Persia or some part of central Asia. The only thing known for certain is that an early form of the game was known in Persia by the seventh century. Called shatranj, it was played on a board identical to that used in modern Western chess, and with the same configuration of pieces, although some of the moves were more limited.
Chess spread westward through the Islamic world until it arrived in Europe in the Middle Ages. At the same time, it travelled into China and thence to Japan, where it took a very distinct form as shogi. There is also a Korean version very similar to the Chinese one. (Further south, the chess of Thailand, which is holding its own as a national pastime, appears to be on a different evolutionary branch.) By the end of the Song dynasty (960-1279), the modern Chinese game was fully developed.
Some authorities insist that China is the birthplace of chess. If this is so, the game must have been exported very early in its development, because the present Chinese game is an obvious improvement on chaturanga/shatranj. What seems more likely is that the prototypical chess converged with one or more native Chinese games. The modern game may even contain traces of an ancient system of divination in which pieces representing celestial bodies were moved about a map of the cosmos, divided by the Milky Way. The Milky Way is called a river by the Chinese, and the chessboard has a river running through it. Charles Kliene gives more evidence of this association in the highly entertaining Preface to his Seven Stars: A Chinese Chess Variation with Three Hundred Endings. See also Joseph Needham’s Science and Civilization in China, vol. 4 pt. 1, pp. 314 ff, and H.J.R. Murray’s A History of Chess (1913), p. 122. A good account of early Chinese references to games that might be associated with xiangqi is given by Peter Banaschak.
Even the name of the game may suggest a connection with some type of astrological tablet. Qi means a strategy game, and xiang is the character that appears on the so-called elephants of the black side. (The equivalent red pieces are called by a homonym that signifies “adviser” or “augur.”) Like so many Chinese words, xiang has several meanings: it can indeed mean “elephant,” but it might equally refer to the ivory from which some sets are made, or it might signify “image” or “symbol” or even (according to Mathews’ Chinese-English Dictionary) “star” or “heavenly body.” Thus xiangqi might be translated “celestial game” or “symbolic game.” Although “elephant game” is a possible translation, it does not seem apt, given the very limited role of the elephant in play; unless the name simply suggests the game’s Indian origins.
It is interesting to compare the evolution of chess in China and the West. The game of chaturanga/shatranj suffered from several weaknesses, and these weaknesses were remedied in very different ways, as follows:
- The pawns in the original game were slow to come into contact with the enemy. In Western chess, this problem was solved by allowing the pawns their initial two-step move. The Chinese solution was to set up the pawns in a forward position.
- The original game suffered from a lack of mobile attacking forces. Among the major pieces, only the rook and knight had their modern moves. The bishop moved only two squares diagonally, the queen just one. In the West, this problem was solved by extending the move of the bishop, by the introduction of castling to bring a rook quickly to the center, and finally by unleashing the modern queen. In China, the queen and bishop became if anything weaker than in chaturanga/shatranj, but two powerful new mobile pieces, the cannons, were added. Moreover, reducing the number of pawns to five, and opening up the second rank of the initial array, created space for the rapid deployment of the rooks.
- Games of chaturanga/shatranj that reached the endgame must often have ended in a draw, because the pawn only promoted to the weak queen. In the West, the extension of the powers of the queen made it easier to enforce checkmate in the endgame. In China, the approach was very different: the king was confined to a small part of the board, making him easier to pin down, and the pawns were promoted earlier, being granted lateral movement as soon as they passed the river at the centre of the board. In addition, the king was given the extraordinary power of striking across the board like a rook against the opposing king, opening up many more possibilities for checkmate with just a few pieces left on the board.
An important part of the game’s history is the development of the problem. Unlike Western chess problems of the “black to move and mate in three” variety, xiangqi problems (perhaps more accurately called studies) usually offer one side an easy forced win, given the first move, but can also be won by the other side if the advantage is reversed. Charles Kliene has documented one such ending, and gives a colourful description of the hustlers (which translates as something like “powers of chess layout”) who would set up such jeux partis at the side of the road and challenge all comers. Evidently this custom is still alive today.