A history of chess

Northern India, around the sixth century

The earliest game that is clearly an ancestor of chess appears in northern India, most likely in the sixth century of the common era. Its name was chaturanga, a Sanskrit word for an army in four parts: foot soldiers, horses, elephants and chariots. Those four parts are still on the board. They became the pawn, the knight, the bishop and the rook, with a king and a counsellor standing between them.

The confident version of this story is more confident than the evidence. No Indian rule book from that period survives. What survives is allusion: a line in a seventh-century Sanskrit prose work by Bana that plays on the word for a chessboard, a similar play on words in Subandhu, later legal and poetic references, and a small number of pieces from excavations whose identification is arguable. None of that tells us how anything moved.

What we actually have is a Persian and Arabic tradition, a few centuries later, that is detailed and consistent, plus Indian texts that are late enough to have been shaped by it. Almost every reconstruction of the original Indian game therefore works backwards from the Persian and Arab material. That is a reasonable method, and it is still a reconstruction. Where the Sanskrit sources describe a piece differently from the Arabic ones, as the elephant appears to be described, scholars have disagreed for a century about which reading to trust.

Rival origin stories exist. A Chinese origin has been proposed more than once, and dice-driven four-player forms have been read as either the ancestor of chess or a later offshoot of it. Neither case is strong. The northern Indian origin remains the best-supported account, not a proven one.

Persia and the Arab world: shatranj

The game moved west into Sasanian Persia, where it was called chatrang. A short Middle Persian tale, sometimes called the Explanation of Chess, describes an Indian envoy bringing the game to the Persian court as a riddle for its wise men to solve. It is a good story and not a record; it was written down long after the events it claims.

After the Arab conquest the name became shatranj, and for the first time the game has a real literature. From the ninth century onwards, players at the Abbasid courts wrote manuals: al-Adli, ar-Razi, as-Suli and al-Lajlaj among them. Those manuals record openings, endgame method, and hundreds of composed problems. Players were even ranked in classes. This is where firm knowledge of the rules begins.

Shatranj was slower than the modern game and different in character. The counsellor beside the king, the ferz, moved one square diagonally. The elephant, the alfil, leapt exactly two squares diagonally and could reach only an eighth of the board. Pawns had no double step and promoted only to a ferz. With no long-range pieces except the rook, the opening was a careful march into a memorised formation, and a game could be decided by stripping the opponent of every piece rather than by mate.

Into Europe

Chess was in Europe before 1100. The clearest routes are through Muslim Spain and through Sicily; a northern route by way of Byzantium and the river trade is likely but harder to demonstrate. Wills, church complaints and finds of pieces put the game in Christian Europe across the eleventh century, spread widely enough that no single point of entry explains it.

The pieces were then renamed to fit a European court. The Arabic ferz became the fers and eventually the queen. The alfil became a bishop in English, a fool or jester in French, and a runner in German, so three languages looked at the same piece and saw three different things. The chariot stayed a rook, from a Persian word, and only later picked up the castle-shaped carving that made English speakers think of a tower.

By the fourteenth century a Dominican friar, Jacobus de Cessolis, had written a sermon-like book that treated the pieces as ranks of society. It was copied and translated across Europe and was among the most widely read books of its age. It tells us how familiar chess had become. It tells us very little about how it was played.

The new rules, and then the rule books

In the last decades of the fifteenth century, in Spain and Italy, the game changed more sharply than at any other moment in its history. The counsellor was given the queen’s sweeping move, and the elephant was given the bishop’s long diagonal. The pawn’s two-square first move, en passant and castling settled into their modern shapes over roughly the same period. Contemporaries knew they were playing something new and said so, calling it the queen’s chess. A book printed in Salamanca around 1497 under the name of Lucena, and the works of Damiano and Ruy Lopez soon after, describe the new game in practical detail.

Why it happened is not settled. The most repeated suggestion links the queen’s new power to the reigns of powerful European queens, Isabella of Castile above all. The timing is suggestive and no source of the period actually says it, so it should be read as a plausible guess rather than an explanation.

Rules then stayed local for another three hundred years. Stalemate was scored differently in England and in France; the Italian schools argued about en passant and about castling long after others had stopped. The nineteenth century forced agreement, because international play demanded it: the London tournament of 1851, Howard Staunton’s codifying work, and the arrival of the chess clock in the 1860s all pushed towards one rule set with one clock and one scoring.

The international federation founded in 1924 took over that job and has kept it. Its Laws of Chess are revised every few years, and the 2023 edition is the one the standard game on this site follows.

Sources

Everything on this page is written from scratch. These are the works it leans on, and what each one settles.

• H. J. R. Murray, A History of Chess (Oxford, 1913)

  Still the foundation of the subject, and the work that assembled the Arabic manuscript evidence into a coherent account of shatranj. This site follows Murray for the shatranj rules and for the reading of the elephant as a two-square diagonal leaper.

• Andreas Bock-Raming, studies of the Sanskrit and Indian chess sources

  Close philological work on the Indian texts, which sets out how thin and how late they are, and argues that they describe an elephant moving two squares orthogonally rather than diagonally. This site records that reading as the alternative in its chaturanga reconstruction.

• FIDE Laws of Chess, 2023 edition

  The current international rule set. It fixes the endpoint of the story told here: the modern rules the standard game on this site is built on.