The history of janggi

A descendant of xiangqi

Chess spread east out of India, was reshaped in Persia and Central Asia, and arrived in China as something distinctly Chinese: pieces on the points rather than in the squares, a general penned into a palace, a river across the board.

Janggi shares with xiangqi the nine-by-ten board of points, the palace, and the whole cast — general, two guards, two elephants, two horses, two chariots, two cannons, five soldiers — in the same starting shape. No two games arrive at that list independently. Janggi descends from xiangqi, and nobody serious argues otherwise.

The cannon is what dates the family. Chess did not travel east with a jumping cannon; nobody invents that piece before gunpowder is an ordinary idea. Both games have one, playing the same part. So the two must have separated after the cannon had already joined the Chinese game, which puts the split well into the second millennium rather than at the moment chess first reached the peninsula.

When exactly the game crossed into Korea, by which route and in whose hands, the record does not say. Guesses exist. Evidence does not.

What Korea changed, and whether it was on purpose

Set the two games side by side and the differences are not adjustments to one corner of the rules. They are spread across every piece:

• The river is gone, so nothing is confined to a half of the board.
• The general begins on the centre of his palace rather than the back rank.
• The palace diagonals became real lines, usable by chariots, cannons and invading soldiers as well as by the general and his guards.
• The elephant leaps a point further and may cross the whole board.
• The soldier steps sideways from its very first move.
• A player not in check may pass, so there is no stalemate.
• The generals facing each other is a draw, not an illegal position.
• The horse and elephant points are chosen before play instead of being fixed.

One at a time, any of those could be drift — a house rule that stuck. Together they point the same way. Each removes a restriction or hands a player a choice, and several fix problems the others create. Give the elephant a longer leap and a fixed back rank leaves it a clumsy piece with two or three targets; let each player choose where the elephants stand and the longer leap becomes interesting. Take the river away and the soldiers need their sideways step from the start, or the middle of the board congeals.

That coherence is the strongest argument that these are local reforms rather than accumulated error. It is still only an argument. No document survives in which anyone says what they were doing, and a rule set can end up looking designed simply because the versions that played badly were dropped.

How thin the record is

Before the Joseon period there is very little to work with, and mentions of a board game are not descriptions of one. A reference in a poem, a game in the background of a painting, an inventory listing a set of pieces: all of these tell you the game existed and none tells you how the elephant moved.

Physical survivals have the same limit. Boards and octagonal piece sets are held in Korean collections and they establish the cast. They say nothing about bikjang, the pass, or whether repeating a check was punished.

The first substantial comparative account in English is H. J. R. Murray’s A History of Chess, published in 1913. It is useful, and it should be read for what it is: Murray worked from correspondence, travellers’ reports and published descriptions rather than from a lifetime at the board, and his Korean material is second-hand. It is good evidence of what was reported about Korean practice around 1900. It is not a Korean rule book, and it did not have one to draw on.

Local practice varied, and it varied on exactly the points you would expect: whether a pass was allowed, what happened when the generals faced, how many repetitions were too many, what to do with a game neither side could win. Those are the rules a club settles by custom and a codifier has to settle by decision. If someone offers you a firm date for the moment janggi took its modern shape, it is worth asking what they are reading.

Codification and the modern game

The twentieth century did to janggi what it did to most traditional games: organised it. Clubs became associations, associations wrote rules, and the written rules became the played rules. The Korea Janggi Association’s published set is the one this site follows.

What codification settled is easy to list, because it is the list of things that used to vary. Setup runs Han, then Cho, then Cho’s move. Bikjang is a draw. A third repetition of the same checking position loses for the player giving check. A game with no capture for 120 half-moves is over. And the points are numbered on one system, so a game can be written down the same way twice.

One further problem was purely administrative. Janggi draws a great many games, and a tournament that produces a column of draws produces no standings. The answer was to price the pieces — chariot 13, cannon 7, horse 5, elephant 3, guard 3, soldier 2 — and award certain drawn endings to whoever has more on the board, with 1.5 added to Cho’s total. The half-point does real work: with every other value a whole number, the two scores can never come out level.

None of that is ancient, and it is worth saying so. The point count is a twentieth-century fix for a twentieth-century problem, and it is the reason a janggi tournament finishes on schedule.

Common questions

Did janggi come from xiangqi?

Yes. The shared board, palace and full set of pieces make the descent certain, and the cannon in both games means the two separated after the cannon had already joined the Chinese game.

When did janggi take its modern form?

Nobody can say precisely. The record before the Joseon period is very thin, club practice varied into the nineteenth century, and the rules were only fixed in one published form in the twentieth.

Sources

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

• Korea Janggi Association rules

  The modern published rule set, and the authority for every specific reading on this site: the Han-then-Cho setup order, bikjang as a draw, the loss for a third repetition of a checking position, the 120 half-move no-progress limit, the numbering of the points, and the tournament piece values.

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

  The first large comparative survey of the chess family in English. It is why this page can say what was and was not written down about janggi before the twentieth century — and, because Murray’s Korean material came to him second-hand, it is treated here as a record of what was reported in his day rather than as a rule book.