The oldest known surviving first hand Go book - Dunhuang Go Manual

Dunhuang Go Manual (敦煌碁經, or as the name written on the scroll 碁經 The Classic of Go) is the oldest known surviving first hand Go book existed (unlike other sources although probably older but mostly second-hand or more, and most don’t have enough details but only a few sentences). I’ve been able to find the full-size digitized image from the International Dunhuang Project database IDP Or.8210/S.5574 Recto


The chapter 1 translation can be found in the IDP page full entry (I haven’t been able to find the full translation yet, but it will be quite difficult to translate due to so many contexts needed to be filled). Although people who can read modern traditional Chinese should be able to read most of the content no problem (provided you are adapt in reading Chinese classical literature, the structure of the sentences was very specific). Most scholars dated it to around 6th century to 9th century AD, but part of the contents are certainly much older since it is a transcribed scroll from various sources and references people hundreds of years ago. And it also references some appendix of Go positions but the diagrams were not found alongside it.

Due to its location being found, it is fairly certain ancient Go had been spread very far and wide when this was transcribed (not just east to Japan and Korean, or south-west to Tibet, but also west towards central Asia), with many terminologies already well established, and strategies exactly as we know them today (like 引征 ladder strategies, and 打劫 ko strategies, even life-and-death shape 直四 straight four 曲四 bent four 花六 flower six 聚五 crossed five, and some very specific end-game life-and-death 角旁曲四 bent four in the corner).

The most interesting thing about it is the Go rules in Chapter 6. It doesn’t mention one of the later common Chinese Go setups started with fixed stone placements (maybe it was in the missing part in the start of chapter 1?), but mostly discuss how the final “score” can be counted and how the game should be finished. And it seemed to recognize some form of score system in the form of auctioning/gambling scores to balance the first move advantage. It placed the worth of the first move (stone) at 3 points, and after the game, every 3 “stones” are worth 1 point (1 stone is 1/3 point). And on face value, this would translate to a modern komi of 9 in area scoring (if the area scoring mechanism is the same). However, the exact value would be determined by how a “stone” is counted after the game.

From various later terminologies, it seems at the time of the text was written, the end game scoring is at a period transition between some forms of area scoring (停道 essentially words for stones and surrounded area, some interpreted it as to fill both territories to end up with the same amount and then just count the “stones” difference, a variation of area scoring method), and fully played out games with stone counting 兩溢, essentially a game played out to fill every point and every territory with just eye space left and nowhere to play on the board, hence both sides “spill” their moves outside the board (imagine in ancient time where most players would be today’s ddk or even worse, and teaching and learning were difficult, how most games would end up). Some suggested that this means at the very basic level since every game is played with even hands without passes, at the end of the game, if there’s nowhere else to play, the pass move becomes giving “stones” as prisoners for your opponents. Thus the counting of “stones” at the end of the game would literally mean counting how many prisoners you have from your opponent. And this makes sense with rituals followed all the way till today like a player can grab a handful of captured stones as a sign of resignation, the ritual of nigiri, and certain features preserved in other variations of Go, where different “prisoners” on the board have different point values. A very high and rounding to multiples of 3 point system would be the equivalent of about 6 to 9 komi (or 9 to 12 komi, depending on rounding up or down), also make sense for some ancient Chinese sources claiming the one who plays first needs to be a better player instead of the other way around.