The Classic of Go

The original manuscript of the Classic of Go[i] is labeled Or.8210/S.5574 (“S” for Stein) and is thought to be the oldest extant manual devoted to the game of Go. In the 1960s, Cheng Enyuan provided a modern simplified Chinese transcription with copious notes. It is his opinion that the manuscript was written in the sixth century A.D., in the state of Northern Zhou, due to the structure of some of the characters and language used[ii]

A revised edition of Cheng’s work was published in 1990. A transcription of the manuscript into modern complex Chinese characters is available on the internet site of Hong Kong’s CCC Yenjing College Go Club. [iii]

I would like to thank John Fairbairn for his close reading and insightful comments on drafts of this translation

Robert W Foster [translator] International Dunhuang Project, 2009.


… Chapter 1[iv]

…[Chen] Ping’s strategies.[v]

If there are places to cut, then attack there, causing weak formations; even if there is no gain, there will be no harm to the overall pattern. If the vanguard becomes strong,[vi]  one can encroach. If strategic advantage[vii]is already lost, then defend the borders. If greedy, one will often be defeated. If timid, one will rarely succeed.[viii]

Like two generals planning against each other, if there is an opportunity, seize it. The ancients said “don’t consider a virtuous heart to be good; one must be capable of subtle deceptions.”[ix]

If your intentions lie in the southeast, then deceptively play in the northwest. Like the Lord of Jin’s invasion of Guo,[x] have ulterior plans; or like Zhuge [Liang’s] implementing the “empty” [city], be able to create good deceptions often. [xi]

Because previous moves cannot be changed, subsequent regrets are truly difficult to bear. Go has innumerable pieces, so situations must be carefully examined. Do not let defeated armies return in rebellion to anger you, to breach the encirclement[xii]and revive. Pieces in these situations are not of a single type: some induce “the ladder” and rush forward like waves; some follow “the ladder” and are instead lost; [xiii]some court death to stay alive; some invite danger seeking ko.

When the armies engage and the two compete, stop the battle and study carefully. Bolster weak groups, since the weaker will be attacked first. Even though the strong groups are self-sufficient, still build up territory. The weak groups must be encroached upon, since encroaching yields profits. Don’t wastefully fortify live groups. Don’t strive to save already dead pieces. Don’t cut two living groups. Don’t link two dead groups. There is no benefit in linking, and cutting loses sente. In Go there is the notion of “losing one” and “playing to gain two.” If you enter [your opponent’s] boundaries and attack his borders, you will often hold sente. [xiv]

Whenever there is a ko situation, the one with the advantage has first created shape and strategic advantage; the one at the disadvantage is the weaker one who has not done so. If the planning is great, then one can rescue [the pieces] from outside; but if there are only minor opportunities, rationally one must stay one’s hand.
Although, this essay is meager and clumsy, it can reflect the patterns of things. Those who play accordingly will stay alive without loss.


Giles, L, Descriptive Catalogue of the Chinese Manuscripts from Tunhuang in the British Museum. London: British Museum, 1957.

This, the earliest surviving Go manual, is dedicated to all those who spread awareness of the dignity and fragrance of the game. Peter and Sheila Wendes Zen Machine.

Physical Info


British Library


Dunhuang Mogao (none)




manuscript, ink on paper

Size (h x w) cm:

15.5 x 240


Information available on the site of the International Dunhuang project;recnum=10528

[i] The Chinese name for the game is weiqi, the first character of the title is a variant of qi, pronounced go in Japan. Since most English speakers familiar with the game have studied it with Japanese terminology, for the sake of clarity, I use Go instead of weiqi, and try to apply the appropriate Japanese terms where applicable.]

[ii] One key being that black pieces are not called “black pieces” 黑子, but “raven pieces” 烏子. Cheng found that “black” was part of the childhood name of one of the Northern Zhou emperors. Traditionally, one cannot use a character in an emperors name out of respect, and so must substitute another character for that which is to be avoided. In Northern Zhou, “Black River” and “Black Peak” were referred to as “Raven River” and “Raven Peak” to avoid the taboo character “black.” Cheng Enyuan, Dunhuang Qi Jing Jian Zheng [An Examination of The Dunhuang "Classic of Go"] Chengdu : Shu Rong qi yi chu ban she : Sichuan sheng xin hua shu dian fa xing, 1990), 95.

[iii] Anonymous, "敦煌棋經," CCC Yenching College,

[iv] All sources agree that there is missing text. Cheng Enyuan suggests that, based upon the average length of subsequent chapters, it is no more than 3-5 lines of text—about 45-75 characters. Cheng, 67.

[v] The reference “Ping” seems to be to the Late Qin-Early Han figure Chen Ping, who was one of Liu Bang’s companions in the wars to establish the Han dynasty. Ibid., 68.]

[vi] Original character is unclear. Cheng suggests it is a Northern Zhou variant of . Ibid., 70.]

[vii]  Here shi () is translated as “strategic advantage” following Roger Ames’s translation of Sun Tzu: The Art of War. The term, though, is multivalent. As Ames writes, “We must struggle to understand how shih can combine in one idea the following cluster of meanings:

1) aspect, situation, circumstances, conditions;

 2) disposition, configuration, outward shape;

3) force, influence, momentum, authority;

4) strategic advantage, purchase.” Sun Tzu: The Art of War, ed. and trans. Roger Ames, The Classics of Ancient China (New York: Ballantine Books, 1993), 73. John Fairbairn has suggested “strong posture” as an alternative translation of this term, making a clearer connection to shapes. The key is that this term indicates a temporary advantage in a specific position at a specific moment. This strategic advantage can be enhanced or lost as the game continues; hence, timing is important. In the chapter entitled “Strategic Advantage,” Sun Tzu wrote, “That the velocity of cascading water can send boulders bobbing about is due to its strategic advantage. That a bird of prey when it strikes can smash its victim to pieces is due to its timing. So it is with the expert at battle that his strategic advantage is channeled and his timing is precise. His strategic advantage is drawn like a crossbow and his timing is like releasing the trigger.” ———, Sun Tzu: The Art of War, 120. Wang Xi, commenting on the term, wrote that the title Strategic Advantage “refers to the fluctuations in the accumulation of strategic advantage. The expert at battle is able to use strategic advantage to seize victory without taxing his strength.” Sun Tzu,  Sunzi shijia zhu [Ten Masters’ commentaries on the Sunzi], vol. 6 of Zhu Zi Ji Cheng [Collected Commentaries on the Various Masters] (Beijing: Zhong hua shu ju : Xin hua shu dian Beijing fa xing suo fa xing, 1954; reprint, 1990), 66. Strategic advantage depends upon preparation of the right shapes that will stay alive while supporting further encroachment. Perhaps underscoring this, the commentaries on the Sun Tzu note that the chapter “Strategic Advantage” logically follows the chapter on “Strategic Positions” (). ]

[viii] Following Cheng, that is a scribal error for , given the parallel phrases in later chapters and other Go texts.]

[ix] This line underscores the tension between Go and the Confucian ideals, which call for straightforward actions based on moral principles. Hence, some Confucians, including Confucius, considered Go a waste of time, or worse.]

[x] In 661 BC, Duke Xian of Jin paid the Duke of Yu for permission to send Jin’s army through Yu to attack the state of Guo, which he did successfully. In 654 BC, the Duke Xian again persuaded the Duke of Yu to give the Jin army passage through Yu to attack and destroy Guo. Upon accomplishing this, the Jin army returned, occupied Yu and annexed it. James Legge, ed. The Ch'un Ts'ew with the Tso Chuen (Taipei: SMC Publishing Inc.,1990), 145-7. The point being that the Duke of Yu did not suspect an ulterior motive for Jin’s request for permission to send its army through Yu.]

[xi]  The manuscript reads “若諸葛之行丘.” Many read as a variant of , hence “like Zhuge [Liang’s] use of soldiers.” Cheng argues that, in fact, an obscure meaning of is “empty,” and that this points to one of the most deceptive and successful strategies employed by Zhuge Liang in Romance of the Three Kingdoms. 成恩元, 敦煌碁经笺证 /Dunhuang Qi Jing Jian Zheng, 79-80. In one instance when his army was routed, Zhuge Liang knew that the city where he was could not repel an attack. So he sat himself above the city gate, playing the qin, opened the gates and had peasants leisurely sweeping the streets. When the attack came, the opposing general thought it was a trap, due to Zhuge’s famed planning, and withdrew. See Guanzhong Luo, Three Kingdoms : China's Epic Drama, trans. Moss Roberts, 1st ed. (New York: Pantheon Books, 1976), 285-89.]

[xii] [It is unclear whether the manuscript reads “encirclement” or “state” .]

[xiii] [The second chapter of this work is devoted to the ladder strategy ().] 

[xiv] [Here the Chinese is the same as the Japanese Go term “sente,” which could be translated as “upper hand,” or “momentum,” or “advantage.”]

Retour Go

Derničre modification le 23/02/2021