A story well told is not necessarily true - being a critical assessment of David H. Li's "The Genealogy of Chess"

by Peter Banaschak


Very recently Professor David H. Li published his monograph "The Genealogy of Chess" with Premier Publishing of Bethesda (Maryland)(ISBN 0-9637852-2-2). When I first heard about him working on a book-size history of Chinese Chess (Xiangqi) and the connection of this game to other chess-games including the Persian variant which evolved into the "international" chess of our time I was very intrigued. So naturally I was quite excited when finally I held his book in my hands.

His work is organised in three parts. The first deals with the origin of chess as presented in Western literature from the 13th century up to 1996; the second part presents his hypothesis that chess was invented in China; and the last part covers the further development and dissemination of the game after its invention.

In the first part of his book he examines the views of some 30 Western researchers of different scientific background, ranging from non-scientists just offering their more or less well-founded views on the matter to specialists on some field or the other. After having rebuked most of them for inconsistencies in their respective hypotheses he casts a vote on the original country of chess for each of them. He counts 13 votes for India and 6 for China (and from 1 1/4 to 1/4 for other candidates from Babylonia to Russia) (p 114). He then proceeds to inspect other writer's definitions of "Chess" and arrives at a rather narrow (but of course, fully acceptable) definition of what "Chess" is. Furthermore he depicts the attitude of the Chinese literati (ie. the learned scholar-officials), and describes the Chinese games of Liubo and Weiqi. Now he produces his hero, the Chinese general Han Xin. This Han Xin is also named as the inventor of Chess in Eyles Irvin's "Account of the Game of Chess, as played by the Chinese".1 He then plunges into a narrative of how Han Xin develops the game out of the principles of Weiqi and Liubo, the military and strategic ideas put forward by the Chinese military classics and his immediate needs as a commander of troops who are encamped during winter. The Chinese proto-Chess which he constructs mirrors all these necessities. In the third part he explains how this Chinese proto-Chess got out of view for some centuries and how it was revived under the Tang dynasty (618-907), and how and why the game was augmented and altered until it gained the appearance it still has today. Finally he reveals how chess was transmitted into the west (first Persia and then India) and the east (Korea and Japan).

In his preface he states that his mission is 'to set the record straight. To wrestle the honor of inventing chess from from ones who are undeserved (Š) and bestow it upon its rightful claimant, whose recognition is denied for centuries.' (p 5)

I would be glad if I could call his work well done and the case he presents a convincing one. His book is well written, it is a pleasure to read, and especially the first part contains some candid views on some of the old boy network of chess historians. All in all is well worth reading.

But: it is (in its present state) in no respect to be called scientific.2 There are some aspects which allow this judgment. First of all, a book on Chinese Chess that does only occasionally contain Chinese Characters makes an awkward impression. Technical restrictions that prohibit the use of Chinese Characters in the text should be countered by providing at least a glossary of Chinese terms together with their rendering in Latin script. This leads to the second aspect. Li does not decide on what transcription system he will consistently use. He mixes Pinyin and Wade-Giles3 (which he occasionally calls Giles-Wade, p 143), sometimes even the Needham variant of Wade-Giles renderings freely.4 Apropos renderings: Li insists on translating most of the Japanese and Chinese titles of his references throughout. This without, of course, giving the original titles in their original script anywhere,5 but for one instance: he gives the title of a paper by Isaak M. Linder in Cyrillic script (although Li confesses that he can't read Russian and quotes Linder from someone else's work, p 105). It is unacceptable that a researcher does not clearly and explicitly state where quotes are taken from. He does not state what edition of older works6 he uses (he only gives the presumed year of publication of the original7). He lists as a reference books of which he himself states that they are lost.8 All this makes the bibliography slightly confusing. In addition he switches the format in which he gives bibliographical data from one entry to another. He does not always check Chinese sources quoted someplace.9 When he does, as with his translation of a seemingly chess-relevant passage of Chuci ("The Elegies of Chu") (p 129/130) he completely disregards the existing standard translations10 of the respective passage (although they more or less agree with him), and even worse, he ignores that there are lots of good Lunyu ("Analects of Confucius") translations (who only in error would dare to translate "boyi" in Lunyu xvii, 22 as "chess"). He mentions works as "consulted but not cited in this book" (p 354), and has in fact quoted from them.11 Moreover he seems not to be familiar with his predecessors who have already written on Chinese chess. His bibliography is very selective (to put it mildly).12

One might say that all this are only technical niceties which are irrelevant to the content. I could reluctantly agree on that. Of course important contributions can be made in unfit form. But again, there is a But: the very core of David H. Li's findings is the declaration that the Western Han (206 BC-23 AD) general Han Xin is the inventor of Chess. This statement is based on virtually nothing. Of course there is the account that Eyles Irwin gives, but subsequent research until now has failed to identify from which source Irwin's informer, Pan Zhenguan, has drawn his information. And exactly this is the crucial point: Although the shred of information provided by Pan Zhenguan through Irwin is as good a starting point as anything else, it is simply not enough. It would have been a great find indeed if Prof. Li could tell us from which earlier source the account given by Irwin comes. Lacking this vital piece of information we cannot follow the tracks of the basic text back through time to decide on its credibility. But as it is, the story Li tells is nothing more: a good story, well told, a story plausible enough in itself, but it is simply not backed by any evidence. He has devised an ingenious explanation for the nearly eight centuries of complete silence from Han Xin's death to the seventh century. According to Prof. Li that is because Han Xin incurred the displeasure of the ruling elites after he was put to death. Li tells us it became too dangerous to openly play the game. But of cause this does not account for the inclusion of Han's biography in the biography sections of the Shiji (the first and the model of the great dynastic histories13) and the Hanshu.

Not only that Prof. Li does not play the scientific, the great game according to its rule, not only that he has practically nothing besides his fertile imagination to back his claim that Han Xin invented chess, Prof. Li commits a few more errors. He leans heavily on Zhang Ruan's Zhongguo xiangqi shi (Peking 1991), but is unable to give Zhang's name correctly (he gives Zhang Anru instead, eg pp 366, 383). He quotes Zhao Buzhi on Guangxiangxi (Broad Xiangqi), but keeps on calling him Yao Buzhi (pp 266, 269, 365). He gives the Chinese word for 'Trigram', gua (today in the 3rd tone), as 'qua' throughout his book (eg pp 157-9, 256-7). All this happens far too often to be just a spelling error or a typo, it's a systematical error. Furthermore, his speculations on why and how the logographs on the pieces (pp 206-207, 230-231) were devised and later changed lack credibility. We simply don't have any old pieces, thus we don't know anything about how the pieces looked like and how they were inscribed. On p 231 he proclaims that the 'earlier' Name of the game was "Game to capture Xiang Qi"14 (Xiang Qi being the enemy of Han Xin) and was changed to "Game to capture Jiang"15 (jiang meaning 'general'). He calls 'xiang' and 'jiang' "almost homonymous". Sinologists16 have reconstructed Middle Chinese (the language of the Tang dynasty (618-907) capital). Xiang Qi (Han's enemy) would have been pronounced sâÿng´ dz'ia`, whereas xiangqi (Chinese Chess) would have been pronounced ziang´ g'ji, and jiangqi (Chinese Chess, but with 'General' as first character) tsiang` g'ji. I find it none too homonymous. But however, why any of these should be "easier to pronounce" (which he thinks accounts for a change of the name of the game) than any other (p 339), remains Prof. Li's secret.

Prof. Li's relevations regarding the further development of the game and its west- and eastward spread would at least need a review of the same size as this present one, but I will spare him, you and myself the labour. In his "Dedication" (p 8) David H. Li states that he showed no part of his work to anyone prior to publication, and acknowledges that all errors are his own. He submits his case to his jury: the readers of his book. Were I prosecutor, I think I had a good case against him. Were I the claimant I would not take him as defender or counsel. As reviewer of his book I think he had better discussed his "findings" with a sinologist.

It remains a fact that the "Xuanguai lu ('Tales of the obscure and peculiar')" by the Tang Minister of State Niu Sengru (779-847) is the first real source on Chinese chess. Until now it has not been convincingly demonstrated that any text or archaeological find is of an earlier date.17


(back to top)

You can jump back to the text by clicking on the respective note's number.

1 in: Transactions of the Royal Irish Academy, Antiquity Section, Vol. 5, 1793, pp 53-63.

2 Li does not call his work "scientific", but on p 5 he says of himself "(...) I unearthed credible evidence, I exercised disciplined judgment, I did my level best (...)", and these are requirements of scientific standard as well (the rest of scientific standards would be about presentation of one's findings).

3 Sometimes (eg p 342) he even uses two transliteration systems to represent one person's name (Pinyin: Yuan Zhuang (whom he insists calling Zhaung), Wade-Giles: Yuan Chuang).

4 Not even to mention the outdated transcription system used by eg Karl Himly. Li constantly refers to the Shuo yüan (pinyin) as Shou wn (eg p 55). Concerning the passage from the Shuo yüan in question, Li makes fun of Himly's incorrect translation (p 55-56), completely ignoring that Himly had only limited access to Chinese source texts and good dictionaries, and there was only a limited philological experience with Chinese texts. This is, of course, also true for the other Westerners, even those in China of the time.

5 Disregarding that this makes its unneccessarily complicated to check quotes (or is this exactly what he tries to do ?).

6 And he is very generous with the term 'Classic' which he bestows on about every old Chinese text.

7 Thereby totally disregarding that completing a work and having it published are not at all the same, and as well ignoring that some Chinese works have very complex publishing and editing histories.

8 eg "Lü Cai (600-665), The Encyclopaedia of Games, in two volumes" (p 364): but "That minister, Lü Cai (600-665), is the author of a two-volume work on The Encyclopaedia of Games. The work, unfortunately, is no longer extant." (p 215, note 4)

9 One simple example: on p 271 he shows the initial array of a Chinese Chess variant, Qiguo xiangqi ("The Chess of the Seven Realms", which he calls Heptagonal Xiangqi) and lists three references, two in Chinese and one in English. But: in Gujin tushu jicheng (101 Vols.) Taipei 1964; (an encyclopaedia first printed in 1726; it compiles older texts under 6109 subheadings in 10.000 books) which a sinologist would surely use, Vol. 59, p 1027, has a nice reprint of the plate with the array and the Chinese text accompaniying it.

10 eg Erkes, Eduard, Das "Zurückrufen der Seele" (Chao-hun) des Sung Yüh, Text, Übersetzung und Erläuterungen, Leipzig 1914; Hawkes, David, Ch'u Tz'u, The Songs of the South, Oxford 1959; Wang Aiguo, Chuci zhaohun bian yanjiu, o.O. 1973.

11 eg he claims not to have cited Leventhal, The Chess of China, Taipei 1978, but in reality he refers to this book on p 271.

12 eg Zhou Jiasen's works are not referred to.

13 As a fact Han Xin is mentioned about 300 times in the Shiji. He is reckoned among the three great military heroes of the Chinese.

14 The dictionaries I have consulted agree on giving the second character with which the name is written as 'ji' (2nd tone).

15 Both of which ideas I heavily doubt, as again they are not backed by evidence.

16 esp. Bernhard Karlgren, Grammata Serica Recensa, BMFEA 29 (1957), Repr. Taipei 1996.

17 Prof. Li mentions (pp 215-218) some interesting events - but these cannot be checked, as he fails to state what his sources are. Until then, they have to be dismissed as fictitious.


Proceed here to the glossary of Chinese terms used on this page. Note that you'll might have to manually switch the character set used for viewing to traditional Chinese - BIG5.

Kontakt zum Autor aufnehmen?