Chinese-Western contacts and chess

Peter Banaschak


This paper deals with some aspects of Chinese-Western contacts, with special regard to the earlier history of chess. I try to hint at some interesting perspectives for further research.

Introductory remarks

This short paper is (loosely) based on a paper delivered to the 5th Symposium of the Initiativgruppe Königstein, held in Hamburg, Nov. 26th to 28th, 1999.

Throughout this paper Xiangqi will be used for `Chinese Chess'. Xiangxi refers to Emperor Wu's game (the so-called ``Astrological game''). Transcription systems used are Karlgren for Middle Chinese and Pinyin for Modern Chinese.

By CHESS I understand any game that belongs to the board game class of Games of the Chess type (in narrower sense).

1  The starting point

The belief that chess originated in India in roughly the sixth century AD is commonly held. If so, the game must have spread throughout the world. Its way to the West is more or less known from the time on the Persians learned the game. In contrast, next to nothing is clear about its diffusion to the East; but the game must somehow have travelled to China. Nevertheless, this is a question which proponents of the Indian origin of chess must be ready to answer. Many researchers into the history of chess have either not bothered themselves with questions on the connections between East Asian chess games as their main interest lay in clarifying the history of what was to become FIDE chess. - Let us assume thus, for argument's sake, that

Starting from this assumption, an early chess game must somehow have been brought to China, e.g. . Although there are quite a few ways of transportation possible, we'll deal mainly with only one (however, the most likely) candidate: the tangle of trade routes known as the Silk Roads and the areas, realms, and regions around it. These Silk Roads link the areas of the Mediterranian Sea and the Caspian Sea to North-West India, Tibet, China, Korea, and South-East Asia.

Click here to view a map of the Silk Roads (map opens in new window).

2  Some remarks on the early history of Xiangqi

Before we start, let me point out that there is an idea that chess originated in China proper. This idea is mainly fueled by texts brought to attention by Joseph Needham. He proposes that the ancestor of all chess games was a game Xiangxi that was supposedly invented by emperor Wu (r. 561-578) of the Later Zhou dynasty (557-581). This is reported in

The first certain references to a game of (possibly) the chess type in Chinese sources date from the early 9th century: these are the short story Cen Shun by Niu Sengru (779-847) and a poem by Bo Juyi (772-846). For the sake of convienence this game is called Baoying-Xiangqi. Nevertheless both texts offer no hints as to how and where the game originated. It has been said that neither text offers a sufficient description (which is true), but if we take both texts into account, we can deduce that in the early 9th century there must have been something not entirely unlike modern Xiangqi.

3  Some ways how caturanga might have come to China

3.1  From India to China?

The origin of chess in India is often linked to the rule of the last Indian High King Harsa. Ainslie Embree mentions that king Harsa (r. 606-647) tried to open diplomatic relations between his own realm (based in Kanauj) and Tang (618-907) China under emperor Taizong (r. 627-650). There was just one embassy, and when the Chinese return mission arrived after Harsa's death, they were attacked by the governour of Tirabhukti. (This led to an attack against and consequent occupation of Tirabhukti by a joint Tibetan-Nepalese army.) Thus, besides of the traffic of merchants, pilgrims, monks, and armies on the Silk Roads, there have been direct contacts between two of the countries which have been credited with the invention/ development of chess, and that during the life times of two rulers who are named quite often in connection with the origins of chess.

There are numerous other contacts between Chinese and Indian merchants and monks in later as well as earlier times, but this large-scale contact should suffice as an outstanding example.

If chess or one of its ancestors had been imported to China it would not be unsensible to expect that this import has left traces. As these traces are not to be found in the shape of the pieces, perhaps there might be linguistic evidence of a kind?

If something is imported from a `foreign' country, the foreign word to name that thing is often imported alongside, but usually this leaves traces in the language. Bernhard Karlgren has reconstructed the Chinese language, more precisely the Chang'an dialect, of about 600 (the so-called Middle Chinese). The Chinese words ``Xiangqi'' and ``Xiangxi'' as well as their Middle Chinese counterparts do not show any influence of neither the Sanskrit word nor its Chinese rendition. The same is true for Persian catrang and Arab shatranj; these words, too, seem to haven't had any influence on the Chinese names of chess games. Of course it might be perfectly possible that someone changed the name of an imported chess game to a more Chinese sounding name. ``Xiangxi'' was already known from emperor Wu's Xiangjing, so (as no one knew what Xiangxi really was) the word might have been adopted to render caturanga.

3.2  From Persia to China?

Accepting the time table recented proposed by Syed and Abka'i-Khavari, that is

a chess game ought to have been imported to China until 846 (the year of Bo Juyi's death). It is in no way superflous to point out that the Middle Chinese predecessors of the words Xiangxi and Xiangqi do not show influences of the Persian word catrang, as there was a continuous flow of traffic on the Silk Roads between Persia and China, not only of caravans carrying goods, but also of diplomatic missions. Chinese sources record that from 553 to 578 ambassadours from

arrived in Chang'an, the then capital of the Later Zhou dynasty. Most of these missions arrived during the reign of emperor Wu (r. 561-578); thus if one of them brought something similar to chess, it may well be that it had influence on the formation of Xiangxi.

Franke and Trauzettel mention that official Chinese-Persian contacts continued in 638, when the last ruler of the Sasanids, Sâhânsâh Yazdegerd III. (r. 632-651) sent a mission to Chang'an (then capital of the Tang dynasty) to request military assistance against the Arabs. His son Piruz in 654 and 661 again appealed to the Tang, but was denied help. He was awarded a formal rank as military governor of Afghanistan. In 674 Piruz fled to the Chinese court where he died in 708 as state pensioner.

Already in 651 Chalif 'Otman ibn 'Affân (r. 644-656) had sent a first mission to Chang'an; from about 700 the Arab and Chinese spheres of influence touched each other in East Turkestan. On the occasion of all these diplomatic contacts an early form of Persian or Arab chess might have been brought to China.

Another interesting fact is that Samarkand, the place where some of the eldest known chess pieces have been found (the so-called Afrasiab pieces), was part of a Chinese protectorate in which Chinese craftsmen are known to have settled. This doesn't prove anything, but it shows some (hitherto unresearched) perspectives.

3.3  Can that be proven?

Nevertheless, questions remain. The first problem being that we don't know whether Persians or Indians played any chess game in China (or vice versa.). Abka'i-Khavari tells us that according to the Sahn¯ame catrang served as a didactic model in the education of princes in late Sasanid time Persia but the text in question dates from the 10th century. Thus one may doubt that it accurately reflects life at the Sasanid court. If we accept that Piruz and/or his retainers knew catrang, they might have made the game known at the Chinese court. The next problem is that (in case the time table presented here holds some truth) we do not hear anything about chess games until the early 9th century. In Niu Sengru's and Bo Juyi's texts Xiangqi appears all of a sudden and without any explanations, as if it were a well-known topic (at least well-known to Niu's and Bo's fellow scholar-officials). A certain diffusion of the game would be necessary for the game to `pop up' so suddenly.

And then there is the question of the gaming material. Neither in Baoying-Xiangi nor in the game mentioned by Bo Juyi we find the four branches of the Indian army. Niu's text explicitly names Horse (resp. Cavalry), General, Chariot, and (Foot-) Soldiers, it can be deduced that there were Crossbowmen and Catapults, in addition there is mention of a King. Bo explicitly names Soldiers and Chariots. The branches of the army that Niu mentions are not identical with those of the Indian army. Then we don't know anything about the shape of the pieces. If we proceed from the Afrasiab pieces dated to 761 Indian/Persian pieces were actual three-dimensional figurines. Niu's text which describes Baoying-Xiangqi possibly speaks of three-dimensional figures. If that really were the case, no traces thereof are known.

The next unanswered question regards the game board. It seems safe to assume that the astapada board was handed down to the Persians along with the pieces. We do not have any information as to the size of the early Xianqi board. It is speculated that the transposition of the pieces from the squares to the lines of the board (as in Xiangqi) triggered the invention of a board of 9 by 9 squares, which in turn triggered the invention of a board of 10 by 10 lines etc. This may be plausible but is utterly unproven. It seems plausible, however, to assume that the transposition from squares to lines in Xiangqi is due to the fact that in other Chinese board games the lines, but rarely if ever the squares were used. The earliest known source on the board size is the preface to Guang xiangxi tu by Zhao Buzhi (1053-1110). There we are told that in Xiangqi 34 pieces are used on a board of 11 by 11 lines.

The question of the initial arrangement of pieces cannot be answered, as we don't even know what kind of and how many pieces were used in early Xiangqi. We maybe may assume that the initial arrangement was modeled after that of chatrang (which in turn we do not know precisely).

It should have become clear that Central Asia and the Silk Roads were under constantly changing influence to which both the Indians and the Chinese contributed. The how, where, and when of these interactions is still not very clear as a sizeable portion of research has focused on religion, less on everyday life, and next to nothing on such questions as board game research.

4  Back to Reality

Returning from the land of what-ifs, we must simply state that too many questions remain yet unanswered to make any definite statement about the possibility of interconnections of early chess games in general, although there are hints that exactly that is the case. It has repeatedly been demonstrated [e.g. by Beauchamp] that not only the moves but also the initial position of the pieces in various older chess games are too similar to explain them through `sheer coincidence' without further research (It will not be enough to glimpse at them and declare them connected (or not connected)). Nevertheless these similarities are not enough to prove that all chess games share a single common origin, and they are not enough to prove that the Asian chess games are somehow connected.

First of all, we chess historians should stop to speculate. It is time to reinspect all known source materials, this time without bias. By ``all known source materials'' I mean all known source materials, even those who are held sacrosanct. Sciences make progress, and something that was deemed true a century ago may today be obsolete. If that has been done we can decide whether additional research in certain directions that have been neglected so far has to be done. As soon as we know what the sources really tell us we can launch a search for the puzzles' missing pieces. If the sources lead us to think that materials that might link certain chess games may be found in a certain area, at a given time, we should devote special care to that given area and time. We might find that we ourselves are unfit to do that work, thus we have to recruit specialists in all the fields needed. Hopefully I have been able to transport the necessity of finding specialists e.g. for the languages and cultures of South East Asia in order to find out more on the chess games of that region. ... and we need a lot of experts. Every opinion has to be checked and rechecked to exclude the possibility of biased opinions. Only then is the time to formulate new working hypotheses, which after being backed by sources may have enough support to become a thesis. We are still far away from any theory on the origin and development of chess (even if there are some who will not like the idea). Up to now it is only benevolent to call the hypothesis of the Indian origin of chess a `hypothesis'; in fact it's not a hypothesis, but a thought, an opinion, a mere speculation backed by only little material. It is time to openly declare that this opinion is nothing more than that; until, of course, unbiased and thorough research makes it possible to formulate the idea as a hypothesis.

File translated from TEX by TTH, version 2.25.
On 30 May 2000, 10:23.

Kontakt zum Autor aufnehmen?