A new paradigm for an "Origins of Chess" theory
by John Ayer

Abstract
This essay argues that the generally accepted scheme for the derivation of the current and disused forms of chess from the original Indian proto-chess is mistaken: that an early variant called Shatranj al-Kamil was the source of chatrang/shatranj/ medieval chess, and also of Shatranj al-Kabir (known as Tamerlane’s Chess), and also of a Chinese game ancestral to the present Chinese hsiang-qi and Korean janggi, and that that game was fused with elements of the Burmese sittuyin to form the Japanese shogi.

Present State of the Problem
It is the present view that chess is descended from a game created in India (we must remember that India’s modern boundaries are, in fact, modern). The reasons are that the name of the game is Indian, and that the earliest references to such a game, coeval in India and Persia, agree that it is Indian. The Persian terminology is partly the original Indian words modified to fit Persian convenience, partly Persian translations of the Indian words. After the Moslem conquest of Iran the terminology was altered further to suit the Arabic language, and knowledge of chess was spread wherever Islam penetrated, and by various routes into Europe and eventually all over the world. As the game traveled, the rules were gradually modified.

It is recorded that chess was transmitted from India into China at an early date. According to Chinese reference works, the Emperor Wên-Ti (ruled A.D. 581–604), while traveling, once found two of his subjects engaged in a game in which two pieces were called emperors, and one bore the sacred title of “white emperor.” The emperor ordered all present beheaded, and forbade the playing of any game in which a piece was called emperor, or was carved to resemble an emperor.(1) To this day, one “king” in Chinese chess is called a governor, and the other a general. From China the game passed to Korea and Japan.

Some have claimed that chess is mentioned in Chinese literature before it is known in India, and that chess was invented in China and traveled from there to India and the rest of the world. This cannot be sustained. The Chinese name for chess, hsiang-qi, can be understood as “elephant game,” “ivory game,” “figurine game,” or “astronomical game,” and is known to have been used for two other games before chess appeared. For this argument to be accepted, therefore, hsiang-qi in such passages would have to be described in sufficient detail to establish that chess rather than some other game was meant. Such literary support has not appeared.(2)

It is sometimes urged that chess is derived from a system of divination. This is entirely possible. Children will play at, and make a game of, almost anything they see their elders doing. They often try to cajole their elders into playing these games with them. They may continue to play these games when grown: paintball comes to mind. The Malagasy have a board game, Fanorona, which they sometimes played for divinatory purposes.(3) As is well known, playing cards are used for divination as well as for a great multitude of games.

One striking fact about chess is the radical difference between the pawns, which are numerous and can only move forward, but capture diagonally forward, and the pieces, which are of several kinds, each kind having only one or two members on each side and possessing radially symmetrical powers of movement, and capturing as they move. Two inquirers have lately based on this contrast a theory that chess originated to the north of India in a fusion of the Greek game poleis and elements of Indian origin. Myron J. Samsin conjectures(4) that Bactria was the scene of a fusion with elements taken from an Indian race-game, perhaps chaupur. Gerhard Josten postulates(5) a date a couple of centuries later, and a location in the Kushan Empire, and proposes further that the Chinese game of liubo was also laid under contribution. We take up the story roughly where they have left off.

Chaturanga, the Early Indian Chess
Chaturanga, the earliest known game in the chess family, is documented from the sixth Christian century. It was played on a square board, eight squares by eight, that was not checkered, but had sixteen squares specially marked. These marks are not used in chaturanga, so the board was probably taken from a previous use, likely a race game for four players. Two miniature armies, one white and one black (or one green and one red), were arrayed opposing each other on this board. On each player’s back row a king and his minister stood side by side on the two central squares. The king moved one square in any direction, the minister one square diagonally only. Flanking them were two elephants, each leaping diagonally to the second square. Flanking the elephants were two horses, which moved as the chess-knight still moves, leaping to the opposite corner of a rectangle two squares by three. Flanking the horses, two chariots stood on the corner squares. They moved as the rook still moves, any number of squares along an unobstructed rank or file. On the rank immediately in front of these pieces stood eight foot-soldiers, moving one step forward but capturing by moving one square diagonally forward. All the other pieces captured as they moved, and, as today, all captured by landing on a square occupied by an enemy piece. A pawn that reached the far side of the board might, depending on circumstances, be promoted to a piece. Elephants, horses, chariots, and foot were the four traditional branches of an Indian army, and “chaturanga,” meaning “four members,” was an established term for an army. The goal of the game is “checkmate,” which means that the opposing king is under attack and cannot escape. To preserve the royal dignity, the king is never actually captured. “Checkmate” is popularly derived from Persian shah mat, meaning “the king is dead,” but the actual origin is shah mand, “the king is helpless.”(6)

Chatrang, Shatranj, and Medieval Chess
Chaturanga seems to have passed along the caravan route from northwestern India to Persia. The names of some of the pieces were translated into Persian—the Indian rajah became a Persian shah—but the Indian minister was replaced by a Persian general, or “ferz,” and the elephant was called the “pil,” which is neither Indian nor Persian. The chariot became the “rukh,” a word that has been remarkably persistent. The name of the game was not translated; it was simply shortened to “chatrang,” which has no Persian root. (7) While the Indian practice was that each player’s king stood just to the left of the centerline and each player’s minister just to the right, the Persians left the white king where he had been, but placed the black king at the far end of the same file, and the general facing the general. No squares were marked. The Persians knew that they had received this game from India—their literature contains some fanciful tales about the event—but they adopted it quite thoroughly; soon the aristocratic heroes of their literature were as accomplished at chatrang as at horsemanship and hunting.

In the seventh Christian century the newly Moslem Arabs conquered Persia, and Arabic became the language of government and of trade with distant parts of the world. In spite of some religious misgivings, the Arabs took up chatrang and spread it far and wide. Lacking the “ch” sound, they called the game “shatrang,” which a later sound-shift modified to “shatranj.” Lacking the “p,” they called the elephant the “fil,” or, with the Arabic definite article, “al-fil.” The ferz became a “firzan” or “farzin.” The Arabs usually played on a cloth, with the playing field marked by embroidered lines, or by rows of embroidered rosettes. All the squares were alike. A player who placed a piece so that it attacked his opponent’s king warned his opponent by saying, “Shah!” This is the origin of our words “check” (in all its senses) and “chess.” The opponent was then obliged to take any legal means to protect his king. If he was unable to escape the attack, the condition was “checkmate,” and he had lost. A pawn that reached the far side of the board was promoted to ferz. The ferz being a weak piece, pawn promotion was not a major factor in chess strategy.

The Christians of western Europe were in contact with the Moslem world along several lines, and may have received knowledge of shatranj from more than one direction. At first they made no changes in the rules, though they didn’t care whether the white king stood to right or to left of his ferz, a word that they kept for some time because they didn’t understand it. Wandering minstrels often carried chess sets, with small and light-weight pieces carried in a bag with an Arab-style chess-cloth. The game became immensely popular in the castles, where there were few diversions. There the board and men were made of wood, so big that there are tales of the boards being used as improvised shields, and the pieces as weapons.

Shatranj al-Kamil, Variation One
The history of chess involves a great many proposed changes to the rules, of which, over the centuries, a few have won wide acceptance. One of the earliest variant games is known from Persia in the eighth Christian century and was attributed to one Khalil ibn Ahmad. The innovation was the introduction of a new piece, the camel, which leapt to the second square along a rank or file. One stood in each wing of each army, and the board was enlarged to ten squares by ten, with ten pawns on each side. This game was called Shatranj al-Kabir, the Great Chess, or Shatranj al-Kamil, the Perfect Chess. To students of chess history it is now widely known as Shatranj al-Kamil, Variation One. It persisted for some centuries, though it never threatened shatranj.

European Great Chess: The Runner Game
The first recorded European innovation in the rules of chess is mentioned in a German poem of 1202 A.D. Each player had twelve pieces and twelve pawns; the board had eight ranks and twelve files. The white king stood just to the left of center on the back row, and the black king directly opposite him. To the white king’s right, and the black king’s left, stood their respective queens, moving one square diagonally; some Europeans had finally decided what to make of the ferz. To the white king’s left, and the black king’s right, stood a wise counselor, which moved one square in any direction, like the king, but, unlike the king, might be placed or left en prise. On the queen’s other side stood a spy, moving one square along a rank or file. Flanking the counselor and the spy was a new piece, known in German as the Läufer, that is, the runner or messenger. This piece moved any distance along an unobstructed diagonal. Flanking the runners was a piece called in German the Bischof, or bishop, which is the old alfil, leaping diagonally to the second square. Flanking these were the knights, and in the corners were the rooks, just like ours. The really distinctive innovation in this game was the Läufer, which is to this day the ordinary German name for the piece we call a bishop. For this reason the enlarged game is known in German as Läuferspiel, the runner game. The game persisted for more than six hundred years. The introduction of the long-distance diagonal move impelled the checkering of the board. There are several depictions of the runner game from the Middle Ages, which all show the board checkered, but some show a player’s right-hand corner square black, and some show it white.

The Rise of Modern Chess
Late in the fifteenth century another modification to the rules of chess was tried out. The old alfil was replaced by the German runner, and the ferz was replaced by a queen with the combined moves of the runner (our bishop) and the rook. The pawn was given the right to move two squares on its first move. This is basically modern chess, though some details remained to be worked out in pawn capture, pawn promotion, and castling. The new rules were adopted almost at once over most of Europe. The last written reference to “chess” that is definitely medieval chess is in 1475. Before the end of that century “chess” referred unambiguously to modern chess, references to “the old chess” having just about ceased.

Chinese Chess
Chinese chess, hsiang-qi, is so unlike European chess that the first Europeans who encountered it wondered whether the two games were related at all. The board is uncheckered, the two halves are separated by a gap one square wide called the river, and a large diagonal cross marks the two central squares in the two ranks at each end of the board. This board is known to have been used for two other games before chess appeared. Play is on the points, not on the squares. The chief piece on each side is not a king; one is a governor and the other a general. Each is assisted by two officials or officers. All of these pieces are confined to the nine positions defined by the X mark at each end of the board, which is considered a fortress. The governor or general moves along the vertical and horizontal lines of the fortress. His two aides-de-camp move along the diagonals, which means that each has only five possible positions. The elephant moves two points diagonally. He no longer leaps; a member of either army on a diagonally adjacent point blocks him in that direction. The elephant cannot cross the river, so he has only seven possible positions. The horse, too, has lost the power of leaping; his move is one square along a line and one more at a 45° diagonal. He can ford the river. The rook, or chariot, is posted at the corners, and still runs any unobstructed distance along a rank or file, and crosses the river without even slowing down. The pawns, foot soldiers, are posted on the fourth rank. There are only five of them on each side, so every second intersection is left vacant. The pawn moves one square forward only, but it can ford the river, and once it has gained the farther bank it can move one square to either side, as well as forward. There is no pawn promotion. On the third rank, and on the second file starting at either edge, is the most distinctive piece in Chinese chess: the catapult, or cannon (these pieces also have different names on the two sides). This moves just like the chariot, any distance along an unobstructed rank or file. However, it captures (still along a rank or file) by leaping over a piece to land on an enemy piece beyond. The piece leapt over is called the “screen.” The screen may be of any rank and belong to either side. It is unaffected by the projectile hurtling over it. Between the cannon’s point of departure and the location of its victim there can be any number of empty spaces on either side of the screen. However, there can only be one screen.

Notice that while the Europeans have developed chess by increasing the powers of certain pieces, the Chinese have decreased the powers of several pieces.

Other Asian Varieties of Chess
Myanmar and Thailand have their own native versions of chess, which are alike in that the elephant, standing on the third file in from either edge of the board, moves one square diagonally or one square straight ahead, moves that to them symbolize the elephant’s four legs and its trunk, its “fifth limb” (this move is also known from India). Each retains the ferz beside the king. The Burmese version, called sittuyin or sitturin, is unique in that it starts with only the pawns on the board, the left four on the third rank and the right four on the fourth. The players then alternately place their pieces, one at a time, behind their pawn-rows; then the moves begin.

Korea has its own national variety of chess, janggi, which is played with equipment that looks remarkably similar to the Chinese chess board and pieces (though there is no central river), but with numerous deviations in the rules: the generals and counselors, though all confined to their respective cross-marked pavilions, have each the power to move in any of the eight directions; the cannon must leap to move as well as to capture; the pawn can move in three directions throughout the game; the elephant ranges across the entire board, and its move is a longer version of the horse’s; etc. The pieces are labeled with Chinese characters. The two kings have different titles.(8)

Japan has a national variety of chess, shogi, which is now known and played around the world. It seems to have been derived from the Chinese, but has features unknown elsewhere. At the corners, instead of the rooks, it has lances that move straight forward only. Inboard of each lance is a horse, and inboard of that is a silver general, moving like the South Asian elephant, and next in is a gold general, moving one square in any of the six directions that are not diagonally backward. In the center of the home row is the jade general, doing duty for the king; the title is written differently on the two sides. On the second rank, in flanking positions reminiscent of Chinese cannon, are a rook and a bishop. The third rank is filled with pawns, which both move and capture straight forward. The pieces on the two sides are identical (except for the kings), and flat, with the rank painted on. Which side a piece belongs to is shown by which way it points. Almost any piece can be promoted, usually to gold general though there are exceptions. A promoted piece is turned over to show its new rank. A captured piece is held by the captor until he chooses to re-enter (“drop”) it on the board as part of his own force. There was formerly another piece, the drunken elephant, which moved one square in any direction except straight backward, but the Go-Nara Emperor (reigned 1526–1557) abolished it.(9)In addition to this standard game, Japan has brought forth a Baroque profusion of variants, some the same size as the standard game, some larger, some smaller.

Himly’s Reconstruction of Early Chinese Chess
Chinese scholars of the Ming period were aware that their forebears had played chess with different rules, but didn’t know what those rules had been. It had explicitly been played on a board of eleven lines by eleven. Karl Himly prepared a reconstruction of one end of the board (reproduced on page 124 of Murray's History of Chess), with the fortress marked by its X, and the king and counsellor aligned inside it. The fortress is flanked on each side by catapult, elephant, horse, and chariot (rook, in the corner). Before them is a skirmish-line of pawns. This diagram contains a subtle but crucial error that apparently misled Murray and all those who relied on his masterful work. As mentioned above, the Chinese chess-board was used for other games before the introduction of chess, and the fortress is a feature of that board. There is therefore no reason to project it onto the board of eleven lines by eleven. Remove it, and we have the unmarked board of eleven lines by eleven, or ten squares by ten, that was used for Shatranj al-Kamil, Variation One, and the same pieces, as well, with the piece that leaps along the rank or file changing its name but keeping a very similar leap. Murray remarked on the similarity between the earlier Chinese chess and Persian decimal chess, as he called it.

Kohtz’s Theory
According to Dr. Ricardo Calvo,(10)

The German historian Johannes Kohtz (1843–1918) supposed that in the proto-chess the Rook was also a jumping figure, with a mobility limited to a third square. So the squares accessible to a Rook in h1 would be f1 and h3, and later in the game f3, d3, d1, b1, b3, b5, d5, f5, h5, h7, f7, d7, and b7. His theory makes a lot of sense (in spite of Murray’s rejection after long arguments by post), because the three jumping pieces (Alfil, Knight, and Rook) represent a diagonal, hook-curved, and rectilinear movement of the same range. It also expresses a perfect ranking order: The King and the Knight are the only pieces which can move to any of the 64 squares. The Firzan has half of the board, 32. The Rook half of that, 16 squares. And the Alfil, half of that, 8....

However, once the Arabs acquired the game from the Persians, the Rook evolved into a long-range piece, becoming the most powerful element in the chess army. This evolution can be explained logically as a necessity once the idea of checkmate has appeared, again according to Kohtz. In the first legend of Firdawsi, the game re-discovered by Buzurdjmir was as follows: “The sage has invented a battlefield, in the midst (of which) the king takes up his station. To left and right of him the army is disposed, the foot-soldiers occupying the rank in front. At the king’s side stands his sagacious counsellor advising him on the strategy to be carried out during the battle. In two directions the elephants are posted with their faces turned towards where the conflict is. Beyond them are stationed the war horses, on which are mounted two resourceful riders, and fighting alongside them on either hand to left and right are the turrets, ready for the fray.”

“Turrets” is a definite mistranslation caused by the shape of European rooks. The word “rukh” now seems to be understood as “champion” or “hero.” We resume:

By the number of pieces it is easy to know that in this game the board was of 8x8 squares, though nothing is said about the rules of movement or the aim of the game. This gap is filled in the second Firdawsian chess legend about two half brothers Gau and Talhend (two typical Persian names), the latter being killed by the former during a civil war. To explain to the queen of “Hend” who was mother of both how her son came to die, the game of chess, which represented a battle, was invented. But it is a different game. The board is 10x10 and had perhaps a dividing line in the middle, as in today’s Chinese chess, because the text says: “This (game) represents a trench and a battle field onto which armies had been marched. A hundred squares were marked out on the board for the manœuvring of the troops and the kings”... This time the movement of the pieces is described. There are three pieces jumping to a third square in diagonal, rectilinear or hook-curved direction. But there is a fourth piece which is the most powerful of all: “None could oppose it, but it attacked everywhere in the field.”

This piece must be the long-ranging Rook, the most powerful figure of the set. Again according to Kohtz’s theory, instead of the previous jumping Rook, the long-ranging Rook was adopted as well in the 8x8 board as a necessity once a checkmate becomes the main goal of the game. Check and check-mate have already appeared, as the beautiful text explains:

“If a player saw the king during the struggle he called out aloud, ‘King, beware!’ and the king then left his square, continuing to move until he was hemmed in. This occurred when every path was closed to the king by castle,

This is still a mistranslation and an anachronism.

horse, counsellor or the rest of the army. The king, gazing about in all directions, saw the army encircling him, water and trenches blocking his path, and troops to left and right, before and behind. Exhausted by toil and thirst, the king is rendered helpless; that is the decree which he receives from the revolving sky.”

Where did the new idea come from? Since the moment when the fate of the king decides the victory in the game, the value of this piece increases enormously because of its “divinization” and inviolability. In a way, chess has become a monotheistic game. The cultural atmosphere in ancient Persia fits well with the implicit idea. In contrast to Greece, where a king was only primus inter pares, basically equal in his human nature to his subordinates, a Persian “Shahanshah” was worshiped almost as a God.

 

The New Unified Theory
Now, assuming this conjecture, we have an Indian pre-chess on an eight-square board with king and ferz in the central files, flanked by diagonal-leaping elephants, curved-leaping horses, and orthogonal-leaping pieces that may have been chariots, with single-step pawns on the second rank. The conjecture assumes that check and checkmate had not yet appeared, so perhaps the battle was to the annihilation of one side. When the Persians introduced checkmate, they also invented the modern rook to make checkmate a realistic possibility. This explains the fact that, while the name of the game is Indian, the name of the winning combination is Persian (even in India). My theory is that the new rooks were added at the corners of the array, with the orthogonal leapers now called camels, and the board was squared off (making a board of ten squares by ten) and two more pawns were added, making the game known as the “Great” or “Perfect” Chess, Shatranj al-Kamil, Variation One. On this hypothesis, Shatranj al-Kamil v.1 appeared considerably earlier than its first mention in the surviving literature,(11) and lasted about as long as Läuferspiel, and from it, I conjecture, developments radiated in several directions.

Since all the rest of the pieces still had moves of one or two squares, the new board was inconveniently large, and one remedy was to delete the orthogonal leapers and move the game back to the eight-square board. (Actually, Al-'Adli , in the ninth Christian century, wrote that the Indian chess-elephant leapt to the second square along a rank or file, which suggests some indecision about how to reduce the game to the smaller board.) Several early writers on chess said that shatranj had been abridged from the greater game.(12) Modern scholars, knowing that the eight-square game is older, reject this tale, but I suspect there is some truth in it. This game, with the new way of winning, then traveled back to India, where it supplanted its predecessor as swiftly and thoroughly as modern chess supplanted medieval chess in the late fifteenth century. As mentioned above, either chaturanga or the pre-chess was carried from India to China, but was convincingly driven out again.

From Persia, Shatranj al-Kamil v.1 traveled along the Silk Road to northwestern China. Perhaps the Chinese knew of the edict against games with emperors, or perhaps it was simply native caution that caused them to replace one king with a general and the other with a governor, or perhaps it arrived in a period of civil war, and they colored the game to match their surroundings. In any case, they moved the pieces to a board with which they were already familiar. The two chief officers found themselves inside a fortress, and were promptly forbidden to leave (the fortress is actually more of a prison, since it does nothing to protect the inmates from outside threats). The elephant was ruled unable to ford the river, and both it and the horse lost their power to leap. For some reason, the camel, now understood as a catapult or cannon, had its power of leaping enhanced, and its power of movement separated from its capturing leap and also enhanced. The pawn was given a slightly enlarged power on the far side of the river, and bidden to capture as it moved. Only the rook remained unaltered. From China the game, not yet in its modern form, traveled to Korea and Japan. China in former centuries had frequent commerce with all the world from Australia to the east coast of Africa, and Japanese ships, sharing in this commerce,(13) could easily have brought Burmese chess to Japan, which would have given them chess with the pieces placed on the squares, the silver general, and the idea of “dropping” pieces onto the board. It would then have been natural to interpolate the gold general and the drunken elephant between the silver general and the “king.” The lance was probably taken from an early form of Chinese chess,(14) and the rooks moved to the position of the cannons, and one them altered to move diagonally instead of orthogonally. The widespread opportunity of promotion reflects the state of affairs before the Tokugawa Shogunate.(15)

The existence of Shatranj al-Kamil I, with a long-ranging orthogonal-rider and an orthogonal leaper, and a diagonal leaper and a piece with a one-square diagonal move, stimulated the development of an even greater chess, Shatranj al-Kabir, which added a one-square orthogonal mover and a diagonal-rider, making orthogonal and diagonal families of pieces, each consisting of a single-step mover, a leaper, and a long-distance runner. The horse was then made the basis of a third family by adding a piece that leapt one square farther (to the opposite corner of a rectangle two squares by four) and a piece that moved one square diagonally and then rode a rank or file.(16) To add to the complexity, each of nine pawns was a miniature of one of these pieces, and would be promoted to its rank on reaching the far edge of the board. The other two pawns became reserve kings, either at once or after further improbable adventures. Timur of Samarkand, alias Tamerlane, the terror of the world in the fourteenth Christian century, was a devotee of this game. It was described in European books of the nineteenth century, which set chess hobbyists to analyzing the elements of the moves of chess-pieces, resulting in an atomic theory of chess and an explosion of new pieces and combinations.

Footnotes & Bibliography:

1. Murray, H.J.R., A History of Chess, Oxford University Press, Oxford, New York, etc., 1913 et seqq., p. 120, note 4. Most of my information on chess not otherwise attributed is from Murray.

2. Professor David Li has lately published a book, The Genealogy of Chess (Premier Publishing Company, Bethesda, Maryland, U.S.A., 1998), in which he states that chess was developed in the winter of 203 B.C. by General Hán Xin to keep his troops alert and in a soldierly frame of mind while in winter camp. Prof. Li describes how Gen. Hán decided on his project, chose the characteristics it should possess, and experimented until he was satisfied.

Unfortunately, the general, who never lost a battle, died in political disfavor, and his papers were destroyed and his game (according to Prof. Li) fell into disuse, from which it was recalled to favor about eight hundred years later, when we first hear of chess in China. This is a fine story, for which Prof. Li produces no evidence. Prof. Li’s further conjectures about the relationships among early forms of chess, while different from mine, deserve attention.

3. Bell, R.C., Board and Table Games From Many Civilizations, Volume Two, Dover Publications, Inc., New York, 1979, page 24.

4. Samsin, Myron J., “Pawns and Pieces: Towards the Prehistory of Chess” ©2002, at http://www.mynetcologne.de/~nc-jostenge/samsin.pdf , accessed August 2006.

5. Josten, Gerhard, “Chess–A Living Fossil” ©2001, at http://www.mynetcologne.de/~nc-jostenge/ , accessed August 2006.

6. Murray, page 159

7. Thomas Hyde in his book Mandragorias (Oxford, England, 1694) derived the name from Persian satrang, the mandrake plant, supposing that its root shaped like a mannikin had suggested the chess-pieces, or vice versa; but linguists now reject this.

8. In fact, according to the Wikipedia at http://en.wikipedia.org/wiki/janggi , accessed June 2006, they are known as “Chu” and “Han” in South Korea, reflecting a period of civil war in China, and as “general” and “minister” in North Korea.

9. Wikipedia at http://en.wikipedia.org/wiki/Shogi_variant , accessed July 2006.

10. Calvo, Ricardo, The Origins of Chess, “Continued Extracts on Gnostic Elements in Chess: 3:7: Linkage of Victory with the ‘Shah-Mat’ Idea” at http://www.goddesschess.com/chessays/calvognosis2.html ,
accessed April 2004. I am grateful to Mrs. Calvo for permission to quote her late husband’s work.

11. Several arrangements of the pieces are known; perhaps Khalil ibh Ahmad invented one of these sub-variations. While I have been forced to the conclusion that this game existed for some time before its first mention, I do not assume any eight hundred years of obscurity, nor do I assume that it was disused during that period.

12. See Murray, pages 211-215 for several examples.

13. Menzies, Gavin, 1421: The Year China Discovered America, Harper Collins Publishers, New York, NY, 2003, page 61. Murray noticed some similarity between Japanese and Burmese chess, but, unaware of the wide-ranging commerce under the Chinese aegis before China and Japan became closed countries, he rejected the idea of a connection as unsupported by the facts.

14. Murray (pages 123-125) documents this move for the chariot in China under the Tang and Sung Dynasties.

15. Wikipedia at http://en.wikipedia.org/wiki/Samurai , accessed November 2006, and numerous other sources.

16. For some reason the diagonal-rider was forbidden to stop on the first square, and the far-ranging member of the knight family was forbidden to stop on either of the first two squares of its orthogonal run. There were yet other complications.