Popular Science Monthly/Volume 14/February 1879/Backgammon Among the Aztecs

by Edward Burnett Tylor

Layout 4

BACKGAMMON AMONG THE AZTECS.

By EDWARD B. TYLOR.

BY backgammon we usually mean one particular game played with dice and thirty draughts, on a board with twelve points on each side. But this is only one of a family of games, whose general definition is that they consist in moving pieces on a diagram, not at the player's free choice, as in draught-playing, but conformably to the throws of lots or dice. It can hardly be doubted that the set of games thus combining chance and skill are all, whether ancient or modern, the descendants of one original game. By a stretch of imagination, it may be possible to fancy draughts or dice to have been fresh invented more than once. But, when it comes to a game which combines the two ideas, it seems to pass the bounds of ordinary probability to suppose, for instance, that a Greek and an Arab and a Birmese were separately seized by the same happy thought, and said, "Go to, let us cast lots, and count them to play at draughts by." If indeed any reader should think such a combination might have happened twice over, he may be asked to look closely into the games presently to be described, so as to satisfy himself that their agreement goes even further, as in the peculiar principle on which the high and low throws are counted, and, so far as one knows, in there generally being in some shape the rule of hitting a blot, that is, taking an enemy's undefended man off the point one's own man moves to. The exact primitive game whence all known games of the class were derived can not now be pointed out, and indeed is perhaps lost in prehistoric antiquity. So we may as well keep to our own word, and call the whole set the backgammon family. It is in this sense that I use the word here, with the purpose of proving that, before Hernando Cortes landed with his invading Spaniards at Vera Cruz, one variety of backgammon had already found its way over from Asia into Mexico, and had become a fashionable amusement at the barbaric court of Montezuma. But, before following the game on its hitherto unnoticed migration into the New "World, let us first glance at its Old World history.

Clearly our English backgammon and the more complicated French trictrac are descended from the Roman game of the "twelve lines" (duodecim scripta), which was played throughout the empire. This is the game which Ovid says has lines as many as the gliding year has months, and he means it where he gives the lover insidious counsel, when his mistress casts the ivory numbers from her hand, let him give himself bad throws and play them ill. Among the Christian antiquities in Home is a marble slab, on which a backgammon-table is cut, with a Greek cross in the middle, and a Greek inscription that Jesus Christ gives victory and help to dicers if they write his name when they throw the dice—Amen. Carelessly scratched as it is, by some stone-cutter whose faith went beyond his trictrac, it shows that the board was like ours even to the division in the middle, which makes the two groups of six points on each side. From ancient Rome, too, we inherit the habit of making the backgammon-board with a draught-board on the reverse side, at any rate the commentators so interpret Martial's epigram on the tabula lusoria:

Hic mihi bis seno numeratur tessera puncto
Calculus hic gemino discolor hoste perit.

Here, twice the die is counted to the point of size,
Here, 'twixt twin foes of other hue, the draughtsman dies.

The very mode of playing the men in classic backgammon may be made out from a fifth-century Greek epigram, commemorating a remarkable hit, in which the Emperor Zeno got his men so blocked that, having the ill-luck to throw 3, 5, 6 (they used three dice, as indeed we continued to do in the middle ages), the only moves open obliged him to leave eight blots. This historic problem, and other matters of Greek and Latin backgammon, are worked out by M. Becq de Fouquières, in his "Jeux des Anciens", with a skill that would have rejoiced the hearts of those eminent amateurs, the old Count de Trictrac and the venerable Abbé du Cornet, to whose teaching history records that Miss Becky Sharp ascribed the proficiency at backgammon which made her society so agreeable to Sir Pitt at Queen's Crawley.

It is not known so exactly what manner of backgammon the Greeks played in earlier ages; but there are various passages to prove that, when they talk of dice-playing, they often mean not mere hazard, but some game of the backgammon sort, where the throws of the dice are turned to account by skillful moving of pieces. Thus Plato says that, as in casting dice, we ought to arrange our affairs according to the throws we get, as reason shall declare best; and Plutarch, further moralizing, remarks that Plato compares life to dicing (κνβείά), where one must not only get good throws, but know how to use them skillfully when one has got them. So with Plutarch's story of Parysatis, mother of Artaxerxes. She was "awful at dice" (δεινὴ κνβενειν), and, "playing her game carefully," won from the king the eunuch Mesabates, who had cut off the head and hand of Cyrus; having got him, she had him flayed alive and his skin stretched. This episode of old Persian history is noteworthy in the history of the game, because Persian backgammon, which they call nard, is much like the European form of the game, which, it has not been unreasonably guessed, may itself have come from Persia. This nard is popular in the East, and orthodox Moslems have seen in the fateful throws of the dice a recognition of the decrees of Allah, that fall sometimes for a man and sometimes against him. It is, said one, a nobler game than chess, for the backgammon player acknowledges predestination and the divine will, but the chess-player denies them like a dissenter. Not to lose ourselves in speculations on the Oriental origin of backgammon, at any rate it was from Rome that it spread over Europe, carrying its Latin name of tabulæ with it in French and English tables. This word has dropped out of our use since the Elizabethan period, but an instance of it may be cited in a couple of lines, conveying another little sermon on backgammon, which the English author no doubt borrowed from the Latin of Terence, even as he had copied it from the Greek of Menander:

Man's life's a game of tables, and he may
Mend his bad fortune by his wiser play.

There is an idea which readily presents itself as to how backgammon came to be invented, namely, that the draughts were originally mere counters, such as little stones, shifted on a calculating board to reckon up the successive throws, and that it was an afterthought to allow skill in the choice of moves. This guess fits well enough with the classic draught being described as a stone, νῆφος, calx or calculus, while in Germany, though now made of wood, it still keeps its old name of stein. Also the playing board on which the stones were moved shares the name of the calculating board, ἇβαξ, abacus. But if the classical varieties of backgammon in this way show traces of the game near its original state, they seem in another respect to have passed out of their early simplicity. They are all played with dice, and indeed the French author lately mentioned seems right in guessing that the division of our board into groups of six points each was made on purpose to suit the throws of cubical dice like ours, numbered on all the sides, from 1 to 6. As to the early history of dice, I have elsewhere endeavored to show ("Primitive Culture," chapter iii.) that the origin of games of chance may be fairly looked for in instruments of the nature of lots, at first cast seriously by diviners for omens, and afterward brought down from serious magic into mere sport. Now, the simplest of such instruments is the lot which only falls two ways, like the shell, white on one side and blackened on the other, which Greek children spun up into the air to fall, "night or day," as they said; or, like our half-pence, tossed for "head or tail." Both in divination and in gambling, such two-faced lots probably came earlier than the highly artificial numbered dice. The kinds of backgammon now to be described seem in general to belong to the earlier stage of development, for it is with lots, not dice, that they are played.

The traveler in Egypt or Palestine now and then comes on a lively group sitting round a game, and in their eager shouts, if he knows some Arabic, he may distinguish not only such words as "two" or "four," but also "child," "dog," "Christian," "Moslem. On closer examination he finds that the game is called táb, and that it is a sort of backgammon played on an oblong checker-board, or four rows of little holes in the ground, where bits of stone on one side and bits of red brick on the other do duty as draughts, being shifted from place to place in the rows of squares or holes. Not dice, but lots, are cast to regulate the moves; these lots are generally four slips of palm-stick, with a green outer side and a white cut side (called black and white), and when they are thrown against a stick set up in the ground, the throw counts according to how many white sides come up, thus:

Whites up:	None	One	Two	Three	Four.
Count:	6	1	2	3	4
	(go on)	(táb)	(stop)	(stop)	(go on)

Notice particularly this way of counting throws, for its principles will be found again in lot backgammon elsewhere. There is evidently a crude attempt to reckon probabilities, giving a higher value to the less frequent throws of all four white and all four black, than to two or three white, which come up oftener. Besides the high count, they have the privilege of a second throw. This, if lot backgammon came first, and was succeeded by dice backgammon, would naturally pass into our rule of giving doubles another throw. The throw of one white, which is called "child," or táb, i. e., "game," has a special power, for only by it may a "dog," that is, a stone or draught, be moved out of its original place in the outer row, and set at liberty to circulate along the lines of squares or "houses," taking an enemy's dog if found alone in its house. While a draught is still in its first inactive, useless condition, they call it a "Nazarene," or Christian; but, when the throw of táb gives it the right to go forth conquering and to conquer, it becomes a "Moslem." It is not needful to go further into the rather complicated rules of moving and taking. Those who are curious may find much about it in Lane's "Modern Egyptians," and in the quaintly learned little book "De Ludis Orientalibus," by Thomas Hyde, who was Bodleian librarian in the reign of William and Mary. But one question suggests itself. Seeing how the modern fellahs delight in táb, one naturally asks, Did they inherit it from the ancient Egyptians? From remote antiquity the Egyptians played draughts on earth, and after death their righteous souls still had the oblong checker-board, and the men like chess-pawns, to amuse their glorified but perhaps rather tiresome life in the world below. But, as Dr. Birch points out, no Egyptian dice have been found earlier than Roman times, nor any plain mention of backgammon. Even if they played like their descendants in the Nile Valley with such things as slips of palm, something about it should be found in the hieroglyphic texts. But at present nothing appears, and there is no reason to add backgammon to the long list of inventions whose earliest traces are found in Egypt. Perhaps the nearest relative of táb is Chinese backgammon, but this is played with dice.

Next, as to India. Here, since ancient times, cowry-shells have been thrown as lots, their "head" and "tail" being according as the shell falls with mouth or back upward. In Sanskrit literature there is an old mention of a game called panchiká, which was played with five cowries, and where it seems that the winning throws were when all the mouths came up or down, as against the commoner throws when some fell each way. That a game of the nature of backgammon was known in India from high antiquity has been plainly made out by Professor Weber. It was called ayánaya, or "luck and unluck"; or at any rate that was a term used as to the moving of the pieces, which traveled right and left through the squares, and took an undefended man from his place to begin his course anew. So, as a Sanskrit riddle has it: "In a house where there were many, there is left but one, and where there was none and many come, at last there is none. Thus Kála and Kâlî, casting day and night for their pair of dice, play with human beings for pieces on the board of the world." Putting these particulars together, it is clearly possible to trace from ancient times the game of pachisi, played in modern India, into which game it will now be necessary for our argument to go more exactly; in fact, to qualify ourselves to sit down and play a game. English backgammon-players will hardly take five minutes to learn it.

Suppose four players to be seated, each at the end of one arm of the diagram or board, of which a figure is here given. Each player will have four little wooden cones as his pieces or draughts, all of one color, to distinguish them. If only two play, each will manœuvre two sets of men. Each player's men start one by one down the middle row of his own rectangle, beginning with the square next the central space, and thence they proceed all round the outside rows of the board, traveling from right to left (contrary to the sun) till they get back to their own central row, and up it home to where they started from, he who first gets all his men home winning the game. A solitary man is taken up and sent back to begin again, by one of his adversary's men lighting upon his square, except in the case of the twelve privileged squares, which are marked with a cross, in which case the overtaking piece can not move. The moving is determined by throwing a number of cowries, which count according to how many fall mouth up; thus, if six cowries are used:

Mouths up:	None	One	Two	Three	Four	Five	Six.
Count:	6	10	2	3	4	25	12
	(go on)	(das)	(stop)	(stop)	(stop)	(pachisi)	(go on)

According to the rules kindly sent me from Dr. Rajendralala Mitra, of Calcutta, the throws of one-up or five-up (das or pachisi) alone can start a man on his first square, or get him off if stuck on his last square. These throws, as well as none-up and six-up, give a new throw. Thus the best beginnings are one-up followed by two-up, or five-up followed by four-up, either of which enters a man and carries him on in safety into a "fort." Seven cowries can also be used, but the primitive game was probably more like the ancient game with five cowries just mentioned, for the name pachisi means "five-and-twenty," and was no doubt taken from the throw when five shells come up. The principles of counting the throws and entering the men are plainly like those in the Arab game of táb, and there are Indian forms with only four cowries which come still closer.

Pachisi is a favorite game in India, and an eager player will carry rolled round in his turban the cloth which serves as a board, so as to be ready for a game at any moment. These cloths, when embroidered with the diagram in colored silk, are quite artistic objects, and one does not often see prettier toys than a set of men in Mr. Franks's collection, little cones (or rather sugar-loaves) of rock crystal, with the colors they are to bear in the game shown by mounting in the top a ruby for red, an emerald for green, etc. There are even stories of yet more sumptuous games, where the board was a courtyard laid out in marble pavement, on which living draught-men clothed in green, red, yellow, and black, walked the circuit and hustled one another off the squares. Our Anglo-Indians sometimes catch the enthusiasm; and there is an often-told tale of that official personage who, when he paid his native servants their wages, would sit down with them to a match at pachisi, and sometimes win his money back. In London toy-shops they sell board and pieces for what they profess to be the game, but these really belong to the modified form of it known in India as chûpur, in which, instead of cowries, stick-dice numbered on the four long sides are thrown, these Indian dice being in England replaced by our common cubical ones. This shows the change from lots to dice in games of the backgammon sort, and it is curious to notice how clearly the new rules for counting by the dice are modeled on the old rules for throws of cowries. Having now sufficiently mastered the peculiarities of pachisi, let us pass from Asia to America, and compare them with the details of the Mexican game of patolli.

When the Spanish invaders of Mexico gazed half in admiration and half in contempt on the barbaric arts and fashions of Aztec life, they particularly noticed a game, at which the natives played so eagerly that, when they lost all they had, they would even stake their own bodies, and gamble themselves into slavery, just as Tacitus says the old Germans used to do. The earliest particulars of the Mexican game come from Lopez de Gomara, whose "Istoria de las Indias" was printed in 1552, so that it must have been written while the memory of the conquest in 1521 was still fresh. He says: "Sometimes Montezuma looked on as they played at patoliztli, which is much like the game of tables, and is played with beans marked with lines like one faced dice, which they call patolli. These they take between both hands, and throw them on a mat or on the ground, where there are certain lines like a checker-board, on which they mark with stones the point which came up, taking off" or putting on a little stone." This may be supplemented from three other old Spanish writers—Torquemada, Sahagun, and Duran. The figure on the mat is spoken of as "a painted cross full of squares like checkers," or as an "aspa," which word means a +, a Greek cross, the sails of a windmill, etc., descriptions which come as close as may be to the pachisi-board. Also, it appears that the stones moved on the board to mark the numbers thrown by the beans were of different colors, one account mentioning twelve stones, six red and six blue, between the two players.

According as the game was played, three to five beans were thrown as lots or dice, and sometimes these beans were marked on one side with a hole, and left plain on the other, while sometimes they seem to have had dots or lines indicating various numbers. If both ways were really used, then the game was known in both its stages, that of two-faced lots and that of numbered dice, just as in India it is played as pachisi with cowries, and as chûpur with stick-dice. As to the way of scoring the throws, only one of the old writers says anything. This is Diego Duran, an extract from whose MS. history I have obtained by the courtesy of Mr. Oak, of the Bancroft Library at San Francisco. He says, as to the holes in the beans which showed how many squares were to be gained, that they were "if one, one, and if two, two, and if three, three, but marking five they were ten, and if ten, twenty." Thus in Mexico we just catch sight of the peculiar trick of scoring, everywhere so characteristic of the game, namely, the advantage given to the extreme throws, which in our own backgammon takes the form of allowing doubles to count twice over. Unluckily, the thought had never crossed the minds of these early Spanish historians of the New World that their descriptions of the Aztec game would ever become evidence of use in tracing the lines along which civilization spread over the earth. Had they seen this they would have left us a perfect set of rules, not such careless mentions of a game which plainly they "did not understand." Still they saw enough of Montezuma's patolli to observe that it was in principle like their own game of tables, while clearly they had never heard of the Indian pachisi, or they would have seen how much closer its resemblance came to that. This touches a point in the history of the game. How did the Mexicans get it? The idea may have already occurred to some readers of this essay. Could not perhaps some stray Portuguese or Spaniard, having lately picked up the game of pachisi in some seaport of the East Indies, have taken his next voyage to the West Indies, and naturalized his newly-learned game on the mainland of America? But there is no room for a suggestion of this sort when it is remembered that patolli was an established diversion in Mexico at the time of the Spanish entry, which followed within three years of the first landing of Grijalva in the gulf of Mexico, and indeed within five-and-twenty years of Colon's first sight of Hispaniola. What seems most likely is, that the game came direct from Asia to America, reaching Mexico from the Pacific coast.

That the remarkable civilization of Mexico as the Spaniards found it was not entirely of native American growth, but had taken up ideas from Asia, is no new opinion. Alexander von Humboldt argued years ago that the Mexicans did and believed things which were at once so fanciful and so like the fancies of Asiatics that there must have been communication. Would two nations, he asks in effect, have taken independently to forming calendars of days and years by repeating and combining cycles of animals such as tiger, dog, ape, hare; would they have developed independently similar astrological fancies about these signs governing the periods they began, and being influential each over a particular limb or organ of men's bodies, would they, again, have evolved separately out of their consciousness the myth of the world and its inhabitants having at the end of several successive periods been destroyed by elemental catastrophes? In spite of Humboldt we often hear Mexican culture talked of as self-produced, with its bronze and gold work, its elaborate architecture and sculpture, its monastic and priestly institutions, its complicated religious rites and formulas. It was my fortune years ago to travel in Mexico and explore its wonderful ruins, and ever since I have held to the view that the higher art and life of the whole Central American district is most rationally accounted for by a carrying across of culture from Asia. Thus it is now a peculiar pleasure to me to supplement Humboldt's group of arguments with a new one which goes on all-fours with them. It may very well have been the same agency which transported to Mexico the art of bronze-making, the computation of time by periods of dogs and apes, the casting of nativities, and the playing of backgammon. What that agency was one can as yet do no more than guess, but too much stress must not be laid on it in speculating on the mass migrations of the American races. Such matters as arts or games are easily carried from country to country; nor can we treat as inaccessible to Asiatic influences the Pacific coast of North America, where disabled junks brought across by the ocean current are from time to time drifted ashore, now and then with their crews alive. The Asiatic communication to be traced in the culture of the Aztec nation may not have been very ancient or extensive; all we can argue is, that communication of some sort there was.

Now one thing leads to another, especially in ethnology. Curiously enough, by following up the traces of this trivial little game, we get an unexpected glimpse into the history of the ruder North American tribes. Having learned about patolli as played in old Mexico, let us take up the account of a Jesuit missionary. Father Joseph Ochs, who was in Spanish America in 1754-'68, and who is here writing about the tribes of Sonora and Chihuahua: "Instead of our cards they have slips of reed or bits of wood a thumb wide and near a span long, on which, as on a tally, different strokes are cut and stained black. These they hold fast in the hand, lift them up as high as they can, and let them drop on the ground. Whichever then has most strokes or eyes for him wins the stake. This game is as bad as the notorious hazard. They call it patole. As it is forbidden on pain of blows, they choose for it a place in the bush; but the clatter of these bits of wood has discovered me many a hidden gamester. To play more safely, they would spread a cloak or carpet so as not to be betrayed by the noise." Here, then, is found toward a thousand miles northwest of the city of Mexico, a game which may be described as patolli without the counters, and which still bears the Aztec name, in a district whose language is not Aztec, so that the proof of its having traveled from Mexico seems complete. The people, being less intellectual than the old Mexicans, have dropped the skillful part of the game and are content with the mere dicing. Nor, by the way, is this the only place where backgammon has so come down, for in Egypt they will lay aside the board and throw the táb-sticks for fun, those who throw four and six being proclaimed Sultan and Vizier, while the luckless thrower of two gets for his reward two cuts with the palm-stick on the soles of his feet.

Yet another fifteen hundred miles or more up into the continent the game is still to be traced. Among the hunting tribes known under the common name of the North American Indians, there is a favorite sport described by a score of writers under the name of "game of the bowl," or "game of plum-stones." The lots used are a number of plum-stones burned on one side to blacken them, or any similar double-convex pieces of wood, horn, etc. They are either thrown by hand or shaken in a bowl or dish, whence they can be neatly jerked up and let fall on the blanket spread to play on. The counting depends upon how many come up of either color, white or black, as is seen in the precise rules given by Mr. Morgan in his "League of the Iroquois." Where six "peach-stones" were thrown, if all six came up, white or black, they counted five, and five up, white or black, counted one, these high throws also giving the player a new turn, but all lower throws counted nothing and passed the lead. It is so curious to find the principle of lot-scoring, which we have tracked all across from Egypt, cropping up so perfectly among the Iroquois, that at the risk of being tedious it is worth while to give in full the mode of counting in the game as played with eight "deer-buttons." The following top line shows how many black or white sides up, with their count below:

Eight Seven Six				Five Four Three				Two One None
20	4	2		0	0	0		2	4	20
$\scriptstyle {\underbrace {\quad \quad \quad \quad \quad } }$				$\scriptstyle {\underbrace {\quad \quad \quad \quad \quad } }$				$\scriptstyle {\underbrace {\quad \quad \quad \quad \quad } }$
go on				stop				go on

In these games there is no board to play on. The Iroquois use beans as counters, the game being won by one player getting all the beans, but perhaps the white men taught them how to do this. So with the game which will occur to English readers who remember it in "Hiawatha," where it is described at full length in prose-poetry as "the game of bowl and counters, pugasaing with thirteen pieces." This game is real enough; indeed, the description of it is taken from Schoolcraft's "Indian Tribes." But there seem to be no early mentions of this Algonquin game with its ducks and war-clubs and elaborate counting, nor of the Dakota game with tortoises and war-eagle son the plum-stones. Thus both may have been lately devised by Indians under European teaching, as improvements on the original pugasaing or "play," which was the simple game with black-and white-sided plum-stones, or the like. This, no doubt, is old, for it is described by the Jesuit missionaries in 1636 under the name of jeu de plat, as a regular sport among the Hurons; and as they clearly did not learn the game from Europe, we are left to argue that it reached them from Asia, very likely through Mexico.

It remains to glance at what may be learned as to the history of the North American Indians from the fact of their gambling with the bowl and plum-stones. It is an interesting question whether "the poor Indian, whose untutored mind" has now and then been too easily credited with the invention of all the arts and beliefs he did not get from the white men, may not really before this have largely taken up in his culture ideas of Old-World growth. It has long been noticed that, looking at the native tribes of what is now the United States and the Dominion of Canada, the tribes on the east side had taken to making pottery and cultivating maize, while the tribes on the west had not, which seems as though there had been a flow or drift of civilization from the Central American district up the eastern half of the continent, which of itself ought to be enough to prevent any ethnologists from looking at the so called red-man of New England or the Lakes as the creator of his whole industrial and social life. Nor is it an unknown thing that the myth and religion of the North American tribes contain many fancies well known to Asia, which the men of the prairies were hardly likely to have hit upon independently, but which they certainly did not learn from the white men, who did not even know them. If we are bound, as I think we are, to open a theoretical road for even a well-marked game to migrate by from Asia into America, then there are plenty of other matters waiting for passage along the route. By such conveyance of ideas it may be easiest to explain why the so-called Indians of North America shared with the real Indians of India the quaint belief that the world is a monstrous tortoise floating on the waters, or why the Sioux Indians share with the Tartars the idea that it is sinful to chop or poke with a sharp instrument the burning logs on the fire. But these considerations lead too far into the deepest-lying problems of the connection and intercourse of nations to be here pursued further. It is remarkable, too, how vast a geographical range the argument on the migrations of a game may cover. The American farmer now whiles away the winter evening in his farmhouse parlor with a hit at backgammon, on the spot where, not long since, the Iroquois played peach-stones in his bark hut. Neither would have recognized the other's sport as akin to his own, though when we trace them through the intermediate stages they are seen to be both birds of one nest. It is by strangely different routes that they have at last come together from their Asiatic home—one perhaps eastward through Asia, across the Pacific, into Mexico, and northward to the St. Lawrence; the other, no doubt, westward down to the Mediterranean, up northward to England, over the Atlantic, and so out into the American prairie.^[1]—Macmillan's Magazine.

1. For special details, copies of original documents, etc., see a paper by the author "On the Game of Patolli in Ancient Mexico, and its probably Asiatic Origin," read before the Anthropological Institute on April 9, 1878.