Chinese Origin of Playing Cards

The American Anthropologist, Volume VIII, January 1895, Pages 61-78

by W.H. Wilkinson

Note: For the most part, this is a direct transcription of the paper by Wilkinson. The paper was scanned from a photocopy of the printed text, edited in a word processor, then created as a Web Page. His spelling and sentence structure have been maintained, but some punctuation has been altered. Words in Chinese and Japanese with diacritical marks are presented within the limitations imposed by HTML. The graphics are slightly edited copies of the sketches accompanying the original text. The size of each drawing is as it appeared in the original paper. Notes about the size of each drawing are taken from the original document, and this information may be viewed as a “tool tip” by placing the mouse cursor on each drawing.

(Page 61) The current Chinese term for both dominoes and cards is p’ai (pronounced “pie” as in Tipperary). There is no essential difference between cards and dominoes in China 1; what are known among foreigners, owing to a superficial resemblance, as “dominoes” are used for the most part precisely as we use cards, and the same game will appear either in the form of a pack of, it may be, coarsely printed slips of pasteboard or as highly finished tablets – “dominoes,” in short – of ivory and ebony neatly fitted (Page 62) into a sandalwood box. Where a distinction in language has to be made, cards of millboard are styled chih p’ai “paper p’ai”, and dominoes ya p’ai, or ku p’ai, “p’ai of ivory” or “bone.”

Literary Chinamen of today will use in writing but not in speech two other terms, yü-p’u, “slips”, and yeh-tzâ, “leaves”. These are the names of two old Chinese gambling games, the first of which was in vogue as early as the third century of our era, while the second was at the height of its popularity about the tenth. Some foreign writers have maintained that both these games were played with cards. Williams, for instance, in his Syllabic Dictionary, s. v. p’u, translates the former term by “an old name for playing cards,” while Schlegel, as quoted by van der Linde2, plainly identifies yeh-tzâ with the cards of today. If either of these authorities is correct, playing cards of some kind can be proved to have been in use in China several centuries before their appearance in Europe. The records of the Tsin dynasty, for example, state that a well-known worthy, T’ao K’an (259-334 AD), “flung into the river the wine cups and yü-p’u of his subordinates, remarking, ‘Yü-p’u is a game for drovers, and swineherds.'” Yang Kuo-chung, brother of the notorious Yang Kuei-fei, mistress of the Emperor Ming Huang, played yü-p’u, with the imperial gambler in the palace AD 750. In 951 T’ai-tsu, the “High Ancestor” of the Later Chou, assembled his nobles to play together at this game for “embroidered rugs and damask and gauze of sorts.” Forty years later yü-p’u was put under a ban and all players of it threatened with the headsman, after which the game appears very naturally to have fallen into desuetude.

Yü-p’u, however, pace Dr. Williams, was not a card game. Originally, so far as we can judge, it was nothing more or less than the modern poker dice, or something closely resembling it. Five dice were used, colored black above and white below (or perhaps having three of the faces black and three white), and one or more of the black faces was marked with a 2-spot, and similarly one or more of the white. The highest throw was five blacks, “the hound,” counting 16; the next was “the cock”, two (Page 63) white 2’s and three blacks, counting 14. Later on yü-p’u would seem to have taken a form (while preserving the poker dice) more akin to backgammon, but in all the notices of it as yet examined it would be an abuse of language to call it a game of cards. Before dismissing it however, it may be worth noting that it is credited, or rather debited, with a foreign origin. Said the old puritan, T’ao K’an, already cited: “Yü-p’u is a foreign game, yet nowadays scholars and officials play it. Can it be that the whole empire is turning foreign?”

There was probably more affinity with cards in the case of yeh-tzâ than in that of yü-p’u. Indeed, a native work of which the date is yet to seek, the Tan-yen-tsa-lu, declares explicitly that the these “leaves” “were like the modern Pasteboard cards.” Chinese authorities differ as to the derivation of the term. The most reasonable of them deride the convenient theory that they (Page 64) were called after the name of their inventor (given by others as Yeh Tzâ-ching). One writer, who dates their use from the middle of the eighth century, explains their origin thus: In the earlier days of T’ang dynasty (say the seventh century after Christ), books were written in the form of scrolls. This was found inconvenient for purposes of reference, and books with leaves were substituted for them (the leaves, however, were more like the tablets in common use in England some few years ago for memoranda – that is to say, they were detached or detachable). Among the books thus made up into tablets were works on dice games. As these were in constant use for reference, “tablets” or “leaves” in this way became synonyms for dice, and finally were used in the place of dice – and thus yeh-tzâ grew into cards. The theory is ingenious and derives considerable support from two circumstances: one, the employment to this day in Japan (which obtained all or nearly all its amusements from China) of card games in which the cards form practically the leaves of a book of poetry; the other, the use throughout western China of packs in which the cards represent, much as in our proverbs, the different words of some well-known sentiment or sentence.

Whatever these “leaves” may have been, the game into which they entered was exceedingly fashionable during the last years of the T’ang dynasty (618-905 AD) and the century following its extinction. China, on the fall, of the Tang, was divided between native princelings on the south and the K’i-tan Tartars on the north – the Tartars from whom China got in Europe the name of Kitai or Cathay. These Tartars took so kindly to the game of yeh-tzâ that one authority declares that “the Kitan were the first to use the game.” He adds, however, that “its origin is not known.” In February, 969, the Tartar (Page 65) prince assembled his lords to a “tournament of leaves,” and in the following month, says the chronicler, he was murdered. Of course the moral is pointed: “Did not evil follow speedily on the use of these unlucky things? Will not the scholars and high officers who now spend their days over them take warning thereby?” What is of far more interest to the western reader, who does not always share the Chinese fondness for a moral, is a scholiast which explains yeh-tzâ succinctly as “Sung money” – that is of the Sung dynasty, of which more presently.

References to cards under their present name of p’ai are rare in the encyclopedias. Besides the sentence from the tan-yen-tsa-lu, already quoted, to the effect that yeh-tzâ were similar to the chih-p’ai of today, the only passage so far found is that unearthed several years ago from the Cheng-tzâ-t’ung by the late Abel Rémusat, an armchair sinologue of some repute in his time. This paragraph is alluded to by, among others, Chatto, Merlin, and “Cavendish”. The latter (quoting, though without acknowledgment from Chatto) wrote in his Card Essays, “It is stated in the Chinese Dictionary, Ching tsze tung, compiled by Eul Koung, and first published AD 1678, that the cards now known as Teen tsze pae, or dotted cards, were invented in the reign of S’eun-ho, 1120. According to tradition they were devised for the amusement of S’eun-ho’s numerous concubines.” In his recent article on cards in the Encyclopedia Britannica “Cavendish” throws tradition to the winds, and boldly fathers the amusement theory on the dictionary itself. The original, however, though quaint enough in its way, is provokingly silent about the concubines; indeed it is strictly seemly, not to say sanctimonious, in its tone. A garbled paraphrase is given by (Page 66) Chatto, but he omits altogether the concluding and most important sentence: “It does not follow that this class of games originated in the period Hsüan-ho” [1119-1126]. A more curious vindication of the tedious maxim about verifying your references could hardly be imagined. Here we have a passage adduced again and again by European writers, from Chatto to “Cavendish,” to prove that Chinese cards “were first invented in the reign of S’eun-ho,” which passage, when carefully examined, distinctly declares that such a conclusion would be unsound. It is perfectly clear, indeed, that all that was done or asked for in 1120 was an imperial decision as to which of several forms or interpretations of the game now known as T’ien-kiu (“Heavens and Nines”) was to be considered orthodox. The game and the cards must have been in existence long before. The passage from the Cheng-tzâ-t’ung runs thus (s.v. pa’i):

Also ya p’ai now the instruments of the game. A common legend states that in the second year of the Hsüan-ho, in the Sung dynasty [i.q. 1120 AD], a certain official memorialized the throne, praying that the ya p’ai (ivory cards) be fixed as a pack of 32, comprising 127 pips [sic, it should be 227, but Chinese printers are careless], in order to accord with the expanse of the stars and constellations. The combination ‘heaven’ [6/6, 6/6] consisted of two pieces, containing 24 pips, figures of the 24 solar periods; ‘earth’ [1/1, 1/1] also composed two pieces, but contained 4 pips, the 4 points of the compass – east, west, south, and north; ‘man’ [4/4, 4/4] two pieces, containing 16 pips, the virtues of humanity, benevolence, propriety, and wisdom, four-fold; ‘harmony’ [2/3, 1/3] two pieces of 8 pips, figuring the breath of harmony, which pervades the eight divisions of the year. The other combinations had each their names. There were four players having eight cards apiece for their hand, and the cards won or lost according as the number of the pips was less or in more the winner being rewarded with counters. In-the time of Kao-tsung [1127-1163] pattern packs were issued by imperial edict. They were known throughout the empire as Ku p’ai, ‘bone p’ai;’ but it does not follow that this class of games, po-sai, Ko-wu, and the rest originated in the reign of Hsüan-ho.

Ku-wu bore some analogy to gobang, po-sai was a form of backgammon. The classing together of these with Ku p’ai (Page 67) curiously resembles the method of Professor Hoffmann in his Encyclopaedia of Cards and Table Games, where dominoes figure among the latter. Whatever may be said of English dominoes. Chinese “dominoes” are, however, most certainly cards, more particularly in the game of T’ienkin, here described.

Cards basing their symbols on dice exist in many different forms throughout China, and it is possible with their aid to trace the gradual evolution of these playthings from knucklebones through dice and dominoes. As, however, these cannot be shown to have directly influenced the form of European cards, nothing further need be said about them here.

It is indeed to yeh-tzâ rather than to the ya pa’i that we must look not only for the origin of most Chinese paper cards, but for that of European playing cards as well; and the modest scholiast already quoted gives the clue when he explains yeh-tzâ by “Sung money”. Of all the Chinese varieties, and they are many, no one kind is so uniformly distributed or so universally popular as that described by Chatto under the style of Tseen-wan che-pae3. Illustrations of these are given by Breitkopf (whom Singer copies), Chatto, Taylor (and his original, Boiteau), and Willshire; and portions of a pack are contained in the British Museum in a box numbered by Willshire O.C. 293. Perfect specimens, obtained from Peking, Tientsin, Chunking, Kiükiàng, Shansi, Honan, Wenchow, Canton, and other parts of China, appeared as Nos. 1 to 17 in a collection of Chinese cards which Mr. Stewart Culin did me the honor to exhibit, at the World’s Fair last year4. Of the descriptions heretofore given of this class of cards, that by Chatto is the most accurate; that by Willshire the most absurd. They are known in central China by the name of kun p’ai, staff or baton cards, or ma chioh, “hempen birds.”‘ A set contains thirty pieces, namely, the ace; 2, 3, 4, 5, 6, 7,8, 9 of three suits, now commonly called ping, tiao, and wan (Page 68) (“cakes,” “strings,” “myriads”), and three separate cards, “whiteflower,” “redflower,” and ch’ien wan, “a thousand myriads” usually styled “old thousand.” The game as sold consists, as a rule, of four of these sets or packets with two, five, or six loose cards, known as “golds,” “flyers,” or “butterflies.” The use of these “golds” is precisely similar to that of mistigris in poker or the braggers in the parent game, Hoyle’s brag – that is to say, they can stand for and take the place of any desired card. They have, however, no vital connection with the game, which may be, and often is, played without them. No specimens other than those at the World’s Fair have so far found their way into western cabinets; none certainly have been described by European writers. As a rule, they are five in number and carry portraits and emblems representing the “Five Blessings,” Luck, Promotion, Longevity, Prosperity, Wealth; but they frequently take other forms. Sometimes they are ordinary cards, just as quinola, the knave of hearts, might, as Cavendish puts it, “be made any card or suit” at one of the oldest of European games, primero. In this latter case, however, they are distinguished by a blot of gilt or other token.

The suit of ping, “cakes,” was originally and properly that of ch’ien “specks” or “cash;” the “strings” are strips of one thousand cash, the “myriads” are myriads of cash. Everyone knows the sapeck of China, a round coin of brass and copper, with a square hole in its center. It is as a rule very badly cast, and it is worth just now about 1/15th of a cent. This coin, then is the unit of the kun p’ai. In fact, the cards of a kun p’ai pack are or originally were bank notes, for which and with which the gamblers played. It may be objected that according to Mayers, bank notes “were invented in China in 1265;” but Vissering 5 has pointed (Page 69) out from Ma Tuan-lin the antiquary (ob. 1325) that “T’ai-tsu of the Sung (AD 960-976) revived the old system of flying cash used under the Tang dynasty” (618-905), the “flying cash” being paper money or bank notes. In Vissering’s monograph is given an illustration of a note of the Ming dynasty (1368-1644) and a still earlier one of the Yüan (1280-1367) has been recently purchased for the British Museum. If any one will examine either of these he will find upon them illustrations of strings of cash the exact counterpart of which appears in the suit of tiao of the kun p’ai packs. Assuming, then, and the assumption seems sound, that the kun p’ai cards were originally banknotes, the question remains: How did these bank-note cards come to affect, if they did affect, the cards of Europe? The question was anticipated by Singer (” Researches,” 1816):

The grotesque appearance of the figures on modern European court cards bears no small degree of resemblance to some representations of the human form in the more rude and early attempts of the Chinese at depicting it. This resemblance has been frequently remarked, but has never, we believe, led to the enquiry whether it was probable that the Europeans obtained the knowledge of cards from thence. As it is certain that they practiced the art of engraving on wood many centuries before it was known in Europe, and as the European card-makers are considered by some to have first introduced that art, a conjecture be hazarded that they obtained both card’s and the means of multiplying them from China by means of some of the early adventurers who, for purposes of commerce, are known to have reached that country as early as the l2th century. Zani says, he adds in a foot-note, “the Abbé Tressan showed him when he was at Paris a pack of Chinese cards and told him that a Venetian was the first who brought cards from China to Venice, and that city was the first place in Europe where they were known.” … This traveller could have been no other than Niccolo Polo, who, with his brother Matteo, returned from China about 1269, or else the celebrated Marco Polo… Still both the printed text and the various MSS. make no mention of such an occurrence.
That Marco Polo, while speaking, as he does, of the paper money of the Great Khan, should have been silent about his playing cards proves little, for to Polo they would not necessarily be as interesting as to Singer or to Chatto. Nevertheless he may well have brought a pack from China with a hundred other trifles of too little value to catalogue in his book. Nor does Merlin’s (Page 70) objection to the small size of the Chinese cards as compared with the earliest specimens in Europe affect the argument. Even at the present day the cards of western China are far larger than those of Canton (the petits griories of Merlin) well in the Polos’ time have been broader still. The narrowness of the modern Cantonese cards is due to the prevailing fashion of holding them in the clenched fist and not spread out fanwise.

The coincidences, indeed, between the first European packs and the Chinese kun p’ai are too numerous to be accidental. The earliest suit-marks in Europe, those of Italy and Spain, were spade (espadas), “swords”; coppe (copas),”cups”; denari (oros), “money”; and bastone (bastos), “clubs.” Spanish packs, remarked Vives (ob. 1541), had no 10-spot (Singer, p.37). The first coat cards were the King, Cavalier, and Servant – Rey, Caballo, Soto. Early European packs contained emblematic cards, the naibis, independent of the rest. In the first games known the cards were used much as in commerce or poker now – to form flushes, sequences, or triplets. Now, if we turn to the kun p’ai we find that the leading principle of the games played with them is the same as that in the old Italian Frusso or Primero, while at the same time, like the Tarocchi, the kun p’ai packs usually include a number of emblematic cards curiously suggestive of the naitis. (Of this more presently.) No Chinese pack contains a ten, the reason for which is clear if we take the cards to have been banknotes, but altogether absent if we call them “cups” and “clubs;” – for let us suppose that the suit of cash represented notes ranging (Page 71) in value from one hundred kopecks to nine hundred (as was probably the cause, since notes for a single speck or for nine specks would be ludicrously small); the suits of strings and myriads on the face of them represent notes for one thousand to nine thousand and for ten thousand to ninety thousand cash respectively. We then have a decimal series of hundreds, thousands, myriads, and a ten of strings would be clearly superfluous, since its place is already taken by an ace of myriads, ten thousands being equal to one ten-thousand.

There is reason, then, in the Chinese suit names, but absolutely none in the European. An attempt has been made to explain the Spanish and Italian suits as emblematic of the four classes of society, clubs standing for peasants; money for tradesmen; swords for the nobility, and cups for the clergy; or, again, as symbols, respectively, of fortitude, charity, justice, and faith. All this is obviously fantastic, whereas if we admit that the first European cards were from a Chinese kun p’ai pack (or from a still more remarkable one, to be presently described, the lieh chih), then everything becomes clear. Rey and Soto find their prototypes in “Old Thousand” and “Redflower” while Caballo represents “Whiteflower” which, Breitkopf observes, “often bears the figure of a horse.” (It is, in fact, a stag, but that Breitkopf took it for a horse makes it easy to understand that his predecessors, Marco Polo’s Venetian friends, should have done the same.) There are three non-numeral cards in the kun p’ai pack; there were three non-numeral cards in those of medieval Europe. Here, again, there is reason in the Chinese game; none in the European. In the Chinese game, which, it should be observed, Messrs. Goodall have lately brought out in English dress under the name of “Khanhoo,” the cards are brought together in combinations of three (suggested probably by the number of the numeral cards, 3X3X3), and hence, whatever the Red Flower, White Flower, and Old Thousand may originally have been, their number was necessarily fixed at three. (In some Kun p’ai packs these coat cards as we should regard them, bear other names, Wang, Ying, Lin Chang, Wu Sung, Prince of Mao, taken from an old romance of the Robin Hood type, the Shui-hu chuan. White Flower, again is sometimes “Sprayflower”, or “White Ash.”

Page 72) But no such reasonable explanation can be given of the restriction of European coat cards to three, whether we call them King, Queen, and Knave, or Rey, Caballo, and Soto; nor is it obvious, however natural it may appear to us now, why the coat cards should be superior to the numeral cards. When we remember that the proper name of Old Thousand is “a thousand myriads,” or, as we should say, “the thousand of myriads,” it is clear that it must necessarily be superior to all numeral cards, the highest of which is the nine of myriads, for a note for ten millions of sapecks must (or at any rate should) be more valuable than one for ninety thousand. The circumstances, too, that one of the original Chinese coat cards should seem to have a special connection with one of the original numeral suits would explain how a set of coat cards came to be attached in Europe to each suit. It is highly probable, moreover, that White Flower (Page 73) and Red Flower were originally marked Ten Myriads and Hundred Myriads, for in a lok pack (a variety of lich chih) from Tapu, near Swatow, cards marked “sprayflower,” “hundred myriads,” “thousand myriads,” and “myriad myriads” are found in addition to the numeral suits. (“Sprayflower,” as we have seen, is another name for “White Flower.”) This then, would explain why one coat card should be considered in Europe superior to another, as the Rey to the Caballo, and not its coordinate, as in “Khanhoo.” Lieh chih, “waste paper,” indeed, throws much light not only on points like these, but also on the remaining difficulty – the fact that the Chinese game has three suits, while the early, like the modern European has four. The normal lieh chih pack consists of thirty-six cards in four suits, arranged thus when the number of players is three:

  • akhs: “hundred sons,” 9, 8, 7, 6, 5, 4, 3, 2.
  • rouleaux: 9, 8, 7, 6, 5, 4, 3, 2, 1.
  • strings: 9, 8, 7, 6, 5, 4, 3, 2, 1.
  • cash: ace of cash, 9, 8, 7, 6, 5, 4, 3, 2.

With four players two additional cards, Prince of Mao and Stagflower, are used and rank as the two highest cash, while the order of the numeral cards in that suit is reversed. In the lok pack just referred to, the place of Stagflower is taken by Sprayflower, a further proof of the identity of both with White Flower.

Now, in the kun p’ai game, of which “Khanhoo” is a faithful copy, the suits are coordinate to one another; no one suit is superior to any other. There is, indeed, no reason why any suit should be superior or any card as the game is now played, for the cards do not take one another, but serve, as in poker or commerce, to form certain combinations. In lich chih, on the contrary, the cards do take one another, and in the very significant order given above – that is to say the 5 of lakhs (500,000 cash) takes the 4 of lakhs and the 1 of rouleaux (10,000 cash) the 9 of strings (9,000 cash). In the case of two of the suits we find the ace counting before the nine, as in many European games. (Page 74) The reason for this is to be found in the kun p’ai packs, where aces are valuable because of their mobility (the feature is copied into “Khanhoo”) and because they enter into so many combinations. Index or “squeezes” marks are met with in most Chinese packs, and it is often the case that the aces are honored with the same index mark as White Flower, Red Flower, and Old Thousand, which three, indeed, are sometimes spoken of as yao, “aces.” In lieh chih “Lin Chung” and “Wang Uing” are frequently substituted for “Prince of Mao” and “Stagflower,” just as we have seen that in certain kun p’ai packs “Prince of Mao” and ” Wang Uing” are occasionally substituted for Red Flower. There can be no question, indeed, that both lieh chih and kun p’ai have the same origin. In the absence of detailed information of early Chinese games it is impossible to say which of the two more nearly represents the primitive form. It is allowable, however, to assume from lich chih that the parent game packs of which found their way westward in the 13th century, had four suits of 9 cards each, and from kun p’ai that it possessed besides three coat cards.

Says Willshire (p. 33): “It is probable that in the perfect sequence there are 5 suits of 9 pieces each suit, the marks of (Page 75) which are bags, money, batons or bows, swords, and a fifth mark not satisfactorily demonstrable.” This conjecture, erroneous as it is, in one way serves a most useful purpose, because it shows what view a European who knew nothing whatever about them would be apt to take of the suit marks of these Chinese games, for Willshire is describing a complete lieh chih pack in the British Museum (O.C. 251). If Willshire in 1878 took the lieh chih suits to be money, bags, batons, and swords, is it surprising if the Italians of 1278 – if that is when they first saw them – took these same suits – to be money, cups, swords, and batons? As regards the suit of cakes or cash, indeed it would require very little imagination to hit upon its meaning. Chatto himself says “the mark of the suit of ” what he calls “Nines Cakes is nearly the same as that of the old Italian danari.” This writer elsewhere observes, “in the 16th century it appears that in Italy the suit of bastoni, ‘clubs,’ was also called colonne, ‘columns’, … merely because the club or mace bore some resemblance to a slender pillar.” Take a number of kun p’ai packs and submit them to various persons who are ignorant of their meaning and it will be found that among the interpretations given to the suit of strings those of “swords,” “bamboos,” “batons,” and “pillars” will be the most common, particularly if the ace or the two is the first card shown.

But what of the “cups””, the copes or coppe? It is precisely this suit-mark, so altogether inexplicable on any other reasoning, that affords the best proof that the early European cards were derived from the kun p’ai or lieh chih of China. Take the first three – the ace, 2, 3 -of the myriad:

The game of Tarocco as played in Italy at the present time is a suggestive compound of the two national card games of China – K’anhu (Khanhoo), played with the kun p’ai, and t’ienkia, played with the “dominoes” – described in the passage from the Cheng tzä t’ung already translated. The Tarocco pack consists of 22 tarots, numbered from 0 to xxi, and of four suits of 14 cards each, namely, the King, Queen, Cavalier, Page, 10,9,8,7,6,5,4,3,2,1 of Money Swords, Cups, and Clubs. The order of the numeral cards varies with the suit, much as in four-hand lieh chih – that is to say, in Swords and Clubs the 10 ranks highest, the ace lowest, while the reverse is the case in Cups and Money. In lieh chih the natural order is preserved in Rouleaux and Strings (the prototypes of Swords and Clubs), the 9 ranking above the 8 and the 2 above the ace, while in Cash (the Italian Money) the ace ranks first and the 9 last. The deal in Tarocco, as in so many Italian and Spanish games, passes round “nidder-shins,” the way the sun does not go, precisely as in all Chinese card games. The dealer deals l9 cards to each player and leaves a stock of 2 cards, which the players successively make use of by exchange to better their hands, much as the stock is used at K’anhu. Having done this, the players then announce any sequence they may have obtained. In making up these sequences the lowest tarot, 0 (il matto, the fool – the joker, in short), is given the privileges of the Chinese “golds” – that is, it can take the place of any required card. So far the principle of K’anhu has been followed. The suit-marks are as we have seen, the K’anhu suit-marks. There is the stock drawn on by exchange, the emblematic joker and the scoring sequence. But after the announcements the game proceeds as in whist, and the cards take one another just as they do at lich chih and t’ienkiu. The tarots act as trumps to the other suits, and by a peculiar rule if the lowest of these, not zero – that is, tarot i, il begatto – takes the last trick it scores double points.

What these tarots were originally has puzzled all writers on cards, nor has any one, so far is I am aware, ever attempted to account for their, number, 22. Yet it is exceedingly probable that in them we have the set of twenty-one natural dominoes which (Page 77) forms the base of nearly all Chinese card games that derive their marks from dice. A domino is, of course, nothing but a pair of dice placed side by side (whence the mark across its face), and the number of possible combinations of two dice is 6.7/1.2, or 21. In the game of t’ienkiu dominoes still remain as dominoes, of wood, bamboo, or bone; but in many derived games they appear as pasteboard. In others, such as hua ho, “flower-matching” (the original of the Japanese hana awaise, praised, rightly but ignorantly, by Sir E. Arnold), they may be had in either form; but wherever they enter the number 21 is always present. Take, for example, hua ho. Here we have the 2l dominoes in three forms: (a) plain, (b) illuminated, (c) doubled; three of the first, two of the second, and one each of the third – in all, 126 cards, or six times 21, together with a varying number of blanks, which serve as jokers. (Portions of the hua ho pack in the British Museum are described with more than his usual extravagance, (Page 78) by Willshire, O.C. 296, 297.) Moreover, each of the different combinations, 6/6, 1/1, 4/4, 1/3, and so on, has its name and fanciful signification, as “heaven,” “earth,” “man,” “the trinity;” and as this nomenclature is thoroughly Chinese, there is no real reason to disbelieve, with Merlin, in the tradition given in the Chêng tzä t’ung. Further, at t’ienkiu the lowest card (the 1/2 for the 1/1 ranks second) is allowed, when it takes the last trick, to score double, just as il begatto at Tarocco.

Is it unreasonable to believe, then, that in the 21 natural dominoes, the base of so many Chinese card games, we have the origin of the 22 tarots? The Chinese dominoes have the emblematic names just as the tarots have, and in their pasteboard form constantly are ornamented besides with full-length figures of men and women.

But it will of course be urged, why are there 22 tarots, then, and not 2l? What, however is the 22d? It is zero, a blank, and its use is to serve as a “joker.” But at hua ho there are, besides the 21 so-called “doubles,” a number (usually three) of blank cards, which are used exactly as il motto is used. If we assume that the Polos, father or son, brought home with them a hua ho pack and a k’an hu pack, with directions, rendered somewhat vague by travel, of how to use them, and that their Venetian friends combined these into a single pack, we have an explanation of the tarots, and the Tarocco game that is at least coherent. The 21 doubles – the most striking of the set – with their full-length figures and emblematic names, would furnish the tarots, the blank (the joker) – becoming il matto, while from the kan hu game (the kun p’ai) would come, as we have seen, the money and the swords, the clubs and the cups. The 10 spot and the queen would be Italian innovations, born of the same ignorance which has added a blank to our European dominoes.


  1. The coloring of the objects illustrated in this paper is represented by heraldic symbolism, the dotted signifying yellow or gold; vertical lines, red; oblique, green; horizontal, blue.
  2. Geschichte des Schachspiel, ii, pp. 381, sqq.
  3. See Chatto: Facts and Speculations on the Origin and History of Playing Cards, 18-48.
  4. Now in the Collection of Games in the Museum of Archaeology of the University of Pennsylvania.
  5. Chinese Currency, Leyden, 1877.

Copied from:

Elliott Avedon Museum and Archive of Games
University of Waterloo, Ontario, Canada