In the original of Dummett's chapter 3, there are numerous words in oriental languages whose transcriptions into Western letters have markings above some of the letters, especially vowels. While in many cases what I have rendered as a "u" with an umlaut (ü) corresponds to what is in the original, in the Chinese word I have rendered as "tzü", it is not an umlaut but more of a small sideways parenthesis mark with the ends pointing up. And in the Arabic words, many of the original transliterations have short horizontal bars over some of the vowels, which I have omitted. For example, the word "Mamluk" has such a bar over its u. Readers who desire Dummett's original transliterations in these cases will have to consult the original pages. Most of the words affected appear on the eight pages I have put scans of at the end of this post.
CHAPTER 3
Europe and Asia
The sudden widespread appearance of playing
cards in Europe in the last quarter, or at most
third, of the fourteenth century is explained by
some as the result of an independent invention of
them, somewhere in Europe, at that time. Even
without considering any positive contrary
evidence, we can recognise this hypothesis to be
intrinsically unlikely. On the face of it, the
structure of the regular European playing-card
pack cries out for explanation: it is not the sort of
thing that would occur to anyone out of the blue.
There is a story of an Eton College Training
Corps platoon, when ordered to number off,
shouting out, ‘One, two, three, four, five, six,
seven, eight, nine, ten, Knave, Queen, King.”
familiarity makes us apt to overlook the oddity of
continuing a numerical sequence with three
human figures. As H.J.R. Murray remarked,
people do not invent games out of wholly
unfamiliar materials: either they adapt the
equipment of one game for use in a new one, or
they use, or represent, some already existing set
of objects not previously part of any game.
1 The
structure of the regular European pack is too
complex, and too unnatural, for it to have been
invented outright in its developed form: no one
can have sat down and thought, ‘If we had fifty-
two equal-sized pieces of paper, marked on one
side with the following designs, then we could
play all sorts of games with them.” One might hit
on the generic idea of playing cards — playing
cards in the sense in which Lexicon cards or
Happy Families cards are playing cards — and
_____________________
1 H.J.R. Murray, The History of Board Games Other than Chess,
Oxford, 1952, pp. 7, 237.
[2nd column]
use for them symbols of some familiar kind or
representations of some familiar objects; or,
having the generic idea, one might think of a new
structure for the pack. But, as we have seen, the
evidence suggests that the regular four-suited
pack appeared quite suddenly in Europe in a
fully developed form: the earliest European
playing cards formed a pack essentially the same
in composition as the four-suited 52-card pack
that we know today. Later — as early as the
fifteenth century
2 — packs of playing cards, in the
generic sense, were devised with a totally
different structure; but it seems certain that the
_________________
2 In 1449 a wealthy Venetian, Jacopo Antonio Marcello,
sent as a present to Queen Isabelle, the first wife of King
René I, Duke of Lorraine, a set of sixteen picture cards that
had originally been painted by Michelino da Besozzo at the
order of Filippo Maria Visconti, Duke of Milan, who died in
1447. The set was divided into four groups of four cards
each, representing Virtue, Virginity, Riches and Pleasure:
each card depicted a suitable classical divinity. The pack
has disappeared, but Marcello’s covering letter has
survived, giving a description of the pack and saying that it
had been painted by Michelino (Michelinum pictorem
elegantissimum adhibuit, alterum hac nostra tempestate Polycletum,
qui totum ludum artificiosissime ornatissimeque depingeret: ‘he
commissioned the very fine painter Michelino, a second
Polycletus of our time, to paint the whole pack most skilfully
and beautifully’). The use of the word ludus shows,
presumably, that this set was really intended to be used to
play some kind of game. See P. Durrieu, “Michelino da
Besozzo’, Mémoires de l'Institut national de France, Académie des
Inscriptions, vol. 38, part 2, 1911, pp. 365-93, especially pp.
373 and 376; also the reports in the Bulletin de la Société
nationale des Antiquaires de France pour 1895, p. 117, and in La
Chronique des Arts, 1895, p. 110; Le Manuscrit, vol. 11, 1895, p.
18, has also a passing remark on the subject in the same
author’s ‘Le Strabon du Roi René’, pp. 2-5, 17-21, which
concerns another present by Marcello to King René himself.
34
Part I: History and Mystery
first appearance in Europe of playing cards in the
generic sense was in the form of the regular four-
suited pack. There is not the slightest trace of any
gradual development leading up to the invention
of a pack with so complex and surprising a
structure; and this fact, of itself, prompts the
supposition that the pack was not invented in
Europe, but introduced from elsewhere in an
already developed form.
If, like the game of chess, the playing-card
pack was an importation into Europe from
outside, it is natural to ask from where it came.
The question from which area European playing
cards immediately derived inevitably leads on to
the further question, which many writers have
striven to answer, where playing cards
originated in the first place; just as chess came
into Europe from the Islamic world, but was
originally an Indian invention, so it is quite
possible that the region from which playing cards
entered Europe was not that which gave them
birth. But, in asking this further question, we
must make just that distinction which does not
arise for the introduction of playing cards into
Europe. The generic idea of playing cards may
well have antedated the invention of the regular
form of pack, or of whatever approximation to it
was the ancestor of the European playing-card
pack: the question as to ‘the origin of playing
cards’ may well have different answers according
as we understand the term ‘playing cards’ in its
broadest sense or only in that narrower sense in
which it applies exclusively to cards of the
regular pack.
Playing cards can be made of other materials
than paper; but it is of paper that they are most
conveniently and typically made, and, indeed,
the very word ‘card’ (
charta, etc.) means
originally simply ‘paper’. Likewise, playing cards
do not have to be printed, but can be drawn or
painted by hand; but printing is obviously the
most efficient means for producing them in any
quantity. It is therefore natural to conjecture that
the land in which both paper and printing were
first invented — China — was also that in which
playing cards, in the generic sense, were first
known. This conjecture is borne out by the
evidence. T.F. Carter gives A.D. 969 as the earliest
certain date for Chinese playing cards: this is on
the basis of an entry for that year in the
Liao shih of
T'o-t‘o, a history of the Liao dynasty (907-1125)
written in the fourteenth century, stating that the
Emperor Mu-tsung played cards (
yeh ke hsi) on
[2nd column]
the night of New Year’s Eve.
3 Dr Gernot Prunner
has cited the
Kuei t'ien lu, a book of anecdotes
written in the eleventh century by the historian
Ou-yang Hsiu, as placing the invention of
playing cards (
yeh tzü ko) in the middle of the
Tang dynasty (618-906), that is to say, about the
time when the earliest books were printed, in the
ninth century.
4 Ou-yang Hsiu’s opinion seems to
be borne out by the title of a poem, mentioning
playing cards (
yeh tzü), by Li Tung, who lived at
the very end of the T'ang period, and by an entry
on ‘golden yeh tzü games’, compiled by Lady
Chou, consort of the last ruler of the Southern
T‘ang dynasty (936-978), included in a catalogue
in the
Tan ch'ien tsung lu.
5 We can therefore safely
regard playing cards as having originated in
China in the ninth century, a date which long
precedes the earliest possible one at which there
is any ground to think that they were known
anywhere else in the world.
From a world perspective, the important step
was the invention of playing cards as such; but,
______________
3 T.F. Carter, The Invention of Printing in China and its Spread
Westwards, New York, 1925, revised edition 1931, pp. 140-1;
second edition, revised by Goodrich, New York, 1955, pp.
184-5; see also the 1925 and 1931 editions, p. 243, footnote.
5, and the 1955 edition, p. 189, footnote 7. ‘Liao’ was the
name of the Kitan dynasty whose empire lay to the north of
the Sung, with its capital at Peking. For the passage from
the Liao shih, see Karl A. Wittfogel and Feng Chia-Sheng,
‘History of Chinese Society: Liao (907-1125)", Transactions of
the American Philosophical Society, n.s., vol. 36, Philadelphia,
1949, p. 257. The reference to the Liao shih is given as
chapter 7, p. 5a of the Po-na edition. T‘o-t'0 includes the
term yeh ke hsi in his glossary.
4 Gernot Prunner, Ostastatische Spielkarten, Bielefeld, 1969,
p. 4, a catalogue for an exhibition which contains many
illustrations and much useful information.
5 Both these are cited by Chin Hsüeh-shih in his Mu chu
hsin hua (Swineherd Games) in the section on playing cards
(chih p'ai); the book is included in the Chao-tai ts'ung-shu
collection, Pieh-chi, chüan 43. Chin Hsüeh-shih flourished
in the second half of the eighteenth century, and his book
contains much valuable information. (The explanation of
the title of the book is that T‘ao K‘an (257-332) had
gamblers flogged and their gaming instruments thrown into
the river, saying ‘Such diversions are fit only for
swineherds.” Among the extremely learned, ‘swineherd
games’ was therefore a general term for games of chance.)
The full title of Li Tung’s poem is ‘Secretary Wei of Lung-
chou played yeh tzü after dreaming of double sixes. I offer
him this poem.” It is in vol. 723 of the Complete Poems of the
T'ang Dynasty. Chin Hsüeh-shih cites Lady Chou’s entry as
included in “Mr Cheng’s catalogue’ in the Tan ch'ien tsung lu,
a work compiled in the early sixteenth century. I owe much
of this information to Dr David Hawkes, of All Souls
College, Oxford, and am much indebted to him for the great
trouble he took to help me with Chinese sources.
Europe and Asia 35
from a Chinese viewpoint, this is a superficial
view. As is apparent from the preceding
paragraph, the term used for playing cards in the
older Chinese works is
yeh tzü, as appears in the
title of the mid-sixteenth-century treatise
Yeh tzü
p'u by P‘an Chih-heng included in the
Shuo fu
compiled by the Ming writer T‘ao Tsung-i. In
modern Chinese, however, the word used is
p‘ai.
This term is applied both to playing cards and to
tiles, such as Ma Jong tiles and dominoes; when
it is necessary to distinguish, the word
p'ai is
qualified by a word for the material (paper,
ivory, etc.) of which the card or tile is made; thus
chih pai (paper
p‘ai) means ‘playing cards’ in our
restricted sense. This terminology reflects the
fact that no distinction of principle is made between
playing cards and tiles. As instruments
of a game, both are alike in looking the same on
one side and different on the other, so as to allow
each player to hold pieces whose value is
unknown to the other players; the game itself is
unaffected by whether these pieces are thin and
flexible or thick and rigid. It is apparent to
anyone who has played Ma Jong that it is
essentially a card game, bearing a general
resemblance to Rummy; but it will reasonably be
felt that there is a deeper difference between card
games and the games usually played in the West
with dominoes. In card games, it has no bearing
on the play which way up a card is laid on the
table; all that matters is which card is played by a
given player at a given time. Thus, although it
was a long time before double-headed cards were
introduced in the West, it was precisely because
the distinction between the top and the bottom of
the card had no significance for the play that
their introduction was a great advantage to
players. Double-ended dominoes, on the other
hand, would, for European players, be an
impossibility; in the usual type of domino game
played in the West, based on the principle of
matching ends, it matters to the game not only
where the domino is placed, but also which way
round it is placed. Judged by this criterion,
however, most Chinese domino games are to be
classed as card games: although the principle of
matching ends is not absolutely unknown in
China,
6 in almost all Chinese domino games all
that matters is the identity of the domino played
at a given time, not where or which way round it
___________________
6 See Stewart Culin, Games of the Orient, Rutland,
Vermont, and Tokyo, 1958 (originally published as Korean
Games, Philadelphia, 1895), p. 141.
[2nd column]
is placed. The idea of double-ended dominoes,
such that one can identify a particular domino as
(say) the 4-3 without being able to associate the 4
with one end and the 3 with the other, is
therefore not at all an absurdity in China. From
the point of view of the games played with them,
Chinese dominoes, like Ma Jong tiles, are to be
classified as playing cards in the generic sense.
Ma Jong itself is frequently played in China,
not with tiles, but with (paper) cards. Even more
common are domino cards, which are now in fact
always double-ended, bearing at each end the
two spot-markings (e.g. 4-3) which together
identify the domino, and sometimes also an
additional design, irrelevant to play, for
decorative purposes. Dominoes are a Chinese
invention, which arrived in the West only very
late, in the eighteenth century.
7 Each domino
represents a possible fall of two dice; there is no
blank in Chinese dominoes, the blank being a
European addition. It is indeed certain that the
early
yeh tzü were either dominoes or domino
cards, as is apparent from the title of Li Tung’s
poem, which mentions the double six, while
Prunner cites the
Kuei t'ien lu to the same effect.
The question whether the earliest ones were
domino tiles or domino cards is as unimportant
from a Chinese viewpoint as the question from
what material the earliest chessmen or
chessboards were made, and without doubt the
term
yeh zü was applied to both. But the question
becomes important when we enquire into the
possibility that the playing cards of other lands
are ultimately derived from those of China. Since
in his poem Li Tung speaks of ‘square ivories
engraved with colours’, he appears to have tiles
in mind, and Prunner believes that it was with
tiles that Mu Tsung played. However, the
erudite writer Yang Shen (1488-1559) said that
the early yeh tzü were probably like the paper
cards of his day,
8 and in this he is followed by
later writers; and since
yeh means a leaf or page of
_____________
7 See p. 530 of S. Culin, ‘Chinese games with dice and
dominoes’, Annual Report of the Board of Regents of the
Smithsonian Institution ... for the year ending June 30, 1893: Report
of the U.S. National Museum, Washington, 1895, pp. 489-537,
and R.C. Bell, Board and Table Games from Many Civilisations,
vol. I, Oxford, 1969, p. 162.
8 Quoted by Chin Hsüeh-shih, op. cit. He actually says
that they were probably like the chih-p'ai chiu-ling, ie.
“drinking cards’, that is, ones not designed for playing an
actual game, but as an aid in a drinking party; but the
important point is that he considered them to have been
paper cards.
36 Part I: History and Mystery
a book (
tzü, literally ‘son’, being used vaguely for
‘piece’), Carter’s opinion that domino cards
preceded domino tiles, the earliest ‘dominoes’
being made of stiff paper, and the more costly tile
form coming into use only later, is highly
plausible.
9
Whichever came first, it seems certain that,
even if we insist on counting only paper cards, it
was in China that playing cards, in the generic
sense, were first invented. It does not follow that
the idea was not hit on independently elsewhere at
a later date; nor does this determine the origin of
the regular pack divided into suits. There are
three principal types of Chinese playing cards
and tiles, of which domino cards and tiles are the
earliest. The latest are the chess cards. These are
packs in which each card bears a character
denoting one of the Chinese chess pieces,
10 divided
into either two or four equal sets —suits, in a loose
sense — distinguished by colour. I know no definite
evidence when packs of this kind were invented,
but have no reason for supposing them to be
earlier than the nineteenth century.
11
The category of Chinese playing cards most
closely resembling European ones is that
consisting of the ‘money’ cards; this category also
comprises Ma Jong cards and tiles. ‘Money’
packs have genuine suits, composed of numeral
cards running from 1 to 9, each suit representing
monetary values. There are two sub-categories,
each comprising several forms: those with four
suits, and those with three. In the Mu chu hsien
hua, a book on games by the eighteenth-century
writer Chin Hstieh-shih, the term chih pai is
reserved for the three-suited packs; while the
four-suited ones are called ma tiao p‘ai; this
terminology is no longer current. Chin cites
various earlier works, including the sixteenth-
____________________
9 T.F. Carter, op. cit. 1925 and 1931 editions, p. 140, 1955
edition, p. 184
10 Chinese chess differs considerably from European chess,
both being descended, by a different line, from the ancestral
Indian game. Each player has a General (corresponding to
our King), two Captains (historically equivalent to the
Queen), two Elephants or Ministers (equivalent to the
Bishops), two Horses (corresponding to our Knights), two
Chariots (corresponding to our Rooks), five Soldiers
(corresponding to our Pawns) and two Cannons, to which
nothing on the European board corresponds. Chinese chess
pieces are white discs, similar in shape to draughtsmen, on
each of which is inscribed the character denoting its rank, in
black or green for one player and red for the other.
11 Chess packs are almost always in the form of paper
cards, but there is a set of rectangular tiles at the Brooklyn
Museum of Art, New York.
[2nd column]
century
Yeh tzü p'u, as referring to
ma tiao p'ai,
while Prunner, who illustrates two seventeenth:
century four-suited packs in his catalogue,
mentions other such sources.
12 A game played
with the three-suited pack is said by Chin to have
been very popular among gentlemen of the
scholar class at the end of the Ming dynasty
(1368-1644); and he himself expresses the view
that the pack must have been formed from the
four-suited one by omitting one of the suits at that
period; that is to say, at the beginning of the
seventeenth century. This appears a very
probable opinion. Chin remarks that, in his day,
the
chih p'ai (three-suited packs) are found
everywhere, and are about ten times as popular
as the
ma tiao p‘ai. The disparity is even greater
to-day; but the four-suited pack still survives in a
version known as
liu hu p‘a (
luk foo pai, six tigers
cards) popular among Hakka-speaking people,
and in a quite different Vietnamese version; the
former is easily obtainable in London.
The four suits vary somewhat in name, but
typically consist of Cash, Strings (of cash),
Myriads (of strings) and Tens (of myriads). In
three-suited packs, only the first three of these
are present. The four-suited packs always
contain just one card of each kind, whereas the
three-suited ones almost always have either two
or four cards of each denomination; this was
already so when Chin was writing. Chin
describes a
ma tiao pack as consisting of 40 cards.
Each suit contains cards from 1 to 9, save that 1
is missing from the Tens suit; in addition, the
Tens suit has three cards higher than the 9,
namely, in ascending order, the Hundred
Myriads, the Thousand Myriads and the Myriad
Myriads, while the Cash suit has two extra ones,
the
Chih-hua (Flower) and the
K‘ung t‘ang. In the
Yeh tzü p‘u of P‘an Chih-heng, these last two cards
are given as
Pan wen chen (‘Half a Cash’) and
_______________________
12 Chin says that the Shu ch'ien yeh p'u of the sixteenth-
century writer Wang Tao-k‘un is still in circulation, and
deals with drinking cards (chih-p‘ai chiu-ling), but that the
chih-p'ai of this work are in fact ma tiao p'ai. He also mentions
the Yün chang ching of Li Sui-ch‘iu and the Ma tiao chiao li of
Lung Tzü-yu, both in the Shuo fu collection. He mentions
two lost works, Ta ma ke chü and Ta ma ke shih, but says that
on the evidence of the Ch'ao ching ta ma ke of the late
sixteenth- or early seventeenth-century writer Wen Hsiang-
feng, also in the Shuo fu, the game of ta ma was not played
with ma tiao p'ai. Prunner, op. cit, p. 8, mentions the
fifteenth-century work Shu yüan tsa chi by Li Jung and the
Chü yi lu of Wang Shih-chen (1634-1711) as both describing
the four-suited pack in detail. The seventeenth-century
packs are nos. I and II of the catalogue.
Europe and Asia 37
K‘ung mei wen respectively. These descriptions and
names tally closely with the seventeenth-century
cards illustrated by Prunner. P‘an also observes
that in the Cash suit the ordinary numeral cards
run from 1 as highest down to 9 as lowest; W.H.
Wilkinson, writing in 1895, confirms this for the
case when the game is played by four people.
13 In
more recent packs, the extra cards, variously
named, have been reduced to three, making a
pack of 38 cards altogether. The three-suited
packs appear from the start to have had just three
denominations of card, again variously named, in
addition to the nine numeral cards of each of the
three suits, so that there are 60 cards in a double
pack and 120 in a quadruple one. In most games
played with these cards, the three extra cards are
not associated with particular suits. The Ma
Jong set itself is simply a variant of the quadruple
three-suited money pack in tile form. The suit
known in English as ‘Circles’ represents the Cash
suit (and is called in Chinese by a name
meaning, in effect, ‘coppers’, as this used to be
used in English for ‘pennies’); the suit known,
quite reasonably from the appearance of its suit-
sign, as ‘Bamboos’ represents the Strings suit,
and is so called in Chinese; and the suit vaguely
called ‘Characters’ is the Myriads suit, each tile
bearing a numeral from 1 to 9 and the character
for ‘myriad’. The four Winds, or simply
directions, in the Ma Jong set are, like the
Flowers and Seasons, an addition, peculiar to
Ma Jong; the three Honour tiles, usually known
in English, but not in Chinese, as ‘Dragons’,
correspond to, though they do not precisely
reproduce, the three extra cards which in the
three-suited money pack do not belong to any of
the suits. An immense antiquity is sometimes
claimed for Ma Jong. In fact, it goes back only to
about 1870, although the general type of game
which it represents, and the money packs of
which the Ma Jong set is a variant, are far older.
The Chinese money packs come close to falling
under our definition of a ‘regular’ pack, but fail
to fit it exactly for two reasons. First, there is no
division into numeral cards and court cards, in
the proper sense. In the four-suited pack, the
extra cards do belong to particular suits, but they
are not shared equally between all the suits and
do not share ranks common to the suits to which
they do belong; rather, they appear to have
________________
13 W.H. Wilkinson, ‘The Chinese origin of playing cards’,
The American Anthropologist, vol. VIII, pp. 61-78.
[2nd column]
originated as continuations of the numeral series
in those two suits. In the three-suited packs, the
extra cards do not belong to the suits, but yet are
not interchangeable with one another, like the
Jokers of a modern European pack. Secondly, not
all the suits have suit-signs. In the three-suited
packs, except the Vietnamese version, the two
lower suits, Cash and Strings, do have suit-signs:
the identity of each numeral card is determined,
in the way familiar to us, by the number of
repetitions of a representation, usually highly
conventionalised, either of a coin (the copper
‘cash’ with a hole in it) or of a string of cash. But,
in the suit of Myriads, there is no suit-sign in the
proper sense: rather, each card has written on it
its number and the name of the suit, e.g. ‘4
Myriads’. Exactly this happens on the Ma Jong
tiles. The same is true of most older surviving
examples of the four-suited pack: the numeral
cards of Cash and Strings are pip cards, with
genuine suit-signs, but those of Myriads and
Tens merely have their denominations inscribed
on them. The two modern versions, Hakka and
Vietnamese, of the four-suited pack have no suit-
signs at all: every card has its denomination
printed on it, in highly distorted characters. Save
for the distortion of the characters, the same
holds good of the seventeenth-century cards
illustrated by Prunner.
Despite this difference, it is tempting to
conjecture that the four-suited Chinese money
pack is the ancestor of the four-suited regular
pack that appeared in Europe in the later
fourteenth century; both Wilkinson and Carter
have contended for precisely this theory.
14 It is,
however, quite evident that it could not have
been the immediate forerunner of the European
pack, as Wilkinson supposed. It is no great
problem that the Chinese suits contain nine
numeral cards, whereas the original number in
the European pack was ten: but the absence of
court cards from the Chinese suits forces us, on
this hypothesis, to assume some considerable
development between the Chinese prototype and
the playing-card pack as first introduced into
Europe. A further difficulty concerns the suit-
signs. It is easy to imagine the Cash suit as the
original of the Latin Coins suit, and the Strings
suit could, by a misunderstanding, very well have
___________________
14 W.H. Wilkinson, op. cit., and T.F. Carter, op. cit.,
1925 and 1931 editions, pp. 139-44, with notes on pp. 241-6,
and 1955 edition, pp. 183-8, with notes on pp. 188-92.
38 Part I: History and Mystery
been the original of the Batons suit. But the
Myriads and Tens suits would not be intelligible
to Europeans of the late Middle Ages: they
would not even be recognisable as suits to people
who could read neither the suit-names nor the
numerals. Particularly is this so because, in
almost all money packs, whether three or four-
suited, the greater part of each numeral card in
the Myriads suit and (when present) in the Tens
suit is occupied by a human face or figure,
essentially the same ground as that on which we
have set aside the hypothesis that playing cards
were independently invented in Europe.
The Chinese money pack cannot have been the
immediate progenitor of the European pack; but
that does not rule out its having been its remote
ancestor, through various intervening
intermediate forms. Whether this is a real
possibility depends upon the antiquity of the
money cards. A single Chinese playing card was
representing one of the characters in the famous
romance
Shui hu chuan (‘The Story of the Water
Margin’). These figures are no longer to be found
in the Hakka and Vietnamese four-suited packs,
and they have been dropped from Ma Jong tiles
and cards; but they are always present in the
Myriads suit of three-suited packs, save for the
Vietnamese one and, of course, Ma Jong packs,
and they were formerly present in four-suited
packs as well. In one of Prunner’s seventeenth-
century packs, figures from the
Shui hu chuan
appear on all the cards; in the other, which is
unusual in its composition also (each suit has a
10 and the Tens suit has a 1), there is a figure on
each card, representing, exceptionally, not a
character from the
Shui hu chuan, but some
historical character from ancient times. The
literary sources, however, consistently state that
the cards of Myriads and Tens are illustrated
with depictions of Sung Chiang and the other
characters of the
Shui hu chuan, and this
association seems so firm that it is likely to have
been a feature of the money packs from their first
invention. Now, although these figures are purely
decorative, and have no relation to the use of the
cards, it would be natural for a foreigner, seeing
the cards for the first time, to take them to be
their principal feature. This would, of course,
lead to a complete misinterpretation of the
Myriads and Tens suits; and this fact constitutes
an insuperable obstacle to the theory that
European playing cards were developed directly
oat of Chinese ones.
We must therefore dismiss the speculation that
Marco Polo or some other of the Europeans who
visited China returned with some examples of
Chinese money cards, which later gave rise to
European playing cards. Had this been so, there
would have been a transitional period during
which the Chinese prototype evolved into the 52-
card pack as described by John of Rheinfelden.
Of any such transitional period, not the slightest
trace remains: the theory is to be rejected on
[2nd column]
essentially the same ground as that on which we
have set aside the hypothesis that playing cards
were independently invented in Europe
The Chinese money pack cannot have been the
immediate progenitor of the European pack; but
that does not rule out its having been its remote
ancestor, through various intervening
intermediate forms. Whether this is a real
possibility depends upon the antiquity of the
money cards. A single Chinese playing card was
found in 1905 by von Le Coq at Turfan. He
tentatively dated it to the eleventh century. It
shows a human figure, which strongly resembles
that which depicts Wang Ying, one of the
Shui hu
chuan characters, and appears, in a number of
money packs, on one of the three extra cards,
inscribed simply ‘Wang Ying’. The Turfan card
has an inscription at the top, in a tilted rectangle,
which, unfortunately, is hard to read, at least in
the illustrations, because it is overprinted by a
seal. If it does represent Wang Ying, the card
cannot possibly be as early as the eleventh
century; Carter assigned it to the fourteenth.
15
Both the sixteenth-century
Yeh tzü p'u and,
according to Prunner, the fifteenth-century
Shu
yüan tsa chi of Li Jung describe the lower two suits
as having suit-signs and the upper two as
showing Shui hu chuan figures; if we assume that
the association with these figures was a feature of
the pack from its first invention, then this
association sets a bound on its antiquity. The
Shui hu chuan is the outcome of numerous
legendary accretions that have attached to a
historical nucleus, the exploits of a Robin Hood-
like band of robbers in the early years of the
twelfth century. The version that is now read
dates from the seventeenth century, but has been
expanded from earlier versions, the first of which,
the
Hsüan-ho i-shih, dates from the early Yüan
(Mongol) dynasty; plays based on the story were
performed at the same period, and portraits of
and poems about ‘Sung Chiang and his Thirty-
six’ dating from the later Southern Sung period
__________________
15 T.F. Carter, op. cit., 1925 and 1931 editions, plate
opposite p. 142, and 1953 edition, plate opposite p. 184.
Carter states that there were fwo Chinese playing cards
found at Turfan, but I do not know any confirmation of this;
I have also been told that there is an early Chinese playing
card in a museum in Bombay, but do not know if this is
true. The Turfan card is also illustrated in C.P. Hargrave, A
History of Playing Cards, New York, 1930, 1966, p. 7, and in
A. Bernhardi, ‘Vier Könige’, Baessler-Archiv, vol. XXIX,
1936, pp, 148-80, plate 9a. Bernhardi gives it as his opinion
(p. 164) that the card is to be dated to about 1400.
Europe and Asia 39
(1127-1279) are well attested.
16 Given our
hypothesis that the association with the
Shui hu
chuan characters is an original feature of the
money pack, that pack can in no case be earlier
than the twelfth century, and can hardly be later
than the fourteenth: its most probable date of
origin lies in the late Southern Sung, that is, in
the thirteenth century.
That would leave just enough time for the
money cards to have spread westwards from
China, undergoing very considerable
transformation in the process, and to have
arrived in Europe about a century after their
invention. But the schedule is extremely tight,
and the thesis is therefore somewhat strained.
We shall do better, at this stage, to try to trace
the connection, if any, in the other direction, by
reverting to our original question, from which
area the European playing-card pack was
immediately derived.
If playing cards arrived in Europe from some
external source in about 1370, they can only have
come from somewhere in the Islamic world;
Europe, at that date, had no contact with any
other part of the world save for those mediated by
Islam. For a long time, many scholars dismissed
this hypothesis on the ground that there was no
evidence that playing cards were known in
mediaeval Islam; hence the attraction of theories
such as that of the direct transmission of playing
cards from China to Europe by the agency of
Marco Polo. In 1939, however, L.A. Mayer
reported the existence in the Topkapi Sarayi
Museum at Istanbul of a fifteenth-century hand-
painted pack of cards from Mamluk Egypt.
17
Mayer’s own analysis of the pack was faulty, but,
when this is corrected, it proves to be an almost
complete pack originally containing 52 cards (of
which 48 survive), consisting of four suits,
Swords, Polo-Sticks, Cups and Coins, each
composed of ten numeral cards and three court
cards headed by the King.
18 Five of the surviving
_________________
16 See Richard Gregg Irwin, The Evolution of a Chinese
Novel, Cambridge, Mass., 1953, p. 14 and p. 20, fn. 40, and
P. Eugen Feifel, Geschichte der chinesischen Literatur,
Darmstadt, 1959, pp. 291 ff.
17 L.A. Mayer, ‘Mamluk playing cards’, Bulletin de
l'lnstitut français d'archéologie orientale, vol. XXXVIII, 1939, pp.
113-18, posthumously reprinted in L.A. Mayer, Mamluk
Playing Cards, ed. R. Ettinghausen and O. Kurz, Leiden,
1971.
18 For a detailed discussion, see M. Dummett and K.
Abu-Deeb, ‘Some remarks on Mamluk playing cards’,
[2nd column]
cards have been borrowed from another pack (or
perhaps from two others), to replace lost cards;
Mayer thought that these five secondary cards
had been specially painted for that purpose, but,
if that were so, the artist would surely have tried,
if not to imitate the general style, at least to make
the shapes of the suit-signs conform to those of
the original pack, which, particularly in the
Swords suit, they do not. The numeral cards are
pip cards in the sense of showing the number of
suit-signs corresponding to their values. The
Swords and Polo-Sticks are very highly stylised,
and their arrangement is similar to that of the
Italian Swords and Batons. In detail, the Polo-
Sticks intersect, forming, on the cards from 4 to
10, a kind of trellis pattern. On the 4 to 10 of this
suit, most of the shafts of the Polo-Sticks form an
obtuse angle with the heads, but the sides are
otherwise straight; but on all the odd-numbered
cards of the suit the odd Polo-Stick is S-shaped,
as are two of them on the 10, and on all cards of
the suit besides the 4 to 10 every Polo-Stick has a
curved head; except on the two cards of the suit
which come from the secondary pack, all curved
heads end in a dragon’s head shape. The two
cards of the Swords suit (the 9 and 10) from the
secondary pack show crescent-shaped swords,
intersecting only once on the 9 but twice, in the
Italian fashion, on the 10 (on the 9 one sword is
straight). In the main pack, however, the Swords
are S-shaped, and are arranged side by side,
extending for most of the length of the card, but
without intersecting; in this they resemble the
exceptional version of the Swords suit mentioned
in the last chapter as found in a few fifteenth-
century Italian packs. On some of the odd-
numbered cards of the Swords suit in the
Istanbul pack, the odd sword is S-shaped, and
on others it is straight.
What most strikingly differentiates this
Mamliik pack from European ones is that the
court cards show no human figures. Instead, they
bear the appropriate suit-sign and, at the
bottom, an inscription giving the rank and suit.
The three ranks are: King (
malik); Viceroy (
na’ib
malik); and Second Viceroy (
thani na’ib). There
are also inscriptions in verse at the tops of all the
court cards and of the 8, 9 and 10 of Swords, the
Ace, 2 and 3 of Polo-Sticks and the 2 and 3 of
Coins (the Ace of Coins is missing); the verses on
______________________
Journal of the Warburg and Courtauld Institutes, vol. XXXVI,
1973, pp. 106-28.
40 Part I: History and Mystery
the secondary cards are in a different metre —
another indication that these cards were taken
from one or more distinct packs, rather than
being especially made as replacements. Closer
examination reveals, however, that, while the
lower inscriptions on the secondary cards were
there from the start, and while all those cards
with upper inscriptions always had such
inscriptions (though not originally those that
they have now), the lower inscriptions were not
part of any of the court cards of the primary pack
when they were first painted. This means that,
when the primary pack was originally produced,
the court cards must have been distinguished
solely by their designs. This is plain in the case of
the Kings, which all bear a prominent square
pattern below the suit-sign. It is only in the Cups
suit that the Viceroy and Second Viceroy bear
anything that could have served as such an
identifying emblem; but it is probably only in
this suit that the original two lower court cards
remain. In Coins and Polo-Sticks what was
originally the Ace, and in Swords what was
originally the 9, have been turned into Viceroys
by the addition of a lower inscription,
presumably when the original Viceroys were
lost; the reason for choosing the 9 of Swords to
make into a Viceroy must have been that it had
an upper inscription, while none of the low-
numbered cards of that suit does. The Second
Viceroy in both Coins and Polo-Sticks is from the
secondary pack, while in Swords it is missing
altogether. The secondary cards were evidently
taken from a pack or packs made at a time when
the practice of identifying the court cards by
means of special emblems had been abandoned
in favour of doing so by means of inscriptions;
when these cards had to be added in order to
reconstitute the primary pack, such inscriptions
were painted on to all the court cards, making it
possible to convert numeral cards lacking the
original emblems into court cards.
19
A curious and so far unexplained feature is the
name of the Cups suit —
taman, which does not
mean ‘cups’. Some have suggested that this be
taken as the Turkish word for ‘trousers’, since,
oddly, a pair of short trousers was a frequently
used emblem in Mamluk heraldry. A more
attractive possibility is that it is the Persian word,
deriving from Mongol or Turkish, for ‘10,000’,
applied particularly to a military unit or a group
_________________
19 See M. Dummett and K. Abu-Deeb, op. cit.
[2nd column]
of villages; if so, there would be a tantalising
coincidence in meaning with the name of the
Chinese Myriads suit. However, neither
explanation accounts for the fact that the suit-
sign is unmistakably a cup; and the second
suggestion conflicts with the fact that the word is
normally written
tuman in Arabic (the Arabic
letters transliterated respectively t and t being
completely distinct), although there is a passage
in ibn Battuta where it appears as
tuman.
20 For
the present, the name of this suit remains a
mystery.
The Istanbul set makes it certain that there
existed in Mamluk Egypt playing cards closely
resembling contemporary Italian ones, and that
different types of design, and different modes of
representation of the court cards, were in use
there. Since the cards date from a period later
than the first introduction of playing cards into
Europe, this does not, of itself, tell us whether the
Mamluk cards were derived from Italian ones or
the Italian cards from the Mamluk ones, and
some scholars, such as Pinder and Hoffmann,
have espoused:the former hypothesis.
21 What it
does do is to alter the balance of probabilities by
removing an obstacle to what is intrinsically most
likely. We have seen that, given the abrupt
appearance in Europe of the playing-card pack in
its fully developed form, it is highly unlikely that
it was invented in Europe, and highly probable
that it was introduced from elsewhere; and, if so, it
is hard to suppose that it came from anywhere
but the Islamic world, since Europe had at that
time no direct contact with either India or China.
Until Mayer’s discovery, the apparent non-
existence of any mediaeval Islamic playing cards
seemed to some scholars an insuperable obstacle,
and so they either fell back on the hypothesis of
an independent European invention, or devised
ingenious but unconvincing acccounts of how
they might have reached Europe from India or
from China without passing through the
intervening belt of Muslim states that encircled
Europe. Once we know that playing cards were
known in at least one part of the Islamic world,
the objection is removed; there is no longer any
____________________
20 See Gerhard Doerfer, Türkische und mongolische Elemente
im Neupersischen, vol. 11, 1963, p. 638, where the passage is
cited from H. von Mzik, Die Reise des Arabers Ibn Bafta durch
Indien und China, Hamburg, 1911, p. 442.
21 Eberhard Pinder, Charta Lusonia, Biberach an der Riss,
1961, p. 14; Detlef Hoffmann, Die Welt der Spielkarte: eine
Kulturgeschichte, Leipzig, 1972, p. 19.
Europe and Asia 41
reason to resist the intrinsically most probable
hypothesis, that it was from one or another part
of the Islamic world that playing cards first
arrived in Europe.
The cards in the Topkapi Saray: are not the
only evidence for Islamic playing cards that we
have. No other pack has been discovered, but
several individual cards, or, rather, fragments of
cards, have been found.
22 If the earliest European
cards were imitations of Islamic ones, it seems
probable that the Islamic packs which first found
__________________
22 There are five other fragments of which descriptions
have been published and which have been definitely
identified as parts of Islamic playing cards (or at least of
designs for cards). The first four of these are all in the
private collection of Dr Edmund de Unger, and the fifth is in
the Benaki Museum at Athens (catalogue no. 9b).
(i) One illustrated in L.A. Mayer, op. cit., 1971, figs. 56
and 57, and in B.W. Robinson (ed.), Islamic Painting and the
Arls of the Book, London, 1976, plate 8. Ettinghausen, in L.A.
Mayer, 1971, pp. 9-10, identifies it as the right-hand half of
a 4 of Cups, but, if so, its arrangement of the cups in a single
row would differ from the arrangement in two rows on the
corresponding Istanbul card, and the shape of the card
would also diverge greatly from the Istanbul ones, which
measure approx. 252 x 95 mm., while this fragment is 119 x
40 mm.; it is therefore much more likely to be the top right-
hand quarter of an 8 of Cups. Ettinghausen considers the
card to be late Fatimid in date, i.e. twelfth century, but
allows that it may come from Ayyubid (1174-1258) or even
early Mamluk times; in any case, it is no later than the
thirteenth century. E.J. Grube, in ‘Fostat fragments’ in
B.W. Robinson (ed.), op. cit., pp. 57-8, who also believes it
to be part of a 4 of Cups, thinks that it is Ayyubid or early
Manlik, probably about mid-thirteenth century and in any
case no earlier than 1200. The catalogue number is 1-27.
(ii) A second is illustrated in B.W. Robinson, op. cit.,
plate 8, and in R. Ettinghausen, ‘Further comments on
Manmluk playing cards” in Gatherings in Honour of Dorothy E.
Miner, ed. U.E. McCracken, L.M.C. Randall and R.H.
Randall, Jr., Baltimore, 1974, pp. 51-78, fig. 22; it measures
120 x 85 mm. It is part of another card of the Cups suit,
probably the Ace. Grube considers it to be of the same date
as (i). Its catalogue no. is 1-28.
(iii) A fragment, measuring 100 x 91 mm., of a card of the
Swords suit, perhaps the 2, has not been illustrated; see
B.W. Robinson (ed.), op. cit., p. 65 (catalogue no. I-51).
(iv) A fourth fragment, measuring 140 x 95 mm., has also
not been illustrated, and is part of a card of the Coins suit,
perhaps the 4; see B.W. Robinson, op. cit., catalogue no.
1-52,
(v) Yet another fragment of a card of the Cups suit,
probably the 5 or 6, is illustrated in M. Dummett, ‘A note
on some fragments in the Benaki Museum’, Art and
Archaeology Research Papers, no. 4, December 1973, pp. 93-9,
fig. 1. It measures approx. 87 x 35 mm.
A further fragment in the Benaki Museum is probably not
a playing card: if it is, it must be from a card of the Coins
suit. It measures approx. 75 x 40mm., and has the
catalogue no. 9c. See M. Dummett, op. cit.
[2nd column]
their way to Europe had actual figures on the
court cards: Even without this hypothesis, it
is a reasonable conjecture that there may have
been Islamic packs of this type: we know from
the Istanbul set that two distinct non-
representational methods of presenting the court
cards were in use, and, in secular Islamic art,
there was far from being a strict adherence to the
injunction not to represent the human form. The
late Dr Ettinghausen indeed identified one
fragmentary drawing as a King of some
undetermined suit, and another as a design for a
mounted Viceroy.
23 These identifications remain,
however, subject to serious doubt; all the
fragments so far identified with any certainty as
mediaeval Islamic playing cards are numeral
cards. One of these has been dated, on stylistic
grounds, to the twelfth century by Ettinghausen
and to the thirteenth by Grube. In respect of the
individual card, both datings may be too early,
despite the unquestioned expertise of these two
authorities. Designs used for playing cards are
notoriously conservative; stylistic features may
be preserved unchanged for centuries. For this
reason, while one may often safely say that a card
could not have been made before a certain date, it
is much more risky to assert, on purely stylistic
grounds, that it could not have been made after a
particular date; all the more so when, as here, one
is dealing with an isolated fragment, constituting
one quarter of a card, of a type of which we have
very few examples. But a doubt, on these grounds,
about the dating of the individual card leaves
unaffected the argument to the early date of the
type which it exemplifies: if Grube and
Ettinghausen are right in thinking the design to
___________________
23 See the article by R. Ettinghausen cited under (ii) of
the foregoing footnote, pp. 70-3 and figs. 29, 33. The
drawing of 2 mounted falconer shown in fig. 33 belongs to
the de Unger collection (catalogue no. 1-31), and is also
illustrated in B.W. Robinson (ed.), op. cit., plate 9.
Especially in view of Grube’s remark in that volume that it
is not the work.of an artist, its claims to be regarded as a
sketch for a playing card may be rated as negligible. Fig. 29
shows a fragment from the Museum of Islamic Art in Cairo
(catalogue no. 15610/4-1). It indeed depicts a seated King,
and it is not difficult to see it as a playing card. Both the ,
King’s hands. are visible, and one holds what Ettinghausen
interprets as a bird, though to me it more resembles a snake.
Ettinghausen implausibly connects this with the Birds suit
occasionally used in Germany in the fifteenth and sixteenth
centuries; until we have definite reason for identifying what
the King is holding as a suit-sign (perhaps some new
deformation of the Polo-Stick), we cannot safely class this
fragment as a playing card.
42 Part I: History and Mystery
have originated no later than the thirteenth
century, then Islamic playing cards already
existed at that time, and we can conclude
without more ado that European cards were
derived from Islamic ones and not conversely.
In addition to the mediaeval Egyptian playing
cards that have been found, one fifteenth-century
Italian card and two sixteenth-century ones have
been discovered amongst mediaeval fragments
from Egypt, indicating that there was a trade in
playing cards with Italy that extended even into
Ottoman times (Mamluk Egypt fell to the
Ottoman Selim the Grim in 1517).
24 That this
trade was not in one direction only is attested by
several European references. In the inventory of
the goods of Louis de Valois and his wife
Valentine, née Visconti, Duke and Duchess of
Orleans, begun at the order of their son Charles
on the day of his mother’s death in 1408, there
are listed
ung jeu de quartes sarrasines and
unes
quartes de Lombardie (‘one pack of Saracen cards;
one cards of Lombardy’).
25 F.M. Graves
concluded that the former was a regular Italian
pack and the latter a Tarot pack;
26 but since the
Tarot pack had almost certainly not been
invented by 1408, let alone by 1389, when Louis
and Valentine were married, and when she very
likely brought the cards with her from Milan,
there is no reason not to take this entry as
meaning exactly what it says. A number of
fifteenth-century inventories from Barcelona also
mention Islamic cards. Two distinct ones, both
dated 1414, have entries ‘j
joch de nayps moreschs’
and 'j j
och de nahyps moreschs’ respectively
(‘one pack of Moorish playing cards’); a third, of
1439, has the entry ‘x
jochs de naips moreschs; iij
altres jochs de naips plans petits’ (‘10 packs of
Moorish playing cards; 3 other packs of ordinary
small playing cards’)
27 Yet another, of 1460, has
_________________
24 The fifteenth-century card is a 3 of Cups in the Benaki
Museum (catalogue no. 9a) from the same pack, probably
Venetian, as sheets I-1009 and II-1010 in the Cary
Collection at Yale University; see the article by M.
Dummett cited under (v) of footnote 22. Of the sixteenth-
century ones, one belongs to Dr Ettinghausen, and is
illustrated (figs. 23 and 24) in his article mentioned under
(i) of footnote 22; it is part of an 8 of Batons. The other is a 6
of Swords in the collection of Mr Sa‘d Khadam of Cairo,
and is referred to by Ettinghausen (p. 65).
25 F.M. Graves, Deux tnventaires de la Maison d’Orléans,
Paris, 1926, p. 49.
26 F.M. Graves, op. cit., p. 134.
27 See ‘El Joc de naibs a Catalunya’, unsigned, Bibliofilia,
ed. R. Miquel y Planas, vol. II, Barcelona, 1915-20,
[2nd column]
the entry ‘
jochs de nayps plans, y altres jochs moreschs’
(‘packs of ordinary playing™ cards, and other,
Moorish, packs’).
28 The Mamluk cards at
Istanbul were not an isolated rarity, as some have
supposed; in the fifteenth century, Islamic
playing cards were perfectly well known, even in
Europe.
So far, three references to playing cards are
known from mediaeval Arabic literature. One is
from the
Thousand and One Nights,
29 despite the
frequent assertions to the contrary in modern
writings on playing cards. Another is from the
sixteenth-century writer Ibn Hajar al-Haytami
(1504-1567).
30 The most important is a passage
in the Annals of Ibn Taghri-Birdi (a history of the
Mamluk rulers from 1382 to 1469) to which
attention was first drawn by Mme Laila
Serageddin.
31 In this, the future Sultan al-Malik
al-Mu’ayyad is recorded to have won a large sum
of money in a game of cards in about the year
1400 (the date is not precisely given). This
confirms that playing cards were known in
Mamlik Egypt at a date not long after their first
appeararicé: in Europe. But by far the most
______________________
columns 181-207; also José Maria Madurell Marimés,”
‘Notas Documentales de Naiperos Barceloneses’, Documentos
y Estudios, vol. 1V, Instituto Muncipal de Historia,
Barcelona, 1961, p. 59.
28 J Brunet y Bellet, Lo jock de Naibs, Naips ó Cartas,
Barcelona, 1886, pp. 80, 111, 113 and 160. A possible,
though less likely rendering of this entry is ‘packs of
ordinary playing cards and other Moorish games”.
29 See Enno Littmann (trans.), Die Erzählungen aus den
Tausendundein Nächten, vol. III, Wiesbaden, 1953, p. 693. The
reference occurs in the story of the clever slave girl
Tawaddud, 460th-461st nights; Littmann’s translation is
from the text edited by Sir William Hay Macnaghten from
an Egyptian manuscript ‘brought to India by the late Major
Turner, editor of the Shah-Nameh’, and published by him
as Alif Laila, or Book of the Thousand Nights and One Night, four
vols., Calcutta, 1839-42, our passage being in vol. 11, p 354.
30 This reference is given in Franz Rosenthal, Gambling in
Islam, Leiden, 1975, p. 37, fn. 146, p. 44, fn. 178, p. 45, fn.
180, and p. 63, as az-Zawdjir (‘an iqtiraf al-kaba'ir), vol. II,
Cairo, 1951, p. 191. The reference to the Thousand and One
Nights is also due to Rosenthal. My thanks are due to Dr
Rudi von Leyden for bringing Rosenthal’s book to my
attention.
31 For an English translation, see Ibn Taghri-Birdi,
History of Egypt 1382-1469 A.D., trans. by W. Popper,
University of California Publications in Semitic Philology, no. 17,
Berkeley and Los Angeles, part III, 1957, p. 50. An
acknowledgment to Mme Serageddin was made by Dr
Ettinghausen in his article cited under (ii) of footnote 22, p.
74. For comments on this article, see the review by M.
Dummett and-Kamal Abu-Deeb in the Joumal of the Playing-
Card Society, vol. III, no. 3, February 1975, pp. 43-52.
Europe and Asia 43
celebrated reference to mediaeval Islamic cards
comes once more from Europe, namely from the
Chronicles of Viterbo. There are three fifteenth-
century Chronicles of that city, all surviving in
manuscript form. Two of them — those by Fra
Francesco d’Andrea da Viterbo and by Niccolo
di Niccola della Tuccia — have been published;
32
the third, by Giovanni di Juzzo di Covelluzzo,
has not, but the relevant passage from it is cited
in F. Bussi’s
Istoria della Citta di Viterbo of 1742.
(Writers on playing cards often give a confused
account of this matter, referring to ‘the’
Chronicle of Viterbo, as if there were only one.)
All three chroniclers state that they are relying,
for their pre-fifteenth-century narratives, on
earlier Ghronicles, and, in particular, for the
fourteenth century, on one by Cola di Covelluzzo
(a relative of Juzzo’s); Fra Francesco and della
Tuccia also cite another fourteenth-century one
by Maestro Geronimo. These earlier Chronicles
have disappeared: but since all three of the
fifteenth-century ones contain slightly garbled
versions of what is evidently the same entry for
the year 1379, it is apparent that they must all be
quoting, directly or indirectly, from the same
fourteenth-century source, almost certainly the
lost Chronicle of Cola di Covelluzzo. When
reconstructed by comparison of the entries in the
three Chronicles, the original statement can be
seen to have read, ‘
Anno 1379. Fu recato in Viterbo il
gioco delle carte, che in Saracino parlare si chiama Nayb'
(‘In the year 1379 there was brought to Viterbo
the game of cards, which in the Saracen language
is called nayb’). Juzzo has expanded the last
clause to ‘
che venne de Seracinia, & chiamasi tra loro
Naib' (‘which comes from the land of the
Saracens, and is called among them naib’): this
remark cannot be credited with fourteenth-
century authority, but at least shows how Juzzo
(who died in 1480) interpreted his source.
(Although one late MS. of della Tuccia’s
Chronicle substitutes for the remark about the
‘Saracen’ language the information that cards
were brought to Viterbo by a Saracen named
Hayl (
da un saracino chiamato Hayl), this is almost
_________________
32 The Chronicle of Fra Francesco, edited by Francesco
Cristofori, was published in Archivio storico per le Marche e per
Umbria, vol. IV, Foligno, 1888, pp. 261-338; that by della
Tuccia, edited by Ignazio Ciampi, in Cronache e Statuti della
Città di Viterbo, Florence, 1872. See also Pietro Egidi,
‘Relazioni delle Chroniche Viterbesi del secolo XV tra di
loro e con le fonti’, Scritti vari di Filologia (a volume of essays
for Ernesto Monaci), Rome, 1901, pp. 37-59.
[2nd column
certainly a copyist’s blunder.) Thus the entry in
the original fourteenth-century Chronicle does
not seem to have explicitly stated that playing
cards came from the Islamic world; but it did, in
the very year of their arrival in Viterbo; and only
two years after the very earliest reference to them
from anywhere in Europe, connect them with the
Islamic world, and show an awareness of their
existence there. If playing cards had in fact been
invented in Europe only a few years before, it is
unlikely that they would yet have been taken up
by the Muslims, or that an Italian chronicler
would have been aware of the fact if they had.
33
The Viterbo reference is of great etymological
interest. In most European languages, the word
for ‘card’ is, as in English, a derivative of Latin
charta, meaning ‘paper’, or of its diminutive
chartula. In fifteenth-century Italian, however,we
very often find, alongside
cartule, carticelle, etc., the
quite different word
naibi used for ‘playing cards’,
although after that century the word went out of
use; and from Marseilles and from Spain we find
forms like
nahipi and
naips, the ordinary Spanish
word for ‘playing cards’ being
naipes to this day.
The entry in the Viterbo Chronicles supplies the
only plausible etymology for these words that has
ever been proposed. It is very unlikely that
playing cards in general were ever called in
Arabic anything like
na’ib, the plural of which is
in any case
nuwwab; but, since this word,
meaning ‘deputy’, is the first word of the title of
the second céurt card in.the Mamluk pack, it is
perfectly possible that, if Italians or Spaniards
first became acquainted with playing cards of
Muslim origin, they would apply the word to
playing cards as such. It is much less likely that
they should have done so had playing cards been
a European invention, subsequently adopted by
the Muslims. The use of the words
naibi and
naipes thus combines with the remark of the
Viterbo chronicler to give great weight to the
hypothesis that playing cards came to Europe
from Islam.
The word used for ‘playing cards’ in all three
of the Arabic sources cited above is
kanjifah; and,
as pointed out by Mme Serageddin, the same
word appears in the verse which forms the upper
inscription on the King of Swords in the Istanbul
pack, which runs ‘kanjifah brings joy and
_________________
33 See M. Dummett and K. Abu-Deeb, ‘Some remarks on
Mamlak playing cards’, Journal of the Warburg and Courtauld
Institutes, vol. XXXVI, 1973, pp. 112-14 and 128, for a more
detailed discussion.
44 Part I: History and Mystery
evermore delights hearts’ (
kanjifatun tajlibu al-
surra wa lam tazal tashrahu al-sudara). This word,
too, is of great interest etymologically. It is not in
origin an Arabic word, but is adapted from
Persian
ganjifeh; according to the prevalent
opinion,
ganjifeh in turn was borrowed from some
undetermined source. From its appearance on
the Istanbul card and in the three literary
sources, it is reasonable to suppose that
kanjifah
was the regular mediaeval Arabic word for
‘playing cards’; it is not generally used in modern
Arabic, but, in Iraq, the form
jinjifa is usual,
while
ganjifeh is the normal term in modern
Persian for playing cards of any kind. Thus in
fifteenth-century Egypt the word for ‘playing
cards’ was one derived from Persian, while in
Spain from the earliest times to the present, and
in Italy from the earliest times until the end of the
fifteenth century, the words
naipes and
naibi,
apparently of Arabic origin, were used. (The
word
naibi appears in the Sienese ordinance of
1377, and, in the forms
naibbe and
naibbi, in the
very earliest certain European reference, the
Florentine edict of the same year.)
The resemblance between the Mamluk pack
and the Italian one is too close to be coincidental:
one must have been derived from the other. The
evidence that we have reviewed — from surviving
cards, from documentary references and from
etymology — does not, even taken together,
constitute an irrefutable demonstration that
European cards were derived from Islamic ones:
we shall not have such a demonstration until
some reference is found to Islamic cards from
before playing cards were known in Europe, or
some Islamic cards are discovered that can
beyond question be dated to before that time.
But, even on the evidence that now exists, the
proponent of the thesis that the four-suited pack
was invented in Europe has to accept a series of
wild improbabilities. According to him, playing
cards must have been invented, somewhere in
Europe, in the second half of the fourteenth
century. By the turn of the century, they were
known and used in Egypt; but, since they were
there called by a name deriving from a Persian
word, they cannot have reached Egypt directly
from Europe, but must have travelled there via
Persia. And what, now, on this theory, can be
made of the Viterbo reference? If the chronicler
was right in thinking that playing cards were
already known in the Islamic world in 1379, and
in thinking that the word
naibi derived from
[2nd column]
Arabic, then either they must have made the
journey from Europe to Persia and from there to
the Arab world in a remarkably short time, or
playing cards must have been known in Europe
for a considerable time before the earliest
references to them. If, on the other hand, he was
wrong, the words
naipes and
naibi become once
more quite inexplicable; moreover, it becomes an
extraordinary coincidence that, while the Viterbo
chronicler was mistaken in deriving the word
naibi from Arabic, the word
na’ib should later
have come to be adopted as part of the Arabic
name for two of the court cards in each suit. One
cannot say that it is impossible: one can say only
that, on the evidence now available, no rational
considerations can make a thesis that demands
such improbabilities seem in the least degree
plausible.
Some considerations about the suit-signs in the
Italian and Mamluk packs, while not of great
weight in themselves, tell in the same direction,
and have been emphasised by Dr Rosenfeld,
who, to his great credit, has long argued for the
deriyation of European cards from Islamic
ones.
34 The generally curved shape of the
Swords, which, in Italian-suited packs, is what
principally differentiates them from the Batons,
makes better sense if we suppose that the suit-
sign originated in the Islamic world. Batons
seem, in turn, to be rather characterless objects
to choose for a suit-sign, whereas polo-sticks were
highly significant in the Islamic world, and were
used as emblems in Mamluk heraldry. Polo-
sticks would not have been recognised as such in
Europe, where the game of polo was unknown,
and so it is understandable that they should have
been converted into batons; but it is worth noting
that in the mid-fifteenth century ‘Liechtenstein
pack’, which has five suits, four of them the Latin
ones, the Batons somewhat resemble polo-sticks.
Particularly interesting is the word
denari still
used in Italian for the Coins suit. Playing-card
Coins have always been gold pieces, in the Italian
and Spanish packs as well as in the Mamluk one,
and indeed the name of the suit in Spain and in
Sicily is
or (‘gold’) and likewise in Portugal
ouro.
In the Istanbul pack, however, the Coins are
called
darahim, the plural of
dirham, although the
___________________
34 See particularly H. Rosenfeld, ‘Die Bezichungen der
europaischen Spielkarten zum Orient und zum Urschach’,
Archiv für Kulturgeschichte, vol. XLIII, 1960, pp. 1-36, and
‘Zur Vor- und Frihgeschichte und Morphogenese von
Kartenspiel und Tarock’, ibid, vol. LII, 1970, pp. 65-94.
Europe and Asia 45
dirham was a silver coin; the word is less odd
than the Italian denari, which properly signified
‘pennies’, that is, low-valued copper coins. As
Rosenfeld has remarked, the explanation is
presumably that the coins on the Islamic cards
originally represented dinars (
dananir in Arabic)
and were so called, and that the Arabic name
was replaced in Italian by the etymologically and
phonetically similar
denari or
danari. Since in
Egypt the gold dinar was replaced as currency by
the silver
dirham in 1346, this reinforces the
hypothesis that playing cards were known in
Egypt at a date preceding that at which there is
any evidence for them in Italy, and thdt it was in
the form of Egyptian cards that people in Italy
first encountered them.
35
____________________
35 The literature on playing cards is full of misstatements
of fact, given without the references that would enable a
persistent reader to correct them. In this book, most of these
are passed by; but a few, occurring in recent publications,
are indicated, lest they ensnare the unwary. The thesis of
the derivation of European playing cards from Muslim ones
receives quite unwarranted support from the statement in
André François, Histoire de la carte à jouer, Ivry, 1974, p. 36,
that there is in Palermo a fresco showing Arabs playing
cards, with a reference to an article by Signor Vito Arienti
in La Voce del Collezionista for February, 1973. This statement
is erroneous in every detail. In La Voce del Collezionista, anno
XVIII, no. 1, January-February, 1973, there is, on pp. 41-3,
the first part of an article, ‘La verita nella leggenda del
“Naibi”', by Signor Ferdinando Provenzano, dealing with
an Arabic manuscript discovered by him, which he
attributes to the traveller Ibn Jubair, and interprets as
referring to a card game played by Arabs in Sicily. In The
Playing-Card Information Circle, no. 26, July 1973, p. 3, its
editor, Mr Emanuel S. Newman, stated that there was on p.
41 of La Voce del Collezionista for February, 1973, an article by
Signor Arienti (a well-known collector of playing cards)
describing a fresco in Palermo showing Arabs playing cards.
In fact, there is no such article by Signor Arienti in that or any
other number of La Voce del Collezionista, and the article by
Signor Provenzano makes no mention of any such fresco; on
the basis of Mr Newman’s statement, I myself made a
determined but unsuccessful effort to find such a fresco in
Palermo. Subsequently, Mr Newman’s statement was
reprinted in Le Jolly Joker, no. 14, September, 1973, p. 17; a
letter of correction by me was printed in no. 15, May, 1974,
pp. 11-12, but too late to prevent the incorporation of the
error into François’s book. Signor Arienti is entirely innocent
in the matter: he did not write the article in question, which
in any case did not contain the alleged statement, and has
consistently disclaimed any assertion about a fresco in
Palermo. His name has nevertheless been attached to such a
claim in a widely circulated book. A correction of the error
by myself was printed in the Journal of the Playing-Card Society,
vol. IV, no. 2, 1975, pp. 32-3; since it is one of a kind likely
to become one of those spurious items of evidence endlessly
reiterated from one book to another, I thought it best to
repeat the correction here.
[2nd column]
The overwhelming probability, then, is that, in
about 1370-75, playing cards came to Europe,
very likely through Venice, from Mamluk Egypt,
where they had been known for some time. The
etymological evidence suggests that they had
reached Egypt from Persia, just as chess had
spread throughout the Islamic world from Persia,
whither it had arrived before the Muslim
conquest. From this point in the trail, however,
the evidence becomes patchy.
The type of indigenous Persian playing cards
to be found in all museums and private
collections is probably irrelevant to our enquiry.
These are a special kind usually referred to in the
literature on playing cards and card games as As
Nas cards, from the name (
as-nas or
asands) of the
game played with them. They consist of four or
five copies of each of five picture cards:
as (Ace)
or
Shir-va Khorshid (Lion and Sun);
Shah or
Pideshah (King);
Bibi (Lady);
Sarbaz (Soldier);
and
Lakat (a trifle - the card usually shows one or
two dancing girls). The different ranks are
always distinguished by background colour.
36
The game of As Nas very closely resembles
Poker: five cards are dealt to each of the four
players, and, though there is no draw, there is a
betting procedure just like that of Poker; there is
no flush, since there are no suits, and there is also
no straight, but otherwise the scoring
combinations are just like those of Poker,
__________________________
36 Examples are known of packs specially made for playing
some game in which suits are irrelevant, usually a gambling
game or game of chance, which have been modified from
ones divided into suits either by repeating some one suit or
by mixing up the suits. An example of the former are the
Japanese packs used for playing Kabu, a game of the
Baccarat type: in these, some one of the suits of a
Portuguese-derived pack has been selected and then
repeated four times; see S. Mann and V. Wayland, The
Dragons of Portugal, Farnham, 1973. An example of the latter
type is the pack made in Bishnupur, West Bengal, for
playing the game of Naksha, a game of a similar general
type in which the object is to obtain a total of 17 points. This
pack cansists of four copies of a sequence of twelve cards,
consisting of a King or saheb, riding on an elephant, a bibi or
Lady, riding on a horse, and ten numeral cards, each
showing an appropriate number of objects, the objects being
differént from card to card; see Pandit Haraprashad
Shastri, Proceedings of the Asiatic Society of Bengal for 1896, 1897,
p- 2, and R. von Leyden, ‘The Naksha Game of Bishnupur’,
Journal of the Playing-Card Society, vol. VI, No. 3, 1978, pp. 73-
82. It is possible that the origin of asandas cards is similar,
that is, that they are a derivative of some pack divided into
suits, but that the suit-signs have been suppressed because
irrelevant to the game; but if so, it is unclear what the
prototype can have been.
46 Part I: History and Mystery
including the hand known in Poker as ‘full house’
(a three and a pair).
37 Poker itself originated in
the United States in the nineteenth century. R.F.
Foster, who made a detailed study of its origins,
believed that it was a direct development from As
Nas as played with a 20-card American pack, the
Persian game having reached the United States
by way of New Orleans; the name ‘Poker’ he held
to be a mispronunciation of ‘Poque’, the name of
a French gambling game which French settlers in
New Orleans had transferred to the Persian
game when it was introduced.
38 Foster held As
Nas to be a ‘very ancient’ game, and in this he
has been followed by most later writers.
39 John
Scarne invokes the authority of a Mr Shapour
Fatemi for the preposterous assertion that it is at
least 2,000 years old in Iran, and adds that ‘some
rather ambiguous historical references suggest
that it may be 5,000 years old’.
40 After this, the
statement of Mr Alan Wykes that there are
references to the popularity of As Nas in
fourteenth-century Persian literature seems quite
sober.
41 Actually, however, it appears to be as
groundless as his purported quotation from a
diary of one of Columbus’ seamen, indicating
that the game was played by them on board ship
_____________________
37 There is a detailed and well-informed account of As
Nas by General A. Houtum Schindler, of Teheran, on pp.
928-9 of S. Culin, Chess and Playing-Cards, Philadelphia, 1897
(constituting pp. 665-942 of the Report of the U.S. National
Museum for 1896). Probably the earliest description of As
Nas published in English was that by ‘Aquarius’ (Louis
d’Aguilar Jackson) in his ltalian Games at Cards and Oriental
Games, London, 1890, pp. 58-9.
38 R.F. Foster, Foster’s Complete Hoyle, 2nd and 3rd
editions, New York, 1909, 1914, p. 208.
39 R.F. Foster, Poker, New York, 1901, p. 9.
40 John Scarne, Scarne on Cards, 2nd edition, New York,
1955,p. 231.
41 'There was a Persian rummy game of the fourteenth
century called as-nas, which has been adapted by every
nation without altering the structural basis of dealing a
hand of five cards to each of four players. (Poker and gin
rummy are the Western world’s most popular variants of as-
nas.) €olumbus’s seamen took the game from the Old
World to the New ... and one of them gives an account of it
in a journal he kept for his wife: “After the Hail Mary at
nightfall those not at watch played at cards, the game
having been showed them by sailors from Cathay ... It is a
game in which skill adapts to chance, each player having
five cards only and these to be matched in every respect to
the coat [court] cards and the as [ace].” It seems very likely
that the sailors from Cathay ... had taught Columbus’s men
as-nas. There are certainly several references to the
popularity of this game in fourteenth-century Persian
literature’, Alan Wykes, Gambling, London, 1964, p. 164.
[2nd column]
is spurious.
42 In fact, I have been unable to find
any evidence that As Nas is any older in Persia
than the earliest surviving As Nas cards, that is to
say, than the eighteenth or possibly the
seventeenth century.
43
Foster identified the first mention of Poker in
print as one ‘in Green’s
Reformed Gambler, which
contains a description of a game of poker played
on a river steamer in June, 1834.'
44 He is here
alluding to Jonathan H. Green,
The Reformed
Gambler (Philadelphia, 1858), on pp. 140-7 of
which there is a story about a game of 20-card
Poker played on that date in the course of a
steamer trip (though not actually on the steamer,
_________________
42 When I enquired of Mr Wykes, he informed me in April
1971 that his notes and manuscript for Gambling had been
sold to an American university, but did not specify the
institution fortunate enough to obtain this material. After
further enquiry, Mr Wykes stated in July 1971 that as-nas
was a game favoured by the fourteenth-century poet Hafiz,
and referred me to Muhammad ‘abd al-Ghanf, A History of
Persian Language and Literature at the Mughal Court
(Allzhabad, 1929-30), where he thought I should find
references to Hafiz playing the game; he said also that he
thought he had got the quotation from the journal of one of
Columbus’s seamen from Hakluyt’s account of William
Hawkins’s voyage to Brazil. When I informed him that both
these references had proved completely incorrect, and that a
systematic search (by myself in English translation, and by
a graduate student in the original) had failed to uncover any
reference by Hafiz to as-nas, he finally disclosed, in May
1972, that much of the research for his book had been done
by ‘professional researchers’ to whom he had simply
telephoned enquiries and put the answers into the narrative
when they had been discovered for him. I have the authority
of Professor S.E. Morison, the leading expert on
Columbus’s voyages, for saying that the quotation given by
Mr Wykes is spurious. A method of composition under
which the information supplied by a so-called research
agency is simply passed on to the reader without having
been checked by the author exposes such agencies to the
temptation of merely inventing information; readers ought
surely to be entitled to feel assured that an author has done
his best to make certain that the information to which he
attaches his name is correct.
43 I have not devoted any serious study to As Nas cards;
being convinced for a long time by the unanimous assertions
of so many writers, and particularly by the circumstantial
statements of Mr Wykes, that As Nas was very considerably
older than the earliest surviving cards, I concentrated on
trying to discover literary references to substantiate this.
Most of the As Nas cards in the museums are of the
nineteenth century, but some are of the eighteenth; there
may be some as early as the seventeenth century, but,
having neglected the question, I should not like to make a
definite assertion. It is quite certain, however, that, if there
are any seventeenth-century As Nas cards, they are rare,
and that there are none of any earlier date.
44 Foster's Complete Hoyle, 1909, 1914, p. 208.
Europe and Asia 47
but in a house at Louisville); in the copy of this
book-in the Bodleian Library in Oxford there is a
postcard sent on 9 December 1903 to Frederick
Jessel by R.F. Foster, saying that he had known
Green personally. Green was what the title of his
book implies; converted from his evil ways, he
devoted the rest of his life to campaigning against
gambling (though, according to Foster’s
postcard, he also invented the rubber-tipped
pencil). Actually, there is an earlier and much
fuller reference to Poker by Green in his
Exposure
of the Arts and Miseries of Gambling (Philadelphia,
1843). In this, after remarking that ‘there is no
mention of this game in Mr Hoyle’s treatise on
games, and I am of the opinion that it was not
used in his day’, Green says that it is usually
played, by two, three or four players, with a pack
of 20 cards, consisting of the Aces, Kings,
Queens, Jacks and 10s, though he later refers to
the game of full deck Poker, played with the full
pack of 52 cards.
45 He also includes a long story
which refers to Poker as being played as early as
1829.
46 Foster stated that Poker took a long time
to get into print, and in 1897 wrote that it ‘is not
mentioned in the American Hoyles of forty years
ago’,
47 but in this he was mistaken: what is
claimed to be the first account of the game, as
called either ‘Poker’ or ‘Bluff’, and played with
52 cards and any number of players less than ten,
is to be found in
Hoyle’s Games (Philadelphia,
1845), with a brief account of ‘20-Deck Poker’,
played with 20 cards.
48 The form with only 20
cards is indeed strikingly similar to As Nas, and,
from Green’s observations, it seems likely to have
been earlier than the 52-card version.
As Foster remarked, ‘triplets, fours and
sequences were among the earliest recognised
combinations of cards and are to be found in all
the oldest games’.
49 A great many early card
games bear out the truth of this assertion. An
example is the Italian game of Cricca, called after
the term used in it for a set of three of a kind, a
__________________
45 4th edition, Philadelphia, 1847, pp. 59, 60, 70; on p. 83
is a story concerning a game played in New Orleans in 1835.
The title of the book was changed to Gambling Exposed for the
1857 edition, but the pagination is exactly the same as in the
1847 one.
46 4th edition, pp. 20-59; the references to Poker are on
pp. 37-9.
47 Foster’s Complete Hoyle, 1897, p. 175. ‘In Cooncan
(Conquián), New York, 1913, p. ix, Foster says that ‘Cooncan,
like Poker, took a long time to get into print’.
48 Pp. 260-3.
49 R.F. Foster, Poker, New York, 1901, p. 7.
[2nd column]
term we shall later find employed in the same
sense in the Tarocco game played in Bologna.
This game was already mentioned in the sermon
of about 1470 quoted by Robert Steele, and it
was forbidden in Bergamo in 1491; there are
several sixteenth-century references to it.
50
Cricca may or may not have been identical with
the game of Gleek, as played in England and
described by John Cotgrave in 1662; this three-
handed game began, like Piquet, with the players
making scores for ‘ruff’ (having the largest
number of cards in a single suit) and then for
‘gleeks’ and ‘murnivals’ (sets of three or four of a
kind), and ended, again like Piquet, with the
hands being played out in tricks.
51 Gleek was an
importation from France, where it was known as
Glic; there it was referred to as early as 1397, and
there are several references to it from the
fifteenth century, including ones by Villon and
by King René.
52 Again, a game called in Italian
Flusso, Flussi or Frussi was also played in France
under the name Flux and in Germany as Flüssen.
Its name is cognate with our term ‘flush’, and it is
played to this day as a drinking game in
Muotatal in Switzerland, in the canton of
Schwyz, under the name ‘Fliisslen’. It, too, is
__________________
50 R. Steele, ‘A notice of the Ludus Triumphorum’,
Archaeologia, vol. LVII, 1900, pp. 189-200, contains the text
of the sermon on gaming: it cites crica as ludus trium cartularum
(a game of three cards), and goes on to say that ludit hoc ludo
dando cartulas a 3 a 3 (one plays this game, dealing the cards
in threes). For the Bergamo edict, see W.L. Schreiber, Die
Ältesten Spielkarten, Strasbourg, 1937, p. 79. In the sixteenth
century it was referred to by Francesco Berni in 1526 in his
Capitolo della Primiera, by Cardano in his book De ludo
alearum, by Garzoni in 1585 in his Piazza universale and by
John Florio in his dictionary of 1598. Berni also mentions it
in his version of Boiardo’s Orlando Inamorato, Book 3, Canto
6,stanza 53:
Sembran Costor due giocator’ di cricca,
Ch’abbian il punto tutto due in bastoni.
The punto is like the point in Piquet (that player wins who
has the most cards in any one suit); this increases the
likelihood that the game is identical with Gleek or Glic.
51 John Cotgrave, Wit's Interpreter, 2nd edition, London,
1662, pp. 365-8.
52 The 1397 reference is from an entry in the account-book
of Louis de Valois, Duke of Orleans; see W.L. Schreiber; op.
cit., p. 160. That from Villon is from Le Grand Testament, line
1705 (the Ballade between stanzas 158 and 159); see Les
Oeuvres de Français Villon, ed. Pierre Messiaen, Paris, 1946,
p. 136. The Grand Testament was written in 1461. The
reference by King René is from L’Abuzé en Court, written in
1473; see Oeuvres du Roi René, ed. le comte de Quatrebarbes,
vol. 4, Angers, 1846, p. 108. See also Godefroy, Dictionnaire de
l'ancienne langus française, s.v. ‘glic’.
48 Part I: History and Mystery
mentioned in the sermon quoted by Steele, and it
was prohibited at Ferrara in 1470, although in
the 1490s it was very popular at the d’Este court,
and it was referred to by Lorenzo de’ Medici,
who called it an ‘accursed game’ (
giuoco
maladetto); there are again many sixteenth-
century references.
52 A game called Sequence
was played in France and Italy during the
sixteenth century; if, as Schreiber believed, it is
to be identified with that known variously in
German as Quentzen, Quentzel, Quentzlen and
possibly also Schwentzlin, it figures, as a game
usually permitted, at least when played for small
stakes, in various fifteenth-century ordinances,
including one of 1443 from Augsburg and of 1448
from Balgau.
53 The recognition of threes and
fours of a kind, of flushes and sequences, as
scoring combinations, is plainly as old as the
______________________
53 See Steele, op. cit., For the Ferrara ordinance of 1470,
see G. Campori, ‘Le Carte da Giuoco dipinte per gli
Estensi’, Atti e Memorie delle RR. Deputazioni di Storia Patria per
le provincie modenesi e parmensi, vol. III, 1874, p. 124. For the
game as played in the d’Este court, see F. Malaguzzi-Valeri,
La corte di Lodovico il Moro, vol. 1, Milan, 1913, p. 575, A.
Luzio, I Precettori d’Isabella d’Este, Ancona, 1887, p. 22, and
A. Luzio and R. Renier, Mantova e Urbino, 1893, p. 64. The
game is again mentioned by Berni, Cardano, Garzoni and
Florio. For the Lorenzo de’ Medici poem, see Tutti i Trionfi
Carri, Mascherate ó canti Carnascialeschi andati per Firenze
(collected by Antonio Francesco Grazzini, called il Lasca),
Florence, 1559, p. 7, where the game is called Frussi. For
French references, see Godefroy, s.v. ‘glic’ and ‘séquence’;
E.S. Taylor, History of Playing Cards, cites another, by the
Sieur de Cholieres in 1586, and the game is also mentioned
in Thomas Varnet and Natalis Beda, La Puite dyableric,
Paris, c. 1510 (a translation of St Bernardine’s sermon on
gaming), and by Rabelais in 1534, among Gargantua’s
games. Flissen was given by Fischart in 1575 in his
translation of Rabelais; Heinrich Rausch, Das
Speilverzeichnis im 25. Kapitel von Fischarts Geschichtsklitterung,
Strasbourg, 1908, cites Charles Schmidt, Historisches
Wörterbuch der elsässischen Mundart, 1901, p. 231, as in turn
citing a text of 1524 in which the game is mentioned. As
flossen or flössn, the game is also referred to in two poems of
Hans Sachs, the first written in 1539; see Hans Sachs, ed.
Adalbert von Keller (later vols. by E. Goetze), vol. V,
Bibliothek des Litterarischen Vereins in Stuttgart, Tübingen, 1870,
pp- 31, 225. For further references and information about the
game as now played in Muotatal, see Peter F. Kopp,
'“Flüssliss” — Vom politischen Kartenspiel der Mächte zum
Trinkspiel der Muotataler’, Zeitschrift für Schweizerische
Archäologie und Kunstgeschichte, vol. 35, 1978, pp. 101-7.
54 The game of Sequence is mentioned, under various
cognate names, by Rabelais, Cardano, Garzoni, Fischart,
Florio and the Sieur de Cholières; see footnotes 45 and 48.
It is also given by Randle Cotgrave in his French dictionary
of 1611. For the ordinances relating to Quentzen, see W.L.
Schreiber, op. cit., pp. 36, 39, 45-6 and 153.
[2nd column]
fifteenth century and perhaps as old in Europe as
playing cards themselves.
One of the distinctive features of Poker is, of
course, that it is not merely a game in which such
combinations score, but that that is all there is to
the game. But games of this general type are also
extremely old; Flusso was just such a game, and
one of the most famous was the Spanish game of
Primero, immensely popular in the sixteenth
century both in England and, as Primiera, in
Italy; each player was dealt four
cards, and there were three possible winning
combinations, the lowest being the prime or
primiera (one card of each suit) and the highest
being the flush (four cards of one suit); if any
player had one of the scoring combinations in his
original hand, there was an immediate show-
down, otherwise there was a draw. In Primero,
however, the stakes were all placed before the
deal. The excitement of Poker, on the other hand,
derives from the fact that there is a competitive
raising of the stakes after the deal and the draw.
But this feature is also not original with Poker. It
is found not only in As Nas, but also in the
English game of Brag, much played in the United
States in the nineteénth century. It also
characterises the French game of Bouillotte; as
Foster observed, ‘the blind, the straddle, the
raise, the bluff, table-stakes and freeze-out are all
to be found in Bouillotte, which flourished in the
time of the French Revolution’.
55 Bouillotte has
continued to be played in France down to
modern times; it was introduced at the end of the
eighteenth century as an improvement on the old
game of Brelan, in which the stake was placed
before the deal. Brelan is another game of great
antiquity; it appears in an edict of Lille of 1458,
and was also mentioned by Villon in 1461.
56 The
term
brelan is still used in Bouillotte and in Poker,
as played in France, to mean a set of three of a
kind. In the passage quoted above from Foster,
he goes on to remark that ‘the draw from the
remainder of the pack existed in the old French
game of Ambigu’; Ambigu, much played in
France in the eighteenth century, was a
derivative from Primero, in which the draw
already featured.
From all this it is apparent that there is no
reason whatever to assume that As Nas had any
_________________
55 R.F. Foster, Foster’s Complete Hoyle, 1st edition, New
York, 1897, p. 175.
56 See the reference to Villon in footnote 47, and W.L.
Schreiber, op. cit., p. 64.
Europe and Asia 49
direct influence on gambling games of the
general type to which Poker belongs, as these
were played in Europe. Scarne, taking As Nas to
be 2,000 years old, infers that it was the direct
ancestor of the European games of this kind; in
fact, however, the latter are much older than we
have any reason to suppose As Nas to be. It is
much more probable that the influence went the
other way, that the invention of As Nas was
inspired by the example of some European game
of this type. General Schindler remarked, in
support of this supposition, that the word as is
not an indigenous Persian word, and is
presumably derived from some European word
for ‘Ace’.
57 Scarne’s theory appears to be that
Poker was an immediate development out of the
French games, and that these in turn were
descended, through Italian games, from the
Persian one. He informs us that ‘the first mention
of the game — or even the word — Poker has been
traced by the author’, i.e. Scarne, ‘to Jonathan
H. Green, a reformed gambler, who in certain
writings dated 1834 described the rules for a
game then being played on the Mississippi
steamboats’; possibly he was assisted in this feat
by Foster’s identical previous discovery,
mentioned in his
Complete Hoyle, a book quoted
elsewhere by Scarne in order to jeer at Foster’s
remarks about the game of Conquin.
58 He goes
on to say that the game did not appear in the
American
Hoyle until thirty or more years later;
as we have seen, this was also Foster’s opinion,
but it is mistaken. When the game did appear in
Hoyle, Scarne tells us, it was called, not ‘Poker’,
but ‘Bluff’, and the rules were those given by
Green, i.e. a maximum of four players using a 20-
card pack. I have not myself seen an American
__________________
57 S. Culin, Chess and Playing-Cards, Philadelphia, 1897, p.
929.
58 Scarne on Cards, New York, 1955, p. 230. I am not
suggesting that Scarne did not look up Green’s books; he
remarks, for instance, that Green noted that the American
Hoyle then current neglected to mention the game, which
must be an allusion to Green’s comment about Hoyle,
quoted above, in the Exposure of the Arts and Miseries of
Gambling (not in The Reformed Gambler cited by Foster). All
the same, he looked at them in a remarkably cursory
fashion. There are no writings of Green’s dated 1834; 1834
is the date referred to by him in The Reformed Gambler, and
later cited by Foster. Scarne makes the same trifling error as
Foster in thinking the game of 1834 to have been played on a
steamboat; he misinterprets Green’s remark about Hoyle,
which appears to concern Edmond Hoyle’s original book;
and, like Foster, he overlooks Green’s mention of the 52-
card form (possibly Foster did not know the Exposure).
[2nd column]
Hoyle in which ‘Poker’ is not the principal name
used, though the 1845 book cited above uses
‘Bluff’ as an alternative; it also treats the 20-card
version as a minor variant. Scarne asserts,
however, that it was only about forty years later,
i.e. after 1900, that
Hoyle first called the game
‘Poker’.
59
In arguing for the immediate derivation of
Poker from French gambling games, and its
ultimate derivation from As Nas, Scarne says, of
the game described by Green, that some features
are of French origin, but that its structure is
Iranian. Later he states that the earliest
published Poker rules, which, by inference from
his earlier remarks, must have appeared in the
1860s, established its French antecedents: it used
the Piquet pack, the pack was cut to the left and
dealt to the right, and the French terms
brelan
and
tricon were used; the draw was taken from the
French game of Ambigu. He goes on, in a
passage strikingly reminiscent of Foster, to state
that ‘Bouillotte, which antedated the Revolution,
developed ... the blind, the straddle, the raise,
table stakes, the freeze-out and (back to Iran!)
the bluff’. Scarne’s remarks about what he takes
to be the earliest published rules of Poker are
very surprising: I have not attempted to make a
complete search through the vast number of
nineteenth-century American
Hoyles, but I know
of none which gives a version of Poker (or Bluff)
played with the 32-card Piquet pack, or in which
the cards are dealt counter-clockwise or three of a
kind called
brelan or
tricon. Indeed, Scarne’s
claim appears to contradict his earlier statement
that the rules, as given in the earliest accounts,
were the same as Green’s. The explanation seems
to be that, as in so many other observations of
Scarne’s, he is here, without acknowledgment,
following Foster, and that, this time, he has
misunderstood his source. Foster, commenting
______________
59 The Hoyles were originally reprints of books issued in
London as later editions of Hoyle’s famous work, more and-
more games being added. After a time, the American
editions came to include games played in the United States
but unknown in England; eventually, they became quite
independent of the volumes published in London. In this
way, the word Hoyle came to be a mere generic term for a
book with rules of games, and to this day American writers
on card games frequently include it in their titles. Although,
as usual, the American Hoyles, published in Philadelphia,
New York, Baltimore, Boston, Chicago and no doubt
elsewhere, copied from one another and reprinted accounts
of games unchanged through many editions, they do not
form a single series; there were several different publishers,
whose selection of games differed somewhat.
50 Part I: History and Mystery
on a game called Whiskey Poker which he
described in his
Complete Hoyle of 1897, says that
it has ‘really little or nothing in common with the
true spirit of poker, and is simply the very ancient
game of Commerce, played with five cards
instead of three. The descriptions of this game in
the earliest Hoyles betray its French origin:
particularly in the use of the Piquet pack; the
French custom of cutting to the left and dealing
to the right; and the use of the words “brelan”
and “tricon”. In later descriptions of the “new
form” of Commerce, about 1835, we find 52 cards
are used, and dealt from left to right, and the
names of the combinations are changed to “pairs-
royal”, ‘“sequences” and “flushes”’.
60 When
Foster speaks here of ‘the descriptions of this
game in the earliest Hoyles’, he means by ‘this
game’ neither true Poker nor even Whiskey Poker,
the latter of which he may have been the first
to describe in print,
61 but Commerce, which is
distinguished from Whiskey Poker, as he says, by
the fact that each player receives only three
cards. The first ‘Hoyle’ to contain a description
of Commerce was probably the improved edition
of Charles Pigot’s
New Hoyle (London, 1805), and
the first known to me to do so, of those published
in the United States, was
Hoyle’s Improved Edition
of the Rules for Playing Fashionable Games
(Philadelphia, 1838).
62 Foster was trying to
_____________
60 Foster’s Complete Hoyle, 1897, pp. 174-5.
61 Foster does not claim priority for his description of
Whiskey Poker; the earliest account known to me in any
other book is that in The Standard Hoyle, New York, 1899, p.
147, Hoyle's Games: Autograph Edition, New York, 1914, p.
189, remarks of Commerce that it is the parent of Whiskey
Poker.
62 Foster’s observations about the old and new forms of
Commerce are inaccurate. Most of the Hoyles, both British
and American, distinguish between two forms of
Commerce, and some, including the works of 1805 and 1838
cited in the text, say that one is the new form and the other
the old one; a few, such as Hoyle’s Card Games: an improved
edition, Glasgow, 1827, describe only the ‘new’ form, without
calling it such. But it is the form with the French
terminology that is said to be the ‘new’ one. In this, a set of
three of a kind is called ‘tricon’ (the term ‘brelan’ is not
mentioned), and the deal is counter-clockwise; in the ‘old’
form, the term is ‘pair-royal’ and the deal is clockwise. The
word tricon and the counter-clockwise deal do characterise
Commerce as played in France; the latter is not the usual
French practice, so possibly the game was of Italian origin. I
have not found a mention in any of the Hoyles of Commerce
being played with the 32-card pack. I do not mean to
question the thesis that Commerce reached England in the
first place from France; but Foster has oversimplified the
history — there must have been a second wave of French
influence on the playing of this game.
[2nd column
establish the French origin of Commerce, the
immediate progenitor of Whiskey Poker, a game
which, as Foster said, was not really a form of
Poker at all, but an adaptation of Commerce to
the use of Poker hands; Scarne’s argument for
the French parentage of Poker embodies a
confusion.
Foster’s own theory was quite different. The
recognised combinations of cards were ancient;
the principle of the betting procedure was well
known from Brag, and its details may have been
derived from Bouillotte; the draw may have been
derived from Ambigu. But, for him, the crucial
new feature, distinguishing Poker from other
games of its general type, consisted in each
player’s being dealt five cards, making possible a
hand of two pairs or that of Full House. In
Primero and Ambigu, a hand consisted of four
cards, in Brelan, Bouillotte, Brag and
Commerce, of only three, as also in Flissen. The
great innovation of Poker was the distribution of
five cards to each player; and it was therefore the
origin of this that needed explaining. Impressed
with the similarity of the early twenty-card Poker
to As Nas, Foster maintained that As Nas had
reached the United States directly from Persia.
His full reasons for taking this view lie buried in'a
New York daily paper for which he was the card-
game correspondent, and in which he reported
his studies of the origin of Poker;
63 on the face of
it, a direct contact at that time between Persia
and the United States, with no European
intermediary, seems somewhat unlikely.
Whether there is any compelling reason to
assume it is unclear. Scarne, who tacitly agrees
with Foster on so many points, scornfully rejects
his derivation of the name ‘Poker’ from ‘Poque’:
‘I can’t endorse this kind of breezy scholarship,'
he declares; ‘I never played Poque, I have never
seen it played, I don’t know its structure, and I
never encountered anyone better informed on the
subject than I’.
64 Alan Wykes, likewise rejecting
________________
63 An article in the New York Sun of 22 May 1904, which I
have not seen, and which Foster refers to in the 1909 and
1914 editions of his Complete Hoyle, p. 208.
64 Scarne on Cards, 1955, p. 232. I have devoted attention to
Scarne because there is particular danger that his remarks
will mislead those readers who are unable to check the now
much less accessible works of R.F. Foster. The danger
springs from Scarne’s calculated contrast between his own
great knowledge of the subject and the ignorance of
supposed experts, conveying the impression that the latter
have contented themselves with ill-informed guesses, while
he has engaged in genuine research. It is plain that a large
Europe and Asia 51
the derivation, tells us that ‘such information as
there is about poque suggests that it was a game
of the whist family’.
65 Scarne’s challenge is
childishly simple to meet, and Wykes’s statement
to refute, since Poque is still described in current
French card-game books; the earliest description
of it of which I know is from 1714.
66 Played with
the 36-card pack, Poque has several extraneous
features: there are fixed stakes, advanced before
the deal, to be gained by the holder of each of the
top cards of a suit to be determined by turning a
card, and, after the show-down, a form of ‘Stops’
play takes place (of the kind that occurs in a game
such as Newmarket). But the core of the game has
a fair resemblance to Poker: betting, after the
deal, on who has the best combination of cards,
the players standing, raising or folding {(dropping
out). The combinations recognised are single
pairs, two pairs, threes and fours of a kind: the
crucial similarity lies in the fact that each player is
dealt a hand of five cards. Foster’s idea was that
French settlers in New Orleans recognised the
similarity of the newly introduced As Nas to their
own game of Poque, and transferred the name to
it; but an alternative possibility is that Poker was a
direct development out of Poque, arrived at by
suppressing the extraneous features and
introducing new winning combinations. In the
latter case, the resemblance to As Nas must have
been accidental; readers must judge whether the
twenty cards and the Full House together form too
great a coincidence. Poque itself, though not very
ancient, is closely related to the German game of
Poch or Pochen, whose existence Scarne does
recognise; if the conjecture of Schreiber is correct,
that the card game prohibited under the name
bocken in many German municipal ordinances of
the fifteenth century is to be identified with
Pochen, it, too, is among the earliest European
games.
67
_________________
part of Scarne’s researches into the history of Poker
consisted in an excessively hasty reading of Foster. Foster
himself, though no more incapable of making mistakes than
any of us, really was a scrupulous and painstaking
investigator of the history of card games, one of the very few
such in that field. It would be pity if his conclusions were to
be obscured or his reputation tarnished by groundless attacks
or by others taking the credit for his discoveries.
65 Alan Wykes, op. cit., p. 171.
66 See, e.g., Frans Gerver, Le Guide Marabout de tous les jeux
de cartes, Verviers, 1966, pp. 240-2; the 1714 account is from
the Académie universelle des jeux for that year.
67 For a modern description of Pochen, see Roland Gööck,
Freude am Kartenspiel, Gütersloh, 1967, pp. 64-6. For bocken
[2nd column]
As remarked, without a knowledge of Foster’s
detailed evidence, it is impossible to form a
judgment about the theory that As Nas was the
direct progenitor of Poker. Whether that theory
is sound or not, it is plain that As Nas and As
Nas cards tell us nothing, on the evidence now
available, about the transmission of playing
cards and of card games from the Islamic world
to Europe; until any contrary evidence comes to
light, it would be helpful if writers on the subject
would cease to reiterate the baseless assertion
that As Nas is of great antiquity. A different type
of Persian playing cards, much more important
for our purpose, is a great deal more elusive. In
his
Voyages en Perse, published in 1686, the
Chevalier Chardin wrote, ‘There are cards
among the common people, which they call
ganjaphé (
kandjaféh). They are made of wood; very
well painted. The pack is of ninety cards with
eight suits’.
68 Chardin was evidently not very
interested in these cards, and his statement must
involve some error, since 90 is not divisible by 8.
Romain Merlin, in his book of 1869, asserts that
the Persians had a 96-card pack, known as
Ghendgifeh, divided into eight suits of twelve cards
each, and illustrates, as examples of such Persian
cards, one from a pack in his own collection.
rectangular and made of ivory, stating that other
examples are in the Douce collection and in that
of a M. Didot. He adds, rather oddly, that they
generally have a round shape among the Persians
of India, but a shape like European cards among
the Persians of Teheran.
69 D’Allemagne, the
great historian of French playing cards, says in
his travel book about Persia, published in 1911,
that there are two types of Persian playing cards,
the As Nas cards and a 96-card pack with eight
suits of twelve cards each. The latter, he says,
are of Indian origin, but the Persians have sub-
stituted a rectangular shape for the circular one
used in India; the oldest Persian cards are painted
on ivory, and several specimens of' these have
survived, but later ones are made of lacquered
cardboard. In illustration, he reproduces two
_______________________
see W.L. Schreiber, op. cit., pp. 36, 38-42, 44, 148; alsogthe
edition of Hans Sachs cited in footnote 53, vol. V, pp. 31 and
47, and the work of Heinrich Rausch cited in the same
footnote.
68 Vol. III, chapter 12, ‘Des exercises et des jeux des
Persans’, of the section ‘Description générale de la Perse’, p.
451
69 R. Merlin, L’Origine des cartes à jouer, Paris, 1869, plate 69
and pp. 122-4.
52 Part I: History and Mystery
plates from his book
Les Cartes à jouer, showing
different packs, both said to be made of ivory; the
first is assigned to the nineteenth century and
said to be reproduced from Chatto’s book, and
the second to the eighteenth or nineteenth
century, with an unstated owner.
70 The mention
of Chatto is a mistake: the cards are actually
reproduced from S.W. Singer’s fine book of 1816,
where they are captioned ‘Oriental Cards on
Ivory in the Cabinet of F. Douce Esq.'
71 The
Douce collection is now in the Bodleian Library
at Oxford, but no longer contains the cards
illustrated by Singer and D’Allemagne, though it
has some round ivory cards from a 96-card pack,
and, in a box labelled ‘Persian Cards’, some
rectangular lacquered cardboard ones very
similar to, though not identical with, those
shown in D’Allemagne’s second illustration;
Francis Douce lived from 1757 to 1834. The
cards shown in D’Allemagne’s second
illustration, like the rectangular ones now in the
Douce collection, have a decorative arch at the
top of each card; the same is true of a single
rectangular numeral card, of lacquered
cardboard, in the Fournier Museo de Naipes in
Vitoria, listed in the catalogue as Persian.
72
These arches are lacking in the ivory cards
illustrated by Merlin, and likewise in those
illustrated by both Singer and D’Allemagne; it is
very probable that D’Allemagne was mistaken in
saying that those shown in his second illustration
were of ivory.
The 96-card pack is well known as one of the
types of indigenous playing-card packs, usually
known generically as
ganjifa, used in India: it is
still being made, although all the traditional
types of playing cards and of card games have
been to a large extent ousted by those of
European origin. Traditional Indian playing
cards are all hand-painted, and are usually
circular. Merlin and D’Allemagne were wrong,
however, to imply that a rectangular shape
indicates a non-Indian origin: there are plenty of
surviving rectangular Indian Ganjifa cards. In
particular, the card in the Fournier Museum,
those now in the Bodleian Library and those
shown in D’Allemagne’s second illustration all
_______________
70 Henry-René D’Allemagne, Du Khorassan au pays des
Backhtiaris: trois mois de voyage en Perse, vol. I, Paris, 1911, pp.
163-5, and Les Cartes à jouer, vol. 1, Paris, 1906, pp. 4, 8 and 9.
71 Samuel Weller Singer, Researohes into the History of Playing
Cards, London, 1816, facing p. 16 and p. 49.
72 Catalogue no. Persia-4.
[2nd column]
closely resemble 96-card packs made in the
Deccan and illustrated in Rudolf von Leyden’s
catalogue of Indian cards in the Spielkarten-
Museum at Leinfelden.
73 Furthermore, the style
of the ivory cards shown by Merlin and by Singer
and D’Allemagne appears to be Moghul rather
than Persian. It seems very likely, therefore, that
all the cards from 96-card packs claimed as
Persian are in fact from India; it is noteworthy
that D’Allemagne, who was of course deeply
interested in playing cards, does not claim that
he actually saw any such cards during his
Persian travels. I am unsure whether the game of
As'Nas is still played in contemporary Iran, but
it is certain that no 96-card pack is any longer to
be found there; European playing cards have
been adopted, and the word
ganjifeh or
ganjifeh
ordinarily refers to them. In face of this, one
might well doubt whether the 96-card pack was
ever known in Persia. Nevertheless, as we shall
see, such doubts would be misplaced: the
statements of Merlin and of D’Allemagne that
such a pack was in use in Persia are perfectly
correct, and can be supported by more than the
flimsy testimony of Chardin. It is strange that, so
far as I am aware, no cards from a Persian 96-
card pack have been preserved.
Since they have not, it is best to start with ones
of which there is a plentiful supply, traditional
Indian cards. With the negligible exception of the
cards from Mysore known as Chad cards, of the
special Naksha cards mentioned in footnote 36,
and of certain packs based on Portuguese
prototypes, all traditional Indian packs have an
even number of suits, each consisting of ten
numeral cards and two court cards. There are
three main types: one with twelve suits, whose
iconography relates to the famous epic, the
Ramayana; the
Dasavatara pack, with ten suits,
each representing one of the incarnations of
Vishnu; and the 96-card pack, with eight suits,
used principally by Muslims and sometimes
referred to as the Moghul pack, with a purely
secular iconography.
74 The word
ganjifa, or some
corruption of it such as
ganjpa, is commonly used
in India to refer to playing cards of all three
types, and, equally, to the principal family of
games played with them. In all the types, each
card has a suit-sign, and the numeral cards, from
______________________
73 R. von Leyden, Indische Spielkarten, Leinfelden-
Echterdingen, 1977; see, e.g., no. 42.
74 Nowadays a twelve-suited pack is made in Orissa
consisting of the Dasavatara pack with two additional suits.
Europe and Asia 53
1 to 10, are pip cards, their values being
indicated by the number of occurrences of the
suit-sign; often the suits are further distinguished
by background colour. Of the two court cards,
the higher normally represents a King, and is
called
Padshah, Mir or
Raja; in the
Dasavatara
pack, it represents an incarnation of Vishnu, and
is accordingly often called
Avatara. The lower
court card is called
Wazir (Vizier) or
Pradhan
(Minister).
Although these Indian packs have a larger
number of suits, a smaller number of court cards
per suit, and different suit-signs from the
European and Mamluk packs, they qualify as
regular packs under the definition of that term
given earlier. Despite the obvious differences, the
Indian packs are so closely related to the
European ones that a connection is undeniable:
the question is only what it is. The existence of a
connection between European cards and Indian
ones becomes even more evident when we look at
the family of games generally known as Ganjifa,
the more important of the two main families of
Indian card games. The other such family,
known generally as Naksha, comprises games of
chance, of the same general type as Vingt-et-Un,
in which the player’s aim is to come as close as
possible to a total value of 17 points on the cards
he holds. A preliminary report on these games
was given by Dr Kaushal Gupta in 1978; they are
often played with ordinary Indian playing cards,
and Dr Gupta has discovered a reference to them
in the
Humayun-Nameh, relating to the year 1545,
in which, perhaps significantly, the word used for
‘playing cards’ is not
ganjifeh but
waraq. In
Europe, perhaps the oldest known game of this
general type is Thirty-One, a game also
mentioned in the Steele sermon and by Hans
Sachs. Though such games are immensely
widespread, being known also in Japan and
Korea, their history has been so little studied that
no deductions can be made concerning the
direction of influence between these various areas.
The Ganjifa family, however, consists of trick-
taking games played without trumps and sharing
a number of special features. Whichever type of
pack is used, one invariable feature is the same
peculiarity of card order that we noticed as
prevailing in fifteenth-century Italy and Spain. In
all the suits, the Kings rank highest and the
Ministers the second highest; but, in half the
suits, the 10 follows the Minister, and then the 9,
and so on down to the Ace, whereas, in the other
[2nd column]
half, the Ace follows the Minister, and then the 2,
and so on down to the 10. It is absolutely
unthinkable that so bizarre a practice should
have arisen independently in different parts of
the world.
Playing cards can have entered Europe only
with one or more games to be played with them.
In view of the dominant position occupied,
among European card games other than pure
gambling games, by those of the trick-taking
category, it would be natural, in any case, to
suppose that trick-taking games were among
those introduced together with playing cards
themselves. Moreover, despite their enormous
variety, almost all trick-taking.games, not only in
Europe, but also in China and in India, have
certain common features which suggest that they
are all offshoots from a single stem. One such
feature; all but invariable, is the rule that the
winner of a trick leads to the next trick. We are so
used to this that we hardly notice that other
possibilities are quite conceivable. The lead
might steadily rotate, from trick to trick,
irrespective of who won the tricks, just as, in most
European games, the deal rotates, from round to
round, irrespective of who won the preceding
round. Or, again, a player might continue to lead
as long as he won, and, as soon as he lost, the
lead might then pass to the next player in
rotation, regardless of who had won the last
trick: this is the rule for the deal in most Chinese
games. But we never find either of these perfectly
possible rules: the rule is always that the winner
of a trick leads to the next one. Another nearly
universal feature is that, unless a trump is
played, or when there are no trumps, only a card
of the suit led can win the trick. Not only can
we imagine different arrangements, but they are
occasionally found: for instance, in the games of
le Truc and Putt, played respectively in France
and, formerly, in England, only the rank counts,
irrespective of suit, and, if there is no card of rank
higher than the rest, no-one wins the trick; this is
also in large part the rule in the game of Aluette,
played on the western coast of France, and in
that of el Truco, played in Spain and Argentina.
These very rare exceptions serve only to point up
how constant is the usual rule, not only in
European games, but in Chinese and Indian ones
as well. The persistence of these common
features makes it probable that trick-taking
games are ultimately all descended from some
common ancestor.
54 Part I: History and Mystery
The occurrence, both in India and in Europe,
of the curious practice of ordering the numeral
cards in one direction in half the suits and in the
reverse direction in the other half shows a
connection, not merely between the playing
cards of the two areas, but between the card play
of Europe and of India; and, since the principal
Indian games are trick-taking ones, the
connection holds within this category of card
games. Indeed, it is only for trick-taking games
that it is necessary to have a ranking of the cards
within each suit. Admittedly, for a game in which
a player scores for having a sequence of three or
more cards within a suit, some ordering is
necessary; but it need only be a cyclic order — we
have to know which cards are adjacent, but need
have no conception of one’s ranking
higher than
another. There is, indeed, an exception in those
games in which only the player with the highest-
ranking sequence of a given length can score; but
it is precisely in such a case that it becomes
impossibly inconvenient to have the ordering
running in different directions in different suits.
When the 10 ranks high in Swords and low in
Coins, it would be very difficult to recognise at a
glance that the sequence 3, 4, 5, 6, of Coins was
higher than the sequence 7, 6, 5, 4, of Swords.
Any reference to the ranking of the cards found in
early sources may thus be taken as testimony to
the prevalence of trick-taking play.
We can be certain that the connection between
European and Indian playing cards and card
games did not consist in a
direct derivation of
either from the other. Trick-taking games existed
in Europe long before the first direct contact with
India in 1498. Conversely, although Indian
versions of the Portuguese-suited pack, retaining
its three court cards per suit, were occasionally
produced, the 96- and 144-card packs were
known in India from at least the time of Akbar,
and, as we shall see, the former was known in
Persia earlier still; there simply is not time for the
one to have developed out of the other, by
multiplying the number of suits and suppressing
one court card from each suit, with no
consciousness of the recent date of the innovation
or of its derivation from a European prototype.
Moreover, by the sixteenth century the idea of
trumps had been adopted for trick-taking games
in most of Europe, an idea wholly lacking from
Indian Ganjifa games. The resemblance between
European and Indian playing cards and card
games must therefore be explained either by an
[2nd column]
indirect derivation of either from the other, or by
their both being derived from a common source.
Even if we knew nothing of playing cards in
Mamluk Egypt or in Persia, this would imply
their presence in the mediaeval Islamic world, as
the only possible route of transmission to or from
Europe, whichever hypothesis was adopted.
In fact, the Indian Ganjifa games are too
unlike European ones to give any basis for a
hypothesis of the indirect derivation of either
from the other. They share the common feature
of being trick-taking games, and the peculiarity
of card order; but the Indian games have several
features that make them utterly different from
any game known in Europe. The most striking of
these is the universal rule that a player who holds
the highest card not yet played in any suit,
known as
hukm, is bound to play it to a trick to
which a card of that suit is led, on pain of its
demotion to be the lowest card of the suit. This
rule obliterates all the opportunities for tactical
play of the kind that characterises European
trick-taking games, and makes the Indian games
a test of memory much more than of strategic
skill. Another constant feature is the existence of
strong, and often complicated, constraints on
what card may be played by the player who leads
to a trick. In European games, there are usually
constraints, such as the obligation to follow suit,
on the cards which may be played to a trick by
subsequent players; but, with rare exceptions,
the player with the lead may lawfully play any
card that he chooses.
75 Yet another common
____________________
75 A few games have minor constraints on the lead, such
as not leading a trump to the first trick; almost the only one
with strong constraints was the nineteenth-century German
game of Fünfzehnern, a four-handed game without trumps
played with the 32-card German-suited pack. Each player
played for himself and received eight cards; in each suit, the
cards ranked, in descending order, Deuce, King, Ober,
Unter, 10, 9, 8, 7, and the top five cards had, respectively,
point-values of 5, 4, 3, 2 and 1. A player received or paid in
proportion to the number of points he had on cards won in
tricks over or below 15. A player who won a trick was
obliged first to lead out all his master cards and then to lead
again the suit of that trick by winning which he first gained
the lead, if he could. Some also required that, if he could
not, he must then instead lead that suit of the previous trick
in which the lead changed hands. (Suppose that A leads
Leaves to the first trick, and B wins. B now leads two master
cards, the Deuce and King of Hearts; having no master
cards left, he must now lead Leaves again if he can. If he
cannot, he may lead any suit; say he leads Acorns, and C
wins. C, having led out his master cards, must then lead
Acorns again, if he can. Under the strict rule, if he cannot
lead Acorns, he must now go back to Leaves, if he can,
Europe and Asia 55
feature of Indian Ganjifa games is the way the
game begins: the player who holds the King of a
particular suit leads it, together with a low card,
to a double trick, to which the other players all
play two cards; this double trick is automatically
won by the player who led to it, the low card,
called the ‘throne’ of the King, acquiring the
power to beat any other card. This mode of
opening the game is, again, quite unknown in
Europe, where, in trick-taking games
without bidding, the lead to the first trick
usually belongs to the player next in rotation
after the dealer. A number of writers assert that
Ganjifa games resemble Ombre; but, apart from
their both being trick-taking games, and sharing
the peculiarity about the ranking of the cards in
different suits, there is no similarity at all:
Ombre has both bidding and trumps, which
Ganjifa lacks.
In view of this, there is no reason to postulate
any European influence on the games played in
India with the eight-, ten- and twelve-suited
packs; and it is equally doubtful that those games
are even the remote ancestors of the games
known in Europe. Rather, it seems likely that the
games played in India represent a development,
along a different branch, from some more basic
trick-taking game that formed the common
ancestor of the Indian games and of European
games of this category. This does not in itself rule
out the possibility that this remote common
ancestor originated in India.
It is to be presumed that trick-taking games,
probably along with card games of other types,
entered Europe from the Islamic world together
with playing cards themselves. And since, as we
have seen, European playing cards and card play
probably had no influence on the Indian many-
suited packs and the games played with them,
the coincidence about the peculiarity of card
order is most reasonably explained by supposing
that that practice, too, was derived by European
__________________
because that was the suit of the last trick, other than the
trick in Acorns which C won, in which the lead changed
hands.) Finally, if a player held the King and Ober of some
suit, but not the Deuce, he might, when he was legally
entitled to lead that suit, lead the Ober, saying ‘Jwang’
(‘Obligation’): the player holding the Deuce was then
obliged to play it to that trick. This did not apply if he led a
card lower than the Ober or chose not to say ‘Jwang’.
Particularly in the rule requiring a player to lead out his
master cards (a rule not applying to the first trick, however),
this game has a certain resemblance to Ganjifa; but this is
presumably fortuitous.
[2nd column]
card-players from that of Muslim ones. We have
no evidence whatever, at present, for the kind of
card game played in Mamluk Egypt; but one
feature of the Istanbul pack does partially
corroborate the supposition that, in it, the
numeral cards ranked in different orders in
different suits. This is the fact that, while, in
every suit, all the court cards have upper
inscriptions with verses, in Swords it is the three
highest-numbered numeral cards that do so,
while in Coins and Polo-Sticks it is the three
lowest-numbered ones that do (assuming that
the missing Ace of Coins did, or, as is more likely,
that it is the original Ace of that suit that has
been converted into the Viceroy). In the Cups
suit, admittedly, no numeral cards bear upper
inscriptions; but, since most of the inscriptions
suggest that the cards bearing them are good
ones to have, it seems quite probable that, in
Coins and Polo-Sticks, the numeral cards ran
from Ace (high) down to 10 (low), while in
Swords and, presumably, Cups, they ran from 10
(high) down to Ace (low). It has to be conceded
that this does not completely tally with the
practice in Europe with the Latin-suited pack, in
which it was in the suits of Swords and Batons
that the higher-numbered cards beat the lower-
numbered ones, and in Cups and Coins that the
opposite happened; but perhaps this alteration is
one that could fairly easily have occurred in the
course of transmitting a new kind of game.
As we saw, the use of the word
kanjifah for the
Mamluk cards strongly suggests that playing
cards came to Egypt from Persia. We have also
seen that some writers have associated the 96-
card eight-suited pack with Persia as well as with
India. Certainly, this pack, as used in India, has
Persian connections; not only is it sometimes
referred to as the Moghul pack, but the very
names of the suits, in the most usual form of the
pack, are Persian words. These are:
sefid (white);
taj (crown);
gholam (slave);
shamsher (sword);
surkh (red);
chang (lyre);
bharat (document or
cheque); and
gimask (goods or cloth). The names
sefid and
surkh are sometimes replaced by other
words meaning ‘coin’, and the suit-signs on these
two suits are a silver and a gold disc respectively.
We find exactly this eight-suited pack mentioned
in Persian sources. Ettinghausen has drawn
attention to two articles on the subject in Persian
journals.
76 In one, by Dr A.N. Manzavi, there is
_____________________________________
76 R. Ettinghausen, op. cit., pp. 75-7. The articles are:
56 Part I: History and Mystery
printed a set of 96 quatrains by the poet Ahll
Shirazi, who died in 1536, one for each of the
cards in a pack of
ganjifeh: the composition of the
pack is exactly that known to us from the eight-
suited Indian packs. Manzavi also quotes from
two seventeenth-century Persian writers, Mirza
Sadiq and Tughrai Mashhadi, who wrote about
the cards and the games played with them, but
stated that the popularity of these games was
then waning in Persia; at least one of the games
cited is known from later Indian sources:
According to another article quoted by
Ettinghausen, by Dr B.T. Khani, the game was
formerly believed to have been invented by Mir
Ghiyath ad-Din Mansur Shirazi, who died
between 1533 and 1543, and was popular in the
reign of Shah ‘Abbas I (1587-1628), but forbidden
by Shah ‘Abbas II (1642-1667).
It is here that there is a big lacuna in the
evidence now available to us. It is certain that the
96-card pack was in use in Persia in the early part
of the sixteenth century; it is also clear that it
was known in India during the reign of the
Emperor Akbar the Great (1556-1605). Indeed,
the whole Moghul court was predominantly
Persian in cultural orientation, and Persian
writers and artists were regularly accustomed to
visit the Imperial court. We have as yet no
definite evidence that playing cards were known
at all in India before the sixteenth century; but it
is likely that they were known there well before
that. As we have seen, it is also probable that, by
the first half of the fourteenth century or earlier,
playing cards had, entered Egypt from Persia.
The 52-card Mamluk pack, with its three court
cards per suit and its four suits, can, however,
hardly have been derived from the 96-card pack,
with its two cards per suit and eight suits. It
looks, rather, as though, behind both the 52-card
pack and the 96-card one lay some earlier
evolutionary stage of which no trace has so far
been discovered: say a 48-card pack with four
suits and only two court cards in each suit,
which, in the one case, was enlarged by the
addition of a third court card to each suit, and, in
the other, by doubling the number of suits. There
is no reason to assume that the word
ganjifeh
_________________________
‘Ali Naqi Manzavi, ‘Ganjafeh-ye AhIi Shirazy’, part 1,
Danesh, no 4, April 1953, pp. 221-8, part 2, ibid, no. 5,
September 1953, pp. 305-8; 'Ali Naki Manzavi, ‘Ganjafeh-ye
Shahid-e Sadiq’, Danesh, no. 8, October 1954, pp. 459-60;
Behruz Taqi Khani, ‘Ganjafeh’, Yaghma, vol. 13, 1960, pp.
296-300.
[2nd column]
always relates to a pack with eight or more suits;
from the fact that the word was borrowed to
apply to the Mamluk cards, it is quite likely that
it was earlier used in Persia for some pack with
only four suits.
77 There is, for example, a famous
passage in the
Babur-Nameh, which is written in
the Chagatai dialect of Mongol, telling how in
1527 the first Moghul Emperor Babur sent a
pack of
ganjifa as a present; Babur was Turkish
by descent and aspired to be Persian in culture.
78
In view of the date, this is very likely to have been
a 96-card pack; but we cannot take this for
granted.
The hypothesis that the 52-card Mamluk
pack, as introduced into Europe, resulted by the
addition of an extra court card from an earlier
48-card pack gains in plausibility from the fact
that, while the title ‘Viceroy’ (
na’ib malik) was a
standard one under the Mamluk Sultanate, the
title ‘Second Viceroy’ (
thani na’ib or
na’ib thani)
does not appear to have been used for any actual
officials; it looks very much like a completely
unimaginative piece of nomenclature adopted at
a time when it was desired to add a third court
card to the existing two. A further consideration
of the Istanbul pack prompts a more radical
speculation. The cards now completely missing
from the set are the Ace and 6 of Coins and the
Second Viceroy and 2 of Swords. The secondary
cards which have been used to replace cards from
the primary pack lost at an earlier stage are the
Second Viceroy of Coins, the 9 and 10 of Swords
and the Second Viceroy and Ace of Polo-Sticks.
Thus only the suit of Cups retains all thirteen
cards of the primary pack. At first glance, the lost
primary cards, which need not, and probably
were not, all lost at the same time, seem a fairly
random selection, although it might strike one as
________________________
77 R. Ettinghausen, op. cit., p. 75, cites a quotation given
in the Nezam Dictionary of Sayyid Muhammad ‘Ali from a
fifteenth-century Persian work, the Javaher-Nameh of
Muhammad ibn Mansur, of a passage speaking of ganjafeh as
being carved from turquoise. If this were correct, it would be
the earliest Persian reference to ganjifeh so far known; but I
am informed by Dr K. Gupta that the word used was not
ganjafeh and did not refer to playing cards (and also that the
material was actually lapis lazuli)
78 Memoirs of Zehir-ed-Din Muhammed Babur, translated by
J. Leyden and W. Erskine, Oxford, 1921, vol. 2, p. 317. The
passage, which related to June, 1527, says that on the night
when Babur’s party left for Agra, ‘Mir Ali Korchi was sent
to Tatta to Shah Hassan’; Tatta is some fifty miles East of
Karachi. It goes on to say that Shah Hassan was ‘very fond
of ganjifa’, and had asked for some packs, which Babur sent
to him.
Europe and Asia 57
odd that as many as three Second Viceroys had
been lost-at one time or another. However, since
it is quite apparent that, originally, the court
cards of the primary pack had no lower
inscriptions giving their rank and suit, and since
the cards now inscribed as the Viceroys of
Swords and Polo-Sticks bear nothing that could
possibly have served as an emblem to mark their
rank, we may regard it as virtually certain that
these cards originally served as the 9 and the Ace
of their respective suits. Moreover, although the
present Viceroy of Coins does bear a decorative
design that could conceivably have been an
indication of its rank, this is not at all the same as
that which appears on the Viceroy of Cups (and,
in a smaller form, on the Second Viceroy of
Cups); it also differs from the Viceroy of Cups in
having only one suit-sign instead of two. It is
therefore very probable that this card was
originally the Ace of Coins. A plausible further
guess is that, just as the 9 of Swords, as bearing
an upper inscription, was pressed into service as
the Viceroy of Swords, so the original 10 of
Swords was converted into the Second Viceroy,
requiring the 9 and 10 of that suit to be borrowed
from another pack. If we make all these
suppositions, it follows that the cards which
needed in the first place to be supplied, either by
direct borrowing or by converting an existing
card and then borrowing to replace it, were the
Viceroy and Second Viceroy of each of the suits
of Coins, Swords and Polo-Sticks. (On this
supposition, subsequent loss must account for
the absence of the primary 6 of Coins and 2 of
Swords, a secondary Ace of Coins to replace the
converted one, and the primary 10 of Swords
converted into a Second Viceroy. Mayer reported
the presence of a very damaged and
unidentifiable card, which does not appear to
have been seen since. The 4 and 10 of Coins have
turned up since Mayer examined the pack.)
Now it is too much of a coincidence to suppose
that it was just the two lower court cards of three
of the suits that happened to be lost. We are
virtually driven to the conclusion that they were
not present in the original pack at all, but were
added later, in a rather clumsy way, when a
change of fashion had brought in 52-card packs
with three court cards in each suit. This would
give us, slightly oddly, a pack originally
composed of 46 cards, with three court cards in
the Cups suit, but only one in each of the other
three suits. Bizarre as this may seem, it might
[2nd column]
explain the absence of upper inscriptions on any
of the numeral cards of the Cups suit: originally,
there were upper inscriptions on the top three
cards — the court cards - of this thirteen-card
suit, and on the top four cards — the King and the
three leading numeral cards — of each of the
other, eleven-card, suits. Another explanation is
indeed thinkable. This is that the pack, in its
original state, comprised only 44 cards, with only
one court card in each suit, and that the work of
supplementing it was undertaken in the first
place by the original artist, or by one skilled
enough accurately to imitate its style, who died
before completing more than the two additional
court cards of the Cups suit. On this supposition,
the plan of the painter who made the Viceroy and
Second Viceroy of Cups, to indicate the rank of
the new court cards by emblems in the same way
as had originally been done for the Kings, was
abandoned after his death, and the less subtle
method of inscribing their rank and suit adopted.
Since the Islamic cards introduced into Europe
in about 1375 must have had three court cards in
each suit, this might suggest that the primary
pack was originally made somewhat earlier, say
in the first half of the fourteenth century; but I
hesitate to challenge Mayer’s expert judgment
assigning it, on stylistic grounds, to the fifteenth.
That the expansion of this particular pack to one
containing three court cards in each suit might
have been undertaken under Turkish auspices
could be suspected in the light of the fact that the
lower inscriptions, giving rank and suit, on the
(primary) Viceroy and Second Viceroy of Cups
are preceded, respectively, by the words
Qushqult
and
Qaraja, said by Mayer to be Turkish
personal names, and the fact that the lower
inscriptions on the (secondary) Second Viceroys
of Coins and Polo-Sticks are in ungrammatical
Arabic, omitting the article before the suit-name
(the equivalent of writing, say, ‘Jack Hearts’).
There is, however, no other evidence known of
the use of playing cards by the Turks in
mediaeval times. In any case, it is clear that the
Istanbul pack has not yet yielded up all its
secrets.
We have left the realm of solidly-based
deduction and entered that of speculation. This
much seems certain: that the 52-card Mamlik
pack and the 96-card Persian and Indian one
must be connected, and that neither can have
been immediately derived from the other. On the
face of it, there are two possibilities. One is that
58 Part I: History and Mystery
the regular pack was invented in India, perhaps
in a twelve- or ten-suited form, and that, from
this, every other form of the regular pack was
ultimately derived. On this hypothesis, the pack
was reduced to eight suits by the Muslims in
India, travelled from the Moghul court to Persia,
and was then further reduced to a 48-card, four-
suited, form in Syria or Egypt, later expanded to
a 52-card pack by the addition of an extra court
card. The alternative is that already suggested:
an archetypal 48-card pack, spreading from
Persia westwards to Egypt, where it was
increased to 52 cards, while, in its Persian
homeland, the number of suits was subsequently
doubled to yield the 96-card pack, which then
later spread to India. On this latter hypothesis,
the ten- and twelve-suited Hindu packs would be
the result of a yet further expansion of the pack
that took place in India. This second hypothesis
need not involve that Persia was actually the
birthplace of the regular pack: it could have
arrived there from some other area again, say
somewhere in Central Asia, perhaps originally in
a 44-card form with the King as the only court
card.
The crucial question concerning Indian
playing cards thus comes to be whether the eight-
suited Muslim pack is a demythologised form of
the larger Hindu packs, or whether the latter are
Hinduised versions of it. On the first alternative,
playing cards were indigenous to India, and were
transformed by the Muslims, encountering them
there, into a type palatable to themselves and
stripped of the Hindu iconography they would
have found offensive. To accept this alternative
would therefore be to embrace the first of the two
hypotheses outlined above. On the second
alternative, playing cards were introduced into
India by the Muslims, and later adapted by the
Hindus to forms employing a more congenial
symbolism: to accept this alternative is to
embrace the second of our two hypotheses.
Highly relevant to this question is a celebrated
passage in the
A’in-i Akbari of Abu Fazl 'Allami
which forms the earliest known Indian reference
to playing cards.
79 In this, Akbar is first
described as having used a twelve-suited pack
with which to play card games. This pack is said,
vaguely, to have come down from ‘ancient sages’.
Its twelve suits are listed in detail: they do not
____________________
79 Adu Fazl ’Allami, A’in-i Akbari, translated by H.
Blochman, Calcutta, 1873, vol. T, pp. 306-8.
[2nd column]
correspond to those of any otherwise known form
of pack. The iconography of this packis largely
secular, but it seems to contain both Hindu and
Muslim elements, as the King of one suit is said to
be the Hindu god Indra, and that of another to be
King Solomon. The passage then goes on to
describe alterations to the pack said to have been
instigated by Akbar himself. This is always
interpreted to mean that he reduced the number
of suits from twelve to eight, although the
passage reads, rather, as though he merely
substituted different suits for the last eight of the
twelve-suited pack. There is, however, very good
reason for the usual interpretation, since the suits
whose invention is ascribed to Akbar correspond
exactly to those of the standard eight-suited form
that were listed earlier. Since, however, we know
that Al Shirazi, who died twenty years before
Akbar, at the age of fourteen, came to the throne,
wrote a poem about the 96-card pack with just
these suits, we can be certain that the claim that
Akbar invented them is without justification. The
evidential value of the passage for the theory that
the eight-suited pack arose from a reduction of
the number of suits in an indigenous Indian
twelve-suited pack is therefore slight. We may,
indeed, conclude that, in the second half of the
sixteenth century, a twelve-suited pack was
known in India, and was of reputed antiquity;
but this does not help greatly to solve our
problem.
There simply does not seem at present to be
enough evidence to settle the question definitely
one way or the other; there is only, as it seems to
me, a slight but significant tilt in the balance of
probabilities. The evidence for a Hindu origin of
playing cards is extremely weak. One writer on
the subject, Shastri, reports a tradition that the
Vishnupur variety of the Dasavatdra pack was
invented by the Malla kings of Vishnupur (the
Malla dynasty ended in 1201), and supports this
tradition on the ground that the Buddha
incarnation of Vishnu is not given its orthodox
place in the list of incarnations;
80 but traditions
about the invention of playing cards exist
everywhere (they are plentiful in France, for
example), and, as Fausta Nowotny has pointed
out, there is not really any ordering of the suits,
and hence of the incarnations, in the pack.
81
__________________
80 Haraprasad Shastri, ‘Note on Visnupur Circular Cards’,
Journal of the Asiatic Society of Bengal, vol. LXIV, 1895, part I,
p. 284:5.
81 Fausta Nowotny, ‘Die indischen Spielkarten des
Europe and Asia 59
Another writer, Clerk, states bluntly that the
pack was invented in the seventh century or
earlier, without giving any evidence.
82 In fact,
however, we can see quite easily that the
Dasavatara pack cannot have been the original
form of Indian playing-card pack, since, in a
pack in which the highest card of each suit
represents one of the incarnations of Vishnu, the
Minister is iconographically quite superfluous;
an incarnation of God obviously does not have a
Prime Minister, and no one would think of
including one unless he were making a somewhat
forced adaptation of a different idea. Clearly the
general form of this pack was determined by the
need to follow the model of other packs in which
each suit was headed by a King and his Minister;
this supposition is clinched by the fact that the
highest card is often called, not
avatara, but
raja.
Rosenfeld has on several occasions argued that
playing cards were originally derived from
Indian four-handed chess, a thesis advanced by
others before him. The theory has the
considerable merit of providing some
explanation for the invention of the regular
pack; of why (in forms known outside India!) it
has four suits, and of why each suit is divided into
court cards (corresponding to the King and the
major pieces) and numeral cards (corresponding
to the Pawns).
83 The explanation thus provided
does not, indeed, seem very convincing. Chess and
card games work on quite different principles: in
particular, in chess all the pieces of one colour
belong to one player, whereas there is no card
game in which each player is assigned all the cards
of some one suit. To this it may be replied that
pieces used in one game may be adapted for use in
a quite different game: but here the
correspondence between the chess pieces and the
cards seems exceedingly shaky. Not only do we
not know of any indigenous type of Indian
playing-card pack with only four suits; but, in
Indian four-handed chess, each player has a King,
three major pieces and four (undifferentiated)
Pawns, whereas, in every indigenous type of
playing-card pack, each suit has a King and only
one other court card, and ten, differentiated,
numeral cards. (Again, we must make an
exception for the Chad packs of Mysore, which
_________________
Museums für Volkerkunde in Wien’, Archiv für Volkerkunde,
vol. 11, 1947, pp. 1-26.
82 S I Clerk, ‘The art of ganjifa cards’, Modern Review, vol.
66,1946, Calcutta, pp. 435-6.
83 See the articles cited in footnote 34.
[2nd column]
are late and of no historical importance.) Some
writers supply an answer to this objection by
simply inventing the type of pack which, on their
theory, ought to have existed: thus Bachmann, an
earlier adherent of the theory that would have
playing cards to have been derived from chess,
alleges that there was an early form of Indian
pack, with 120 cards, called the Chaturanga (i.e.
chess) pack, a statement that appears quite
groundless.
84 Rosenfeld himself has put the theory
to use to explain, not merely the origin of the
regular pack, but of the very idea of playing cards,
in the generic sense: he states that early Indian
chess pieces were circular, with figures painted on
the top, so that they needed only to be made flatter
to take the form of Indian playing cards.
85 This
statement, too, appears to be without foundation.
Chinese chessmen indeed have this form, save
that they bear, not pictorial designs, but written
names, but neither of the great historians of
chess, Antonius van der Linde and H.J.R.
Murray, makes any allusion to such round
chessmen in India; certainly, by the time that
chess spread from India to Persia, the pieces were
like modern European ones, save that they were
more abstract in design.
The theory that Indian playing cards
developed from chess pieces should be set aside
as baseless. For the rest, we have no direct
evidence that playing cards were known at all in
India before the sixteenth century; but, then, we
have no direct evidence that they were known in
Persia before that century. The riddle of the
origin of the regular pack remains unresolved:
there is no proof or disproof of either of our two
____________________
84 Kurt Bachmann, Die Spielkarte, ihre Geschichte in 15
Jahrhunderten, Altenburg, 1932, p. 7. When writing this
booklet, Bachmann was an adherent of the theory of the
origin of playing cards in the ninth century in what is now
Sinkiang, a theory first propounded by Engelbert Huber in
his ‘Zur Kulturgeschichte der Spielkarten’, Festschrift zum 72.
Skatkongress in Altenburg, Thüringen, Altenburg, 1928, pp. 17-
32, and one that appears to be a complete fantasy.
Bachmann later embraced the theory of an Indian origin. In
articles expounding this belief, ‘Zur Entwicklung der
Spielkarten und der Kartenspiele und deren Beziehungen
zur Skatstadt Altenburg (Thür.)’, in Beiträg zur
Sprachwissenschaft und Volkskunde: Festschrift für Ernst Ochs, ed.
Karl Friedrich Müller, Lahr, 1951, pp. 308-73, and ‘Neue
Forschungsergebnisse zur ältesten Geschichte der
Spielkarten und des Kartenspieles in Asien und Europa’,
Forschungen und Fortschritte, 26. Jahrgang, Heft 5/6, 1950, pp.
63-8, he reiterated the assertion concerning the Chaturanga
pack.
85 This statement is also to be found in Bachmann.
60 Part I: History and Mystery
hypotheses; either involves assumptions we have
no way of justifying. If the regular pack was
invented in India, perhaps in a twelve-suited
form, and so passed to Persia and then to the
Arab world, suffering in the process a reduction
in the number of suits first to eight and then to
four, we cannot place its invention any later than
the early thirteenth century if we are to allow
time for these developments. But, equally, the
alternative theory, that the eight-suited pack was
introduced from Persia to India by the Muslims,
and that the Hindu forms, with yet more suits,
were a later development, involves assuming the
previous existence in Persia of a 48-card four-
suited pack of which all trace has vanished. But if
we have to choose between the two theories, the
latter seems to me preferable. For one thing, the
theory of the Indian origin of the regular pack
involves that the generic idea of playing cards
was invented twice over, in India and in China;
although such independent inventions do occur,
it is better not to postulate them without
necessity. For another, the word
ganjifa, used
throughout India for playing cards of the
indigenous kind and for the games played with
them, appears to be borrowed from Persian and
therefore suggests that it was through Persian
influence that playing cards first came to India;
the same appears to hold good of many of the
terms used in the game.
The theory of the Indian invention of the
regular pack also entails the independent
invention, twice over, of trick-taking games. The
category of trick-taking games does not occupy
the dominant position among Chinese card
games that it does among European ones. Most
Chinese card games, other than pure gambling
games, belong to one or the other of two
categories that have played a comparatively
minor role in the history of the European card
games. One is the category of draw-and-discard
games, of which Ma Jong is only one of many
examples. The fundamental principle is that
each player in turn draws one card from the stock
or from the preceding player’s discard, and then
discards one card, face up: the winner is the
player who first forms a hand wholly composed
of certain special combinations of cards. Ma Jong
happens to be the one Chinese game of this type
that has attained popularity in the West (and in
Japan as well): but there is an immense variety of
such games in China, played with domino cards,
chess cards and money cards as well as with Ma
[2nd column]
Jong tiles; the great Vietnamese game of
t'ô-tôm,
played with a Vietnamese version of the money.
pack, is also of this type. The draw-and-discard
principle will be familiar to every card player to-
day in Europe or the Americas as that underlying
all the games of the vast Rummy family (save
that, in Rummy games, it is often lawful for a
player to ‘go out’, bringing play in that round to
an end, before obtaining a hand wholly
composed of the relevant combinations, i.e.
sequences and threes and fours of a kind).
Indeed, all Western draw-and-discard games
belong to the Rummy family and are thus
connected to one another by descent.
86 The
surprising fact is, however, that it is only in the
present century that the draw-and-discard
principle has become known to most European
card players: when W.H. Wilkinson, in 1891,
induced the British cardmaking firm of Charles
Goodall to issue a special pack of cards to play
the game of Khanhoo, adapted from a Chinese
draw-and-discard game played with the three-
suited money pack, he was able to write in the
accompanying booklet of rules, ‘The principle of
this old game is hardly known in the West.’ It
seems likely that games of the Rummy type were
played in Spain in the nineteenth century; but if
so, they never achieved any prominence there,
and, with the possible exception of Portugal, did
not spread to other European countries.
Nowadays Rummy games are popular all over
___________________________
86 The kind of draw that occurs in Poker is not an example
of what is here meant by ‘the draw-and-discard principle’:
only confusion results from calling Poker and As Nas
‘rummy games’, as is done by Alan Wykes, loc. cit. In
games of the Poker type, there is usually at most one draw
and discard; more than one card may be discarded at a
time, and the corresponding number drawn; the discard
precedes the draw; and the discard is made face down, other
players not having the opportunity to claim an opponent’s
discard instead of drawing from the stock. As the draw-and-
discard. principle is here being understood, the following
features are essential: at his turn, each player draws and
discards only one card, and must do so as long as play
continues; he discards after drawing, not before; he discards
face up, and an opponent has the option of taking his last
discard instead of drawing from the stock, at least if he can
make a scoring combination with it; and play continues in
this manner, in rotation, until one player ‘goes out’. Poker
would be tame under a system of fixed scores for winning
combinations: its interest lies in the choice the players have
whether to raise, stay in or drop out. This is not at all true of
draw-and-discard games, in which there is almost always a
fixed schedule of payments, depending either on the hand
of the player going out, or on those of his opponent or
opponents, or both.
Europe and Asia 61
Europe, many countries having developed their
own national form: but they originally arrived as
an importation from the United States in the
period after 1910 and particularly in the 1920s, a
new wave of invaders arriving from South
America in the 1950s in the form of the Canasta
games, a branch of the Rummy family.
The form from which all other Rummy games
have derived was one known in the United States
as Cooncan: this was a two-handed game played
with a 40-card pack, omitting the 8s, 9s and 10s,
and treating the Jack as consecutive to the 7; a
player could take his opponent’s discard only if
he could make a sequence or set with it, play
ended only when one player, without making a
final discard, had disposed of all of his cards in
sequences or sets of three or four of a kind, and
there was a simple score for thus going out,
unaffected by the cards remaining in the
opponent’s hand. Cooncan was first described in
print in 1897 by R.F. Foster in the first edition of
his
Foster’s Complete Hoyle, under the name
Conquian;
87 he, too, remarked that it was a game
‘quite different from any other in its principles’.
According to Foster, it had been popular in the
South-Western States since about 1860.
88 In
1900, Foster published an article about it in a
New York paper,
89 and remarked later that the
letters from correspondents which it provoked
showed that the game was gradually becoming
known in the Eastern United States. On Foster’s
account, the next development took place in
about 1908 ‘somewhere in the West’; the game
was adapted for four players, using the full 52-
card pack, and the new form was named ‘Rum’
or ‘Rummy’. Foster published another
newspaper article about this variation,
90 which
was reprinted in several English papers; the
game was taken up in England with even more
enthusiasm than in the United States, and Foster
reports that by 1912 the Bath Club in London
was playing it with two full packs and two Jokers,
and had printed a set of rules for that form.
91 The
variations called Poker Rum and Poker Gin, the
__________________________
87 Pp. 448-52. The next account, under the name Coon
Can, seems to have been in The Standard Hoyle, New York,
1899, pp. 502-3.
88 R.F. Foster, Cooncan (Conquián): a Game of Cards also
called ‘Rum’, New York, 1913, p. ix.
89 The New York Sun for 23 December 1900.
90 The New York Sun for 10 September 1911.
91 All this will be found in R.F. Foster, Cooncan (Conquián),
New York, 1913, p. x.
[2nd column]
latter being the game later immensely popular as
Gin Rummy, were described by Foster in the
1914 edition of his
Complete Hoyle; I do not know if
he was the first to describe them, but they cannot
have been invented long before.
92 Their names do
not indicate any derivation from Poker, but were
based on a faint analogy: their novelty consisted
in the fact that a player’s sequences and sets were
not revealed until the final show-down.
Foster stated that Cooncan had entered the
United States from Mexico, its name being a
corruption of the Mexican name ‘Conquián’.
Almost all American authorities, for instance
Geoffrey Mott-Smith and A.A. Ostrow, have
accepted Foster’s account of the historical origin
and development of Rummy.
93 We are indeed
fortunate that so careful an observer recorded
that development at the very time it was taking
place.
94 The origin of Conquián itself remains
__________________
92 Foster’s Complete Hoyle, 1914, pp. 691-2. A.H. Morehead
and G. Mott-Smith, Hoyle’s Rules of Games, 20th edition, New
York, 1962, p. 113, date the invention of Gin Rummy to
1909, which seems rather early, if Foster’s dating of the
invention of 52-card Rummy is correct; they ascribe it to
Elwood T. Baker, of New York.
93 See, e.g., A.H.-Morehead and G. Mott-Smith, op. cit.,
p. 103. The story, found in some sources, e.g. Richard L.
Frey, According to Hoyle, New York, 1963, p. 78, that the
name ‘Rum’ was originally adopted in England, is at
variance with Foster’s account, and is almost certainly
wrong.
94 The one dissentient voice is that of John Scarne. He
denies that Rummy is derived from Cooncan, which he
writes ‘Coon Can’; he further maintains that the name
‘Conquián’ is a corruption of ‘Coon Can’, rather than vice-
versa, and that the game itself originated in the United
States. Despite Scarne’s sneers at Albert Morehead and
A.A. Ostrow, the derivation of Rummy from Cooncan is
hardly to be called a hypothesis: it is plain to see in the record
provided by the card-game books. As already stated,
Conquián was first described by Foster in 1897, and Scarne
observes that, as Cooncan, it appeared in an American Hoyle
of 1900; the form for three or four players using a 52-card
pack, was first described, under the name ‘Rum’, in Foster’s
newspaper article of 1911, the U.S. Playing-Card Company
shortly afterwards issuing a set of rules for it under the same
name. Scarne triumphantly cites the description of Whiskey
Poker from the American Hoyle of 1875 as revealing the true
ancestor of Rummy (Scarne on Cards, New York, 1955, p. 37).
Foster was well aware of Whiskey Poker, describing it in all
editions of his Complete Hoyle. Though, in his booklet on
Cooncan or Rum, he does not mention it, he apparently
thought it had played a part in the development of Rummy
out of the original Cooncan, since in the 16th edition of
Official Rules of Card Games, Cincinnati, 1912, edited by him,
he included an account of Rum with the note ‘This is a
combination of Conquian and Whiskey Poker’ (p. 15).
Actually, it seems unlikely that Whiskey Poker had
62 Part I: History and Mystery
something of a mystery. Most card games played
in Mexico are of Spanish origin, and the
presumption is that this is true of Conquián also;
but I know of no direct evidence for Conquián’s
having been played in Spain. The Spanish-
playing-card manufacturers, Heraclio Fournier,
issue two booklets,
Fuegos de naipes españoles and
Juegos de naipes extranjeros, two Rummy games
being included in the former under the names of
Remigio or Rabino and Chinchón. The non-
Spanish origin of Remigio is evidenced by its
being played with 52 cards; Fournier even put
out a hybrid 52-card Spanish-suited pack, with
English indices (the Caballo thus bearing the
index ‘Q’), for playing Rummy and Poker.
____________________
anything to do with the matter; its resemblance to Rummy
is quite superficial, and the essential principles were already
present in Cooncan. Scarne quotes the 1914 edition of
Foster’s Complete Hoyle, and jeers at his derivation of the name
‘Cooncan’ from ‘Conquián’: he asserts, on the contrary,
that Coon Can was created in the American South. The
reason why Coon Can, ‘born in Dixie’, has turned up in
Mexico is ‘because it has followed the American Negro
wherever he goes’ (p. 36). Scarne offers no evidence that it
was in ‘Dixie’, rather than, as Foster says, in the South-
West, that Cooncan was first played in the U.S.A;
moreover, ‘the American Negro’ has not, so far as I know,
gone to Mexico very much. Scarne is obviously hinting that
the name ‘Coon Can’ is connected with the exceedingly
offensive racial epithet ‘coon’. It would be incredible that
black people, in the American South or elsewhere, should
employ for a favourite game a name incorporating such a
term; that Scarne could make so tasteless a suggestion
demonstrates his utter insensitivity equally to people and to
probabilities. He further comments, ‘As for the game’s
being Spanish because played with a deck of forty cards,
rubbish!”, and goes on to remark that card games are played
in the United States with a great variety of shortened packs.
If he understood the first principles of card-game history, he
would realise that this is because the U.S.A. contains
immigrants from many European countries, who have
brought their games with them. There are games played in
France with the 32-card pack which are nevertheless of
Spanish origin, although the 32-card pack is unknown in
Spain; this is because the games have been adapted to the
French form of shortened pack. But whenever we find, in
some country, a game played with a variety of shortened
pack that is generally unfamiliar in that country, but well
known in certain other countries, it is a sure sign that the
game originated in one of those other countries. It strains
credulity past breaking point to be told that, with no
influence from any Latin country, Americans chanced to
invent a game in which the 8s to 10s were omitted and the 7
was treated as in sequence with the Jack, and would do so
even if we did not know the game to be played in Mexico,
which Scarne does not deny. The argument dismissed by
Scarne as ‘rubbish’ is in fact conclusive as showing Foster to
have been right in saying that Conquián came from Mexico
to the United States and not conversely.
[2nd column]
Chinchón, on the other hand, is played with a 40-
or 48-card pack, and thus may antedate the
Rummy invasion from the United States.
Conquián is closer to the Chinese draw-and-
discard games than are the later forms of
Rummy; but it is hard to be sure whether the
draw-and-discard principle was invented
independently in Mexico or in Spain, or whether
it was originally derived from Chinese games. If
the former, Conquián may be of Mexican rather
than of Spanish origin, and in either case may
not have been invented before the nineteenth
century.
95 If the basic principle of the game was
derived from China, the Spaniards may have
learned it from the Portuguese, probably in the
sixteenth or seventeenth century. Even in the
latter case, it is plain that the idea did not make
much impact on card play in the Iberian
peninsula, and none anywhere else in Europe,
until Rummy went into a nova state in the early
years of this century. The fact that, until the
present century, European card players were
almost entirely ignorant of the draw-and-discard
principle,the basis of an enormous number of
Chinese card games, corroborates our earlier
conclusion that it was not through direct Chinese
influence that playing cards entered Europe.
The other main category of Chinese card
games is that which we may label ‘Fishing’
games, after the name usually used for them in
China. In these games, a number of cards lie face
up on the table, forming a pool; if a player, at his
turn, has in his hand a card matching one of
those in the pool, he uses it to capture that card,
setting both on one side; if he cannot make a
capture, he has to contribute a card from his
hand to the pool. At the end of play, each
player’s score is determined from the cards he
has captured, according to the values of the
individual cards, often with premiums, in
addition for having special combinations of cards
among those captured. The mechanics of
Chinese fishing games are always the same. At
the outset, a fixed number of cards are laid out
face up to form the pool, each player is dealt a
hand, and the remaining cards are left face down
to form the stock. At each player’s turn, he first
plays a card from his hand, either to make a
capture, or to the pool, as already described; he
__________________
95 Albert A. Ostrow, The Complete Card Player, New York
and London, 1945, p. 540, states that Conquián is ‘Spanish
in origin and is said to have a history dating back at least
four hundréd years’, but gives no authority for this.
Europe and Asia 63
then turns up the top card of the stock; if it
matches a card in the pool, he captures that card
with it, and adds both to those he was won; if
not, he places it in the pool. The number of cards
in the stock is always the same as the total
number of cards in the players’ hands taken
together. A great many fishing games exist in
China, played with different types of cards, and
also in Japan, where they include Hachi-hachi
and most other games played with the
kana fuda
(flower cards).
96 Certain games of this
category are played in Europe, such as Casino,
moderately well known in Britain, Scopa,
immensely popular throughout Italy, Callabra,
and a few others; some appear to have originated
in the Middle East. The mechanics of the
European games differ from those of the Chinese
and Japanese ones in that there is no drawing
from the stock; instead, there are repeated deals
throughout the game whenever the players’
hands become exhausted. Despite this difference,
the analogy is recognised by the Chinese; when
Casino was introduced into China in this
century, it was immediately named ‘Fishing’, like
so many indigenous games.
97 It does not appear
that this category of games is very old in Europe;
certainly it has played little part in the history of
European card games.
[2nd column]
Both draw-and-discard games and fishing
games are played in China with every variety of
playing cards: with chess cards, with money
_______________
96 The hana fuda are by far the best known of the
indigenous types of playing cards still used in Japan. They
form a 48-card pack, divided into twelve sets of four, each
set representing a month and symbolised by a flower or
plant. In the fishing game played with it, a player may
capture from the pool a card of a given month with another
card of the same month. The origin of the pack is uncertain;
a pack of the same general type, but with a greater number
of cards, is to be found in the Cary Collection in the
Beinecke Library at Yale University. It is at first sight
puzzling that many Japanese accounts of the game state
that the pack is derived from the 48-card Portuguese pack. It
is, however, undoubtedly the case that playing cards were
first introduced into Japan by the Portuguese, that the
earliest Japanese playing cards were of the Portuguese type
and that descendants of the Portuguese pack are still
produced and used in Japan to this day. If we suppose that
fishing games (themselves of Chinese origin) were first
played in Japan with some form of the 48-card Portuguese-
derived pack, a card in the pool being able to be captured by
one of the same rank, it becomes intelligible that, later, a
pack with the structure of the hana fuda pack should have been
devised to play the same game, replacing the four-suited
Portuguese one
97 I owe this information to Dr David Hawkes.
[2nd column]
cards and with dominoes and domino cards.
Trick-taking games, on the other hand, exist in
far fewer numbers and do not constitute a
characteristic form of Chinese card play. Despite
this, one of the oldest and most celebrated of
Chinese games, that of
T'ien chiu (Heaven and
Nine), which is played with dominoes by four
players and goes back to at least the early
twelfth century, belongs to this category.
98 This
seems surprising, because dominoes do not fall
naturally into suits; but, for the purpose of this
game, the Chinese domino set is divided, by pure
convention, into two unequal suits, the military
suit and the civil suit, within each of which a
conventional ranking is imposed. The game is, of
course, played without trumps, but the two
characteristic features of Indian and European
trick-taking- games both prevail: the winner of
each trick leads to the next one, and a trick can
be won only by a domino of the suit led. Like
most European games, and unlike the Indian
ones, the player who leads to a trick is subject to
no constraints restricting which domino he may
lead.
It is conceivable that the idea of trick-taking
games was independently invented in India and
in China; but it is more probable that it had a
single origin. If so, then the theory of an Indian
origin either of the regular pack or of playing
cards in the generic sense must be finally
rejected. It remains just possible that the Chinese
four-suited money pack was the ultimate
ancestor of the regular pack: that it spread
westwards from China, acquiring on the way
definite suit-signs for the upper two suits, a tenth
___________________
98 On p. 66 of his ‘Chinese origin of playing cards’, The
American Anthropologist, vol. VIII, 1895, pp. 61-78, W.H.
Wilkinson gives a translation of a passage from the
seventeenth-century dictionary Cheng tzü ch‘ung which, as he
remarks, had often been cited by earlier Western writers
who completely misinterpreted it as referring to the
invention of playing cards. In this passage, it is given as a
‘common legend’ that in 1120 an official appealed to the
Emperor for a ruling on the game of T'ien chiu. The passage
goes on to give, as fact rather than as legend, a cursory
description of the game as played at the time when the
Imperial ruling was supposedly asked for. It further remarks
that, in the time of Kao Tsung (1127-1163), pattern sets
were issued by imperial edict, to regulate the composition of
the set of tiles used for the game; ‘these’, it says, ‘are now
known throughout the Empire as ku p'ai’, that is, as bone
p'ai or dominoes. As Wilkinson stresses, it concludes with
the warning that it should not be inferred that it was at this
period that games of this kind were invented. T'ien chiu thus
dates from at least the early twelfth century, and possibly
earlier yet.
64 Part I: History and Mystery
numeral card in each suit, and first one and then
two court cards.
99 Since that leaves only about a
century for such an extensive evolution to have
occurred, this does not seem the most likely
possibility. A guess that seems to me more
plausible is that, with the spread of the Chinese
invention of paper, the use of the domino cards,
and, with them, the idea of trick-taking games,
travelled to somewhere in Central Asia, or to
northern Persia, and that it was there that the
idea first arose of devising a set of cards
displaying a natural division into four suits in
place of the artificial division into two suits
imposed on the dominoes. There is no way of
saying what the original suit-signs would have
______________
99 A special type of playing cards was formerly used in
Korea, though they are now apparently obsolete, They are
amazingly long and thin, and form a pack of 80 cards,
divided into eight suits, or sometimes only of 60, divided
into six suits, each consisting of nine numeral cards and a
General. The suits and ranks are indicated by highly
distorted characters. In the later versions of the pack, the
suit-designations have disappeared, having no significance
in the gambling games for which the pack was principally
[2nd column]
been. Perhaps one was a coin, and another a
symbol representing a military commander of a
tümen or unit of 10,000. Such a pack could then
have spread further west, to Persia and on to
Egypt and eventually to Europe, meanwhile
giving place, in Persia, to the eight-suited pack
which then went, with the Moghuls or with
earlier Muslim rulers, to India. It could also, at
the same time, have returned to China, to be
adapted, by a misunderstanding or a deliberate
reinterpretation of the suit-symbols, to the
money pack. But here we are once more deep
into the realm of speculation; as already
observed, the problem of the origin of the regular
pack is still a long way from being solved.
____________________
used. Culin considered this pack the archetypal form of the
regular pack, giving rise to all others, but this must surely be
dismissed as a fantasy. There is no correspondence between
the names of the suits in this pack and in the Chinese money
pack, and its origin must be left in obscurity; but it may be
regarded as indicating that it might have been natural to add
a single ‘court card’ to the nine numeral cards of the money
pack suits. Note that in both Chinese and Korean chess, our
King is replaced by a General.
Below are pp. 34-35, for the Chinese transliterations, and 42-45 and 56-57 for the Arabic and Indian, as mentioned at the start of this post.
GameofTarotCh1&2&3-1_page-0001.jpg
Viewed 1559 times 2.79 MiB
GameofTarotCh1&2&3-1_page-0002.jpg
Viewed 1559 times 2.71 MiB
GameofTarotCh1&2&3-1_page-0003.jpg
Viewed 1559 times 2.72 MiB
GameofTarotCh1&2&3-1_page-0004.jpg
Viewed 1559 times 2.71 MiB
