Playing card family tree

Various kinds of playing cards from around the world are related to each other by common descent, forming a family tree with ever diverging branches (and in rare cases converging ones). The illustration below includes the major European types together with other descendands of the same ancestors, but not their non-European descendants.

French-suited tarot
TarotFrench
SpanishSwiss
PortugueseGerman
ItalianLuxury
MamlukGanjifa
MǎdiàoTujeon

The leaves represent fairly broad groups with significant internal development. For the purpose of determining relationships between different groups, the earliest forms within each group are of the greatest interest. These are more often than not partially or entirely lost, but can be reconstructed on the basis of their surviving descendants within the group.

The lozenges represent hypothetical types that are not directly attested, but must be assumed in order to explain relationships between different groups. To a certain degree they can be reconstructed based on shared features in their descendants and ancestors.

Prototype Z – Mamluk Sultanate

Besides a fragment of a card, only a single deck from the late medieval muslim world is known today, but replacement cards originally from two other decks demonstrates some of the variety once present. Apart from having abstract representations rather than pictures of people on the court cards, this is strikingly similar to early Italian decks. Though it is later than these, there is enough evidence to rule out that the Mamluk cards descend from the Italian ones, and the similarities must be due to a shared ancestor, Z.

The features of this lost intermediary form can be reconstructed based on the relationship between its descendants – the two main branches of European cards and the Mamluk deck – and its ancestor, prototype Y. The latter is in turn only known through its descendant groups, which bedsides prototype Z only includes the north Indian Ganjifa; and even more indirectly trough other desendants of its ancestor, prototype X, such as the Chinese Mǎdiào.

Features

  • Four suits: coins, polo-sticks, cups and swords (shared by Mamluk and early European cards, against eight suits only partially matching these in Ganjifa; coins is also found in Mǎdiào cards, some of the other three can be assumed to be related).

  • Each suits contains thirteen cards: one to ten, plus two officials and a king (shared by Mamluk and early European cards, against only one official in Ganjifa for a total of twelve cards).

  • The court cards presumably have images of people (shared by Ganjifa and early European cards, against the geometric patterns in Mamluk cards).

  • In one half of the suits, the number cards are ranked in reverse (shared by European, Mamluk, Ganjifa and Mǎdiào cards; in Mamluk and Mǎdiào cards as well as the European ones that have this suit, coins is always among these, while the oldest known Ganjifa cards have two different coin suits, one reversed and one not).

  • The highest cards in some or all suits are marked by writing or symbols independent of suit and value indicators at the top of the cards (found in Mǎdiào, Ganjifa and Mamluk cards, but not in early European ones).

Prototype Y – Persia

From historical and literary records, Ganjifa cards are known to have been in use in Persia during the European renaissance. Actual surviving cards are only known from northern India from a later period. While these usually are round, rectangular cards are also known, and this must have been their earliest shape. The decks had eight, ten or twelve suits; the former is the earliest type. Each suit consists of twelve cards: one to ten, plus a vizier and a king.

Even this oldest form reachable by internal reconstruction must have developed differences from its shared ancestor with the Mamluk and European cards, prototype Y. This can be reconstructed based on its descentant groups: prototype Z (itself reconstructed) and Ganjifa, and its ancestor prototype X, as reconstructed through other descendant groups.

Features

  • Four or eight suits (Mǎdiào, Mamluk and early European cards all have four, while Tujeon and Ganjifa both have eight in their earliest forms).

  • Each suit contains twelve cards: one to ten, some sort of official and a king (so only in Ganjifa; Mamluk and early European cards have two officials, Mǎdiào and Tujeon does not have tens, Tujeon only has one court card per suit, a general, while Mǎdiào has an asymmetrical system with differences between the suits).

  • Half the suits are reversed with lower numbers beating higher ones (Ganjifa, Mamluk and early European cards, in Mǎdiào at least for one suit, unknown in Tujeon).

  • The highest cards in some or all suits are marked by writing or symbols independent of suit and value indicators at the top of the cards (found in Mǎdiào, Ganjifa and Mamluk cards, but not in early European ones).

Prototype X – China

The oldest prototype, X, is not unsurprisingly the hardest to reconstruct. The distance in time between its existence and any surviving descendants means that even the relationship between its three descendant branches is uncertain.

Several different types of playing cards have evolved in China, forming a number of major groups. The group that is presumed to be the oldest, the so-called money suited cards is the only one that is clearly related to the other groups discussed here. This group is divided into two varieties, with three and four suits respectively. While it is not entirely certain which of these were earliest, the four-suited variant is the one most closely related to the other groups, and therefore must be used as the basis for the reconstruction of X.

The oldest known money suited cards with four suits is known as Mǎdiào (馬吊), but even the earliest surviving cards of this type might have developed differences from the common ancestor. However, it must be assumed that any feature found both in Mǎ diào and in non-Chinese cards must have been have been present in X. Features common to Mǎ diào and money suited cards with three suits or other, more distantly related types might also be assumed to be shared with this prototype, but here one must also consider the possibility of cross-branch borrowings.

Korean Tujeon (투전, also called Htou-tjyen) had eight suits in the oldest known decks, each with nine numbered cards and a general. This structure is intermediate between Mǎdiào and Ganjifa cards. Formally, this does not necessarily mean that they are developmentally intermediate; the same could be the case if they were a common ancestor of both, or that all three had a lost common ancestor. Lacking more specific information, the latter is assumed in the illustration.

Features

  • Four or eight suits (Mǎdiào, Mamluk and early European cards all have four, while Tujeon and Ganjifa both have eight in their earliest forms).

  • Each suit contains cards from one to nine (so in Mǎdiào and Tujeon, versus Ganjifa, Mamluk and early European cards which also had tens).

  • The numerical value of each cards was marked either by repeating a suit symbol the appropriate number of times, or by a number beside the suit symbol like the corner indices in modern cards (Mǎdiào has two suits of each type, Tujeon only the latter and Ganjifa, Mamluk and early European cards only the former).

  • The deck also contained non-numeral cards, each of which was part of one of the suits (asymmetrically in Mǎdiào, one general per suit in Tujeon, king and vizier in Ganjifa, king and two officials in Mamluk and early European cards).

  • Some suits are reversed, meaning that lower numbered cards beat higher (unknown in Tujeon, present at least in the suit of coins in Mǎdiào, present in half the suits of Ganjifa, Mamluk and early European cards; in the latter two coins were among these).

  • The highest cards in some or all suits are marked by writing or symbols independent of suit and value indicators at the top of the cards (found in Mǎdiào, Ganjifa and Mamluk cards).

Prototype X does not necessarily represent the ultimate ancestor of all playing cards, but without any further evidence, nothing further back can be reconstructed.