The Mamluk Fifty-Two
The earliest European mentions of playing cards appear in the 1370s, and the direct ancestor of those cards is a game called Kanjifah used in the Mamluk Sultanate (Egypt and Syria, 1250–1517). In 1931 — some sources say 1939 — the scholar Leo Aryeh Mayer discovered a nearly complete Mamluk pack in the Topkapı Palace Museum in Istanbul. Roughly 43 to 48 of the original 52 cards survived; the pack dates to the early 15th century, though the tradition reaches further back. That pack is still held at Topkapı and is the physical anchor for every claim about this family tree.
The structure of the Mamluk deck is startlingly close to a modern playing deck: four suits, thirteen cards each, fifty-two in total. The four suits are Jawkan (polo sticks), Tuman (cups), Suyuf (swords), and Darahim (coins). Each suit runs Ace through Ten, plus three court cards titled malik (king), nā'ib malik (viceroy or deputy king), and thānī nā'ib (second deputy). Because Islamic tradition avoids figural representation, the Mamluk court cards used elaborate Arabic calligraphy and floral arabesques in place of portraits. European makers, inheriting the structure, began painting kings and knights into those slots.
Worth being clear about: this does not prove tarot originated in an Egyptian temple. The Mamluks were a medieval Islamic dynasty, not the pharaohs, and their cards were everyday gaming tools, not priestly ciphers. When the 18th-century Frenchman Court de Gébelin declared tarot a lost manuscript of the Egyptian god Thoth, he was spinning romance — no connection to the actual Mamluk record.
Arrival in Europe
Europe first clearly mentions playing cards in the 1370s. Over the next hundred and fifty years, the imported deck was reshaped, branched, and standardised — eventually splitting into tarot and the modern 52-card pack as two distinct fates.
First European Mentions
A 1371 Catalan rhyming dictionary lists naip as a playing-card term (derived from the Arabic nā'ib, the Mamluk viceroy — the number-two court card). In 1377, Florence banned card games, and in the same year the Swiss monk Johannes of Rheinfelden wrote the first systematic European treatise on cards. Within a decade, records appear across Italy, France, Spain, Switzerland, and Germany.
Milan Adds Twenty-Two Trumps
Between 1440 and 1450, manuscripts in Milan, Ferrara, Florence, and Bologna record a new variant called carte da trionfi — "triumph cards" — in which 22 allegorical paintings (the Fool plus 21 ordered trumps) were added above the original four suits. The surviving Visconti-Sforza deck, commissioned by Duke Filippo Maria Visconti of Milan, is the most famous early example. Those 22 trumps became what we now call the Major Arcana.
The French Deck Takes Shape
Around 1480, French cardmakers simplified the German suits (acorns, leaves, hearts, bells) into clubs ♣, spades ♠, hearts ♥, diamonds ♦, and began producing them through stencil printing — the first time cards in Europe became a mass-reproducible commodity. In the same period, the French deck dropped the 22 trumps and dropped the Knight, arriving at 52 cards with three court ranks (King, Queen, Jack) — the pattern we still use.
Marseille Tarot Standardised
The Marseille cardmaker Nicolas Conver published a woodblock tarot in 1760 that later came to be called the Tarot de Marseille — consolidating three centuries of regional variation into a relatively stable figure set. His pattern remained the dominant French tarot design through the 19th century and is the basis of every modern Marseille reissue.
Rider–Waite–Smith Published
A. E. Waite, working with the painter Pamela Colman Smith, released the Rider–Waite–Smith deck through London's Rider & Co. For the first time, all 78 cards — including the pip cards from Two through Ten, which had previously been mere suit-and-number — were rendered as narrative scenes. This single change made the minors legible by image rather than memorised verse, and a century later, most contemporary decks still trace their visual language to RWS.
Four Suits, Four Languages
Once the Mamluk deck entered Europe, different national card-making traditions translated the four suits into four different visual vocabularies. The table below traces the fork from Arabic originals through Italian tarot, French playing cards, and the German pack.
| Tarot | Mamluk | Italian | French ♣♥♠♦ | German |
|---|---|---|---|---|
| Wands | Jawkan (polo sticks) | Bastoni (batons) | Clubs ♣ | Eicheln (acorns) |
| Cups | Tuman (cups) | Coppe (cups) | Hearts ♥ | Herzen (hearts) |
| Swords | Suyuf (swords) | Spade (swords) | Spades ♠ | Blätter (leaves) |
| Pentacles | Darahim (coins) | Denari (coins) | Diamonds ♦ | Schellen (bells) |
French suits did not simplify directly from Mamluk — they re-simplified from German suits, which is why the French ♣ is called trèfle (clover) in French while the German cognate is Eicheln (acorns): shared plant-growth semantics, rewritten visually in two steps. The Mamluk polo stick becoming an Italian upright baton and then a French leaf-form is the same semantic thread walking across three centuries.
Why the French Deck Dropped the Trumps
Between the late 15th and the 16th centuries, European popular gaming tastes shifted. Italian tarocchi was a complex game — 22 trump cards with a strict ordinal ranking, trick-taking, elaborate scoring. What caught on in French courts and city households were lighter games organised around the four suits alone: piquet, triomphe, later whist and bridge. The trumps were deadweight in that register.
Mass stencil printing was the other pressure. Removing 22 cards cut about 28% from the production cost of a deck; dropping the four Knights shaved another 5%. French cardmakers consolidated the Page's role with the Knight's, keeping three court ranks (King, Queen, Jack) rather than the Italian four (King, Queen, Knight, Page). The move was commercially obvious.
This was not tarot's extinction — tarot continued as a game in Italy, southern France, Switzerland, and southern Germany throughout this period, only drifting from gaming to occult usage in the late 18th century. From roughly 1500 on, the two decks ran on parallel tracks: playing cards served games, optimising for rule simplicity and mass reproduction; tarot served narrative and symbolism, preserving the full 78-card structure.
78 − 52 = 26
Lay both decks on the same table and the difference decomposes cleanly into two terms:
- Tarot 78 = 4 suits × (10 pips + 4 court: Page / Knight / Queen / King) + 22 trumps (Major Arcana)
- Playing cards 52 = 4 suits × (10 pips + 3 court: Jack / Queen / King) + 0 trumps
- Difference 26 = 22 trumps + 4 Knights (one per suit)
When Tarot Became Divination
For roughly its first three hundred years — 1440s to 1780s — tarot was almost exclusively a complex trick-taking card game, with no systematic divinatory use. The pivot happens in late-18th-century France. In 1781, Antoine Court de Gébelin published the eighth volume of Le Monde primitif, in which he declared that tarot was a surviving manuscript from the Egyptian god Thoth. This was the first public claim binding tarot to the occult — and it was wildly wrong. He knew nothing of the Mamluk record, and no Egyptian evidence supported him.
Shortly after, the Parisian cartomancer Etteilla published the first tarot designed explicitly for divination (1780s–90s). In the 19th century, Eliphas Lévi grafted tarot onto the Kabbalistic Tree of Life. The early 20th century produced Rider–Waite (1909) and Thoth (Crowley and Harris, 1944), which set the occult iconography we still read today. Put plainly: tarot divination is a use about 250 years old, practiced with a deck already 650.
The playing-card branch, meanwhile, kept walking down the gaming path — 18th-century whist crossed from England to the Americas, 19th-century whist evolved into bridge, and poker developed in the first half of the 1800s out of poque and primero. Same underlying cards, bifurcated use: one tradition serving the mathematics of play, the other serving the symbolism of narrative.
The Two Traditions Today
Today, a tarot deck and a poker deck still share about 95% of their structure: four suits, Ace through Ten, King and Queen in every one. The only difference is whether the 22 trumps and the 4 Knights are present. This means anyone who has studied tarot can half-read a poker deck on sight, and anyone fluent in poker will recognise the geometry of a tarot minor. The split is in use, not in structure: playing cards serve the probability of games, tarot serves the cadence of story. Two fully legitimate uses, growing from one tree.