Dice and the Minor Arcana: Outlining the challenge

In past posts, I have discussed possible ways of pairing the 21 Tarot trumps with the 21 possible rolls of two dice, concluding in the end that the "Air Hexactys" system was the best. In this system, dice rolls are ranked first by total (e.g., any roll that totals 7 outranks any roll that totals 6); and then, among rolls with the same total, by the higher of the two numbers rolled (e.g., among rolls totaling 7, 1-6 outranks 2-5, which in turn outranks 3-4). The dice rolls, thus ranked from lowest to highest, were then paired with the trumps from 1 to 21. This system seems to "work" and to make symbolic sense, as briefly discussed in my post "The root trumps of the Air Hexactys."

Now, the main reason for assuming a link between dice rolls and Tarot cards is the fact -- unlikely to be a coincidence -- that there are 21 possible rolls of two dice (2d6) and 56 possible rolls of three (3d6), corresponding precisely to the 21 trumps (the practice of numbering the Fool and counting it as a trump is a relatively recent development) and the 56 suit cards (Minor Arcana). The next step, therefore, is to try to find a dice-to-cards mapping that works for the Minor Arcana as well as the Air Hexactys works for the trumps.

Here are some of the challenges facing anyone who would attempt such a mapping:

⁂

Four suits: The trumps have a simple linear structure, being numbered from 1 to 21; all that is required is to establish a linear ranking of the dice rolls, which is not difficult. The Major Arcana, on the other hand, are organized in four suits. There is no obvious way to divide up the 56 possible rolls of 3d6 into four equal categories.

⁂

Rank within each suit: In most traditional games played with Tarot cards, the long/black suits (swords and clubs) are ranked, beginning with the lowest, A 2 3 4 5 6 7 8 9 10 J C Q K -- the same ranking used in most modern card games, except that aces are low. However, the round/red suits (cups and coins) rank the pips (but not the court cards) in reverse order: 10 9 8 7 6 5 4 3 2 A J C Q K. Although there are a few exceptions (most notably the French jeu de tarot), I think we can take this idiosyncratic ranking system as a very old and probably original feature of the Tarot pack.

⁂

Relative rank of suits: Due to the quirk just mentioned, it seems that if we want to rank all 56 Minor Arcana linearly, we have to group them by suit first and then by rank within each suit. It wouldn't make sense, for example, to begin with the four aces, then the four deuces, and so on, because long aces are low but round aces are high. I suppose we could begin with the long aces and round tens, then the long deuces and round nines, etc., but this seems very unnatural.

The problem with ranking by suit first is that Tarot games do not rank the suits. Players must always either follow suit, play a trump, or discard a card. In no case can cards from two different suits be played in the same trick, so the question of whether, say, the King of Swords outranks the King of Clubs never arises.

Modern "esoteric" or divinatory tarot is not a trick-taking game, so "rank" as such is not an issue. Nevertheless, it has become customary to think of coins/pentacles as the "lowest" suit, followed by swords, then cups, and finally clubs/wands as the highest. Waite's chapters on the Minor Arcana in The Pictorial Key to the Tarot, for example, begin with the King of Wands as the highest card and proceed down the ranks to the Ace of Wands; then the Cups, from King to Ace; then the Swords, and finally the Pentacles. (Waite makes tens high and aces low for all suits, contrary to the Continental tradition of which he was perhaps ignorant; Tarot has never been a game in England.)

Historically, the four suits do have a natural rank, since they can be traced back to Chinese suits representing different denominations of money. Coins, which have preserved their character as single coins, are the lowest denomination. Next come Clubs, which were originally strings of 100 or 1000 coins. (Ancient Chinese coins had a hole in the center and were strung together.) Cups derive from the Chinese character for 10,000; and Swords, from the character for ten -- meaning, in context, ten myriads, or 100,000. This ancient ranking has been preserved in many modern Anglo-French-suited games, where Spades (Swords; cf. Italian spade) are the highest-ranking suit, followed by Hearts (Cups), Clubs, and finally Diamonds (Coins). (Bridge, in which Diamonds outrank Clubs, is an exception.)

⁂

Relation to trumps, especially root trumps: Because each of the trumps has been associated with a roll of 2d6, each roll of 3d6 will be associated with between one and three of the trumps. For example, the roll 1-4-6 would be associated with 1-4 (Lover), 1-6 (Hanged Man), and 4-6 (Sun) -- or, alternatively, with the three root trumps associated with 1, 4,  and 6 -- namely, Magician, Death, and World. For rolls where the same number occurs three times, only one of the trumps will be linked; for example, 4-4-4 should be related to Death. (The Four of Swords would seem a natural choice for this roll.)

⁂

Sparse symbolism: Compared to the trumps, the Minor Arcana contain relatively little imagery. Prior to Waite's innovative "scenic pips," a card like the Seven of Swords portrayed nothing more nor less than seven swords (stylized as arcs in the Marseille tradition). Such cards have no immediately obvious meaning, and the traditional meanings ascribed to them vary widely. This makes it hard to judge whether or not a particular dice-card pairing is a "good" one. This can be viewed either as a bug (because it offers little guidance) or as a feature (because it allows the cards to accommodate a wide variety of systems).

A rationale for Waite's orientation of ROTA

Yes, yet another post (see the others) on this very minor bit of Tarot iconography.

In a past post (qv), I wrote that "of the four possible orientations of the ROTA, the only one I can't think of any good reason for is the one actually used by Lévi and Waite, with T at the top!" Now a possible rationale for it has occurred to me.

Notice that the wheel is apparently turning counterclockwise: the snake on the left is descending, and the cynocephalus on the right is ascending. So the letters ROTA correspond, respectively, to the low point, the descent, the high point, and the ascent. Lévi had already identified the A and O of ROTA as corresponding to alpha and omega -- representing the beginning and the end. It seemed to me that the T and R ought to correspond to some similarly fundamental opposition, but nothing came to mind until today. T and R correspond to the Hebrew words tov (טוֹב, "good") and ra (רַע "bad/evil"). Obviously, the top of the wheel represents good fortune; and the bottom, bad fortune. One starts at the bottom, rises (alpha, the beginning), reaches the top, and then falls (omega, the end).

Oswald Wirth's version of ROTA

(For context, see my past posts on the Wheel of Fortune Tarot card.)

I found this diagram near the beginning of Oswald Wirth's Le Tarot, des Imagiers du Moyen Age.

Éliphas Lévi, who was the first to introduce into Tarot iconography a wheel marked with letters ambiguously reading either ROTA or TARO, always put T at the top of the wheel. Wirth's diagram is the first example I have found of someone adopting Lévi's idea but changing the orientation of the letters. (In my post on the orientation of ROTA, I spoke favorably of this orientation, with A at the top, because it corresponds with the orientation of the Tarot suits used by both Lévi and Strieber, but still consider R at the top to be the correct orientation.)

Arranging the 22 Major Arcana, rather than the four Minor suits, around the wheel is a little strange, 22 not being divisible by four, but it is central to Wirth's system. Each card is considered to be related to those horizontally and vertically (not diametrically) opposite it on the wheel, as shown in this diagram, also from Wirth.

I haven't yet read and assimilated Wirth's commentary on these proposed mappings, but at first glance the system seems just slightly off. My initial reaction is that I would rotate the cards clockwise 8.18 degrees, so that the horizontal axis lined up perfectly with 0 and 11.

My proposed modification of Wirth's schema

The resulting mappings seem much more intuitively correct -- although, as I say, I have not yet looked in any detail either at Wirth's system or at my proposed alternative.