I am ready to comment on Nathaniel's hypothesis, now that I think I understand it. Your theory overlaps with what I have been promoting in several ways, in that it postulates three basic components, namely the game of
imperatori, the most important virtues, and Petrarch's
Trionfi, as well as the transformation of the theologicals into the celestials. The main difference is in how much of the various components - specific cards and their order - of each and when.
Your original sequence, ur-tarot, is Love, Chastity (in a chariot), Death, Fame (in a chariot), Time, First Eternity (Angel), Second Eternity (World), Temperance, Justice, Fortitude (with Prudence somewhere among these cardinal virtues), Faith, Hope, Charity.
You say this is hard to remember, compared to what comes later. It seems to me that Petrarch's sequence is quite easy to remember. It is not even necessary to have read the poem: it can be explained in one sentence. Likewise the four cardinals, from Aquinas's teachings (if Prudence is high), again explained briefly, as he did (
Summa Theologiae II.II. q. 123, art. 12, ans.). Surely priests would have based sermons on it. For Fortitude above Justice, Aquinas himself cites Ambrose, who puts Fortitude high because of "a certain general utility." There is also Wis. Sol. 8:7. All good texts for sermons. The three theologicals everyone would remember, from I Cor. 13:13 (talk about luck!). It is harder to remember the sequences as later known. In minchiate Hope is before Faith, with Prudence in between. Fame and Time are both before Death.
The order of Angel and World is something that is there in your ur-tarot, so I will discuss it next. Nathaniel wrote, about the hypothesized change in Ferrara and Milan from Angel last to World last:
But it seems very unlikely that such an unintuitive deviation would spread to two cities, if it was not part of the original order. The best and easiest way to explain the existence of this "unintuitive" order, and especially its dominance in two of the three early centers of tarot, is through the Petrarchan interpretation I have outlined above: If the World was in fact originally a representation of everlasting life in Heaven after the Last Judgment, then placing it at the top of the trump order makes perfect sense, and even more so if the Sun, Moon, and Star were not present in the deck at the time.
You say that putting the Angel last is counter-intuitive, because heaven comes after the Judgment. But Christ looking down on the world, as in Pesellino, is the world in time looked at from the perspective of eternity. That comes before the Last Judgment, from which Christ and the Saints already in Heaven are spared. They are working from above in this world to help sinners find the way. The lady on the Charles VI World card has a similar function. She exactly fits the description of the narrator's guide in
Amorosa Visione (I.35-42), guiding the narrator in this world so that he will win Glory in the next.
I grant you two cards for Eternity. But if one of them shows our world on the bottom of the card, in time, I think it naturally precedes one that shows the beginning of the end of time. The A/B World card is our world, in time, in which some have risen to, or never lost, eternal glory. The PMB World card, on the other hand, does not show our world, but rather the New Jerusalem, perhaps also heaven, since it is on top of the card. In that case, yes, it can intuitively be after the Angel card. There has been a switch. Likewise for the French version, with the four evangelists and Jesus, presumably in heaven. And the Ferrara
order, with World last, seems influenced by that of Sforza Milan, even if its cards follow the A order
designs. Perhaps the d'Este deck was made in Florence, as the recent Issy catalog contends, and the Metropolitan/Budapest followed the same conception. If any order is counter-intuitive, it is that of Ferrara, with an A order this-worldly design on its World cards, nonetheless last in the order.
Then there are the two chariots. You justify your two "original" cards with chariots by saying that it appears as both Fame and Chastity in the early decks of Milan. But these are different decks. Where do two chariots occur in the same deck? They are so similar that indeed they would be easy to confuse: one has a sword and four attendants, the other a stick, and in the right vs. left hands, while both hold out golden orbs with the other hand. The red vs. whitish horses suggest the same Platonic allegory as the rearing vs. calm horse of the Cary-Yale, which in Plato was surely for Chastity rather than Fame (as you say, the Issy could be Chastity). But the Issy's sword, orb, and chariot correspond to Boccaccio's description of Worldly Fame in
Amorosa Visione (VI.65-72). The Catania, without attendants and with similar horses clearly corresponds to Fame, as does the Charles VI, Rosenwald, and Bolognese. But these are details in cards that are still too similar.
It seems to me that the card must have been called "Chariot" from the beginning, because otherwise there wouldn't have been the variations in portrayal. I defy you to find any suggestion of deck with two chariot cards. It is one subject, Chariot, portrayed with different Petrarchans in mind in different decks.
I think that the reason for changing the card from Pudicizia/Chastity to Fame was that there were already Temperance and Fortitude, which were close enough in meaning to Pudicizia that the latter would be overkill. So the cardinal virtues go where Pudicizia was, allowing the Chariot to be reinterpreted as Worldly Fame, sooner or later with a male on top, except in Minchiate, where the lady, naked except for a banner, clearly is not Chastity, but might still be derived from her.
Which was first? It requires some thought to realize that you don't need a special card to represent Petrarch's Chastity, a reason for the Chariot to have been originally Chastity and the World originally Fame. But the original designer might have already thought beyond Petrarch, and decided to put first Fame and then Time before instead of after Death, Fame now being the Chariot. So I am dividing your step 1 into two parts, a and b. 1a has the Petrarchans in Petrarch's order, 1b is the same but with Death moved to second to the end, and two cards for Eternity, one Eternity in Time (Providence, guidance, lessons), the other the beginning of of the end of Time. There are still 6 Petrarchan cards, but one of Petrarch's 6 is no longer represented, and one is represented by two cards. Whether 1a is a real deck or just the first draft is unknown. If there was a 1a, it had a very restricted run It need not have been in the same city as 1ab. If there aren't 7 Petrarchans, of course, there is no symmetry to be had with 7 virtues. That could be one reason for adding the Wheel, precisely so as to make 7. That might have been the first real deck, too, starting from 1a. Call it 1ac, or, if using the 1b order, 1bc.
In step 2 you add the 8 Imperatori. I wish to challenge that, to say that the tarot might have borrowed fewer than 8 at first, as part of step 1, and added more later.
Why in step 1? Karnöffel and Imperatori were, besides, Marziano's game, the earliest games with trumps, perhaps even earlier than Marziano, although we don't know. There is the probable playing of Karnöffel in Milan of 1420, then the record of
VIII Imperatori in 1423 Ferrara, ordered from Florence, and
Imperatori again in 1434.
In Karnöffel one was the Emperor, another the Pope, etc. How far back that goes back we don't know; these names may be imports from other games, such as tarocchi. Yet an Emperor is the most natural candidate for an imperatore, followed by an empress, because emperors and empresses rule over kings. So do Popes, in the Guelf universe. The virtues and Petrarchans rule over them.
I have suggested that Imperatori differed from Karnöffel in that Imperatori had permanent trumps, as opposed to a designated suit by the luck of the draw. In that way it doesn't turn the world upside down, making low ranking cards high ranking, thus avoiding Filippo's prohibition. It may well be that in Imperatori none of the cards were actually emperors. By analogy to emperors, these cards, presumably, rule over kings and below. Of course if the Rothschild cards are from an Imperatori deck, then at least one was. But we don't know that. It may be that the word "imperatori" functioned in the same way as "trumps," i.e. cards that were superior to kings and below. Yet an Emperor is the most natural candidate for an imperatore, followed by an empress, because emperors and empresses rule over kings. So do Popes, in the Guelf universe. The virtues and Petrarchans rule over them, the one as obligations, the other as life situations.
Nathaniel wrote,
The highest ranked of these special cards (called the Karnöffel) was able to beat all the other cards in the deck; one or two others could beat all the numeral cards, and the others had only restricted trump powers: some were able to beat only numeral cards, while some could beat one or two of the lower court cards as well, but not the kings.
This is not quite right. Dummett in 1980 surveyed a great many variations, the earlier ones in obscure language, and came to this conclusion:
But, in the earlier form, there are just three cards of one suit, the Karnöffel (Unter), Pope (6) and Kaiser (Deuce), lifted out of their natural order to become fully-fledged trumps, three more, the 3, 4 and 5, which serve as partial trumps, and one, the Devil (7), which plays a special role. (p. 190, at
viewtopic.php?f=9&t=1175)
The special role for the 7 was that it beat any other card, or perhaps any other card except one or two (Karnöffel, Pope). So we have four
Königstrechers, as they were called. A later version had one more, a Sow.
I am inclined to agree with your intuition about Imperatori, that some of the cards besides the 4 papi were inspired by that game or by Karnöffel. One of my thoughts about them, which I have kept to myself up to now, is that there would have been 4 "good" imperatori and 4 "bad" (just as there are three "good" Petrarchans and 3 "bad"). So in Karnoffel there is the Devil, and the Fool could be the tarocchi counterpart of the Karnöffel, the insensitive, dimwitted lout, but reduced to his appropriate level of power. Then we have a con man (bagatella) and a traitor, so four that fit the bill. The Ferrara court would have liked such a game. Half of 8 is 4, so two could have a special relationship to each suit, just as rows of gods had special relationship to suits in Marziano's game (even if how that special relationship was reflected in the rules is not said in so many words). The two were perhaps one good and one bad, or maybe the "bad" ones attached to two of the suits and the "good" ones to two of the other (as seems to be implied by Marziano).
But would all 8 have been added at once, in your step 2? I see no particular reason why, and several reasons why not. One reason for thinking that it was 4 imperatori at first (in the tarocchi), the "good" ones, is my principle that in general, the more variation there is in the placement of a subject in the order, the more likely it is to be early rather than late, the only exceptions being those at the divide between the sections that seem to have been respected (as Dummett discovered), i.e. the Pope as the highest of the 4 imperatori, and Death as the dividing line between the here and now and the far away, in space or time. The Popess, Emperor, and Empress, are in all sorts of variations. But the Fool, the Bagatella, the Hanged Man, and the Devil are always in the same place (or lack of place, in the case of the Fool). So these "bad" imperatori would be later, the "good" ones earlier.
Another reason for thinking the "bad" ones wouldn't have been there at first is that in a sense making them trumps does put them above the rest of society, except for the Devil. So a card maker who made them trumps is at least risking the wrath of the ruler, especially Filippo, and also is giving the legislative body, in the case of Florence, a reason to withhold legal status for the game, which it did not do for tarot until 1450 there. Having a Devil card at all is suspect, too, because people could use it in black magic, and its high position suggests a lot of power. Legality is good for sales. Once the game is legal, however, having such cards would be good for sales.
Literally speaking, there was only 1 imperatore, or 2 if the masculine plural is taken to include both the Empress and Emperor, so a deck like the Cary-Yale could have gotten away with just them. But there is good allegorical reason for adding the Pope, to assert his authority over both. There are also good allegorical interpretations of 4 imperatori. One is pope, antipope, emperor or emperor-elect approved by the pope, excommunicated emperor. Another is Pope plus Emperor in Rome, Patriarch plus Emperor in Constantinople. Another reason is the division into 4s and pairs already discussed.
Allegorically speaking, the best candidate for the first of these interpretations is Bologna, where it was in fact made later. And it is a good explanation for the "equal papi" rule (in my view the East/West interpretation is not, because religion was put higher than secular authority, and by the end of the council the pope was highest of all). Bologna is also a good candidate for the second, as Bologna was the pope's originally favored site for the unity council (Ross has done some good research on that.) But of course Ferrara and Florence were also favored, later.
I do not agree with your discounting of Bologna as a place of origin. There is an early and long-standing tradition that the game began there, with no one, there or elsewhere, disagreeing. It was known for its playing cards at an early date, even if triumphs aren't mentioned in Bernardino's published sermon. He might well have done so in person, but changed it to "and so on" (as in the published version), to avoid giving a vile new game free publicity. That there is no evidence is not persuasive, because you have to consider the places where early evidence has been found. Mainly, it is arrest records in Florence (none before 1443, despite Franco's diligent search) and ruling family records in Ferrara. There do not seem to have been ordinances against card games in Bologna, just taxes on the decks (I seem to remember reading somewhere). The ruling family's palace was burned and demolished in 1506-1507; it seems unlikely that the fleeing Bentivoglios would have taken the pains to save their purchase records from 60 years earlier. That Giusti's diary was saved, and then copied a long time afterward, with one of the copies containing the mention of triumph cards (and the archaic language used suggesting that it was not a later addition), is a fluke.
With 4 imperatori, there is the possibility of a nice symmetry, with 4 virtues. Added to 6 Petrarchans, we have our desired number of 14. There is also room for flexibility. A deck with one imperatore is fine, and so are two and three, depending on how many are needed to add up to the desired number, whatever it is. From what we know, we cannot deduce that the principle governing the whole was a particular ratio of trumps to regular suits, whether 1:1 or 3:2. There may have been more than one principle. One of them might have been to have the same number of trumps as Marziano, who after all may well not have had 4 court cards per suit, since he only mentions Kings.
For just the 4 cardinal virtues at first, one argument is that they go with 4 suits and 4 imperatori (the 4 virtues take the place of the 4 "bads"). The attributes for the cardinals have distinctive correspondences to the four suit-signs: cups for Temperance, swords for Justice, sticks and stick-like columns for Fortitude, and round objects, mirrors, for Coins. Such correspondences are found in two sources that Moakley found: a funeral oration for Giangaleazzo Visconti and a "game of the king" devised by Innocenzio Ringhieri. In that way there is a correspondence between trump and suit, somewhat similar to Marziano's correspondences between god and suit.
Another argument is again my criterion of greater disarray meaning earlier, except for natural dividing points between Dummett's three sections. The theological virtues do not seem to have been involved in any such drastic rearrangements, if indeed they changed into celestials, because the celestials are all together everywhere, as are the theologicals in minchiate (that prudence is with them can be explained by a desire to put the different cards of that deck all together, flanked on both sides by those the two decks have in common.
This is probably a good place to talk about the Cary-Yale (Modrone, CY, Visconti). You say 22, and it just happens that the missing cards are precisely the six imperatori and nothing else. That seems implausible to me.
I think Cary-Yale was likely 2 + 7 + 7, and yes, there is a symmetry between virtues and "Petrarchans", as well as a 1:1 correspondence of trumps to cards per suit. But in this case there is the Wheel as a kind of extra "Petrarchan", known from the Brera-Brambilla. The Visconti had a large fresco of one in their palace on Lake Maggiore. That makes the Cary-Yale a step 1ac" or step 1bc, as I previously defined them. In other words, a possible "original deck." But 16 cards per suit was unusual, a bit much. When there are 14 cards per suit, 3 imperatori + 4 virtues + 7 "Petrarchans" also works, for the triumphs. That is my hypothesized structure for the Brera-Brambilla. So might I have no longer just one candidate for an ur-tarot, depending on whether the theologicals were there, and how many trumps, 14, like the number of suit cards, or 16, like Marziano? Well, it is possible.
I have one more argument for 4 virtues, which involves going to a later stage, when the Bagatella and Hanged Man have been added. and then inferring what was likely there earlier. I think that the Bagatella was added in Ferrara, because only the spelling "bagatella" has the precise double meaning needed, between "trifle" and "prestidigitator". There is also the incredible awkwardness, grammatically, of the Steele Sermon's "El Bagatella." Elsewhere the card was called by a word ending in -o.
I observe the residue of a pattern in the Ferrara order, assuming only the Bagatella and Hanged Man added:
B order:
Temperance, Love, Chariot, Fortitude, Wheel, Time, Hanged Man,
Death, Angel, Justice, World.
The pattern would seem to start out every third card a virtue. But then there are five cards until the next one. If the Hanged Man were replaced by Prudence, the pattern would have repeated twice. That may well be the way it was. Not only that, but if Love is trump 6, the Hanged Man will be 12, the number associated with him in the gospels and by Pope John XXII in his shame painting against Muzio Attendola Sforza. That the Hanged Man substituted for Prudence is also suggested by Imperiali's
Risponsa to Lollio's
Invettiva.
In the C order, the pattern is without the Wheel, every 2nd card a virtue:
C order:
Love,
Justice, ,
Chariot, Fortitude,
Time, Hanged Man,
Death, Temperance.
Again, the hole is where the Hanged Man is. For testimony we have Piscina's
Discorso, which associates both it and the Old Man (and only them) with Prudence. If the Wheel were there, its place would cause an exception to the rule in that one space, but again the Hanged Man would be 12. Since the Wheel certainly was in Milan by the time of the Brera-Brambilla, the pattern without it either existed earlier, or else for some good reason the Wheel needed to be added in Milan in the first place. Elsewhere I have suggested that the reason might have been to satisfy the requirements of a Marziano-style grid, when the theological virtues are absent.
What these patterns tell me is that in the popular deck of Ferrara, there were no more cards between Death and Angel, and there were 16 trumps in all by the time of the Hanged Man and Bagatella. Indeed, the five from Devil to Sun are always in exactly the same order everywhere. That does not have to be true in either C or A, however. So one or more of them (including theologicals) may have existed in one of those places, before all 5 were put in a row everywhere.
It also tells me that the order of the virtues was decided in each place at an early stage 1, 1+, or 2 (meaning a change in the order from that elsewhere, or the addition of cards other than the cards from Devil to Sun in all the cities). This fits the general preference of players not to change the cards' order relative to one another very much.
If so, the question is, how did the theological virtues get changed into celestials, if they weren't even in the deck before the celestials' appearance?
Well, it's just a theory that the celestials replaced the theologicals. But I think it's a good theory. My answer is that there were two types of deck in Milan, one with and one without the theologicals, and the same in .
In Florence 7 virtues plus 4 imperatori plus 6 Petrarchans equals 17, 1 too many for a 16 triumph deck, 3 too many for 14. So either there was just 1 imperatore, for 14, or there were 3, on the principle of having the same number as Marziano had, or the practice started in Milan, in a deck with 16 cards per suit, i.e. the Cary-Yale. Which came first I can't say: each has good points.
There are a few scenarios. One is that the game with 4 virtues and 4 imperatori is first, in either the C or A region, from which it goes to the other. In Milan Filippo gives it its characteristic spacing of virtues, adding the Wheel and dropping one imperator, perhaps to fit a Marziano-style grid with virtues corresponding to suits. In that form it goes to Ferrara, which gives the sequence a more consistent spacing of the virtues (although if my spacing idea is wrong, it could have used Florence's 14, with 4 imperatori and no Wheel). The Wheel is then added in Florence. The three theological virtues are added there, while in others, or in Bologna, it is the celestials instead (or the theologicals are changed to celestials). In the former decks, the celestials are put after the theologicals, with Prudence among them after it is removed from the tarocchi. Before or after Florence, Filippo adds the theologicals in Milan, later changed to celestials, Milan emulating Florence or vice versa. Meanwhile other cards have been added in Ferrara and/or Bologna, emulated in the others.
I think that the theologicals were more likely included in Florence before Milan, because the Star in Milan resembles Hope, and the Moon resembles Faith. That suggests the order Hope first, then Faith, which is also their order in minchiate. This is counter-intuitive, because it is not St. Paul's order. Yet it is the order seen on a tomb painting in Florence thought to have been painted by dal Ponte (I have uploaded a picture of it somewhere on THF), whose workshop has long been associated with the Rothschild cards, recently reaffirmed in the Issy catalog. So likely dal Ponte is the source.
Another scenario, similar to Nathaniel's, would have all 7 virtues to start with and 1 imperatore (in that regard different from Nathaniel), in either Milan or Florence, more probably Florence (where our Spaniard would get the idea), then going to the other and Ferrara, each acquiring its characteristic placement of virtues, before or after acquiring the Wheel. This is followed by more or all of the rest of the 8 imperatori, some perhaps in Ferrara, changing Prudence to the Hanged Man, and the theologicals changed to celestials, except that some people in Florence like having those removed cards in the deck and leave them where they were, but moving Prudence so as to keep the order otherwise the same in the two decks.
I don't like this latter scenario as much as the former, because either it doesn't observe the special characteristics of the cardinal as opposed to theological virtues (the spacing in Milan and Ferrara, the repeating 4s of virtue, imperatori, and suits, etc.), or delays spacing the cardinals until the theologicals have been removed, where the game has already been played for a while. To get around that problem one can postulate two milieus for the game, early elite and late popular, as Nathaniel does. This can work with the Cary-Yale, I think (as I postulated when assuming the former Beinecke order and suit assignments for that deck), because of its special composition and his extreme isolation, but not otherwise, because there is nothing special otherwise about the deck early as opposed to late, and because in Ferrara, Bologna, and Florence, there wasn't nearly the separation of elite vs. popular as in Visconti Milan.
There remains the question of the Tower, and perhaps the addition of the Devil. I have many theories about those two cards. As for Nathaniel, for me the Tower is a bridge. To the theologicals, they are the antidote to the Devil and God's lightning. To the celestials, Piscina supposed that the two cards represent the spheres of air and fire, as the two Platonic means between earth and the heavens. Lightning is fire from the sphere of fire, and it strikes towers before other structures. There is the
Divine Comedy, which had the Devil's Hell and the purging fires of Purgatory before the celestial spheres, the Mountain of Purgatory as a kind of tower. There is the Apocalypse narrative, with its devils, lightning-toppled steeples (in the medieval illustrations), hailstones, and fireballs, before the woman clothed with the sun, etc., appears. Another possibility is the geography of Bologna, with the Devil and Star in its Bolognini Chapel (Magi and Hell frescoes), and the famous towers, leftovers of an earlier age, outside. There are also a few Platonic allegories that might have provided bits of imagery, Plato's allegory of the cave, with its increasing light, or in Plutarch a Hades between the earth and the moon, lightning that hurls people downward, and Clotho with her distaff governing either the moon or the sun, depending on which work you read, "On the Genius of Socrates" or "On the Face that appears in the Orb of the Moon".