Many people have spent many years reflecting on the possible origins of the tarot. I too have been interested in this rather neglected topic and although the genesis of the tarot is not the only problem which I have investigated recently, it is nevertheless one of the most complex. A game having undefined origins cannot be considered as exceptional - the same thing indeed occurs with many games including chess, draughts, and even such recent widespread card games such as bridge and rummy. The fact is that for tarot, the occult tradition associated with these cards has rendered their origin even more obscure than the origins of other games. In my opinion, the occult tradition is based on nothing plausible and is essentially a fantasy created in an attempt to carry back the origin of tarot to the ancient civilisations. In the printed images of the cards there may be traces of older figurative details, but the origin of tarot must be dated to the Italian Renaissance. Despite this attribution, I am however, ready to accept an early date, down to the turning of the 14th into the 15th century.
Before entering a discussion of the possible sources, let us examine the pack and focus our attention on what is to be sought for.
1 - STRUCTURE OF THE PACK
Commonly, each tarot deck is composed of 40 numeral and 16 court cards, divided into four suits. The distinguishing feature of tarot decks is the presence of an additional series of cards, generally numbered from 0 to 21 - the so-called triumphs. Let us consider the three groups separately.
1.1 Numeral cards.
The numeral or pip cards usually have an ascending numerical order in each of the suits without any further apparent meaning. In particular, they have no evident "social" counterpart; they may be compared with chess pawns before James de Cessole (see section 3.3) - if they had to represent common people they did it in a undifferentiated way. In chess, the renowned Liber de moribus hominum assigned a social role to each pawn - a similar process can be traced for instance, in the Ambraser Hofamterspiel, where each numeral card is associated with a particular social profession. For tarot cards, this pattern of association is however much more the exception than the rule with social professions generally represented, if at all, in court and triumphal cards.
1.2. The Court cards.
The three court cards of a traditional deck of playing cards are: a Page, a King and an uncertain intermediate card which today is usually a Queen but more often was in the past a Knight. This knight - as also occurs for chessmen, was either represented as mounted on a horse or sometimes as a horse alone. Such cards can be found in older packs in which the courts have a military character and in which no woman is depicted. Indeed, the German series of Unter-Ober-King maintains a strong affinity with the original oriental series. This militaristic order was transformed into the now familiar Jack-Queen-King court series by a similar process to that occurred in chess, where the appearance of a Queen, along with bishops etc. was a result of the transition from the earlier militaristic hierarchy to that of a monarchical court.
By contrast to playing cards, the tarot pack has characteristically, both Knight and Queen along with the Page and King, together constituting the quartet of court cards appearing in each of the four different suits making a total of sixteen. It is not certain that the sixteen 'court' cards did really originate as intrinsic components of the four suits. Originally, part or all of the sixteen personages depicted on the tarot court cards could have belonged to a single discrete aristocratic series. The personification of the figures depicted on the cards might well have been introduced only after a particular suit-mark was ascribed to the figure at a later date. As an alternative it must be considered that the four tarot court cards were intercalated into the pack having originated from amongst the triumphal cards (q.v.).
1.3. Triumphal cards.
In the tarot pack there is present, a group of cards at least in part similar to courts but distinguished from them by not belonging to either of the four suits. There are 21 such cards in the 78 card deck each depicting human, spiritual and temporal themes plus an un-numbered extra card called 'The Fool'. They are used in the game as major cards, counting above the "usual" cards. This Triumphal series, which is discussed in more detail below, has stimulated a lot of discussion because of the persistent uncertainty on their origin, meaning and order(1).
2 - STRUCTURE OF THE TRIUMPHS
The sequence of the triumph cards must be an ordered one for playing most games. Even where numbers are not written on the cards, the correct order must be remembered and enforced. Thus, particular sonnets were passed down through generations - by learning them by heart, one could learn the value of the card. There are various such orders recorded, and it is hard to establish which, if any, is the original one. To complicate matters further, it is not certain if a particular triumph sequence had a similar order when it was first introduced, nor even that such an order originally existed. In the following sections, possible arrangements and sequences of the triumphal cards will be discussed.
2.1. A Single Series in a Continuous Sequence.
I understand that it is possible to infer a significance of the cards by proceeding continuously from tarot 0 to tarot 21 or in another sequence(2). "Experts" have done it more than once, with the suggestion of provenance as old as the Kabbalah and the Egyptians(3).
I do not exclude the possibility that a plausible meaning once existed for the whole tarot series and that its source has yet to be discovered, however, I must set several criteria on such an interpretation. The original source - let us assume a book - could not belong to the ancient civilisations: the date of its circulation should be fixed to the end of the 14th - beginning of the 15th century, because this is when the triumphal series first came into use. This book would contain a story or a philosophical treatise, or something similar, with the rather surprising quality of being familiar at the time and practically forgotten nowadays. These two hardly compatible conditions are necessary to explain both the large popularity of this sequence of images at the time and our difficulty in recognising it now.
We also have to make the assumption that the triumphal sequence as inferred is in deed a complete one. Should it, however, prove to be only a part of a more complex set of traditional images, then one has to consider how an isolated fraction of the many topics or images present in an ancient encyclopaedic work would be interpreted today in the absense of the bulk of the information.
A sequence containing spiritual and temporal images together is not incompatible with contempory imagery, because it was then common to deal with angels together with humans and the natural elements. More difficult is the task to identify the source in a work with precisely, or almost precisely, the figures of the triumphal cards. In looking for this source, we have further to discuss whether or not the order of the events recorded in the "story" or book mentioned could be different from that supposed for the cards. We have thus, in principle, two kinds of sources even assuming a single sequence with roughly the same items, 1) sources with all the items without any correspondence of their order to that of the triumphal cards, 2) sources with all the items following one of the orders recorded for play, or a very similar one.
2.2. A Random Series.
It has been proposed that tarot triumphal figures were randomly selected from the stocks of standard images available to painters and engravers of the time and this was once suggested by no less a person than Michael Dummett. I do not know whether the bold assertion that the triumphal cards could be adopted by card players without any specific meaning or source was only meant to provoke readers or to convince them. If images were chosen at random, why are there no representations of more common subjects such as inn signs, town banners or coats of arms? These often show everyday tools (sword, cross, key, horn, ladder, hammer, scissors, balance), animals (eagle, lion, wolf, cock), flowers (rose, lilies, fleur-de-lis), and so on, which could be additionally distinguished by detail or colour. However, nothing of this kind can be seen in the tarot.
On this basis, I do not believe that the triumphal cards could be gathered at random from stocks of current images and, consequently, I do not think that our task is simply to find a figurative or literary source for each single triumph subject, independent of the others.
2.3. A Series of Several Groups.
If we observe the actual series of tarot triumphs, it could seem to correspond to an intermediate case between the two indicated above: neither a single sequence nor a set of independent items, but formed by several groups of subjects gathered together. For instance, the groups of the four elements, the four virtues, and so on, could initially be classed together, and then have lost part or all of their systematic character due to later modification of the series.
It may be worth noting that the triumphal figures usually show either a single person with an insignificant background or by contrast, the reverse situation in which a detailled scene is shown with the people having only a minor role. This distinction is not always present in ancient packs. There is a possible hint suggesting that personages similar to courts could once have been more emphasised than they are now. For instance, in Visconti tarots Sun, Moon, and so on, were personages painted with the celestial object in their hands, very similar to court personages holding a suit sign. Some more direct link may once have existed between triumph and court personages. It seems possible that within cards of this kind, we can recognise a fifth suit containing elements of both pip and court cards and maybe a few real triumphs, even higher.
In the original meaning of triumphal cards, we would expect positive personages, with favourable influence or meaning, as indeed are the virtues, and so on. In tarot these "positive" personages do exist, but they are accompanied by others which have a neutral meaning and especially by a few which cannot be considered positive at all (Death however was actually included among the triumphal subjects).
As a tentative conclusion, let me propose that the series of tarot is too significant for having been formed at random but at the same time, appears too random to have been purposefully designed unless markedly changed by later modifications.
3 - POSSIBLE SOURCE-BOOKS
3.1 I Trionfi.
The first possible source book is Petrarch's 'I Trionfi' written around 1360. We are dealing with a great poet and a renowned work, which was copied many times. To associate in detail tarot with this book and with actual processions and festivities of the time, has been the research and interpretation of Gertrude Moackley(4). There is no need to repeat here the pros and cons, since it is a well-known and debated theory. However, the triumphal processions were common facts both before and after the writing of the poem - it is thus the idea of triumphant personages which should be examined(5).
Is it possible to set each triumphal card in correspondence to a known triumphal character? Is this sequence of events meaningful? In my opinion, only a part of the triumphs may reasonably well be put in correspondence with the canonical subjects of triumphal scenes (to begin with the limited number of triumphs explicitly considered by Petrarch). No actual record of performances or processions is known to me with the same order of characters, nor with all of these characters.
In my opinion, this is a theory which can easily explain a part of the evidence. For the remaining part, I have to think to other works, which might have been written by the same author, or by authors of the same time - for instance by Boccaccio - which could have been collected together. Certainly, if we take the great Florentine literary tradition of the 14th century completely into account, it becomes an easy task to find a literary reference for all the personages and the scenes present in the triumphs (Popess included!) - even though to put the whole set in order in a reasonable way may still be hard.
3.2 Tractatus De Deificatione Sexdecim Heroum.
The pack was painted for Filippo Maria Visconti by Michelino, but the actual treatise was written by Marziano da Tortona in the 1410s. This case is unique among the several indicated here - the connection with tarot is not a speculative one: we are in the presence of the earliest tarot pack or at least a quasi-tarot pack(6), here both book and cards are real, even though somewhat out of the ordinary and deserve further comment. We have sixteen personages, all gods or mythological personages, which are originally structured: they belong to four suits; however, their hierarchical order is not dictated by these suits, but proceeds continuously across them. We can thus forget that there are four groups of four cards and consider the set as a new series of orderly triumphs increasing in power from 1 to 16.
We have here a systematic set, but it does not cover the whole standard series of tarot 0-21 triumphs. The six lacking cards could not be added by following the same scheme because it is not possible to add three halves of court cards to each suit. The overall correspondence with the standard sequence is not yet satisfactory, but this source-book may provide an explanation for the manner in which the triumphal cards were inserted into the pack(7). We can think of several evolutionary steps leading to tarot. For instance, eight emperors added as higher courts; a different pack with a fifth suit consisting of ten pips and eight higher cards; four of the higher cards entering among the standard courts; a temporary pack consisting of five suits of 14 cards each; the addition to the latter of another set of eight major triumphs. The fact is that initial packs of 70 cards are documented(8), as well as the peculiar Marziano series of 16 personages.
3.3 Liber de Moribus Hominum et Officiis Nobilium.
This book was written by James da Cessole in the end of the 13th century but its success remained high for a couple of centuries. It has many of the qualities we were looking for: a book which is very seldom reprinted nowadays and a book which was very familiar at the time. Its great popularity was due to the successful comparison between chessmen and social characters, its moral teaching, its being the only work on chess before the first technical handbooks appeared.
The number of extant manuscripts is very great; many of them have illuminations or drawings representing social characters which in some cases are strictly reminiscent of triumphal cards. These characters were painted in a rather standard way. Several manuscripts show blank spaces where a specialist painter had to insert the standard figures after the writing of the text.
Why not suppose that the same painter could draw similar characters for card players? We have to remember that, both in chess and in playing cards a change from a military to a court representation took place. Thus, we can think to a straightforward connection with courts, and then to an extension to higher cards, following a similar procedure or environment. For instance, we can imagine (with some help from the Visconti di Modrone tarot) a pack which has left almost no trace: plausibly only eight pip cards corresponding to the pawns, four male courts headed by the king together with four corresponding female courts headed by the queen (obviously we have here to multiply them by the four common suits, instead of two).
There are alas, several complications. To begin with there is the lack of a similar meaning for the pips cards, which might correspond to minor personages of the court. In a socially-oriented configuration, which cards could be considered as triumphal? Obviously the highest components of social life, noblemen and so on. The problem is that most of these personages, including queens and kings, were already inserted among the "common" court cards! It is, therefore not surprising that in the triumphal cards we find even higher social figures, such as Emperor and Pope, which are present on the Earth as single specimens, as they are in the tarot. The habit of having each king accompanied by his queen lets a wife be present both near the Emperor (even if the Empress seldom was a powerful personage) and near the Pope (and this remains somewhat surprising).
Still more surprising is the appearance among these superior cards of a couple of characters, such as the Fool and the Bagatto, which clearly appear as personages associated with a court series but not among the top cards. They look as if they belong to a series of pre-existent court cards, which included higher and lower components and afterwards was distributed among the four suits. Thus courts and triumphs were partly mixed together without producing an orderly system. The problem is that it is impossible to consider figures such as Death, Tower, Star as extraordinary court cards. I am thus convinced that - supposing that some social and court characters were at the origin of court and triumphal cards - we can explain some triumphal cards, but not all of them and that again at least a second source has to be discovered.
3.4 Libro Imperiale.
The date of this work is mid 14th century; the author is Giovanni Buonsignori from Città di Castello, but other attributions have been suggested(9). The work seems to be in the style of Boccaccio, to whom its authorship has been assigned in the past. By the way, Boccaccio himself, probably more than Petrarch, has contributed to describing famous men, women (Popess included!), gods, especially in his Latin works. This book answers to many of our requirements.
The possible link between the Libro Imperiale and the tarot just begins with its name and title(10). Very striking to me has been the beginning of the story with the several triumphal processions of Caesar. In subsequent days he was celebrated wearing on each day a different clothing, either military or religious. I could not help thinking of Caesar so dressed in those occasions as to be easily identified with tarot triumphs since he alone could appear as nothing less than Emperor, Empress, Pope, and Popess! Moreover, a correspondence with Imperatori (and/or Kaiser) cards and games may be inferred. "Carte da imperatori" were named special cards which - among other quotations - were bought in Florence for use at the court of Ferrara(11). It is now known that these packs included eight images of emperors and were available in Florence as early as 1423(12).
The story of the book is a classical romance with many adventures, and with people passing from throne to slavery and vice versa. A final section, only present in some of the manuscripts, is precisely devoted to the series of the Roman emperors. The book contains apparently, everything we were looking for, much more besides. However, the succession of the tarot triumphs again is not a representation of this complicated story. When I think of the story and look for cards highlighting its contents, I do not find the essential facts represented. Also, when I consider the cards and look for analogies in the romance, I have likewise difficulty in placing all the items. Once again, the whole tarot series finds here no satisfactory correspondence.
The date of this work is about 1382 - the author is usually indicated as James Palladini (1349-1417) from Teramo(13). Again the popularity of the book is what we are looking for - once very popular but soon forgotten. This book is seldom mentioned and - different from most of the other items considered - an ancient specimen is not easy to find neither in hand-written nor printed form(14). One possible reason is because the book was strongly disapproved of by the church and any copies of books and manuscripts found, were destroyed. As it often occurred with forbidden texts, the contents were not so bad, but they differed with the prevailing canonical thinking and this was the worst of the deviations of the time.
In this book, we find a fight between Good and Evil, between angels and demons, and especially the process to Jesus carried out by the Devil. It is not surprising here to find images that can be put in correspondence both with the positive and negative cards of the tarot! Again, however, this cannot be the exact source of the tarot sequence, in any order - only a couple of images can be transferred directly to the tarot series and they can at most, provide some useful components.
Before considering any source for the triumphal series of tarot, a few fundamental questions should be answered. Do we consider the sequence as single, complete, ordered, or must we put limitations to these requirements? A particular detail concerns the connection, if any, with the court cards, the intermediate group between the numeral and triumphal ones; at least some of the triumphs seem to be connected with the courts.
Having investigated the possible origin of the tarot images from contemporary sources, I have outlined a few candidate works which can be added to the interpretations based on triumphal processions and Petrarca's Trionfi. When I first encountered each of these sources, I had the impression that it could contain the explanation I was seeking. Now, I must admit that these attributions have only the same value as those based on the triumphal processions with only a few cards finding a reasonable correspondence.
Different explanations have different merits, but none can provide the complete solution of our problem. In particular, no correspondence can clearly be found with the whole ordered sequence of the tarot triumphal cards. However, it may be not required to search for an exact correspondence between the order of the canonical succession of tarot figures and the images of the source because nobody has yet proved that the tarot sequence was a single set, born complete and in that order.
The works cited in this article could have at most contributed some cards to the pack but they may now suggest to somebody a similar work which exactly followed - why not, in principle - the sequence of our tarot series. Good luck for the search!
My conclusions favour two alternative interpretations. 1) The real origin has not yet been found but one source-book - not very different from those indicated here - does exist and we only have to discover it. 2) No such book did ever exist and our task only is to collect the corresponding images from several sources.
I agree that adding further hypotheses where only one should be enough is running against Occam's razor. However, I greatly prefer that the razor cuts contemporary works of the time rather than others arising from many centuries before. Setting aside Egypt, Israel, gypsies, and so on, we have at least accomplished a remarkable change of focus, toward the end of the 14th century. Here some other possible traces might be found! I am going on keeping my eyes peeled in search of possible sources but I am certain that for finding THE source, then readers must keep their eyes peeled too.
I now think that the second hypothesis above is the more probable. If this is the case, it becomes less important to find the various sources and the real aim is to understand the way in which different groups of personages were inserted within and above the four suits.
1. M. Dummett, The Game of Tarot. London, 1980.
2. J. Shephard, The Tarot Trumps. Wellingborough, 1985.
3. R. Kaplan, Encyclopedia of Tarot. New York, 1978.
4. G. Moackley, The Tarot Cards Painted by Bonifacio Bembo. NewYork, 1966.
5. F. Pratesi, Kortleiken, 4 nr 7 (1992) 7-11.
6. F. Pratesi, The Playing Card, 18 (1989) 28-38.
7. M. Dummett, The Playing Card, 18 (1990) 73-75.
8. G. Ortalli, Ludica, 2 (1996) 175-205.
9. Dizionario Biografico degli Italiani, vol. 12. Roma, 1970, pp. 407-409.
10. A rather late edition (Venetia, 1510) has for title: Imperiale che tratta gli triumphi honori & feste ch'ebbe Julio Cesaro ne la citta di Roma - con altre belle historie.
11. F. Pratesi, As de Trèfle, 54 (1995) 16-17.
12. A. Franceschini, Artisti a Ferrara - vol. I, Ferrara 1993, p.120. (quoted by Ortalli)
13. D. Heubach, Der Belial. Strassburg, 1927.
14 J-C. Brunet, Manuel du libraire. T.5. Paris, 1864, coll. 801-804.
Pictures: selected and inserted by Lothar Teikemeier
1. Chess figure "King" c. 13th/14th century
2. Cary Yale Tarocchi, card "Emperor", c. 1440, commissioned by Filippo Maria Visconti
3. according Schachzabelbuch from 1463, Oberschwaben (Cod. Pal.germ. 463)
4. according Pierpont-Morgan-Bergamo Tarocchi, c.1452, commissioned by Francesco Sforza, Milan
5. according Schachzabelbuch from 1463, Oberschwaben (Cod. Pal.germ. 463)
6. according Pierpont-Morgan-Bergamo Tarocchi, c.1452, commissioned by Francesco Sforza, Milan
7. from unknown Cessolis edition, likely De ludo scachorum, Straßburg, Heinrich Knoblochtzer, September 1483
King / Emperor
1. old Chess figure,
King with accompanying persons
2. Trionfi card c. 1440,
Emperor with accompanying persons
3. Chess pawn "innkeeper"
(pawn according Cessolis' Chess interpretation)
4. Trionfi card c. 1450, "Bagatello",
occasionally addressed as "innkeeper"
Player / Fool
5. Chess pawn "Player or Messenger"
(pawn according Cessolis' Chess interpretation)
6. Chess pawn according Cessolis' Chess
interpretation, badly clothed
7. Trionfi card "Fool" c. 1450,