Ross's chart

Over a month ago, on a different thread, Marco asked me to consider some points about Ross's graph on 15th century reports of tarot activity (viewtopic.php?f=11&t=917&p=13527#p13540)

This is for me a difficult subject, and it's taken me a while to think about it. I also needed to do some reading: Umberto Eco (Interpretation and Overinterpretation, Limits of Interpretation ch 1-3)), Panofsky (Meaning in the Visual Arts ch. 1), Gombrich (Symbolic Images ch. 1, 5, 9)). The thread went on to other things. But I want to say a few things about Ross's chart. So I start a new thread. First, here's Ross's graph (from viewtopic.php?f=11&t=917&p=13527#p13527):

Essentially, Marco thinks that the graph shows that the tarot was most likely invented a little before 1440. I think it shows no such thing.


Marco wrote
it seems that we agree that Ross' chart shows that something happened just before 1440: before that date we consistently have nothing, after that date we consistently have documents. No significant change occurred in the rate of destruction of documents before and after 1440, so I don't think this aspect is relevant.
On the question of the lack of floods, etc., that would destroy documents and so bias the results on the chart. I was thinking of two particular events: the destruction of the Visconti castle in 1447 and the destruction of the Bentivoglio Palace in Bologna in 1506. These events would have wiped out many of the records and decks of the major consumer/commissioner in both these places, Milan up to 1447 and Bologna up to 1506. So it is not surprising that there are only two reports from Bologna on the chart, despite other grounds for saying that there was considerable tarot activity there. Ross, I believe, still maintains that tarot was played continuously in much the same way with much the same deck from its beginning there, by 1442, throughout the century. Also, it is not surprising that there are no documents from Milan despite the existence of two decks and iconography. Any decks in the Visconti palace would likely have been destroyed; while special care was taken to save illuminated manuscripts before the palace was destroyed, that may not have pertained equally to card decks, in an emergency situation such as occurred, or the purchase records of a dying regime. I would guess that the two decks that survived (CY and BB) were not taken from the palace. They were gifts to people who didn't live there.

There may be other special acts of destruction that need to be taken into account, which then would affect all cards and records up to that point in that location. For example, fear of the Inquisition may have led to selective destruction. Inquisitors in different areas, from different orders, acted differently. Also, some cards were more dangerous than others, e.g. the Devil and the Tower. Such destruction would have applied not only to that particular time but to anything saved from an earlier time.

Allowing for these special acts of destruction doesn't mean that we can say what we like about what happened in the times affected. But they must be taken into account. They make for a less than level playing field.

Another factor that would tend to reduce the number of records would be playing card prohibitions that don't exclude triumphs. In general, people will want to avoid being denounced to the authorities. Unlike the previous types, this factor only applies for the specific period of prohibition. Here it is as much a question of records never being kept as of them being destroyed as soon as practicable.


Marco wrote
it seems that we agree that Ross' chart shows that something happened just before 1440: before that date we consistently have nothing, after that date we consistently have documents. ...If the something that happened is the invention, we provide an answer to the question "when was tarot invented" that is compatible with the available evidence. So, we have a theory that is supported by the available documentation.

It is possible that "something else" happened just before 1440: for instance, the number of decks was very small before that date, but it had an explosion just before 1440, so that we consistently have documents after that date and none before. This scenario is more complex than what Ross proposes, because it involves two events:
1) the invention of tarot
2) the "something else" (e.g. the "explosion" of the number of decks)
Moreover, it is not a clear theory until a documented date is indicated for the creation, and a clear description and an explanation is provided for the "something else". Given the same explanatory power, a simpler theory is to be preferred. In this case, the simpler theory also seems to explain more.
I want to give some concrete examples of games where these alternatives (1) and (2) would apply, from our own time, and then get to the idea that "a simpler theory is to be preferred."

First I need to make a change in terminology, mostly a quibble. It seems to me that there are almost always two steps. That is, first there is the invention, then the production. Generally the invention is worked out by someone, either as a proposal or a sample, before its production for use or display. Then after production starts, either it catches on quickly and is noticed by those who keep records or collectors' items that last for centuries, or there is a period of small production, unnoticed, and later a boom, a bust, or continued small production. So what Marco is saying is that given the available evidence, a two step theory is preferable to a three step one, because simpler.

This is just a terminology change, for clarity, because "invention" and "production" are different things. Where Marco has "invention", I say "invention and noticed production shortly after". There is certainly at least a short lag between invention and noticed production, because in 1440, when Giusti uses the word "triumphs" he is using a word already in use. Otherwise he would have said more, if "triumph" was a word new to him, to help him remember what he had bought. When Marco says "invention plus something else" I say "invention and unnoticed production, followed by something else, which results in the invention being noticed by those who keep records or collectors' items that last over the centuries."

I think it helpful, for clarity, to say what the "something else" might have been. One possibility is that famous people started buying tarot decks, so that it became a sought after item. There may have been an event that attracted famous people to the place where someone was trying to sell them. The Council of Florence comes to mind. Or if not famous persons, then a larger number of lesser known but still influential people, at the same event. Another possibility is a change in the rules of play, so that it became more interesting to play. Another possibility is that it started getting produced by a recognized craftsperson in a place recognized by many--and so a prestigious, well-trafficked producer or shop finally takes it on, after creeping along elsewhere.

For concrete examples, we might look at games in our own day. Trivial Pursuit is an example of Marco's type 1, games that catch on quickly; invented in 1982, it was extremely popular by 1984. Skateboarding is an example of type 2, an invention with a definite boom after a period of slow production. Although it existed from the 1880s according to my friends, its documentable profile started in the 1940s, getting slowly bigger, a mix of homemade boards and manufactured ones, but attracting pretty much only people in surfing areas until the early 1970s, when the invention of polyurethane wheels created the boom. Actually, a surprising number of games fit this pattern. Monopoly, also type 2, limped along for 25 years before a change in production company and game-name helped it to boom. Scrabble was the same way, but only taking 10 years.

It seems a safe generalization that there were these same types then, not to mention games that never got noticed sufficiently for us to know about them at all.

In the case of tarot, we know that it did take off, so we can eliminate that last scenario. Between 1 and 2, the "simplicity" argument says to take 1. The problem is that it will eliminate 2 in every case. Since it is a reasonable generalization of human experience that then as now, there are cases of scenario 2, perhaps even more than scenario 1, there is a contradiction. The argument from simplicity doesn't apply.

Let me put this in another way. If the argument from simplicity rejects type 2 in the case of tarot, it will reject it in every case. In other words, in the 15th century, there were no games which became popular, after a period of slow production of which there is no trace, and no games of which we have no record at all. This is contrary to experience and common sense.

The criterion of simplicity might work in the case of, say, Kepler's laws of planetary motion vs. Ptolemy's principle of epicycles, although even here there are many complications (see That is a different situation. It is not a case of: given that some planets move in epicycles and some in ellipses, which is to preferred in this particular case in front of us? Those two theories are mutually exclusive in every instance. We cannot say that about games, which are of at least types 1-3, of which for tarot only type 1 can be ruled out.

You might want to argue a reductio ad absurdum against me. Since there is no evidence of Ross's sort at any time before 1440, I can't even say that tarot didn't exist in ancient Egypt, or in fact didn't originate in a teapot floating in outer space. I can exclude the latter possibility on the general principle that there is no reason to suppose that such a thing is the case. If there is no reason for believing something, then it shouldn't be believed or even postulated. Egypt is a little different; some games from Egypt did survive, such as a form of backgammon, and there are parallels between the cards and aspects of ancient Egypt, as de Gebelin observed. But now I can use empirical generalizations to discount this hypothesis. generalizations that apply to all games like tarot and all sequences like the tarot sequence. The longer the time of invention before 1440, the less chance it has of not going extinct after not catching on, and the more chance it has of being found in surviving records. The chance quickly becomes quite remote, especially given the interest taken by history- and pedagogy-writing clerics in games and sequences that uses Christian imagery with no noticeable hint of heresy. Also, much of the imagery is not seen before the 14th century, or the 13th century at least. Conceptually some parts of the sequence are not in the educated mindset in the time before Petrarch's schema for the Trionfi was known. I don't see a pictorial illustration putting a beggar, jester, king, pope, and other categories of people together until c. 1400, illustrating the De Remediis. Much of the tarot does have a "virtue vs. vice" structure, which is very ancient, but it does not conform visually or conceptually to any psychomachia known before the 14th century. (I give Giotto's of 1305, the benefit of the doubt, although it is not very close.) There are other predictions we could make that aren't fulfilled, predictions the result of generalizations based on experience and perhaps subject to the requirement of parsimony, or at least being part of a system of generalizations subject to that requirement. I hesitate to say because the application of the test of simplicity is a very complex subject (again, see. for the basic issues). But the force of the reasons that I have given for excluding ancient Egypt applies to the tarot whether it developed as type 1 or as type 2.

There is also the question of arbitrariness in making correlations. Many of De Gebelin's correlations work individually, if we have a broad enough selection of texts and pictures deriving from ancient Egypt (mostly Greco-Roman ancient Egypt). But there have to be correlations as a sequence, not merely individually; and in a form that they could be accessed later so as to result in their appearance in 1440.

And to the extent that the sequence is embedded in a card game, there are other considerations, such as when cards were introduced into Europe. It is hard to imagine them not being noticed in surviving records for very long, certainly not more than a generation or two.

In general, for inventions like tarot, medievalists say that one must allow at least 20-25 years between invention and first record, according to Andrea Vitali. So we get to Marco's last point.


Marco wrote,
Vitali's statement that the invention must have occurred twenty years before the earliest document seems to me to be unverifiable, because it is based on the lack of evidence. This theory says that one must expect a twenty/twenty-five years gap between the creation of an artifact and its first documented evidence. But Ross's diagram clearly shows that there are no twenty years gaps in tarot documents.
to which he added:
I must correct my previous post: Vitali presents the 20/25 years statement as an "assumption" ("a thing that is accepted as true or as certain to happen, without proof" - or a "conjecture" (i.e. "a proposition that is unproven" - wikipedia). This confirms my impression that the statement is not verifiable (it was not meant to be). In the second sentence of my post, I inappropriately called this statement a "theory" ("a well-substantiated explanation of some aspect of the natural world, based on a body of facts that have been repeatedly confirmed through observation and experiment" -

"As the oldest known documents about the game of triumphs dates back to 1440 (Florence) and 1442 (Estense Court), by historical assumption they must date back to at least twenty/twenty-five years earlier, a period which matches with the Prince being in Bologna. This type of conjecture, with reference to practice of use, is commonly confirmed by the historians of the middle ages. Specifically, professor Rolando Dondarini, teacher of medieval history at the University of Bologna, is an agreement with the writer."
First, as Marco realizes in his 2nd post, it is not a question of twenty year gaps in documents. It is a question of a twenty year gap between invention and its first known documentation. I will attempt to explain.

I have been in correspondence with Andrea Vitali for clarification. Marco has, too, in Italian (to reduce the problem of miscommunication due to translation), with cc to me .

The first question that comes up for me is, how connected is this alleged 20-25 year lag to Andrea's theory that the tarot was invented by a Count in Bologna c. 1410? On the one hand, the lag is something attested to by numerous professors whose names Andrea supplies and also, he says, a conference of academics in Genoa. To that extent, whatever its merit, it would seem unrelated to Andrea's particular views.

On the other hand, he, and perhaps they, are affirming the lag in a particular context. Andrea says that this principle does not apply to transmission through the courts, which can be very fast. His example is that of hair styles. If a painting shows a great lady with a new hair style, and that painting is taken to another area, others may copy that hairstyle very quickly

Andrea's thesis, for him, applies to merchandising outside the courts. For him, the context was a merchant in Bologna selling to the Ferrara court.. So first we have to ask: Is it reasonable to assume that there was merchandising outside the courts prior to the documented instances we now have? If we include production in Florence for a condottiere's court in Rimini as court production, there is so far no evidence of merchandising outside the courts in its early years, before 1450. There was the use of the word "triumphs" already, but perhaps that was within a court setting, too. So clearly it is not necessary that there be a 20-25 lag between invention and first documentation. Also, it seems to me that an enterprising Bolognese merchant might indeed sell a new invention to a Ferrarese court before 20-25 years were over. But perhaps there are other kinds of necessity.

In particular, Andrea says that what is necessary is to assume a 20-25 year lag. There's that word "assume" or "assumption". Marco looks up the word "assumption" in a dictionary. Dictionaries differ as to the primary meaning of the term. The Free Online Dictionary gives as the first meaning of the word
Assumption. 1. The act of taking to or upon oneself: assumption of an obligation.
The idea of accepting "without proof" doesn't enter in until its 4th definition.I think that dictionary reflects our ordinary usage.

Even when we do have "proof", it is often relative to certain standards, about which all we can say is that they work. In Euclidean geometry, there are axioms, accepted without proof. But they are not arbitrary conjectures. They seem like they don't need proof. And they work. They are good assumptions, in the sense of principles to act on without proof.

It seems to me that Andrea's use of "assumption" is in accord with the Free Online Dictionary's primary meaning, as an operating principle (but not as strong as an axiom), based on previous experience and subject to correction if more facts are learned. This is my own assumption of what he means, in my case meaning something I will proceed on, even though I have limited information.

Andrea, in his emails, gives an example which might clarify his meaning. A preacher in 1306 Florence said that eyeglasses had been invented in Italy "not yet 20 years" earlier. The preacher even met and talked with the man who invented them. This sermon is cited today on numerous websites. And in fact although many have searched diligently, no report has been found of eyeglasses, meaning lenses attached to frames and mounted on the nose, before this one in 1306.

I do not have Andrea's source, a book in Italian on inventions, but for my purposes a paragraph in Wikipedia ( will do as well:
the first eyeglasses were made in Italy about 1286 or a little after, according to a sermon delivered on February 23, 1306 by the Dominican friar Giordano da Pisa (ca. 1255–1311): "It is not yet twenty years since there was found the art of making eyeglasses, which make for good vision... And it is so short a time that this new art, never before extant, was discovered... I saw the one who first discovered and practiced it, and I talked to him."[8] Giordano's colleague Friar Alessandro della Spina of Pisa (d. 1313) was soon making eyeglasses. The Ancient Chronicle of the Dominican Monastery of St. Catherine in Pisa records: "Eyeglasses, having first been made by someone else, who was unwilling to share them, he [Spina] made them and shared them with everyone with a cheerful and willing heart."[9] By 1301, there were guild regulations in Venice governing the sale of eyeglasses.[10]
From this example, it seems to me that the 20-25 year lag is a generalization rather than a conjecture. It is not a prediction of a 20-25 year lag between documentations. It is rather a generalization from cases where it is pure chance that we have some very meager instance of documentation, or cases where such a lag is actually indicated in documents ("It is not yet 20 years"), to cases of no prior documentation It says that based on experience we can't assume, without more evidence, that in this time period an invention was invented later than 20-25 years earlier than the first documentation.

This example of eyeglasses is instructive on that account. There is actually a debate on whether eyeglasses were invented in Pisa, where the preacher was from (although he hadn't lived there for 22 years), or in Venice. (Of course there is the statue in Florence of their supposed Florentine inventor; this can be ignored, although it is something that was believed by many for years. Claims made 350 years after the fact require evidence; claims made 20 years afterwards based on first-person testimony are more credible.) As is often the case, Wikipedia has oversimplified (as does, I think, Andrea's source).

Wikipedia gives a reference, to Vincent Iliardi's 2007 book Renaissance Vision from Spectacles to Telescopes, at ... &q&f=false. On p. 8 Iliardi says, referring to two scholars in Venice and one in Padua,
These scholars argue in varying degrees for a Venetian priority in the invention itself, which they claim was exported to Pisa by Venetian glass workers and/or Dominican friars from the Veneto. (13)
This whole page fortunately is on the Internet, so I won't type out the footnote.

The reason for the claim is that Venetian guild regulations talk about "discs for vials and for eyes" and say that they are repeating the regulations of 1284. It's the "for eyes" that gives the opening. Iliardi also says that making the frames for such discs is the simplest part of the process. It is the invention of lenses put close to the eyes that is different from what had been done for centuries, with the "reading stones" that Arab texts had talked about since the 10th century, held close to the object.

So do we go with the simplest theory or the more complex? This is another case where the principle of simplicity does not apply; no generalization about human experience or nature will exclude one or the other. The preacher might have assumed that the person he talked to was from Pisa, when he was actually from Venice. Or he was there without permission of his Venetian workshop and wanted to disguise the fact. Between Venice and Pisa, it is unclear, given what I have said, which has priority. More information and reasoning is needed. One might wonder whether Pisa would have had the precise glass-making facilities of Venice prior to 1286. Iliardi mentions them only for the early 14th century. Inventions do not typically spring from nothing. Also, Arab sources would have to be checked, because "reading stones" had been invented by them centuries earlier. In the end, we might find that the balance is in favor of

We might still favor one vs. the other, once we have accounted for the available facts; but not, if we are researchers as opposed to Wikipedia writers, on the basis of simplicity.

In the case of eyeglasses, we are dealing with an invention of great use to clerics, which you would expect them to mention, in sermons and other documents that were carefully kept. Yet there is still a 16-20 year lag. How much of a lag should we allow between invention and documentation in the case of tarot? In part, it depends on how many places we have to consider. Iliardi notes that there is much more documentation in Florence than in Venice. And Venice is a place that escaped the destruction that occurred in Milan and Bologna. If everywhere is ruled out but Florence, perhaps we can expect less of a lag than elsewhere. But of course Florence was the main center for artists and artisans. Tarot could have been invented in any of many other places and brought to Florence at the time of the Council. There is also the question of the type of thing we are talking about. Is it something regulated by guilds, or would the artists be on their own? My impression is that for tarot the artists would be on their own, even when they belonged to particular guilds. So no help there.

It might be good to research the history of consumer goods development generally for that period, to understand conceptually the multitude of variations in what actually happens and its documentation and lack thereof. Lynn White Jr.'s article, "The Act of Invention: Causes, Contexts, Continuities, and Consequences" (1962) is a good place to start (perhaps Michael has written about it, since he cites other works by White). White's focus on military inventions is a little off our subject, but note particularly his account of the invention of the piston; the Portuguese who developed it in Europe never mentioned the Malayans they almost certainly got it from.Another interesting invention is buttons. Their first documentation is in Koln, Germany, in the 1230s, where they can be seen on church reliefs. Yet they also have been discovered recently in 9th century Hungarian tombs. That is a case of a 370 year lag at least. Of course buttons are of little interest to the clerics who write things down, compared to eyeglasses and card games.

For tarot production in the period 1410-1440 (allowing somewhat more than 25 years to be safe), direct reports are best, of course: merchants' or notaries' records (it is not coincidental that the earliest non-court records are from notaries; it is their business to keep records); a letter talking about a new game using a bunch of extra cards described in a certain way, hopefully not just "triumphs", would be very nice. Few of us are in a position to do this research; but there are other indicators. If there is a spike of interest in Petrarch's Trionfi, as indicated by manuscript commissions, in the 1420s, for example, coinciding with a bunch of triumphal processions and theatrical productions featuring triumphs, or some tarot-sounding subjects (22 would be nice), that is suggestive. If there are illuminations in manuscripts suggestive of tarot subjects by then, that helps. There might be particular texts that the tarot triumphs seem to reflect. And so on. As of now we don't have nearly enough relevant information in one place compared to what is out there, scattered and in a variety of languages. And new books and articles on the period are being published every year. These are all indirect means of filling the "black hole" before 1440--and also the hole with not a lot of light after that date. However even then, it will only apply to invention/production in the courts. Before that, unless prior non-court production is somehow unlikely, the time of origin is still unknown--although better known than otherwise.

How important is it to arrive at the truth about the origin of the tarot? I think that its importance is overrated. The meaning of particular cards is determined by specific historical contexts and conventions operating at the time of specific productions, between producers and consumers, including the various ways the cards were used (gambling, instruction, meditation, poem-writing, dialogue-writing, essay-writing, inspiration for pictorial art, fortune-telling, etc.). As such, to interpret particular aspects of cards, it would be helpful to know exactly where and when such imagery started. But we already know quite a bit about the period in general, enough to interpret most aspects. In many particulars, meanings change slowly over time. It matters little whether the CY has 15. 16, 20, 21, 22, 24, or 25 cards, whether the PMB has 16, 20, 21 or 22, or whether the Charles VI has a Popess or Devil. (I do have trouble with other numbers for these decks.) And meanings do change with time, leading to an increasing variety of interpretations.


I think it is important to distinguish between explanation and interpretation--even though in ordinary language the terms are often interchangeable. An explanation, in the 1967 Webster New World Dictioanry's third definition, is to "account for; state reasons for"; To explain something that happens is also to give its causes, I see on one website. For artworks, I think, these may include conditions of which the actual author may not have been aware. I choose this definition because it seems as though the explanations we are looking for are more than "clarifications": they answer "why" and "how" questions for the facts at hand, not just "what". It is only then that there could be any similarity between tarot theories and scientific theories of the sort that the "simplicity" argument would apply to.

To interpret something is to explain its meaning. Besides this basic definition, there is also, in Webster's third definition, "bring out the meaning of, esp. to give one's own conception of, in performing, criticizing, or producing a work of art". Meanings are a matter of the relationship between a text or image and someone's understanding of it. Meanings take on a life of their own independent of their creator's intention, especially in the case where the creator's intentions are unknown. To understand the creation it is not necessary to fathom the creator's intentions. Judicial laws, for example, do not mean whatever their creators intend them to mean; poorly constructed laws, for example, sometimes mean something quite different from their author's intentions. Even in the best of cases, they do not mean what those who created or adopted them intended them to mean: they have meaning, and are intended to have meaning, in situations far beyond anything an original legislator would have imagined.

In the case of explanations, it is important to consider whether one's explanation actually explains anything. For example, if I say that the Cary-Yale deck could have been made in 1468 as an exact copy of one made before 1447, I might advance that as part of an explanation of why there are the particular banners there are on the CY Love card: in 1468 a Visconti duke married a Savoy duke's daughter. But that is not relevant to the question of why there were two banners on the original deck before 1447: it is that fact that needs to be explained. A 1468 deck does not explain a pre-1447 deck. There might be facts that a 1468 deck would explain--if, for example, the paint used wasn't in use until the 1460s--but there is no point in explaining hypothetical facts, unless they in turn explain some relevant known facts. Removing non-explanatory "explanations" is not the same as requiring the simplest explanation.

The two concepts, explanation and interpretation, overlap. A text or image is made for the purpose of being understood in a certain way or ways, and perhaps even not understood in certain ways. But as I say, the meaning of the work has a life of its own, beyond the intention of the creator and even beyond the conventions of his or her time, which may be forgotten while the work itself survives. In such cases, however, it is important to distinguish between what a work meant at such and such time from what it meant at a different time, or means now; and to distinguish among various audiences of the work, including even audiences of one person, a particular interpreter writing (or painting) at a particular time. There may also be primary meanings, inferred from conventions, and secondary meanings, inferred from particular contexts, both within the work and outside it, in other texts, images, and historical conditions, including innovations that are in no sense conventions.

In interpretation, even more than explanation, there is no requirement of simplicity. This has nothing to do with deconstructionism; one has only to read William Empson's Seven Types of Ambiguity, 1930 ( ... y_(Empson)). In literature, the point seems so obvious that when Umberto Eco repeatedly invokes the principle of "economy" to dismiss an interpretation as over-interpretation, I do not understand what he could mean by such a principle. If a principle, such as Kepler's laws, applies in one case, it applies universally. One cannot pick and choose, without a special argument for exemptions, where it applies and where it doesn't. In general, it doesn't. Literature would be very dull indeed without metaphors, puns, etc., all of which would seem to be eliminated by the principle of "economy." I perhaps am misinterpreting his principle of "economy".

In the Renaissance, multiplicity of meanings, even opposite meanings. was in many contexts considered very desirable, and not only by the interpreters but by the artists themselves. That is not to say that anything went. Even if the meanings were unlimited, that would not mean that anything went. An example from arithmetic: there may be an unlimited number of prime numbers, but very few numbers are allowed the term "prime". Even in the Renaissance, however, we are not dealing with a large number of possibilities. In the tarot, I doubt if in this particular time and place there are more than, say, 20--to pick a number that is about three times what even I can think of. I will give examples to explain what I mean, soon, over on the "meaning of the first six triumphs" thread.

Who is online

Users browsing this forum: No registered users and 9 guests