Tarot History Forum

I’m not denying Jewish persecution or the importance of Ferrara/Avignon communities: those are givens. The issue is evidentiary:

No phonetic bridge. We still lack any documented pathway from Torah/Tora to tarocchi/taraux. A surface resemblance isn’t enough; we’d need texts showing Renaissance Italians or Provençals actually reshaping the word. And “Torah” was already a stable, recognized form in both French and Italian, with no sign of it morphing into new variants.

Chronology doesn’t fit. The 21+Fool scheme predates 1492 (Sola-Busca, Boiardo, Steele) & persists afterwards. The name tarocco appears at Brescia in 1502, then Ferrara/Avignon in 1505, and spreads rapidly into regions with little if any Jewish presence. That looks vernacular, not esoteric.

Suffix problem. -occhi vs -aux needs a real explanation.

By contrast, the “fool” etymology has direct lexical support (tarocco = foolish/peevish; taroccare = to play the fool) and ties neatly to the Fool’s unique role in play.

As for the “ChatGPT suffix adaptation” idea, it’s speculative, not evidence, and ignores the fact that “Torah” was already and remained a known and stable loanword.

Also, your “coincidence stacking” (Torah = Tarot, 22 letters = 22 cards, Tav legs in Noblet, Jewish centers in Ferrara/Avignon) is circumstantial. And more to the point: Torah was sacred. Naming a card game after it would have been not only implausible but irreverent in the extreme.

SteveM

1- Jews and Jewish persecution. It is great we can agree that Jews actually did exist in Ferrara and Avignon. That's actually a big concession as the entire opus of Tarot history has a truly weird Judaic blind spot.

2- No phonetic bridge. I agree there is no documented bridge just a strong phonetic similarity. If we had a phonetic bridge this would be fact and not theory.

3- The chronology: The timeline fits perfectly. How could it possibly fit better? If the word Tauraux / Tarochi arrived in 1400 or 1600 that would be a strong point of criticism. I think the issue is that you think the exile of 1492 was a spontaneous surprise event. It wasn't. The persecution of Jews had been building up for many decades before hand. Example: The exile of French Jews started in the 1300s.

I do agree that nearly all decks that copied the original Torah card template had no idea what they were copying. I also don't see how Torah cards could become a mass hit.

4- Suffix problem: Agreed, I don't have the language skills to tackle that in earnest.

5- The fool tarocco etymology: Again: Can you please explain how this word was entirely unknown to Dummett? Was Dummett's Italian really that bad? How did a word so rare become such a popular name for Tarot cards? And, again, what is its phonetic bridge to Tauraux that needs to occur in the same exact year?

In short: Any phonetic or etymological criticism against Torah / Tarot is doubly relevant against Taroch given that Torah was a far more common word.

6- Coincidence stacking. Yes, the coincidence stacking is entirely circumstantial. Let's review the circumstances again:

22 Hebrew letters 22 Major arcana
1475 Abraham ben Ḥayyim opens a Hebrew printing press in Ferrara
I483 in Soncino the first Talmud tractate was printed.
1492 Ferrara is openly declared a safe haven for Crypto-Jews
1500 Avignon is the home of the 'Pope's Jews' and even absorbs another wave of exiles from greater Proveance.
~1500 The Duke of Ferrara grants a gambling license to the Jews of Ferrara
1505 Tarochi and Tauraux make their grand debut in Ferrara and Avignon right when the region has a large influx of Jews fleeing Spain
1520 The Christian printer, Daniel Bomberg, prints the first complete Talmud in Venice.

Here is an example of a Jewish scholar from Avignon who ended up translating Hebrew texts for the Duke in Ferrara around 1500: https://en.wikipedia.org/wiki/Abraham_Farissol

This is the historical context in which we are saying "Huh, these two words sound an awful a lot alike!"

I understand the refutation of all of the above is:

"Oh, there was this very rare word that Dummett never knew about, and it explains close to nothing, but, you know, this other theory got here first so all of the coincidences above can be safely ignored."

But I concede all of the above is not proof. The Torah / Tarot theory is just a theory.

For true proof, I'd have to produce a truly unambiguous 100% confirmed Jewish symbol in a historical deck of Tarot.

Right?

Stav, the argument you're trying to make here will not persuade any sensible person who is well informed on this subject.

Here are some reasons why, in increasing order of importance:

- Currently , the earliest known use of the name from which "tarot" is derived is its appearance as "tarocho" in the lyrics of a jocular song known as a barzelletta, published in Brescia in 1502 and brought to our attention by Thierry Depaulis in "Entre farsa et barzelletta: jeux de cartes italiens autours de 1500," The Playing-Card 37, no. 2 (2008), pp. 89-102. Thierry correctly observed that the Brescia edition could not have been the first edition, as it lacks one line that appears in the later editions (p. 92). So the song is probably from a slightly earlier date than 1502. The publisher of the Brescia edition, Bernadino Misinta, was active only in Brescia, Cremona, and Verona (p. 92-4), and the song indicates that the tarot game was played using the "Type B" order of northeastern Italy. For that and other reasons that I won't go into now, I think the song is most likely to have originated somewhere in the general region of Verona. In any case, this publication reduces the significance of the Avignon/Ferrara references from 1505, and strengthens the already-strong hypothesis that the name originated in Italy as "tarocco/tarocchi", with the "taro" form being a later variation, formed by French speakers dropping the "k" sound at the end.

- The historical evidence makes it pretty clear that the standard number of trumps in the tarot deck was established at 22 long before 1500, i.e., long before the "tarocco" name appeared:

A non-standard tarot deck with 22 trumps was designed by the poet Matteo Boiardo most likely sometime around 1480. Boiardo died in 1494, and a detailed description of the game was written by one Pier Antonio Viti of Urbino in a letter to a lady in Urbino; Viti was in Bologna in the 1480s, studying medicine, before returning to Urbino in 1490 or 1491, where he died in 1500. So his letter was probably written before 1490, and the game must have been invented sometime before that, which makes a date around 1480 probable.

There is also the evidence of the Steele Sermon, which I have argued was probably written in the 1460s; it is generally acknowledged to have been written before 1500 at least. viewtopic.php?p=27238#p27238

Finally, as others have observed, it is extremely likely that the deck had its standard set of 22 trumps before the tarot game began to spread to many different places, because it is virtually inconceivable that tarot players in all of those different places would have adapted their gameplay and their decks to the standard set of 22 if that set had been created somewhere at a later date. Players in different cities even developed their own local order for the trumps, and did not willingly copy the trump orders used in other cities, so they are exceedingly unlikely to have adopted any radical change to the number of trumps made in some other city. Even the removal of the Popess in Florence, which reduced the trumps there to 21, was not copied elsewhere – players in several other cities shared the Florentine animosity towards that card, but they replaced it with a different subject rather than just dropping it with no replacement at all.

The Minchiate deck did constitute a radical change in the number of trumps, but that was a special deck made for a particular game, and its designs, trump subjects and their order were established before it spread to many places. The standard 22-trump set had evidently undergone many changes in different cities by the early 16th century, both in the designs and order of the trumps. That indicates independent development over a lengthy period of time, not the spread of a new 22-trump game in the same manner as the spread of Minchiate.

- There is absolutely no evidence of any historical connection at all between tarot and Jewish people, culture, or religion. There are no records of Jewish people associated with the game or the deck, no Jewish authors writing about it, no evidently Jewish imagery on the cards. Nothing.

- The argument you are trying to make is, in essence, a linguistic one, but you neither have the linguistic expertise yourself to make that argument, nor have you consulted anyone who does.

As Steve has already pointed out, it seems extremely unlikely that the word "torah" could have become "taro" through any normal kind of linguistic evolution in the Romance languages. You find examples of common phonetic phenomena such as consonants being swapped (in the same way that some people pronounce "ask" as "aks") or consonants being dropped and vowels merged (such as "aqua" becoming "eau", pronounced "oh", in French) or vowels being altered. But you do not see examples of two vowels being swapped in a word while the consonants around them remain intact. Steve suggested that such a thing might be done as an "intentional disguise", but he was being overly generous to you by allowing that possibility, because that is not really credible to anyone who is not a hopelessly irrational conspiracy theorist. If you wanted to obfuscate the origin of something by changing its name, you would obviously change the name entirely – you wouldn't do something as ludicrously transparent as swapping the two syllables around. The Holy Inquisition was not staffed by morons.

If you had any linguistic expertise at your disposal, you would probably try to rescue your argument by noting that in the Semitic language family, words are often based on roots consisting of three consonants (or sometimes four or even five), with new related words being created by inserting different vowels between and around those three consonants. Hebrew is a Semitic language, so it is imaginable that Hebrew speakers could take the three consonants of "torah" (t-r-h) and change the vowels to create a new, related word "taroh". But in order to make that argument with any kind of credibility, you would have to get the assistance of someone with expertise in Hebrew to tell you how likely it would be for that to have been done.

From what little research I have been able to do, I can't find any instances of Hebrew speakers in any era using the word "torah" as the basis for creating new words in this way. Moreover, it seems that the word "torah" does not have any close relatives in Hebrew at all, which is to say, there are no other words using "t-r-h" that are related to it linguistically. It doesn't even seem entirely clear which consonantal root was the origin of the word "torah" in the first place.

So it looks like you would have difficulty arguing that the word could have become modified in this way even in Hebrew, let alone in the Romance languages, even if you did have the linguistic expertise to build that argument.

Hello everybody!

This hypothetical evolution of Torah into Tarot isn't natural but an intentional play on words. Therefore, I believe it's inappropriate to speak of etymology when proposing this hypothesis.

We have studied and discussed at length some possible etymological origins in this thread:
viewtopic.php?t=334&start=510

*********

Huck once suggested an explanation for the coincidence of two mentions of the Tarot game in 1505, in Ferrara and Avignon. He mentioned the Duke of Ferrara in connection with Avignon, but I didn't understand it. I asked about it later, and I don't think anyone has answered it.

HUCK, please !!

How did you connect Ferrara and Avignon in 1505? Did the Duke of Ferrara go to Avignon or order a deck of cards there? What was your explanation for this possible connection?

*******

On the other hand, I'm curious to know when the word "tarot" (ending in t) was first documented. Anyone know? It seems to be a fairly late word.

I hope I don't need to point out that the word 'Torah' and the 22 Hebrew letters existed long before the 1400s.

And the phenomenon of Crypto-Judaism existed long before 1492.

Of course the foundation of the argument that Torah is not the origin of Tarot is the claim that there are zero decks with any Judaic content.

But this is actually an error. I have conclusive evidence showing the exact opposite.

We have a deck of Tarot that is very clearly a deck of Torah cards - which makes the connection between the two words quite substantial.

Magic Moon wrote: 12 Sep 2025, 17:34

Huck once suggested an explanation for the coincidence of two mentions of the Tarot game in 1505, in Ferrara and Avignon. He mentioned the Duke of Ferrara in connection with Avignon, but I didn't understand it. I asked about it later, and I don't think anyone has answered it.

HUCK, please !!

How did you connect Ferrara and Avignon in 1505? Did the Duke of Ferrara go to Avignon or order a deck of cards there? What was your explanation for this possible connection?

:-) .... I say ....
"Avignon was dominated by France in 1505.
In 1505 Alfonso d'Este, duke of Ferrara ,assisted France in a militaric matter."

The KI of the moment , which not always says the truth, states to my combined sentences:
" The first statement, "Avignon was dominated by France in 1505," is true, as Avignon was a papal territory but was closely aligned with France. The second statement is also true; in 1505, Alfonso d'Este, the new Duke of Ferrara, became an ally of France and later distinguished himself as a general for the French during the Italian Wars, notably in the Battle of Ravenna. "

I definitly remember a historical statement about a cooperation between Alfonso and France, but I don't remember, what this precisely was.

Alfonso became duke of Ferrara in 1505. Ercole d'Este, his father, had died (probably) January 25 in 1505 . Alfonso had to make political decisions quickly..

Here is something about this theme ....
viewtopic.php?t=503

As far I remember, Alfonso had a journey to Northern countries, when his father became sick and Alfonso interrupted his journey. I don't find a reference to this journey in the moment.
Generally Ferrara was associated often with France in 15th century by an assumed French origin of the Este family, which brcame part of the texts of Boiardo and Ariost later. The "Orlando" topic.

Thanks, Huck. I find nothing in what you say to suggest that Alfonso would have stimulated an interest in the game, under a French name, in Avignon. What is more likely is that the French invasion of 1502 had stimulated an interest in the game of tarocchi among the French, which would have continued in 1505. The contract in Avignon was to supply "taraux" to the town of Pinerolo, a Savoy outpost in Piedmont that on the map looks like it would have been on the road between France and Italy. I have translated the contract at viewtopic.php?p=27152#p27152. Somehow I doubt that these packs of "taraux" would have been for eager Jewish gamblers in that backwater town. They were either for the local population or for soldiers passing through. Piedmont, of course, is where the word "tarochus" appears a little before then, in the 1490s. It is not much of a coincidence if this word gets to Brescia by 1502, now as the name of a very specific card game, and Ferrara and Avignon by 1505. It's to be expected that the word would get to those more distant places (from Piedmont) by around the same time, with or without Jews.

It would be useful, however, to know when, if Alfonso was traveling, he might have happened to visit Avignon - or Pinerolo - or also Lyon, also mentioned in the contract.

Stav wrote

We have a deck of Tarot that is very clearly a deck of Torah cards - which makes the connection between the two words quite substantial.

What is needed is not just any deck - there are many decks with Hebrew letters in them - but one used around 1500, preferably in the region bounded by Piedmont, Avignon, Brescia, Verona, and Ferrara.
There is actually one very good candidate, namely that represented in the Cary Sheet. It seems to me that the existence of a 2 of Coins found in the wall of the Sforza Castle with the date "1499" and the name of a Milanese card maker, the design as much like that of later French cards as the Cary Sheet is, is close enough (it is in Kaplan vol. 2). The Cary Sheet is not far in design from the Noblet; however, the Vieville is closer. Another candidate is the Budapest/Metropolitan sheets, although they are from the eastern edge of this region, Venice, and a little later. I look forward to your analysis, if you can extend it to any of these three.

It is quite possible that the Budapest Star card does represent a Jewish figure, namely David, who may by then have been associated with a "Star of David." However, if so, the model is likely to have been Michelangelo's statue. No Jewish source is required.

That is probably not what you have in mind. You write

Strangely, by coincidence In the Noblet Tarot we see the letter Tav, the 22nd letter, in the shape of the Fool's legs.

And Tav is, by yet another strange coincidence, the first letter in Torah.

Perhaps you will want to say that the Noblet is evidence of a tradition in which the 22 tarot images somehow reflect the 22 letters of the Hebrew alphabet, in order, by picking out details of the images that conform to certain details of the letters. Of course, the last letter of "Torah" is not the first letter of the alphabet, Aleph, at least as it is spelled today. But perhaps Aleph is somehow reflected in the World or Judgment (both were sometimes the last trump). Hopefully, what you say will apply to the decks I have mentioned. Even if it does, however, there are difficulties. For example, while all the Fool cards known did show a man with two legs, the same can be said for the Bagatto. There is also the Devil card in the Budapest set, who also has two legs, which perhaps could have evolved into the two small devils on the Noblet. It may be true that these legs do not match up as well with the letter as the Fool's; but I don't think the match is that close anyway. I look forward to your analysis, with correspondences that don't look arbitrarily selected.

Another problem: Even if such an analysis can be done, why would that suggest a connection to the name "Torah"? The Torah was the most sacred book of the Hebrew religious tradition, the first four books of the Hebrew Bible. True, it is made up of different combinations of the 22 letters, but calling any sequence of those letters, divided among three or four players in a game, by that name surely would have been denounced by some Jews as blasphemy and a mockery of the Jews' most sacred book. Not only that, they are woodcuts, which are precisely examples of "carved images" expressly prohibited as a means by which to represent God, which the Torah surely was understood as doing (God's message to His chosen people). Surely there would be written evidence of at least accusations of heresy among Jews, if not excommunications, if it was members of their congregations that were promulgating such a thing.

This is not to say that it is pure coincidence that the 22 cards correspond in number to the 22 Hebrew letters. That God''s original language had been Hebrew was reflected in the writings of several Church Fathers, most fully in some [changed from "the"] non-heretical writings of Origen which only in the 1460s were known in the West , due to Bessarion (in the book Bologna and the Tarot, see p. 3). The number 22 could have been chosen precisely so as to give the impression that there was something supernatural about how they would be distributed to players: good cards would go to those rewarded by God by His control over the cards (or that of devils under the control of the right magician). In short, it could be a marketing ploy by Christians, taking advantage of the common view of Jews as in possession of special and ancient forms of magic involving combinations of Hebrew letters. In fact there was a revival of interest in Raymond Llull's ars combinatoria around that time, among both Jews and Christians (which actually had nothing supernatural about it). There might also be the suggestion that "random" combinations of subsets of these cards could reveal the "destinies" of those who picked them, in the manner of Folengo's introduction to his tarot sonnets of 1527 Mantua - mysteriously appropriate card combinations, as though they were manipulated by unseen intelligences. But hardly expressions of the Torah, or even needing knowledge of Hebrew to interpret.

Sta wrote,

In Italian, the same word could be regularized into a form like tarochi, since Italian often pluralized borrowed words with -i and inserted a hard “ch” before i to preserve the consonant sound. In both cases, speakers who did not understand the Hebrew original adapted the sound of Torah into endings that felt linguistically “correct” within their own tongues."

You lost me here. Was "torah" pronounced with a hard c sound at the end in Italy but not Provence or France? That surely requires evidence.

On the other hand, Nathaniel wrote,

...the "taro" form being a later variation, formed by French speakers dropping the "k" sound at the end.

That, too, is without evidence, as far as being a pattern. Depaulis wrote an article suggesting just the contrary: words ending in -cco never, that he could find, simply dropped the -cco or -cchi; they became -que in French, e.g. "Barocco" turning into "Baroque" in French ("Des 'cartes communement appelees taraux', 2ieme partie", The Playing-Card 32, no. 6, p. 248). On the other hand, Depaulis adds, there was a massive borrowing from French into Italian in the 16th century; in particular, he says, the suffix -cco is "a classic augmentative", adding "This, linguists say, is a process much easier to explain." However, he gives no evidence for that assertion. Instead, he just says it is the opinion of the German-Swiss linguist Walter Haas, apparently a personal communication. Perhaps there are some examples in his source for the -cco to -que shift, vol. 1 of Lexical borrowing in Romance languages, by T. Hope - unfortunately, I cannot find it online.

What there is evidence for is a parallel development out of the Arabic "tarh" with a hard "h" or "ah" at the end, like the "ch" in "Bach", which got dropped in some words, both in Italian and Provençal, but not in others, while the "a" sometimes became "o". In Provençal there was the "-o" sound without the hard c ending. See Frederic Mistral, Lou Tresor dou Felibrege; ou, Dictionnaire provencal-francais, embrassant les divers dialects de la langue d'oc moderne, vol. 2, p. 959 in archive.org, entries for "TARO" "TAROU" and "TAROUS". The latter is said to be equivalent to the French "taré", meaning "crazy". It is a clear counter-example to Nathaniel's claim that "you do not see examples of two vowels being swapped in a word while the consonants around them remain intact," [added next day: well, not of a swap, but of the relevant vowel variation at the end]. The reason is that the word that both derive from is the Arabic "Tarh" with a dot under the h, meaning it is aspirated, also given as "tarah". I suspect that medieval Arabic may have been like Hebrew, with the vowels poorly indicated. Arabic speakers themselves may have varied in their pronunciation, [added: and of the first vowel, too].

In Italian, the corresponding word is "tarato", with the same meaning, from the verb "tarare", first person "taro", third person "tara". That is another way that vowels change while keeping the consonants: in conjugations. (I do not propose that operating here; it is just another counter-example.) Whether Arabic-derived words vary among Romance languages in vowels other than their last I don't know.

This is not to deny that the word meant "fool". All those words - tarous, taré, tarato - mean "crazy." I am not sure how "tarou" or "tarous" would have been pronounced, but it might well have been a long "o", as in the Cajun pronunciation of "bayou," which in Hank Williams' song "Jumbalaya" rhymes with "my-oh" and "gumbo".

There is a ready explanation for why such a word - tarocco in Italian, tarous in Provencal, taraux its French spelling - would have been chosen as a new name to supplant "trionfi" - an important rule change involving the Fool card. It seems to me quite possible that the card changed from being the lowest trump to being the "excuse," playable at any time. That fact would make it very much the "game of the Fool."

If it could be shown that something similar happened with the Hebrew word "t-r-h", ending in an aspirated "h" pronounced in two different ways and the vowels rather vague, Sta might have something, especially if the second letter was sometimes pronounced "a" instead of "o". However, the onus is on Sta to provide the documentation. It does not necessarily require knowledge of linguistics. Etymological dictionaries would be good enough for me. I can't find the necessary variability in the vowels as the word appears in the different languages, however.

Nathaniel wrote,

Finally, as others have observed, it is extremely likely that the deck had its standard set of 22 trumps before the tarot game began to spread to many different places, because it is virtually inconceivable that tarot players in all of those different places would have adapted their gameplay and their decks to the standard set of 22 if that set had been created somewhere at a later date.

I am at a loss as to why this should be so. What difference does it make to game play if there are 14 trumps and 56 suit cards vs. 22 trumps and 48 suit cards, for example (to take the example of the 70 card trionfi decks of 1457 Ferrara)? Or 78 cards vs. 70? With 72 cards, I can see that in one case, there is a remainder after the deal, and in the other case not. But that is not a very significant detail. Also, "many" is rather vague. What would be excluded by "not many"? Brescia, Avignon? Please explain.

mikeh wrote: 15 Sep 2025, 14:14 That God''s original language had been Hebrew was reflected in the writings of several Church Fathers, most fully in the non-heretical writings of Origen, which only in the 1460s were known in the West, due to Bessarion (in the book Bologna and the Tarot, see p. 3).

I just want to point out that some of Origen's writings survive in Latin translations made by Jerome and Rufinus of Aquileia. Specifically homilies on I Samuel, the Song of Songs, Isaiah, Jeremiah, and Ezekiel. His two Homilies and a Commentary on the Song of Songs survive only in Latin. For critical texts of these see Wilhelm Adolf Baehrens, Homilien zu Samuel I, zum Hohelied und zu den Propheten; Kommentar zum Hohelied; in Rufins und Hieronymus' Übersetzungen (1925). https://archive.org/details/origenes-we ... 3/mode/2up

The Commentary on the Song of Songs was well-known in the Middle Ages. For example, it is cited frequently in Saint Bernard's and Thomas the Cistercian's own sermons and commentaries on the Song of Songs.

I haven't found any reference to the Hebrew alphabet in those works, though. I don't know if you are only referring to that, or to the knowledge of Origen in general.

Thanks, Ross. What I wrote was misleading. It is not that all of Origen's non-heretical writings were unknown in the West before Bessarion, just some of them, in particular the collection that had his thoughts on the Hebrew alphabet. I have changed "the" to "some" in my previous post. I was thinking of the Philocalia, which is an edited collection of Origen's writings brought over by Bessarion. The piece on the alphabet is very near the beginning. For the details, see my post at viewtopic.php?p=26166#p26166. (I had forgotten that I had written about it here.)

I have also edited my comment on Nathaniel's reply to Sta, which also had a misleading component.

mikeh wrote: 15 Sep 2025, 14:14 Depaulis wrote an article suggesting just the contrary: words ending in -cco never, that he could find, simply dropped the -cco or -cchi; they became -que in French, e.g. "Barocco" turning into "Baroque" in French ("Des 'cartes communement appelees taraux', 2ieme partie", The Playing-Card 32, no. 6, p. 248). On the other hand, Depaulis adds, there was a massive borrowing from French into Italian in the 16th century; in particular, he says, the suffix -cco is "a classic augmentative", adding "This, linguists say, is a process much easier to explain." However, he gives no evidence for that assertion. Instead, he just says it is the opinion of the German-Swiss linguist Walter Haas, apparently a personal communication. Perhaps there are some examples in his source for the -cco to -que shift, vol. 1 of Lexical borrowing in Romance languages, by T. Hope - unfortunately, I cannot find it online.

What there is evidence for is a parallel development out of the Arabic "tarh" with a hard "h" or "ah" at the end, like the "ch" in "Bach", which got dropped in some words, both in Italian and Provençal, but not in others, while the "a" sometimes became "o". In Provençal there was the "-o" sound without the hard c ending. See Frederic Mistral, Lou Tresor dou Felibrege; ou, Dictionnaire provencal-francais, embrassant les divers dialects de la langue d'oc moderne, vol. 2, p. 959 in archive.org, entries for "TARO" "TAROU" and "TAROUS". The latter is said to be equivalent to the French "taré", meaning "crazy". It is a clear counter-example to Nathaniel's claim that "you do not see examples of two vowels being swapped in a word while the consonants around them remain intact," [added next day: well, not of a swap, but of the relevant vowel variation at the end]. The reason is that the word that both derive from is the Arabic "Tarh" with a dot under the h, meaning it is aspirated, also given as "tarah". I suspect that medieval Arabic may have been like Hebrew, with the vowels poorly indicated. Arabic speakers themselves may have varied in their pronunciation, [added: and of the first vowel, too].

In Italian, the corresponding word is "tarato", with the same meaning, from the verb "tarare", first person "taro", third person "tara".

In the article by Depaulis that you cite here, he also says that if a French term "tarau" was borrowed into Italian and replaced the earlier Italian name "trionfi" for the tarot game, this would be a "paradoxical swap" (échange paradoxal) that would be difficult to account for, given that the tarot game was clearly invented in Italy. This is the big difficulty that I too have with the idea that the "tarot" name could have been French originally: the game must have been far more firmly and widely established in Italy than in France at that time, and the name "tarocco" actually appears even earlier there than in France. So it's very hard to believe that the Italians imported a French name for the game.

But you comments about Provençal, French, and Italian words derived from Arabic طرح (ṭaraḥ) are entirely valid. The Italian word "tara" seems to be the earliest attested of these derivations: that goes back to at least the 13th century, and is believed by at least some linguists to be the origin of the French term "tare" (see e.g. https://www.cnrtl.fr/definition/tare); if so, then it is probable that the Italian also predates, and was the origin of, the Provençal terms.

But French "tare" goes back to at least the 14th century. So the age of these terms in Italian and French doesn't help us much to solve the question of which came first, the French/Provençal term "tarau" or the Italian "tarocco".

Given that "tarocho"/"tarochus" is attested in the late 15th century in Piedmont and "taraux" quite early in Avignon, the most likely answer is perhaps that the name originated somewhere in the Duchy of Savoy, either on the Provençal-speaking side of the duchy or on the Italian-speaking side, and then passed very rapidly to the other side.

If some derogatory word like "tarocho" already existed at that time on the Italian-speaking side (=Piedmont), then it's likely that a matching derogatory word like "taro" already existed on the Provençal-speaking side, and that bilingual people on both sides would have thought of them as equivalent. So it's easy to imagine how this new name for the game, derived from that derogatory word, would have passed very quickly from one side of Savoy's linguistic border to the other.

For what it's worth, I still think the name is most likely to have come from the Italian-speaking side – indeed, it may have been imported to the Provençal side along with the game itself, maybe sometime in the late 15th century. In other words, that may actually have been the moment when the tarot game first started to be played in what is now France.