Forgive me if you feel this is not the best forum to ask this question on, but it seems the most apposite one out of those currently available.

The inscription that appears on the wall in Daniel 5:25 (mene, mene, tekel, upharshin) appears in Rembrandt's picture in columns, rather than rows.

That this arrangement was put forward in Jewish commentaries as a possibility, to explain why others might not have been able to decipher it, is noted in an article by French archaeologist Claremont-Ganneau, who references this source, although he does not give page numbers, so I have not been able to look into this further.

So there seems to be a precedent for the arrangement that Rembrandt used.

A couple of sources (on The National Gallery's website, for instance, and this article exploring Rembrandt's links with Jewish culture) state that he got one of the characters wrong. The latter source notes that he had originally painted a 'zayin' as the last letter, when it should be a 'nun sofit', but overpainted it.

There is more precision in Michael Zell's book, 'Reframing Rembrandt: Jews and the Christian Image in Seventeenth-century' in which he shows the x-ray evidence relates to a shift of the vertical descender from the right to the left (p 61). However, Alan Cohen argues that the shift may have been deliberate.

Is there a mistake, or isn't there?

  • The French article (possible a horrible translation of a French article) refers to a German work to support the contention that some "rabbins" prefer the up/down arrangement but without a source I can read or look up, I can't know if it is truly a precedented idea. The placement of the hand doing the writing makes it tough to see which letter-form of the nun will emerge.
    – rosends
    Commented Apr 29, 2015 at 13:35
  • Based on my limited knowledge of the history of Hebrew character forms, the last letter in the painting (bottom left) looks like the modern "Zayin" and nothing like a "Nun" or "Nun Sofit". As all the characters in the painting resemble their modern forms, I'm lead to believe that that (i.e. the Zayin/Nun) is the letter being discussed.
    – Lee
    Commented Apr 29, 2015 at 13:38
  • The Forward.com article I cited also reads a truncated 'nun sofit' as a 'zayin', but the tops of the letters are totally different, so I find this unconvincing - and you do say 'modern' Hebrew, which wouldn't apply in Rembrandt's case. Commented Apr 29, 2015 at 13:41
  • 3
    @Lee Zayin and Nun Sofit look exactly the same except the length of their legs.
    – Double AA
    Commented Apr 29, 2015 at 14:44
  • 1
    @DoubleAA Yishai's link below ("this chart") supports your point as regards STa"M script (the third chart in his link). As regards 'modern' script (the first chart in his link), IMHO the Zayin's head juts out to the right whereas the Nun Sofit's head does not.
    – Lee
    Commented Apr 29, 2015 at 14:49

3 Answers 3


The discussion is in the Talmud Sanhedrin 22a. The background is the disagreement among the Rabbis if the Torah was originally in Ivri or Ashuri. The Talmud says that according to the view that it was in Ivri, Ashuri script was first seen when the Angel wrote it on the wall, thus the Jews were not familiar with it - this is why they couldn't read it.

However, according to the view that Ashuri was always the script of Torah scrolls, the reason they couldn't understand it is that a letter substitution scheme was used. The first suggestion is Atbash, which would spell potentially plausible words that could be mistaken for having their face value meaning.

The next option (that of Shmuel) presented is the closest to the Rembrant picture. He says it was three words as Rembrant has them, but neither the Talmud nor Rashi say anything about them being one on top of another, instead of in a line, but that doesn't preclude such a reading. The Abarbanel in his commentary on the Book of Daniel Maynei HaYeshua 7:1 says that they were written one on top of the the other, exactly as Rembrant represents it. He died about 100 years before Rembrant was born, and he credits no one else with the idea, so that would appear to be the source for the exact representation.

There are additional similar letter possibilities suggested which don't seem relevant to your question.

I don't see any mistakes in the script Rembrant used. A modern reader might think that the Samech is close to a final Mem (compare with the one in this chart), but in fact some checking of some more contemporaneous printing (courtesy of here - see image below) finds that in fact the Samech lines up perfectly, but the final Mem of those days is how a Samech is currently printed. In any event those letters are very similar in shape anyway. The last Nun is a plausible, even reasonable final Nun (and a final Nun is what appears in the Talmud) - compare it to the image below. I originally though it was at an odd angle, but zooming in to the image convinces me that the visual effect is that the entire left-most column is spacing is wider than the other columns, and the nun starts higher than the adjacent letters from the same row, creating that effect.

1492 Hebrew printing

  • Thank you so much - I note from the link to the alphabet chart that while both forms of the 'nun' are written in 'Ktav Ashuri' with horizontal top line and descender created in one stroke, and the 'zayin' as 2 strokes; in 'Sta”M', both these letters have separate top lines and descenders, which confirms the dilemma I previously intuited and elegantly restates the question, or at least leaves it open. If this was an angelic hand, would the 'Sta”M' script not have been used? Commented Apr 29, 2015 at 14:50
  • 1
    @LeonConrad - Here is the Ketav STa"M Wiki article if it helps. STa"M is an acronym for Sefer/Sifrei (Torah), Tefillin and Mezuzot.
    – Lee
    Commented Apr 29, 2015 at 14:52
  • 1
    @LeonConrad The link provided is Ashkenazi Sta"m script ("Ksav Beis Yosef"). The writing in the painting is much more reminiscent of Sefardi Sta"m script ("Ksav Vellish").
    – Double AA
    Commented Apr 29, 2015 at 14:55
  • 2
    @LeonConrad, Rembrant wasn't really using the fancy STa"M script you see in the chart. He was using the printer script - which was around in his time, which as Double AA noted is very similar and modeled after KSav Vellish. You can see the differences between the different STa"M styles here.
    – Yishai
    Commented Apr 29, 2015 at 14:59
  • @DoubleAA, considering the prestige of the S&P kehilla in the Netherlands, I would be unsurprised. Commented Apr 29, 2015 at 22:44

To summarize from Yishai's answer, the Talmud says there was something funny about the way it was written; "in columns" is one possible interpretation. Assuming Manasseh ben Israel gave Rembrandt a sketch of what the letters should look like, I'd find it far more likely that Rembrandt was faithful to the sketch he was given (i.e. it was in columns) than that Rembrandt corrupted it.

From Jewish Virtual Library:

... there is no doubt that Rembrandt and Manasseh knew each other. In Rembrandt's 1635 Balthazar's Feast (National Gallery, London), a mysterious hand writes the words: Mene Mene Tekel Upharsin in Hebrew letters on the wall. Rembrandt may have consulted Manasseh about the script and in what manner the writing should be arranged. He wrote the words from top to bottom, according to an old Jewish tradition, which was later quoted in Manasseh's book De Termino Vitae.

As for how precisely the final letter is written, Rembrandt was a non-Hebrew reader who was most likely working off a Hebrew sketch someone had given him. It's certainly close enough that it would look fine to all but the most discriminating eye.


Assuming the writing in the painting is Ketav 'Ivri, which I humbly propose is quite clear, the incorrect letter in the painting is the final one. The verse in Daniel ends with the word "ופרסין" ("ufarsin"), spelled with a final Nun ("Nun Sofit") to indicate a plural noun. The Nun Sofit, in Ketav 'Ivri, should resemble the rightmost letter in the following chart:

enter image description here

A careful examination of the painting shows that the final letter more closely resembles the letter Zayin in Ketav 'Ivri. This is most clearly indicated by the tail on the right of the letter's "head" (see image below highlighting the tail in the painting and compare to the image immediately below it depicting a Zayin in the Ketav 'Ivri; it being the leftmost of the two depictions).

enter image description hereenter image description here

A Nun Sofit, as depicted in the rightmost letter in the chart above, does not include such a tail (i.e. the letter's "torso" begins from the rightmost portion of the letter's "head"), whereas the letter Zayin does. I therefore propose that the final letter in the painting is incorrect.

If one would like to claim that the final letter's tail is unintentional, I suggest looking at the second Alef in the painting. The second Alef's tail is smudged in a similar fashion to the painting's final letter. But, it is quite clear that an Alef in the Ketav 'Ivri also has a tail on the right portion of its head (as depicted in the painting's first Alef).

  • Ksav Ivri doesn't have final letters. I think you mean something else?
    – Yishai
    Commented May 6, 2015 at 13:21
  • Apologies if that's not its academic name. I'm referring to the alphabet that Wikipedia (Hebrew) calls "Alef-Bet 'Ivri", which is linked to in the beginning of the answer.
    – Lee
    Commented May 6, 2015 at 13:23
  • Two anonymous downvotes?
    – Lee
    Commented May 10, 2015 at 8:53

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .