Sunday 10 February 2019


An early manuscript of Rashi's commentary found under the cornerstone of the old synagogue in Mainz when it was demolished in 1850 to make way for the new synagogue.


In Torah literature, we sometimes find that a work ascribed to a certain author, was in fact authored by someone else. Other times, the author is not the exclusive author and the work may have been the result of multiple authors.

We see this with classical books like Halachot Gedolot[1] and even the more contemporary Mishna Berura[2].

It is, however, most surprising to discover that perhaps the same may be said about sections of Rashi’s commentary on the Torah.

In this article, I draw extensively from the research of Dr Deborah Abecassis[3] who spent five years researching various versions of Rashi’s commentaries in over thirty different Rashi manuscripts and in over fifty manuscripts of the Tosafists which contained texts of Rashi. I also draw from Rabbi Dr Shnayer Leiman[4] and his profound research on the Rashi texts.

The research shows that Rashi’s commentary may be a lot more complicated than most would imagine. It appears that Rashi’s commentary continued to expand and develop during the two generations after Rashi’s death, and that significant portions were apparently added by the Tosafists who succeeded him.

[The term Tosafists generally relates to the rabbis of Northern France and Germany. Their period of influence lasted about two hundred years after Rashi - encompassing the 12th and 13th centuries, and ending with R. Meir of Rothenburg (d. 1293).]


Unfortunately, Rashi’s original manuscript does not exist.

The earliest extant manuscript of Rashi’s commentary is dated as late as 130 years after his death.
This gap of 130 years is concerning because the direct link back to Rashi is not there.


Rashi’s popularity did not help matters either. So many people wanted to read his commentary and as a result, manuscript copies spread far and wide in a very short time. More than any other commentary text, Rashi manuscripts were well circulated in all communities.

Many of these copies were not intended to be accurate versions of the commentary because they were used for practical study purposes - as opposed to transmission documents. Some included the private notes of his students and other scholars. As the process progressed, scribes continued to produce more copies, until a point was reached where “additional comments and explanations were often blindly incorporated into the body of the text.”

This created a problem. Rashi’s original writings were becoming less and less distinguishable from the additional commentary.

The challenges presented by enthusiastic students and scribes were obviously not unique to Rashi but they became more extreme and acute because of the popularity and ubiquitous nature of his commentary.

Professor Yaakov Spiegel[5] writes that the people of Ashkenaz (Germany) were known to have taken great liberties when it came to copying texts, even those of the Talmud itself and “the possibility of losing the original texts of these works was a genuine fear.”

This was obviously an issue because even before Rashi’s time, Rabbenu Gershom (950-1028) - who lived in Mainz, Germany and who headed the Ashkenaz community - felt motivated to issue a decree that no one should add to or ‘correct’ a text they were copying.


Following the invention of the printing press in the 1400s, Rashi’s commentary on the Torah was the very first Hebrew book to be printed. This added to his popularity.

If the earlier proliferation of manuscripts created problems, the mass production of printed versions created even more difficulties because it highlighted the innumerable differences between the printed versions and the hundreds of earlier Rashi manuscripts.

And, the first printing of Rashi took place more than 350 years after his passing.

Rabbi Leiman illustrates just how variant the early printed versions of Rashi were, by comparing nine editions of the first printings of Rashi on the same verse:

(The dates of printing and the names of the manuscripts are on the left, with the various Rashi texts on the right:)

Rabbi Leiman makes the startling point that in this case, no two editions printed the same text!

The printer, R. Moshe Alkabetz[6] wrote, in the colophon[7] of the 1476 edition of Rashi’s commentary, that he ‘relied on logic’ to eliminate errors. Unfortunately, there is no record of what changes he made and this again added to the complexity of an already complex matter.

Along similar lines, in 1482, R. Yosef Chaim who was the son of R. Aaron Strassbourg wrote is his colophon:

“I was careful to correct the commentary of Rashi, to restore it to its pristine glory, as much as possible...Because the words which were obscured...from so many errors, will now a light.”

Again, we just don’t know what changes were made, because he didn’t inform his readers. Nor it is clear just how he knew how to restore all of Rashi’s commentary back it to its ‘pristine glory’.

Dr Abecassis points out that while a student using a manuscript would clearly see corrections or amendments written by hand in a different ink or style, these would no longer be apparent in the printed versions. And, to compound matters even further, the wide distribution especially of the printed Rashi texts, would have lent them an air of unmistakable authenticity:

“Despite its tenacious popularity, few of the commentary’s readers are aware of the questionable nature of the text...

Even the modern printed editions that minimally provide alternate versions in the notes, do not express to the reader the true degree of uncertainty and unreliability in the text.

Most students of the printed commentary had and have no concept of how a text evolved through its transmission...”

[For more on the early printing of Torah books, see Daniel Bomberg – The Story behind the Tzuras haDaf.]


Rashi’s grandson, Rashbam (and others), attest that Rashi’s original commentary had maps and diagrams. These only exist, to some extent, in a limited number of printed versions.

Professor Mayer Gruber[8] explains that originally the scribe would have used a copy of Rashi’s commentary which included diagrams. In some manuscripts, the scribe would write ‘kazeh’ or ‘like this’ before leaving a blank space for an artist to later fill it in. However, that space was often left blank. Then, when they were copied again, the spaces were left out and the word ‘kazeh’ was removed as it was now redundant.


Professor Hananel Mak from the Department of Talmud at Bar Ilan University shows, as an example, how a well-known section of Rashi’s commentary may not have existed at the time of Rashi.

In Genesis 32:5, Jacob, tells his twin brother, Esau, that he had dwelt or sojourned with their mutual uncle Lavan.

Rashi comments that ‘sojourned’ (garti) comes from the root ger which means ‘stranger’.  Thus Jacob is subtly suggesting that he did not rise to a position of great status while living with Lavan and therefore there is no reason for Esau to hate him because their father’s blessing (that Jacob will dominate Esau) was not fulfilled.

Then the commentary continues with a ‘davar acher’ (another interpretation): ‘Sojourn’ (garti) has the numerical value of 613 (corresponding to the 613 commandments) – implying that although Jacob sojourned with wicked Lavan, he did not learn from his evil ways and he continued to maintain his righteousness by observing all the commandments. This was Jacob’s implicit warning to his brother Esau not to attack him because his (Jacob’s) righteousness was still intact.

Hananel Mak shows, however, how the second segment of Rashi’s comment could not have been part of his original commentary. The second section, beginning with ‘another interpretation’, did not exist in six of the most authoritative Rashi’s manuscripts which Mak consulted. It only began circulating during the 16th century (which is four centuries after Rashi’s passing).[9]

Yet, once it got printed in the Mikraot Gedolot editions of that same century it became indistinguishable from the other Rashi commentaries.


Although much of the populace was generally blissfully unaware of issues with the printed versions of Rashi’s commentary, there have been individual scholars who have attempted to study the earlier manuscripts to get closer to the original Rashi.

In the early 1500s, Avraham Baqrat collected early manuscripts and tried to collate them but his methodology was rather haphazard. He had access to manuscripts that may no longer be extant but the criterion for his selection of the variant texts is unclear.
In 1866, R. Avraham Berliner also tried to find a more accurate Rashi text and published a new edition of the commentary.

In 1982, Mosad haRav Kook, in an attempt at improving on R. Berliners’ work, published a new edition of Rashi’s commentary under R. Charles Chavel. This edition had access to editio princeps or first publications of Rashi, which the Berliner edition, although earlier, did not have.

These attempts acknowledged that Rashi texts were problematic in that so many different versions existed. However, it has been suggested that the most accurate way of trying to reconstruct a text as close as possible to Rashi’s actual text would be to examining the writings of the Tosafists.


Generally speaking, the Tosafists wrote anonymously[10]. Many of them were part of Rashi’s family, or students of students. They considered their work to be the extension of their teacher Rashi, and felt comfortable with adding comments here and there in furtherance of the culture of scriptural elucidation.

This created a milieu which over time made it very difficult to know exactly what Rashi wrote himself and what was added later and by whom. This, as mentioned was compounded when the printed versions began to appear.

Abecassis writes:

“Patterns in Rashi’s use of peshat and derash [literal and Midrashic][11] interpretations -  when he included both forms of exegesis, or when he cites only one, and why one precedes the other – cannot be determined from the printed editions...”

Put very poignantly:

“Rashi the eleventh-century French exegete, will remain an enigma until the text of his commentary is restored to a version as close as possible to the one he wrote.”

According to Abecassis,

“...the text of Rashi utilised by the Tosafot was significantly different from the [later][12] printed versions.

Examples show that portions of the printed interpretations attributed to Rashi are actually explanations and criticisms offered by the Tosafot that, through various processes, were attributed to the master himself.”

Interestingly, although the Tosafists inadvertently contributed to much of the confusion surrounding the original Rashi manuscripts, they could - nevertheless - be part of the solution to recapture aspects of the authentic Rashi texts.

It is possible to reconstruct a version of Rashi’s commentary, based on the citations of Rashi which are found in the writings of the early Tosafists. These citations would record the most accurate versions of Rashi’s commentary as they are the closest to his lifetime.

Rabbi Leiman makes a similar point that if one wants to sort out the interpolations from the original Rashi, then, quoting Leopold Zunz:

“[one] must make a comparative study of all the early manuscripts...and the citations in related commentators.”[13]

But of course, even if and when this work is done comprehensively, it is unlikely that it will ever gain acceptance in the wider community as they have come to trust the text they have always known in the popular printed editions they are familiar with from childhood.


Part of the challenge of understanding Rashi is that although he said that he only came to expound on the simple and literalpeshat’ of the Torah[14] - the fact is that around three-quarters of his text is Midrash (non-literal).

One way to reconcile what Rashi says about only expounding on the peshat, with what is actually printed, is to suggest that some of the later commentators included Midrashic material which later got conflated with his original commentary.

However, it’s not as simple as that because Rashi’s grandson, Rashbam writes that his grandfather told him that had he had more time, he would have written other comments more in keeping with the literal peshat!

This implies that he was aware and fully cognizant of the large Midrashic content within his commentary - unless he was hinting that already in his time he knew of the ‘additions’ to his text.

This last suggestion may not be so farfetched because as Rabbi Leiman writes: “...numerous glosses were added by others to all the...manuscripts of Rashi’s commentary (a practice already initiated during his lifetime).”

Either way, the true story behind the original text of the most popular commentary on the Torah, remains to be discovered.

But sadly, as Rabbi Leiman puts it: “ is unlikely that we will ever know with precision what Rashi really wrote.”

[1] Where, although it is assumed that Shimon Kayarra compiled it - it may instead have been authored by Yehudai Gaon. See here.
[2] Where, although it is assumed that the Chafetz Chaim authored the entire Mishna Berura exclusively, his son contributed certain sections as well. Just how much is uncertain.  See here.
[3] See: Reconstructing Rashi’s Commentary on Genesis from Citations in the Torah Commentaries of the Tosafot, by Deborah Abecassis.
[4] See: Was Rashi’s Torah Scroll Flawed?  by Shnayer Leiman.
[5] Amudim beToldot haSefer haIvri, by Yaakov Spiegel.
[6] Not to be confused with the author of Lecha Dodi, R. Shlomo Alkabetz.
[7] A colophon was used in the early printed books and still followed the tradition of handwritten manuscripts where the printer attested to the authenticity of the text.
[8] See: What Happened to Rashi’s Pictures? The Bodleian Library Record 15.2, by M. Guber.
[9] According to Mak, the Midrash about the numerical value of garti only begins to appear three Midrashic works (Lekach Tov, Bereshit Rabati and Midrash Aggada) from the 11th or 12th century. And although other pre-sixteenth-century commentators quote it, not one of them claims to have read in Rashi’s commentary, although they usually acknowledge Rashi by name.
[10] Besides the well-known Tosafist commentators like Rashbam, Yosef Kara, and Bechor Shor.
[11] Parenthesis mine.
[12] Parenthesis mine.
[13] Geschichtr und Literatur, Zunz, Berlin 1845, p.64.
[14]vaAni lo bati elah lipeshuto shel mikra


  1. No surprise. Almost all popular text from the middle ages have various deviations. It would be good to know in percentage how high the deviations are according to the manuscripts. The student of R. Berliner, R. Bamberger writes in his introduction to the German translation of Rashi that there is a letter of R. Tam cited in the HIDA which says that R. Tam was aware of the additions in his grandfathers commentary which are done by some R. Eliezer who compiled it into the Rashi from different sources. And R.Tam was fighting against it. I dont how reliable this citation is. You can find it here:

  2. Interesting and points to the fact that the true text by Rashi will never be known. Thank you for this blog as always...

  3. All popular texts until the invention of the printing press have various deviations. In particular, this is true of the Torah.