The Real Origins of the Rashi, Rabbenu Tam Tefillin Dispute

The Dead Sea Scrolls Tefillin and the Making of an Urban Legend

Dr. Yehudah Cohn

In a renowned medieval dispute, the French Tosafist Rabbenu Tam (1100-1171) disagreed with his grandfather, the famous Bible and Talmud commentator Rashi (R. Shlomo Yizḥaki; 1040-1105) on the correct ordering of the four Torah passages found inside head tefillin (shel rosh).

To this day, some Jews don both versions of tefillin during morning prayers to ensure that they fulfill the mitzvah properly. Since the discovery of the Dead Sea scrolls inside the caves at Qumran (just off the present-day road from Jerusalem to the Dead Sea) and elsewhere in the Judean desert, many of which date to late Second Temple times, it has become a staple of encyclopedia entries and other such discussions to report that the findings from the Dead Sea included both competing versions of tefillin, previously unattested before the middle ages. On closer investigation, however, it turns out that such claims are little more than a persistent urban-legend.

In this article I will show why the ancient “evidence” does not in fact support a connection between the ancient tefillin discovered at Qumran and the medieval dispute. Instead, I will emphasize that the origin of Rashi and Rabbenu Tam’s debate centers on the interpretation of a talmudic passage.

The Rashi and Rabbenu Tam Tefilin Dispute

The dispute between Rashi and Rabbenu Tam about the order of the four biblical passages in the tefillin concerns the order of the two sections from Deuteronomy: Should Deut 6:4-9 appear as the final section (reading from right-to-left), as Rabbenu Tam suggested, or should Deut 11:13-21 be last, as Rashi ruled, and thus the order of all the passages follows the biblical order?

Rashi’s Order
D.  Deut 11:13–21
והיה אם שמוע
C.  Deut 6:4–9
B.  Ex 13:11–16
והיה כי יביאך
A.  Ex 13:1–10
קדש לי

Rabbenu Tam’s Order
C.  Deut 6:4-9
D.  Deut 11:13-21
והיה אם שמוע
B.  Ex 13:11–16
והיה כי יביאך
A.  Ex 13:1–10
קדש לי

How does this dispute match up with the archeological evidence?

Before answering this question, a caveat is in order: The dry climate in the Judean desert caves did a remarkable job of preservation for two millennia, but there was – inevitably – some measure of deterioration over time. Some tefillin cases had apparently rotted away entirely, even though their parchments had not. Alternatively, the Bedouin shepherds who first discovered the Dead Sea Scrolls, starting in the nineteen forties, may have unpacked the contents of some cases. In any event, most ancient tefillin parchments were no longer in their cases when discovered, so their original order cannot be reconstructed.1

The Two Tefillin Discoveries

Fortunately, two of the ancient head-tefillin cases found in the Judean desert still contained decipherable parchments at the time that the archaeologists examined them. One of the tefillin had (A) קדש לי (Ex 13:1–10) in between (B) והיה כי יביאך (Ex 13:11–16) and (D) והיה אם שמוע (Deut 11:13–21), whereas these ought to be adjacent in Rabbenu Tam’s view, or, according to Rashi, separated by (C) the שמע (Deut 6:4-9).

The second was a head-tefillin purchased by Professor Yigael Yadin from an antiquities’ dealer, shortly after the Six-Day War in 1967, which also did not correspond to the two types known from the middle ages. In the Yadin tefillin, the שמע was found in between the two passages from the book of Exodus, which ought to be adjacent to one another according to both Rashi and Rabbenu Tam.

Although a great archaeologist, Yadin was no dry academic. He had also served as the IDF Chief of Staff, and in later life was a politician. And we know that he saw virtue in trying to harmonize archeological findings with issues of continuing halachic concern.2 As for the tefillin he had proudly brought to light, Yadin may have confused subsequent scholars by making the observation that the order of the texts “is nevertheless most similar to that of Rashi.”3 The statement, however, is entirely without foundation – there are twenty-four possible ways of ordering four objects, and the ordering found can hardly be described as more “similar to that of Rashi” than any of the other twenty-three.4

The Origin of the Legend and
the Tefillin at Wadi Murabba’at

What’s at Stake?
It is easy to understand why the public would welcome an ancient precursor to Rashi and Rabbenu Tam’s famous debate.  Some years ago, the great Talmudist, David Weiss Halivni, told me that if the claims were true that the ancient tefillin correspond with the medieval disagreement, it would be the only known instance of a medieval dispute dating to the Second Temple era. Such an unbroken tradition of halachic disagreement, even as a singularity, would be of considerable significance for understanding the development of halachah. In addition, it would help resolve the mystery of how such a dispute could suddenly appear for the first time in the twelfth century. Surely, one would think, the two versions of tefillin must have pre-existed the two medieval masters, who might otherwise have resolved their differences by the simple expedient of opening up a tefillin case to see what it contained!

If neither of the two sets of parchments found inside the Judean desert tefillin cases corresponded to either Rashi or Rabbenu Tam’s view, where did the oft-repeated idea originate that both these types were found? The extent to which Yadin dwelled on the dispute in his publication may be partly to blame. More significant was his repeating a suggestion that had earlier been made separately by two renowned Dead Sea Scrolls scholars, Geza Vermes and Josef Milik, to the effect that another set of tefillin parchments, found in a Judean desert cave at Wadi Murabba’at that dated to the second century CE Bar-Kochba era, were effectively Rabbenu Tam tefillin.5

Wadi Murabba’at Tefillin: Two Parchments
In the Wadi Murabba’at tefillin, one parchment contained the two Exodus passages followed by והיה אם שמוע (Deut 11:13-21), while another much smaller parchment only contained the שמע (Deut 6:4-9). It could have been argued that this somehow constituted a prototype for Rabbenu Tam tefillin, in which והיה אם שמוע is also adjacent to the passages from Exodus.

However, this view is problematic, if for the simple reason that these tefillin were written in a single column, whereas the dispute between Rashi and Rabbenu Tam is about the correct right-to-left order of the four passages.6 In the Wadi Murabba’at tefillin, והיה אם שמוע may have been written on the larger parchment along with the Exodus passages and the שמע on the smaller one for practical reasons, namely that there was not enough room for והיה אם שמוע, which is far longer than the שמע, on the smaller of the two pieces of parchment available to the scribe.7

Be that as it may, next the then chief rabbi of the Israel Defense Forces (and later Chief Rabbi of Israel), Rabbi Shlomo Goren, took Vermes’ and Milik’s claim at face value, referring in an article he wrote to “the researchers of the findings in their book published in Oxford, in the Murabba’at volume.”8 With Yadin embracing it as well, the story of the Rabbenu Tam tefillin found among the Dead Sea Scrolls took on a life of its own.

Was Rashi’s order in evidence? On the one hand, Rashi simply follows the biblical sequence, which is hardly surprising. Yet given that multiple text sections in Judean Desert tefillin were written in a single column, they cannot really be viewed as confirming Rashi’s views on right-to-left order.9

At this point, it is fair to say that some of the Judean Desert tefillin contradict both Rashi and Rabbenu Tam’s views, with none found that provide any support for either of the two.

A Precedent of Ancient Tefillin Serving
as Evidence for Halachic Disputes

The modern scholarly claim that the Rashi / Rabbenu Tam dispute is confirmed by ancient tefillin in fact continues a pattern dating back to medieval times. In fact, the Tosafot passage that cites the view of Rabbenu Tam (b.  Menachot 34b) already links the dispute with Rashi to the venerated figure of R. Hai Ga’on.

“… and so too R. Hai Gaon … who would wear והיה כי יביאך and והיה אם שמוע as the innermost, adjacent to each other.”

The nature of R. Hai’s view was contested by other medieval rabbis, as the dispute turned into claims for competing traditions, which provided support for each of the opinions.10

Maimonides recounts being notified that R. Hai Gaon’s own set of tefillin had been opened, and found to contain the passages in their biblical order.11 R. Moses ben Jacob of Coucy relates similar results that had reportedly been obtained by checking tefillin found after a structural collapse at the Prophet Ezekiel’s grave.12 (According to R. Shem Tov ben Abraham ibn Gaon, the tefillin concerned had been found under Ezekiel’s skull.13) The Piske Tosafot impart information on two pairs of tefillin that were allegedly found in Neharde’a and Jerusalem – one like Rashi’s view and the other like Rabbenu Tam’s.14

The Actual Origin of Rashi and
Rabbenu Tam’s Debate

As is often the case with medieval halachic disputes, the disagreement between Rashi and Rabbenu Tam is not preserved in a self-standing source, rather it is found in competing interpretations of a few lines in the Babylonian Talmud (Menachot 34b). The passage in question concerns the right-to-left sequence of the scriptural passages contained in tefillin.

The Talmud reads as follows:

ת”ר: כיצד סדרן?

קדש לי והי’  כי יביאך מימין, שמע והי’ אם שמוע משמאל.

והתניא איפכא!

אמר אביי,  לא קשיא: כאן מימינו של קורא, כאן מימינו של מניח

והקורא קורא כסדרן

The Rabbis taught:  How is their sequence?

(A)  “קדש לי” (Ex 13:1-10) and (B) “והיה כי יביאך” (Ex 13:11-16) on the right, (C) שמע (Deut 6:4-9) and (D) “והיה אם שמוע” (Deut 11:13-21) on the left.

But has the opposite not been taught (in another baraita)?

Said Abbaye:  There is no difficulty.  Here (in the orientation of the earlier statement) it is “on the right” for a reader (i.e. a person facing a tefillin wearer), there (in the orientation of the other statement) it is “on the right” for a wearer.

And the reader (facing a tefillin wearer) reads in their order.

Rashi interprets the Talmud as stating that the individual passages in tefillin are placed in their biblical order:

והקורא קורא כסדרן : כסדר שהן כתובין בתורה מוקדם מוקדם ומאוחר מאוחר
And the reader reads in their order: In the order in which they are written in the Torah …

Rabbenu Tam
However, Rabbenu Tam understands the talmudic passage to mean that they are out of biblical sequence, with a reversal in the order of the two Deuteronomy passages:

והקורא קורא כסדרן :  … וקשה לר”ת דאמאי פלגינהו שקורא שתיהן הראשונות מימין וכל האחרונות משמאל … ומפרש ר”ת קדש והיה כי יביאך מימין של קורא ומשמאל של קורא הוי שמע מבחוץ ואחריה והיה אם שמוע מבפנים וניחא השתא מה שחלקו…
And the reader reads in their order: … And it (i.e. Rashi’s interpretation) was problematic to Rabbenu Tam. For why split them, so that one reads the first two on the right and the last two on the left … And Rabbenu Tam explains: (A) “קדש” and (B) “והיה כי יביאךfrom the reader’s right, and from the reader’s left (C) שמע is on the outside, and after it (D) “והיה אם שמוע” on the inside, and now the way they split it makes sense …

The view of Rabbenu Tam seems to be motivated by interpretive considerations, rather than from familiarity with an alternate version of tefillin that contained a different order.

Re-Reading the Text

To understand Rabbenu Tam’s objection to Rashi’s interpretation, let us re-read the text of the Talmud through Rashi’s eyes, again “lettering” the texts named there as A, B, C and D in their biblical order:

Rashi’s Order
D.  Deut 11:13–21
והיה אם שמוע
C.  Deut 6:4–9
B.  Ex 13:11–16
והיה כי יביאך
A.  Ex 13:1–10
קדש לי

The Rabbis taught: How is their sequence?  A and B on the right, C and D on the left – i. e. DCBA viewed from right to left.

But has the opposite not been taught (in another baraita)? – namely A and B on the left, C and D on the right.

Said Abbaye:  There is no difficulty.  Here (in the orientation of the earlier tannaitic statement) it is “on the right” for a reader (i.e. a person facing a tefillin wearer), there (in the orientation of the other tannaitic statement) it is “on the right” for a wearer.

It is clear from the language of his solution that Abbaye viewed one statement as representing the mirror image of the other. Therein lies a problem. For in a discussion of sequence in a language written from right-to-left, the phrase C and D on the right implies that C is to the right of D.

Thus, according to the opposite tannaitic view, C is on the far right. This cannot be the mirror image of the first alternative (according to Rashi’s view), where the placement of D on the far right would be the only way to reconcile the two alternative statements as referring to mirror images of each other.

Rabbenu Tam, on the other hand, solves this problem by reading the Talmud as follows:

The Rabbis taught: How is their sequence? A and B on the right, C and D on the left – i.e. CDBA viewed from right to left.

But has the opposite not been taught (in another baraita)? – namely A and B on the left, C and D on the right…

According to Rabbenu Tam, the “opposite” statement refers to ABDC. This view lends itself to Abbaye’s reconciliation, as ABDC is indeed the mirror image of CDBA.

Rabbenu Tam’s Order
C.  Deut 6:4-9
D.  Deut 11:13-21
והיה אם שמוע
B.  Ex 13:11–16
והיה כי יביאך
A.  Ex 13:1–10
קדש לי


If my analysis is correct, Rabbenu Tam emerges as a bold figure who determines halachah by interpreting the Gemara in his own way, even if this means flying in the face of received tradition – including the view of his revered grandfather!

As it turns out, there is little point in trying to find ancient historical proof for Rabbenu Tam’s view, as some mid-twentieth century scholars attempted in the wake of archaeological discoveries. Instead, Rabbenu Tam’s position likely arose from a radical reshaping of traditional custom. Yet, this too teaches us something about the development of halachah.  Namely, that in order to establish correct halachah one does not necessarily look into earlier custom, rather at new forms of interpretation.15


Yehudah Cohn completed a D.Phil. in Oriental Studies at the University of Oxford in 2007, and also has an MA from the Jewish Theological Seminary of America. Originally from London, Dr. Cohn had a career as a commodities trader in New York prior to beginning doctoral research. His dissertation was transformed into a book, for Brown University’s Judaic Studies series, and was published by the Society for Biblical Literature in 2008 under the title Tangled Up in Text: Tefillin and the Ancient World. More recently he co-authored a Handbook of Jewish Literature from Late Antiquity (135-700 CE) together with Fergus Millar and Eyal Ben-Eliyahu, published for the British Academy by Oxford University Press in 2012. He is currently a research associate at New York University’s Institute for the Study of the Ancient World.

  1. Even when tefillin cases were found to contain parchments, these were not always decipherable, and once again could not shed light on the order of the passages that they contained. Hopefully, the forthcoming investigation of previously indecipherable tefillin by Yonatan Adler will provide some additional data.
  2. Famously, for example, he reported at length on a visit to Masada by Rabbi David Muntzberg, an expert on mikva’ot, who pronounced that a pool found there was a kosher mikveh. See Yigael Yadin, Masada: Herod’s Fortress and the Zealots’ Last Stand (Jerusalem1966): pp. 164-166.
  3.  Yadin, Tefillin from Qumran (XQPhyl 1-4) (Jerusalem 1970): p.14
  4. These tefillin were also quite dissimilar to rabbinic ones in content, as two of the individual parchments combined sections from both Exodus and Deuteronomy, and three included additional text not prescribed by the rabbis.
  5. Geza Vermes, “Pre-Mishnaic Jewish Worship and the Phylacteries from the Dead Sea,” Vetus Testamentum 9 (1959); J.T. Milik, “Textes littéraires,” in Discoveries in the Judaean Desert, Vol. 2* (texte) (Oxford 1961): p. 81.
  6.  Indeed, even if the order of the biblical texts was entirely halachic, the Wadi Murabba’at tefillin, which were written on two pieces of parchment, could not possibly have been rabbinic tefillin, which are either on four pieces (head-tefillin) or a single piece (arm-tefillin).
  7.  Here it should be noted that Judean Desert tefillin seem to have been written on parchment scraps, so that this rationale would not be at all strange, even if it might at first seem so to a contemporary reader. Indeed, the phenomenon of writing on such tiny pieces of parchment is documented in Mishnah Shabbat 8:3, which mentions “(enough) parchment for writing the small passage in tefillin, which is Shema Yisra’el”.
  8. Shlomo Goren, “Hatefillin Mimidbar Yehudah Le’or Hahalakhah,” in Torat Hamo’adim (Tel Aviv 1964): pp. 496-511.
  9. In a lone exception, where the Deuteronomy passages were partly written over two columns, the bottom part of the right-hand column wraps under the left-hand one to form a single column – again, presumably, because והיה אם שמוע has so many more words than the Shema.
  10. See Yaakov Gartner, “Ha’im Heniach Rav Hai Ga’on Tefillin Shel Rabbenu Tam?,” Sinai 107 (1991).
  11. Blau, Teshuvot Harambam, Vol. 2 (Jerusalem 1960): p. 543.
  12. Sefer Mitzvot Gadol, mitzvat aseh 22.
  13. Migdal Oz commentary to Maimonides Hilchot Tefillin 3:5.
  14.  Piske Tosafot to Menachot, siman 92.
  15.  As for what the ancient Dead Sea Scroll tefillin can actually teach us about the history of halachah see my book Tangled Up in Text: Tefillin and the Ancient World (Providence 2008).
Print Friendly, PDF & Email