marcus jastrow, - chagumoed.weebly.com€¦ · web viewalthough, elazar ben azariah speaks last...

The All-Night Seder in Bene Beraq: A Literary and Cultural HistoryJay Rovner

Introduction

The Seder in Bene Beraq at which five second century rabbinic sages reclined has inspired generations of

Jews who recount that event with admiration at their own Seders, year after year. They enthusiastically

narrated and examined the Exodus and the birth of the nation all night long. This how the standard

Passover Haggadah tells story.

�י �רב �יה ו �י א�ל�עזר ב�ן עזר �רב �הושע ו �י י �רב �ר ו �י אל�יע�ז מעש�ה ב�רב�י טר�פון �רב עק�יבא ו

�ין ב�ב�ני ב�רק סב מ� היו ש��לה י ל אותו הל �ר�ים ב�יצ�יאת מ�צ�רים כ ספ מ� �היו ו

�מן ק�ר�יאת ש�מע �יע ז , ה�ג ותינו לה�ם: רב �אמ�רו ידיה�ם ו �מ� תל עד ש�באו.ש�ל שחר�ית

The dramatic effect of this narrative is heightened by its placement in the Haggadah text. For it comes to

illustrate the propositions that every Jew is obligated to recount the Exodus, and that in doing so they are

worthy of praise (paragraph 2 below). Furthermore, it anticipates an exegesis of one of those fabulous

five, Elazar ben Azariah, who claims that the Torah itself mandates telling of the Exodus at night

(paragraph 4 below).

יה.1 �טו �רוע נ ב�ז מ�שם ב�יד חזקה ו �י אלהינו י וצ�יאנו ל�פר�עה ב�מ�צ�רים, וי ים היינו . עבד� ב�ני בנינו ו בנינו ו מ�מ�צ�רים, הרי אנו א א�ת אבותינו � הו ך לא הוצ�יא הקדוש ברו ו �א�ל ים ו מ�שע�בד�

ל�פר�עה ב�מ�צ�רים. היינו2 לנו כ . ואפ�ילו לנו ים, כ חכמ� לנו �ים, כ �בונ נ יוד�ע�ים א�ת התורה, מ�צ�וה עלינו לנו �ים, כ �קנ ז

�יצ�יאת מ�צ�רים. ר ב ל�ספ�כל המר�ב�ה �ה מ�שבח1ו �יצ�יאת מ�צ�רים - הרי ז ר ב . ל�ספ

�י טר�פון .3 �רב �י עק�יבא ו �רב �יה ו �י א�ל�עזר ב�ן עזר �רב �הושע ו �י י �רב �ר ו �י אל�יע�ז מעש�ה ב�רב�ין ב�ב�ני ב�רק סב מ� היו ש�

�לה י ל אותו הל �ר�ים ב�יצ�יאת מ�צ�רים כ ספ מ� �היו ו לה�ם �אמ�רו ידיה�ם ו �מ� תל :עד ש�באו

�מן ק�ר�יאת ש�מע ש�ל שחר�ית �יע ז , ה�ג ותינו .רב�יה .4 �ן עזר �י א�ל�עזר ב �י כ�ב�ן ש�ב�ע�ים שנה: אמר רב הרי אנ ,

�ן זומא �רשה ב ילות עד ש�ד �צ�יאת מ�צ�רים בל �לא זכ�ית�י ש�תאמר י ,ו�אמר , ש�נ �יך �מי חי ל י ר א�ת יום צאת�ך מא�ר�ץ מ�צ�רים כ �כ �ז �יך , ל�מען ת �מי חי ים- י , הימ� �יך �מי חי ל י כ

ילות- הל .ים אומ�ר�ים �יך : וחכמ� �מי חי �ה- י �יך , העולם הז �מי חי ל י ל�הב�יא ל�ימות המש�יח- כ .

1 Marbeh is a late addition. Just the simple act of story and discussion itself was the sufficient to achieve praise. This will be discussed in the final section, below.

/tt/file_convert/5e0575cb33a17772597ed1da/document.docx 5/15/23 1

The Bene Beraq narrative is of interest from a number of perspectives. As a textual phenomenon, what is

the nature of the tale? This will elicit generic and structural observations. Contextually, what is its

redactional history? This will lead to literary-historical and chronological reflections. The nature of its

redactional history invites a closer look at this story itself: can one trace its evolution and development?

Looking at the details, seeing what is typical in this narrative and what is not, sheds light on how it came

to be.

In the course of the examination, methodological issues will be mooted and some approaches and

claims will be challenged. Unique textual evidence will be utilized to clarify how this narrative originated

and evolved. We will see how its theme of sippur bi-yetsiat Mitsrayim (recounting/discussing the Exodus)

contributed to the development of a new approach to the Seder night’s activity. Where, after the younger

children had been introduced to the significance of the occasion, tradition had developed two conflicting

programs, one based on halakhic discussion, another on aggadic-midrashic expansion of,2 the Bene Beraq

anecdote dramatizes a Seder for adults and older children that features a less technical evening, i.e., a

program focused on narrative and discussion of the events. This popular, non-technical, approach for

adult participants developed in a post-Talmudic Babylonian milieu, probably after the seventh century.

Similarly, a popular literary form, Palestinian Targum, which augmented and rendered Biblical text into

Aramaic for synagogue goers unfamiliar with Hebrew, developed a little earlier, in the fifth-seventh

centuries in Erets Israel, after the completion of the Jerusalem Talmud. (In contrast, classical paitanim

had begun developing their complex and demanding poetic art form during the same period.)3

The Nature of this Tale

This narrative is nicely wrought. The exposition puts five sages in a location on Seder night. Their action

consists of an all-night discussion. This is balanced by another group, their disciples, whose action also

consists of speech, viz., the declaration that it is time to commence with the daily obligation to recite the

Shema. Contrasting themes are presented through a number of complementary pairs: sages versus

disciples; night versus day; unbounded all-night vigil versus time-bound liturgical obligation; the timeless

versus the quotidian. Implied, as well, though unstated, is the change of location, i.e., the house where

sages reclined versus the synagogue or house of study where they will join a quorum for communal

prayer (submerged in this version but present in its predecessors). While, the preceding paragraph (2

2 The first is the toseftan Seder type introduced in the next section; the second is the Scripture the semi-liturgical Miqra Bikkurim Midrash of the Babylonian style Haggadah, e.g., Ernst Daniel Goldschmidt, The Passover Haggadah, its Sources and History (Jerusalem: Bialik Institute, 1969), pp. 120-123.

3 See Avigdor Shinan,”The Late Midrashic, Paytanic, and Targumic Literature,” The Cambridge History of Judaism, iv: The Late Roman-Rabbinic Period (Cambridge: Cambridge University Press, 2006), pp. 691-695.


above), as well as the superior status of sages over disciples lead one to favor the first element in each of

those opposing pairs, the reality of this-worldly obligation expressed in liturgico-halakhic structure

remind one that the second elements are to be acknowledged and respected. Although, Elazar ben Azariah

speaks last in the extended text, the disciples get the last word in bringing the story proper to a close.

This narrative is a good example of a sage story in the form of an anecdote, which is a short

narrative about an incident in the life of a rabbi (here: rabbis) of religious or moral relevance.4 In our case,

this is a tale about a group of sages engaged in sippur yetsiat Mitsrayim; the latter is being promoted as a

praiseworthy activity. In this maʻaseh, the action is brought up short by a surprise ending, slightly

humorous but at the same time serious, which introduces a new consideration and perspective.

Its message is further enhanced by the following remarks of Elazar ben Azariah (paragraph 4).

Citing Ben Zoma he shows through midrash, a rabbinic expository technique, that the Torah itself

mandates that the Exodus should be mentioned at night. This grounding as an obligation5 provides

rhetorical support to the activity and main topic of the preceding story.6

4 Different writers tailor their definitions to the material they are dealing with. Thus, Moshe Simon-Shoshan, Stories of the Law: Narrative Discourse and the Construction of Authority in the Mishnah ((Oxford, New York: Oxford University Press, 2012), pp. 84-85 says that an anecdote is a brief story that focuses on a single incident, generally involving only a few individuals (p. 85), whereas Catherine Hezser, Form, Function, and Historical Significance of the Rabbinic Story in Yerushalmi Neziqin (Tübingen: Mohr Siebeck, 1993), says that they relate a “sequence of events in the life of a rabbi … of moral or religious relevance” (p. 309), consisting of a number of scenes geographically or chronologically separated; the narratives can be elaborated with details, dialogue, and even rhetorical elements (p. 310). Actually, most of the anecdotes she presents consist of two scenes, sometimes augmented by a “pronouncement.” The Bene Beraq story consists of two scenes, the one running into the other. In a sense, the ensuing citation of mBerakhot 1.5 could be seen as supplying the pronouncement in the words of Ben Zoma.

5 In addition to Elazar ben Azariah’s “pronouncement,” the nature of the obligation to rehearse the Exodus was enunciated through the verb, hayav, in the liturgical units connecting Avadim hayinu to the Bene Beraq narrative (unit 2 on page 1 above; table 1, below, units 4-6). That approach is examined below, pp. 28??-30??

6 Critical readers are quick to point out that this text refers to an ongoing nightly situation rather than a Passover Seder. Philip Birnbaum, The Passover Haggadah (New York: Hebrew Publishing Company, 1953), p. 23, and Shmuel and Zeev Safrai Haggadah of the Sages; The Passover Haggadah [in Hebrew] (Jerusalem: Carta, 1998), pp. 118-119, suggest linguistic and thematic connections that may have occasioned this borrowing, but they do not suggest that it produced a new understanding of the Mishnah pericope. Still, the Haggadah liturgist may here be engaging in a sleight of hand performance, albeit an artful and deliberate one. mBerakhot 1.5 sets out a debate concerning the exegesis of Deuteronomy 16, 3, with Elazar ben Azariah citing Ben Zoma to demonstrate that liturgical mention of the Exodus must be expressed at night (even though the paragraph of the Shema that mentions it is not then recited: see, e.g., the commentary of Hanokh Albeck, Shishah Sidre Mishnah, Jerusalem: Bialik; Tel Aviv: Dvir, 1957 [in Hebrew], 1:15; Saul Lieberman, Tosefta Ki-Fshutah, Zeraim, part 1 [New York: JTS, 1955], p. 12); and anonymous “sages” follow with an opposing exegesis of that passage that takes the words in a different direction.

The Haggadah liturgist seems to be continuing Elazar ben Azariah in Bene Beraq into the opening of that mishnah, to make it seem as if the exegesis there was reported at that Seder. (Some Haggadah texts add into to their mBerakhot 1.5 citations, “Elazar ben Azariah said lahem [=to them, i.e., to the other sages with him in Bene Beraq ...”) That could mean that the Shema exegesis was originally reported at this time (Meir Friedmann, Meir Ayin [reprint of ed. Vienna, 655 [1894 or 1895]: Jerusalem: Shai, 731 [1970 or 1971]), entertains the possibility on p. 86, or that it “is incorporated here merely because it discusses the importance of remembering the Exodus” (Joseph Tabory, JPS Commentary on the Haggdah [Philadelphia: Jewish Publication Society, 2008], p. 85) and /tt/file_convert/5e0575cb33a17772597ed1da/document.docx 5/15/23 3

The artistry of this little narrative can be further appreciated when one notices that the episode at

Bene Beraq was not the first all-night Seder vigil in the rabbinic record. The Tosefta7 tells of a group of

our sages’ contemporaries.

מעשה ברבן גמליאל וזקניםשהיו מסובין בבית ביתוס בן זונין בלוד

והיו עסוקין בהלכות הפסח כל הלילה עד קרותהגבר

8הגביהו מלפניהן ונועדו )צ"ל: וניערו(

והלכו להן לבית המדרש

Only one of this company, Gamaliel, is named; the location, Lod, however, is further particularized by the

identification of the host in whose home those sages had reclined.

typical of sympotic literature (idem, pp. 38-39). But many commentators correlate more intensively its Haggadah context with Elazar ben Azariah’s emphasis that Ben Zoma’s exegesis teaches about a nighttime practice (Ben Zoma did speak about both night and day). For example, Joshua Kulp (The Schechter Haggadah [Jerusalem: The Schechter Institute, 2009], p. 205) writes that “the editor has taken this derashah out of its original context (Shema and its blessings) and applied it to a new context (the seder).”

Despite the apparent self-contradictory nature of this construction of the exegesis, Kulp’s point of view can be defended in view of the rabbinic literary background of this text. In considering the echoes and refractions of this reuse, I am cognizant of two factors. One is that, a Passover eve/night hovers in the background, for the original intent of Scripture’s למען תזכר את יום צאתך מארץ מצרים כל ימי חייך (the crux of Ben Zoma’s and his collocutors’ disagreement) is that “the day of your departure from Egypt” is to be remembered/commemorated through the consumption of the Paschal offering along with matzah (Rashi) each year (Ibn Ezra), i.e., not literally “all the days of your life,” but idiomatically “for as long as you live” (New JPS). The second is that recontextualization in order to produce innovative connections, and new significations, is a stylistic technique ubiquitous in rabbinic literature and in piyyut. It is practiced in midrash with biblical phrases, and in Talmudic sugyot with both biblical and rabbinic teachings; indeed, this technique is employed to modify or infuse new meaning into existing halakhic passages. See for example, Shamma Friedman, “A Critical study of Yevamot X with a methodological introduction” [in Hebrew], Texts and studies: Analecta Judaica (New York: Jewish Theological Seminary, 1978), pp. 277-441. Moshe Halbertal, Interpretative Revolutions in the Making: Values as Interpretative Consideration in Midrashei Halakhah [in Hebrew] (Jerusalem: Magnes Press, 1997). It is here put into practice in the post-Talmudic Haggadah through the reuse of the early (mishnaic) exegetical passage.

It may seem somewhat disconcerting to see mBerakhot 1.5, including Sages’ seemingly irrelevant counter-exegesis, cited in full, but it is a habit of orally-oriented traditions and literatures to quote a passage in extenso even though only one section is contextually required (see Safrai 1998, p. 210). This is practice is followed in rabbinic literature. Moreover, the Haggadah contains a lengthy example of this practice: although Psalm (113-)114 would have sufficed to rhapsodize on the Exodus, the remainder of the Hallel (Psalms 115-118) is (re)cited in full, even though it commemorates later historical events and concerns. Still, one can find enriching thematic connections in both cases. Thus, the full Hallel carries the theme of redemption and salvation beyond Egypt into the ensuing Jewish historical experience; and the insistence that the Exodus will be recalled even in the distant future, after yet more spectacular events have culminated in the messianic era. That point opens an alternative possibility of understanding the intent of the Hagggadah’s reuse of mBerakhot 1.5, which will be presented in the section on ??“linguistic and liturgical matters,” pp. 30ff.?? below.

7 tPisha 10.12.

8 As corrected by Saul Lieberman, The Tosefta (New York: JTSA, 1962), Pisha 10: 12 (p. 198, variants and comments) based the reading in the Erfurt manuscript (ונוערו) with comparison to other usages (cf. his Tosefta Ki-Fshutah, Moed, part 4 [New York: Jewish Theological Seminary, 1962) p.656, on lines 34-35.


While it is not as tightly organized as the Bene Beraq narrative, which conveys a complex set of

messages in carefully arranged contrasting pairs, the toseftan tale is also artfully composed. A cascading

chiastic structure lets the distinction between “all night long” and the moment of “cockcrow,” hence the

division and contrast of the nocturnal action and the diurnal obligation, break right at the center (3/3´):

Expositionמעשה ברבן גמליאל וזקניםSages were reclining at Baitos’s home in Lod 1שהיו מסובין בבית ביתוס בן זונין בלוד

Sages discussed hilkhot ha-Pesah (the rules of Passover) 2והיו עסוקין בהלכות הפסחAll night 3כל הלילה

Until cockcrow ´3עד קרות הגברremoved the tables [Servants] ´2הגביהו מלפניהן

Sages aroused themselves and went to the bet midrash ´1וניערו והלכו להן לבית המדרש

Sages and their admirable scholastic engagement – this time hilkhot ha-Pesah (the rules of the

Passover offering and related matters) -- are contrasted with the class of servants,9 whose presence is not

even mentioned except by implication, in the execution of the humble task of clearing away the remains

of the rabbis’ meal, which, of course, was their Passover Seder meal, as opposed to the Passover topics

discussed by the sages. The servants’ undertaking, while not a rabbinic activity, indirectly supports it (2/2

´). The contrast is heightened by the fact the rules of Passover, which are collected and examined in the

tractate Pesahim, are concerned with the Passover evening meal, viz. the laws of separation from

leavening and matzah, the laws of the Paschal offering, and that evening’s ritual.10 The narrator also

focuses our attention on the sages by opening and closing this narrative with sages and their concerns,

from reclining at the Seder to attendance at the bet midrash (1/1´).11

The presence of two rabbinic texts, both similar and different, invites comparison, resulting in

mah nishtannah, a host of questions. Two obvious ones are. Which is the appropriate activity, inquiry into

9 Leberman, idem, idem, even after he corrected his understanding of נועדו/נוערו, seems to think that Gamaliel and his party themselves picked up the (individual) tables on which lay the remains of their meal. However, e.g., mPesahim 7.13 and tBerakhot 5.28 speak of the involvement of a shamash (attendant, waiter) at a meal, the latter in cleaning up. Paul Mandel, The Origins of Midrash: From Teaching to Text (Leiden, Boston: Brill, 2017), p. 193, also infers that “servants” were the ones who “lifted (the eating trays) from before them.”

10 The tractate Pesahim, in which they were probably engaged, covers the areas mentioned; the definite article in the term, hilkhot ha-Pesah indicates “the Passover offering.”

11 Contextually, this would probably be to engage in the morning liturgy with a quorum. Paul Mandel, ibid, pp. 182-190, explains that the bet ha-midrash in tannaitic literature signifies a location where sages, with disciples in attendance, made themselves available to serve the public by, e.g., answering queries, providing instruction and resolving disputes. He points out that this tended to happen on Shabbat and holidays, i.e., times when people would have the leisure to come (pp. 190-196) and cites tPisha 10.12 as an illustrative passage (p. 193 and n.64). Other examples Mandel provides indicate that liturgy was performed there (p. 192), and it is probable that Gamaliel and the elders repaired there to participate in a prayer quorum, after which they would provide an audience for the public. Mandel also quotes tSukkah 4.5, which (anachronistically?) relates a progress from the morning sacrifice to the synagogue, then the bet midrash, followed by “the prayer service of the additional sacrifices,” etc. (p. 195). Perhaps Gamaliel’s Lod did not have as many specialized sites as Second Temple Jerusalem.


hilkhot ha-Pesah or discussion of the Exodus? Why name Gamaliel, but not his followers, while Eliezer

and his whole retinue is enumerated?

There are others. How did the storyteller know who was with Eliezer? If Eliezer is the chief

figure, why did the other sages not come to him in Lod where he dwelt?12 The one sage who resided in

Bene Beraq was Akiva; why not report that the sages reclined at his house as they did for Baitos ben

Zunin in Lod? The Gamaliel episode is recorded in the Tosefta, a collection of early (tannaitic) sources,

while the narrative featured in the Haggadah is not known in any rabbinic composition: where did it come

from?

Many of these questions can best be resolved by a detailed analysis of the genesis and evolution

of the Bene Beraq narrative. But the first, and most important, one, viz., the primacy of halakhot versus

the Exodus can best be settled by a macrocosmic examination. It is the latter, an investigation of the story

in its context in the Haggadah, that can illuminate the process of the struggle for primacy, and who won,

to which we now turn.

The Narrative in Context: the Redactional History of a Section in the Passover Haggadah

There are two milestones in the development of the Passover Haggadah, one encoded in the Mishnah in

the tenth chapter of tractate Pesahim; the second was achieved during the time of the Geonim, from the

seventh to the eleventh century. During that period, there were two major liturgical families, one in Erets

Israel, and another which developed in Babylonia. Each had several versions. Those versions persisted

long after the Geonim had already stated their preferences. Even after one basic form would come to

predominate in each family, Genizah manuscript fragments from eleventh and twelfth century Haggadot

showed that alternative forms were still in use then. Evidence will be introduced below to show that

extreme variations persisted in some Babylonian-derived rituals as late as the 17th century.

The text and context under examination occurs only in the Babylonian branch. The Babylonian

Talmud reports that Rav and Samuel,13 two early Babylonian Amoraim, would introduce the major

section of the Haggadah, maggid, differently:14

רב אמר: מתחלה עובדי עבודת גלולים היואבותינו. ]ושמואל[ אמר: עבדים היינו.

12 bSanhedrin 32b.

13 David Henshke, Mah Nishtannah’: The Passover night in the Sages’ Discourse [Hebrew] (Jerusalem: Magnes, 2016), concludes that neither attribution is reliable. On Rav, cf. pp. 439f. and 445 (and n. 132)-449; on Samuel, pp. 445 (and n.134) and 447-449.

14 bPesahim 116a. The Talmud asks how to begin bi-genut (with disgrace), as prescribed by mPesahim 10.4, and each snippet commences in that vein and concludes with shevah (praiseworthy matter, as advised idem).


Both of those introductions came to be included in the Babylonian Haggadah, although each one was

enhanced by many subsidiary recitations. The Babylonian Talmud provides only the opening rubric. The

Jerusalem Talmud provides a more fulsome version of what Rav may have meant,15 casting some doubt

on whether mi-tehillah was actually part of what was originally to be said.16

רב אמר: מתחילה צריך להתחיל, בעבר הנהר ישבו אבותיכם וגו' ואקח את אביכם את אברהם מעבר

הנהר וגו'.

Babylonian Haggadot combine both approaches. We have no evidence for what the teaching attributed to

Samuel included, but that is important for us because Avadim hayinu leads into the Bene Beraq narrative.

It was certainly not the expansive version in current rites, as presented above, or even that in Saadia

Gaon’s Siddur (see below); a fragmentary text in Ginze Schechter17 undoubtedly presents something

closer to what might have been meant:

עבדים היינו לפרעה במצ]רים[ ויוצ]יאנו[ ה' ממצ]רים[ ביד חזקה ובזרוע נטויה.

That introduction is followed in Gaonic Haggadot by a text drawn from the Mekhilta de-Rabbi

Ishmael,18 viz., the baraita of the four sons. That text privileges hilkhot ha-Pesah over recounting the

Exodus, assigning the former to the wise son (who is mentioned first), and relegating the latter to the

simple son and the one who does not know how to ask. For this hierarchical presentation, the Tosefta’s

all-night Seder would be ideal.19 On the other hand, that baraita’s exposition does recognize that there are

four types of children, and the father should take them all into account in presenting the Seder to his

family. The Bene Beraq episode could be an idealized recreation of what such a Seder night might be

like. It reminds the Seder participants of the purpose of the baraita of the four sons, albeit transformed

15 yPesahim 10.5 (page 37d).

16 See Saul Lieberman, Tosefta Ki-Fshutah, Zeraim 1, Introduction, p. 21, n. 40; Henshke, idem, pp. 439-444. Lieberman explains that mi-tehillah=one should say or do as was originally done, and the scriptural source follows. The intervening tsarikh lehathil (“one should begin” is an explanatory gloss).

17 Louis Ginzberg, Genizah Studies in Memory of Doctor Solomon Schechter=גנזי שעכטער [in Hebrew] (Reprint of 1928 edition: New York: Hermon Press, 1969) 2:252-260 (the selection under discussion may be found on pp. 259-260). The shelfmark is Cambridge, CUL: T-S Misc.36.179 (Alternate number: T-S Loan Collection 179).

18 Mechilta d’Rabbi Ismael, ed. H. S. Horovitz, I. A. Rabin (Reprint: Jerusalem: Wahrmann, 1970), Bo 18 (p. 73).

19 Judith Hauptman compares and contrasts the Tosefta’s version of an earlier form of the Seder with that of the Mishnah in “How Old is the Haggadah?,”Judaism 51 (2002), pp. 5-18, and Rereading the Mishnah (Tübingen: Mohr Siebeck, 2005), pp. 50-63.


into scholars (paragraph 4 above, and compare paragraph 3).20 It also disregards the Tosefta’s privileging

of hilkhot ha-Pesah, concerning itself instead with the theme of the Exodus.

Still, the discussion portrayed in the Bene Beraq story is a sensible precursor to that exposition. It

is time to see how it came to be included in the Haggadah text. The following table helps one visualize

the difference between a Haggadah that contains it and one that does not.

Table 1. Two views of Avadim hayinu in the Babylonian Haggadah

Saadia’s Siddur21 Ginze Schechter Haggadah Natronai Gaon Haggadah22

1 עבדים היינו לפרעה במצרים ויוציאנו ה' אלהינו משם ביד

חזקה ובזרוע נטויה

עבדים היינו לפרעה במצ]רים[ ויוצ]יאנו[ ה' ממצ]רים[ ביד חזקה

ובזרוע נטויה

ל�פר�עה ב�מ�צ�רים, ים היינו עבד��יד חזקה מ�שם ב �י אלהינו י וצ�יאנו וי

יה. �טו �רוע נ ב�ז ו2 ואלו לא גאל המקב"ה

]=המקום ברוך הוא[ אתאבותינו ממצרים

כבר אנו ובנינו ובני בנינו משועבדים היינו לפרעה

במצרים

א א�ת � הו ך לא הוצ�יא הקדוש ברו ו �א�ל ו מ�מ�צ�רים, אבותינו

ב�ני בנינו ו בנינו ו ים עדיין אנו מ�שע�בד� ל�פר�עה ב�מ�צ�רים. היינו

3 ולא את אבותינו בלבד גאלהמקבה אלא אף אותנו גאל23 שנ' ואותנו הוציא משם.

4 לפיכך מצוה עלינו לספר ביציאת מצ'5 ואפילו כולנו חכ' כולנו נבונים וכולנו

יודעים את התורה, מצווה עלינו לספר ביציאת מצ'

ואפי' כול' זקנים כול' ישישים כול'יודעים את התורה,

מצוה עלינו לספר ביציאת מצ'

לנו ו כ ואפ�לו לנו ו ים, כ חכמ� לנו �ים, כ �בונ נ לנו ו �ים, כ �קנ יוד�ע�ים א�ת התורה, ז

�יצ�יאת מ�צ�רים. ר ב ל�ספ מ�צ�וה עלינו

6 וכל המספר ביצי' מצ' הרי זה משובח ר ב�יצ�יאת מ�צ�רים - הרי ל המאריך ל�ספ כ ש��ה מ�שבח ז

7 ור' יהושע אליעזר ומעשה בר'בן עזריה ור' עקיבה אלעזר ור'

,ברק- שהיו מסובין בבני,והיו מסיחים ביצי' מצ' כל אותו הלילה

�י �רב �הושע ו �י י �רב �ר ו �י אל�יע�ז מעש�ה ב�רב�י טר�פון �רב �י עק�יבא ו �רב �יה ו א�ל�עזר ב�ן עזר

�ין ב�ב�ני ב�רק סב מ� היו ש��סח �ר�ין ב�יצ�יאת מ�צ�רים ב�ליל פ ספ מ� �היו ו

20 The aspirational family activity is extrapolated from the comparison to the sages, per unit 5 in table 1, below.

edited by Israel Davidson, Simha Assaf , סדור רב סעדיה גאון = כתאב ג'אמע אלצלואת ואלתסאביח 21and Issachar Joel (Jerusalem: Mekitze Nirdamim, 1941).

ספר יובל Manfred Lehmann in ,סדר והגדה של פסח לרב נטרונאי גאון על-פי כתב-יד קדמון שברשות המחבר 22 ;edited by Shaul Yisraeli et al. (Jerusalem: Mossad ha-Rav Kuk , לכבוד מורנו הגאון רבי יוסף דוב הלוי סןלובייצ'יקNew York: Yeshiva University, 1984), pp .976-989.

23 Deuteronomy 6, 23.


: עד שבאו תלמידיהם ואמרו להם הגיע זמן קריאת שמע של, רבותינושחרית

לה�ם: �אמ�רו ידיה�ם ו �מ� תל עד ש�באו ותינו �יע, רב .שחר�ית ש�ל ש�מע ל�ק�ר�יאת ה�ג

8 אמ' ר' אלעזר בן עזריה: הרי אני כבןשבעים שנה,

ולא זכיתי שתיאמר יצי' מצ' בלילות, עדשדרשה בן זומא,

למען תזכור את יום צאתך 24שנ',מארץ מצ' כל ימי חייך,

- הלילות.כל ימי חייך- הימים, ימי חייך - העולםימי חייךוחכמים אומרים:

הזה, את ימות המשיח.להביא כל ימי חייך -

�יה �ן עזר �י א�ל�עזר ב �י[ כ�ב�ן: אמר רב הרי ]אנ,ש�ב�ע�ים שנה

�צ�יאת מ�צ�רים �י ש�תאמר י �לא שמעת ו�ן זומא, �רשה ב ילות, עד ש�ד בל

�אמר ר א�ת יום צאת�ך מא�ר�ץ, ש�נ �כ �ז ל�מען ת �יך �מי חי ל י ,מ�צ�רים כ

�יך �מי חי ים- י �יך , הימ� �מי חי ל י ילות- כ .הלים אומ�ר�ים �יך : וחכמ� �מי חי �ה- י ,העולם הז

�יך �מי חי ל י ל�הב�יא ל�ימות המש�יח.- כ

9 Barukh ha-Maqom25

10 Baraita of the four sons Baraita of the four sons Baraita of the four sons

The Ginze Schechter and Saadia Siddur versions exemplify two approaches to introducing the

maggid section of the Babylonian Seder by augmenting the text attributed to Samuel. (On the eclectic

Natronai Gaon version see below.) Avadim hayinu developed in two directions. In Saadia’s Siddur, units

2 and 3 were added, and then came 10. That version is designed to foster in those participating at a Seder,

at any point in time, an awareness of the significance of the Exodus event in its implications in their own

lives (2-3). Seeing the theme of freedom from bondage set out in terms of one’s own situation, arouses an

eagerness to find out more about the Exodus. From that vantage, the Seder liturgist then moves to remind

everyone that there are at least four different developmental and personality types to keep in mind as the

examination continues, and that it should be done over the Paschal meal (10).

In the Ginze Schechter family, Avadim hayinu is followed by the bene Beraq complex (7 and 8),

its transitioning introduction (4-6 or, in other exemplars, just 5-6) and the ensuing baraita (10). In this

version, the approach is to respond to the Exodus event by retelling and discussing it (4, 6). Even learned

elders should participate (5). In fact, the most learned and revered rabbis have done so; one session even

went on for the whole night (7-8). On the one hand, this version skips over the implications of the Exodus

for the Seder participants’ own situation; on the other, it has the advantage of anticipating the need for

awareness that there are least four different developmental and personality types to keep in mind in the

ensuing recounting and discussion (10). Moreover, people would come to realize through the process of

narration and discussion the existential significance of the Exodus event.


25 The barukh ha-maqom peroration was added in later textualizations, such as the version ascribed to Natronai.


Later Haggadot combined the two forms, as in the rightmost column above, which borrowed unit

2 from Siddur Saadia, with 5-8 from Ginze Schechter. A Haggdah included in a 17th century Siddur

according to the Persian rite provides an indication of how the transformation could have occurred for

those beginning with a Saadian version. In the body of the text, units 1-3 are followed by 10, as in

Saadia’s Siddur. Items 5 and 6 appear in the upper margin, in small lettering; and 7-8 appear in the lower

margin, also in small letters. One can imagine a later copyist simply copying 4 (or 5)-8 directly into their

base texts. That process is reflected in the state of the Haggadah version preserved in a liturgical

fragment from a different rite that was added by the copyist at end of the Siddur of Shelomoh bar Natan

of Sijilmassa (the Sijilmassa-B version).26 Interestingly, only units 4 and 6 are found there, suggesting that

the Bene Beraq-oriented expansion of Avadim hayinu was originally made up of either units 4 and 6 or 5

and 6 (see table 4?? at the end). Those elements could also be combined, as in the Ginze Schechter

version.

Most of the Haggadot that included Bene Beraq-oriented versions removed item 3, perhaps

because its contents, were already present further in the Haggadah, in the paragraph beginning be-khol

dor va-dor.27

This situation leaves us with a chronological conundrum with respect to the text on which our

inquiry is centered. The Bene Beraq sage story lies between two sets of material from widely varying

periods. Depending on how closely unit 1 reflects the actual wording of the text attributed to Samuel, it

may be from the Talmudic (amoraic) period. Unit 8, imported from mBerakhot 1.5 is, like the following

baraita (10, from the Mekhilta de-Rabbi Ishmael, as mentioned above), earlier, i.e., they are found in

tannaitic-era compositions. Moreover, units 4 (or 5)-6 may well be a late addition to the “Samuel” text.

For example, the expression le-sapper bi-yetsiat Mitsrayim is not to be found anywhere in Talmudic

writings.28 Therefore, unless that phrase is a hapax legomenon in that literature, units 4/5-6 would likely

edited by Shemuel Hagai סידור רבינו שלמה ברבי נתן זצ"ל, אב ביד דין מן העיר סיג'ילמסה, המכונה אלגבאלי 26(Jerusalem: Hagai?, 5755 [1994 or 1995])Pp. 258-259. The manuscript was copied in Barca (now Marj), Lybia, in 1202 (see the online catalogue of the Institute of Microfilmed Hebrew Manuscripts, Jerusalem). It is described in the introduction, p.4- 5; it was evidently copied in an Oriental hand. There is some question about whether Shelomoh bar Natan’s Sijilmassa was the one in the Maghreb or the Middle East (cf. the discussion and bibliography in the introduction, pp. 5-6).

27 It is in the standard text of, e.g., Goldschmidt 1969, p. 125. Found in standard editions of mPesahim 10.5 it is not included in the Erets Israel version of the Mishnah, viz., the Kaufmann Mishnah manuscript, nor is it contained in the original, EI, version of the Haggadah, e.g., that found in Goldschmidt, ibid, p. 81. It seems, therefore, that the Mishnah later absorbed this passage from the Haggadah Liturgy (see Safrai 1998, p 36; Goldschmidt, idem., p. 53, suggests that it originated in a baraita, i.e., a tannaitic text not [originally] included in the Mishnah). Goldschmidt notes that it is sometimes augmented by our unit 3’s Deuteronomy 6, 23 or Exodus 13, 5, and that the former verse has been repurposed elsewhere in some Haggadah versions as well. Moreover, unit 3 is cited in its entirety is many Haggadah versions.

28 Henshke, ibid, pp. 32-33, 392, suggests that this an indication that the Bene Beraq text is a fabrication.


be of Gaonic provenance. Now, they are like our sage story in that they strongly promote speaking about

the Exodus over reviewing hilkhot ha-Pesah (championed in the Tosefta’s story of the nightlong

Gamaliel Seder). Our story’s position could also be tannaitic, for it seems to reflect an approach to the

Seder advanced by the Mishnah,29 in opposition to the Tosefta.30 To be sure, units 4/5-6 seem to have

been formulated as a transition from the Avadim hayinu declaration to the Bene Beraq narrative. So, they

themselves could have been composed anytime between the amoraic period, when that opening rubric

may have been composed, and the time of the Geonim, when the Bene Beraq story was incorporated into

the Seder liturgy and, possibly, when it was composed.

That narrative tells a story about tannaitic sages engaging in an activity encouraged in a tannaitic

passage, so it does seem like it could be a genuine tannaitic text, just like units 8 and 10, not to mention

the similar story of Gamaliel’s Seder in Lod. It is different from these texts, however, in that all the

amoraic or tannaitic units of this group are definite borrowings, i.e., they can be traced back to actual

sources in Talmudic or tannaitic documents. The Bene Beraq story has no antecedent. It cannot be traced

to any late antique rabbinic source. Even though it seems like a tannaitic text, and it is paired with one

(item 8), perhaps its genesis may be connected in some way with units 4-6. That is, it may have been

created along with them as a transition element to item 8 or, slightly antecedent to them and subsequently

intended by the author of 4 or 5-6 as their target.

Further cause for uncertainty is another phrase from this story, kol oto ha-lailah (that whole

night), can be added to le-sapper bi-yetsiat Mitsrayim as unattested in tannaitic literature. It does,

nonetheless, occur in amoraic literature.31 The question then becomes, can the possible hapax and the

amoraic kol oto ha-lailah be indicative of a text formulated sometime during the amoraic period but that

did not make it into Late Antique compositions, only to be found and conserved in the Babylonian branch

of the Passover Haggadah? While unlikely, it is nonetheless a definite possibility that rabbinic texts, even

tannaitic ones thought to have been lost, or whose existence was not even known about, have been

discovered and published throughout the past century. So, it may not be necessary to write off this

possible provenance for the Bene Beraq story.

I will show below that the linguistic concerns are not really conclusive and, when all is said and

done, they are irrelevant; nonetheless, strong evidence can be adduced to demonstrate that the Bene Beraq

episode must be an apocryphal, Gaonic-era creation. Several stylistic problems and discrepancies show

29 Israel Yuval, הפוסחים על שני העיפים: ההגדה של פסח והפסחא הנוצרית, Tarbiz 95:1 (1995), pp. 5-28 argues that this narrative represents a Yavnean Seder. See J. Hauptman’s critique, Rereading the Mishnah , p. 51-52, n. 7.

30 It is interesting, and significant, that, mPesahim 10.4-5 framed by the four question and Gamaliel’s answers, speaks of teaching, expounding and explaining (מלמד, דורש, אמר), appropriate to a didactic setting, whereas tPisha 10.11 prescribes a collegial engagement in (לעסוק) hilkhot ha-Pesah.

31 Henshke, ibid, p.393- 394 and n. 151.


that the narrative, as it stands, clashes with expectations that one would have from reading other rabbinic

sage stories. Full conclusions will be drawn through a textual study of the evolution of this passage. That

study will also shed light on how those stylistic discordances came about.

Microcosmic Examination: The Nature and Evolution of the Bene Beraq Seder Narrative

The above table shows that the formula attributed to Samuel (row 1) developed, in the Babylonian branch

of the Haggadah, in two directions. Saadia’s Siddur added units 2-3 before proceeding with the baraita of

the four sons/children; the Ginze Schechter version added rows 4-8.32 Natronai Gaon’s version, which is

virtually identical with current rites, followed Ginze Schechter, while borrowing unit 2 from Saadia, as

well.

Examination of these texts suggests that even the simplest version, i.e., that of Siddur Saadia,

seems to have evolved in two ways that ultimately merged. In view of the existence of the contrasting

Ginze Schechter form, it seems likely that the Saadia text coalesced from two, probably post-talmudic,

augmentations, as follows.

עבדים היינו לפרעה במצרים, ויוציאנו ה' אלהינו משם ביד חזקה ובזרוע נטויה.

עבדים היינו לפרעה במצרים, ויוציאנו ה' אלהינו משם ביד חזקה ובזרוע נטויה.

ולא את אבותינו בלבד גאל המקבה, אלא אף אותנו גאל,

33 שנ' ואותנו הוציא משם.

ואלו לא גאל המקב"ה את אבותינו ממצרים, כבר אנו ובנינו ובני בנינו משועבדים היינו לפרעה

במצרים.

The two cells in the second row make the same point in two different ways. The right-hand one claims

that we and our descendants would still be slaves had our ancestors not been redeemed, while the left-

hand asserts that we were also redeemed with our ancestors. With the amalgamation of the two versions

the latter contention was put last in order to close with a proof text, a fitting way to bring a passage to a

conclusion.

Close reading reveals that the Ginze Schechter-type version seems not to have been fully formed

in this period. Units 4 and 6 do have the possibly late locution sapper bi-yetsiat Mitsrayim. But it is not

secure. Natronai’s version employs the infinitive le-haarikh in unit 6, and even Ginze Shechter varies in

unit 7 with mesikhin instead of mesapperim/n. The Natronai Gaon version does not contain unit 4, which

is present in Ginze Schechter.

In his comprehensive and exhaustive study of the rabbinic Haggadah,34 David Henshke cited

many problems with the Bene Beraq narrative. He found anomalies on the linguistic level, as mentioned

32 It conflated two formulations in row five; current versions have accepted zeqenim from the second one.


34 ‘Mah Nishtannah, cited n. 13?? above.


above, as well as irregularities in the way that the Bene Beraq story is told. These may be different

aspects of the compositional problem. He noted that sages appear out of order of superiority,35 i.e., Akiva

is presented before Tarfon.36 He claimed that Elazar ben Azariah is place in the middle position, between

Eliezer and Joshua, and Akiva and Tarfon, because of an aggadic tradition that Elazar ben Azariah was

promoted when Gamaliel was deposed.37 In addition, since Bene Beraq is Akiva’s hometown, the

reclining presumably occurred at his residence. Therefore, the narrative should be set out differently: “A

tale of (maʻaseh be…) Eliezer, etc., who reclined chez Akiva in Bene Beraq (she-hayu mesubbin etsel R.

Akiva …).”38 Furthermore, he stipulates, narratives introduced by the phrase maʻaseh be… never feature

more than two or three sages.39

Those peculiarities open up a window onto others, even more troublesome. I would add that

Tarfon is not just out of order; he is totally wanting in the Ginze Schechter recension. Is this just a scribal

omission, or is there an explanation for that? Elazar ben Azariah may be important, but Eliezer, who is

named first, is really the senior person in this assemblage, as well as one of Akiva’s masters, so does

seem that his presence is appropriately emphasized by the order in which he is presented.40

35 Precise biographical details are hard to pin down. Adin Steinsaltz, The Talmud: the Steinsaltz Edition, a Reference Guide (New York: Random House, c1989), p. 33, groups Eliezer (ben Hyrcanus), Tarfon and Yehoshua (ben Hananyah) in the generation of Gamaliel (II) of Yavneh, i.e., the third generation of Tannaim (80-110 CE); and Elazar ben Azaryah with Akiva in the one following (110-135 CE). I follow his lead, but note that an older arrangement compiled by Hanokh Albeck, Introduction to the Mishnah (Jerusalem: Bialik Institute; Tel Aviv: Dvir, 719 [1958 or 1959]), locates Elazar ben Azariah in Gamaliel’s generation (p. 224) and Tarfon in that of Akiva (p. 225). There is some disagreement in our sources with regard to the precedence of Akiva versus Elazar ben Azariah, possibly because some texts may have been influenced by a story in both Talmuds where the latter is chosen to be a temporary nasi, despite Akiva’s senior status (yBerakhot 4.1 [page 7c-d] // bBerakhot 27b-28a; see table 2??, p. 15?? below). Safrai, op cit., p. 208, in an alternative assessment of their sources, explain that Akiva, who follows ben Azaria in the Bene Beraq list, is really the youngest member of the group, but he increased in prominence over time, although he did not serve as leader.

36 Henshke, ibid., pp. 392-393.

37 Idem, pp. 394-395. The aggadic tradition is cited at the end of n. 35?? above.

38 Idem, p. 393.

39 Idem, p. 392.

40 Henshke’s proof , idem, p. 394-395, that Elazar ben Azariah’s place in the center of the list of sages at Bene Beraq is questionable in two respects. It is based upon a late source (Kallah rabbati 7.4; after his survey of previous scholarship and the evidence of Kallah rabbati itself, David Brodsky, A Bride Without a Blessing [Tübingen: Mohr Siebeck, 2006], p. 238, concludes that chapters 3-9 are post-talmudic or, at least, post-amoraic). The senior personage, Gamaliel, was put in the center in the Kallah rabbati narrative because of a unique set of circumstances. The group was approaching a residence, and that way Gamaliel’s would be the first and primary person to be seen by the person receiving them. Maʻaseh narratives that list sages according to age and status always put Gamaliel first, not in the middle. In our story, the leader is not ben Azariah but Eliezer (ben Hyrcanus, ha-gadol), the sage named first. Safrai, op cit., p. 117, and n. 3 and p. 208, conclude that the sequence in our maʻaseh follows the proper order according to status (and cf. n. 35?? above).


The solution to each sage problem is best approached from a different perspective. The inclusion

or absence of Tarfon is a textual problem and goes hand-in-hand with the previously noted linguistic

variations. They can be addressed best through tracing the evolution of this narrative. The issues

involving the numbers and the position of Eliezer are stylistic and formal. They can be understood

through clarifying the nature of maʻaseh be… and mesubbin narratives, i.e., when sages are named, and

who takes precedence.

The generic and stylistic issues raised by Henshke will be clarified and I will add some others to

the one I raised above. In the section after that, those problems and the linguistic issues will be addressed

as the evolution of this narrative is demonstrated, and conclusions drawn.

The Nature of the Bene Beraq Seder Narrative

In order to ascertain whether this narrative could be an authentic Talmudic-era creation,41 the number of

sages present, and the way the location of the gathering is recorded, may be dispositive evidence. The

Bene Beraq narrative names five sages. This would indicate to Henshke that its author did not understand

how a Talmudic maʻaseh works, for maasim never include more than three sages. This surplus of sages

is improper, for the self-identification maʻaseh is present in our story’s opening rubric.

However, while most maasim feature only two or three sages, some do indeed include more.

Thus, an exchange at Puteoli,42 styled as a maʻaseh, included the four sages, viz., Judah ben Bathyra,

Matithiah ben Heresh, Hananiah son of the brother of Joshua, and Jonathan.43 Another maʻaseh finds

Elazar ben Mattiah, Hananiah ben Hakhinai, Simon ben Azzai and Simon ha-Timni in discussion.44 One

more group of four, Gamaliel, Joshua, Elazar ben Azariah and Akiva are featured in no less than four

maasim.45

41 For a summary of those who regard the story as a record of an actual event, and those who do not, see Henshke, idem, pp. 391-392, and notes 139 and 143.

42 The source (Sifri Deuteronomy, Reeh, 80, p. ??) reads פלטום, which Jastrow (entry for that term, p. 1179), suggests is a corruption of פוטיולין (entry found, idem, p. 1140).

43 Sifre Deuteronomy, Reeh, 80 (see previous note).

44 tBerakhot 4.18 (Lieberman says that this was interpolated by the editor of tBerakhot from a different source (see his brief commentary there, and his long commentary, Tosefta Ki-Fshutah, Zeraim, part 1:69, l. 75).

45 mMaaser Sheni 5.9; Sifra Emor, parashah 12, perek 16.2 (ed. Weiss, p. 102c-d)=tSukkah 2.11=bSukkah 41b; Shemot Rabbah Mishpatim, parashah 30.9 (reprint of ed. Vilna: Romm = reprint by Pe’er of ed. Jerusalem, 5700 [1939 or 1940], p. 106a-b); Talmud. Minor tractates. Derekh Erez (Rabbah=Pirqe Ben Azzai), 5.2 (cf. Marcus van Loopik, The Way of the Sages and the Way of the World [Tübingen: Mohr, 1991], pp. 100-102). Note that, while they are credited with speaking in the Shemot Rabbah narrative, they are portrayed as doing so as a group (dareshu, ameru), i.e., probably one spoke words with which the others were in agreement.


The last group, which includes three of the sages in the Bene Beraq session, will be a focus of the

next section, on evolution. And yet another passage, one in which not four, but five sages appear, should

be mentioned, especially because those are the same five sages as in the Bene Beraq narrative.46 Although

the latter text is not styled a maʻaseh, i.e., it is not introduced by the term maʻaseh be…, it is still a

Talmudic-era anecdote or story, and thus relevant to our concerns. Cited as a baraita in bSanhedrin 101a,

this story comes from a tannaitic source text that appears in both the Mekhilta de-Rabbi Ishmael and the

Sifre Deuteronomy.47 The tannaitic document versions begin, רבי 48כבר היה רבי אליעזר חולה ונכנסו Rabbi Eliezer had been ailing for a while49) טרפון ורבי יהושע ורבי אלעזר בן עזריה ורבי עקיבה לבקרו

when Rabbis Tarfon and Joshua, Elazar ben Azariah and Akiva entered to pay him a visit). The story-

teller simply chose to introduce his narrative with an expression that put Eliezer’s ongoing condition

upfront, as opposed to, e.g., maʻaseh be-rabbi … ve-rabbi …, etc., she-nikhnesu le-vaqer et Rabbi Eliezer

she-hayah holeh … (an incident about Rabbi … and Rabbi … , etc., who had come to visit Rabbi Eliezer

who had taken sick). 50 The sequence in this tannaitic text does not necessarily reflect the eclipse of Akiva

by Elazar ben Azariah, who precedes Akiva in the list of sages (see below). Here, Akiva comes last

46 Safrai, 1998, p. 117, observe that groups of “five sages/elders” )"חמ]י[שה חכמים/זק]י[נים") are “quite common” and in n. 4 cites three of them. Henshke correctly deletes “sages,” but ,(מסורות) in tannaitic traditions (שכיח למדי)claims that this designation is (just) “common” in tannaitic “texts” (מקורות) and also cites three (op. cit., p. 392, and n. 141). The following stipulations should be noted. Firstly, the number of such occurrences is not at all large. There are only one in the Mishnah (Eruvin 3.4), four in the Tosefta (Sheviit 4.21, Eruvin 2.16, Miqvaot 7.10 and Tohorot 9.14) and one in Sifre Zuta (ad Numbers 19, 21). (There are also three cases in the Bavli, plus two duplications; and two in the Yerushalmi, one mentioned three times.) Moreover, the sages do not speak as individuals, and they are named only in one occurrence, viz., bSanhedrin 8b. In addition, there is one group of “five disciples” (תלמידים) in the Mishnah (Avot 2.8) and one in the Bavli (bSanhedrin 43a). In the one in mAvot, their names and qualities are listed. Basically, when a group of five is designated, they are anonymous or, at the most, just listed by name (and quality). Thus, although fit is significant that groups of five occur in this literature, such groups are not portrayed in stories; rather, rulings and traditions are ascribed to them as a collective. Finally, and this is significant for our narrative text, these (almost exclusively anonymous) groups are all comprised of Tannaim (except for the list, with names, of Jesus’ disciples, bSanhedrin 43a).

47 Yitro, Ba-hodesh 10 (pp. 240-241) and Va-ethanan 38 (pp. 57-59) respectively. Both editors stipulate that it must have been copied into the midrashic setting from another source. It is also present in two Genizah manuscript fragments of the Sifre (Kahana no. 35-37 [pp. 255-256] and no. 37 [p. 257]). Althought is not styled a maʻaseh, there is an episode where arbaah zeqenim (fours elders) get together to refute Eliezer’s teachings posthumously, and they are joined by a fifth (yGittin 9.1, p. 50a), all of whom engage in the debate.

48 The Mekhilta adds here arbaah zekenim (four elders), and then enumerates them.

49 Marcus Jastrow, Dictionary of Targumim, Talmud and Midrashic Literature (New York, Berlin: Choreb;

London: E. Shapiro, Valentine), p. 609, s.v, II כבר, “a long time since, … already”; cf. entry כבר in Michael Sokoloff, A Dictionary of Jewish Babylonian Aramaic of the Talmudic and Geonic Periods (Ramat Gan: Bar Ilan University Press; Baltimore, London: John Hopkins University Press, 2002), pp. 550-551.

50 Alternatively, one could begin maʻaseh be-rabbi Eliezer she-hayah holeh, ve-nikhnesu Rabbi … le-vaqero. See table 3, on alternative narrations of a dinner party hosted by Gamaliel (p. 25?? below), one of which begins kevar, another with maʻaseh be…, and another simply, keshe-asah…


because he speaks last in the ensuing encounter. The reason he speaks last is because his comment is the

only one to which Eliezer reacts, and that with a request for more details. Thus, he alone elicits a response

from the ailing master, and his ensuing reply is the longest communication in this text. Even if this

episode acknowledges the hierarchical precedence of Elazar ben Azariah, it does so in a way that

celebrates the superiority of Akiva’s hermeneutical insight and his favored status in the eyes of Eliezer;

indeed, that could be sufficient in itself to account for moving Akiva to the final position.51

The evidence of four sages in tannaitic maasim does invalidate the argument that three is the

largest number of sages to appear in a maʻaseh. Indeed, from the perspective of the number of sages, the

Bene Beraq story could definitely be a tannaitic-era tale. This possibility is enhanced by the fact that two

different (and early) versions of this anecdote originally featured only four sages, as mentioned above

(and see the discussion below (page 21??). Furthermore, an episode of five named sages,52 in which four

visit Eliezer, is tannaitc, and our maʻaseh features the same sages, the only difference being that Akiva is

apparently the one being visited.

What about the order in which the sages are named? As I stipulated in the preceding section,

Eliezer is the most prominent person, so he is named first. Tarfon’s position will be addressed in the

following section, on the evolution of the Bene Beraq maʻaseh. What do we know about the other three,

viz., Joshua, Elazar ben Azariah and Akiva? Those three appear in other maasim, along with Rabban

Gamaliel. The latter is the senior figure, so he is always named first (which supports the above claim that

Eliezer is featured in the initial position because of his seniority). Gamaliel’s eminence is further

indicated in some of the narratives, i.e., the ones that begin maʻaseh be-rabban Gamaliel u-zeqenim/veha-

zeqenim … (a story about Rabban Gamaliel and some elders/the elders …). Some go on to enumerate

them. Rabbi Joshua, who is also from Gamaliel’s age-group is the next to be named in the narratives of

four, to be followed by the other two, who are members of the following generation. The order there

differs, and the Bene Beraq text follows the majority.

Table 2. Precedence of Akiva over Elazar ben Azariah in naratives

Note: Citations preceded by an asterisk are from narratives not introduced by maʻaseh be…

Akiva named first Elazar ben Azariah named first*mBerakhot 1.253

mMaaser Sheni 5.9 (Gamaliel + zeqenim) 54=

51 Similarly, Henshke, op. cit., pp. 392-394, explains that some texts present Akiva before Tarfon, despite the latter’s pre-eminent status (see n. 34 above), for reasons specific to the circumstances there (n. 145). But below, see n. 58??

52 For Safrai, 1998., p. 117, the (apparently) frequent citation of groups of “five sages” or “five elders” in tannaitic traditions (they provide references n. 5), supports the historicity of the present account. Henshke, ibid., p. 391-392 is not persuaded of the historicity of our narrative by such phenomena. Cf. n. 45 above.

מ' ר' יהוד' פעם אחת הייתי מהלך אחר ר' עקיבא ואחר ר' אלעזר בן עזריה והגיע זמן קרית שמעא 53 .


yMaaser Sheni 5.4 (p. 55c)tShabbat 3.3

*Mekhilta de-Rabbi Ishmael, Ba-Hodesh 10 (pp. 240-241) = Sifre Deuteronomy, Va-ethanan 32 (pp. 57-58) = bSanhedrin 101a*Mekhilta de-Rabbi Ishmael, Ki Tisa, Shabbeta 1: KiSifra Emor, parashah 12, pereq 16.2 (Gamaliel + zeqenim; ed. Weiss, p. 102c-d)=tSukkah 2.11 (sages are not named) = bSukkah 41b*Sifre Devarim, Ekev 43, 16:Ve-akhalta (p. 94)55

*Sifre Devarim, Ki tetse 269: Keritut she-yehe (p. 289) // yKeritot 9.1 (p. 50a)56

ySukkah 2.4 (page 52d)57

Derekh Erez Rabbah (Pirqe Ben Azzai) 2.158

Shemot Rabbah, Mishpatim 30.9 (p. 106a-b)

In two maasim that feature just Akiva and Elazar ben Azariah, Akiva is mentioned first in tShabbat 3.3

but last in ySukkah 2.4 (page 52d). Each is typical of the group of texts with which that source clusters in

the above table. It seems likely that the editorial history of the preponderant texts reflects the conception

behind the narrative of the deposition of Gamaliel, in which Elazar ben Azariah was accorded precedence

54 In this story, Gamaliel gifts Joshua and Akiva, and then Joshua gifts Elazar ben Azariah.

.וכבר היו רבן גמליאל ורבי יהושע ורבי אלעזר בן עזריה ורבי עקיבה נכנסים לרומ 55לאחר מיתתו של רבי אליעזר נכנסו ארבעה זקנים להשיב על דבריו רבי טרפון ורבי יוסי הגלילי ורבי אלעזר בן 56 Note that the list in the yKerotot version varies. Note also, that Yehoshua, who is not listed .עזריה ורבי עקיבהthere, speaks first, and following amoraic interventions, Yose ha-Gelili and Akiva speak last.

57 Gamaliel precedes Akiva in the parallel in bSukkah23a , which is another indicator that the prominent person is conventionally named first in these stories.

58 Note that it begins maʻaseh be arbaah zeqenim and, as it progresses, Gamaliel takes up a position between Joshua and Elazar on the right, and Akiva on the left.


over Akiva,59 whereas the Mishnah and the Tosefta, which were not affected, preserve the memory of

Akiva’s original precedence.60

The matter of precedence renders moot the question of halakhic propriety of a student or disciple

reclining in the presence of his master.61 That concern does not seem to have affected the practices of

sages described in mesubbin texts found in tannaitic documents. Gamaliel, the nasi, reclines with many

contemporaries, and they are very aware that he is in a class by himself in respect of themselves.62 But we

also find a younger Akiva reclining in the presence of Gamaliel63 or another elder, Tarfon64 (together with

Akiva’s contemporary, Yose ha-Galili).65

59 Alluded to by henshke, p. 394-395 (and cf. n. 56?? above). The story of the deposition of Gamaliel, in which Elazar ben Azariah was preferred over Akiva (and Yehoshua), may be found in yTaanit 4.1 (p. 67d) = bBerakhot 27b-28a. On this narrative and its traditions, see Goldenberg, Robert, “The Deposition of Rabban Gamliel: An Examination of the Sources,” Journal of Jewish Studies 23 (1972): 167–90; Shapira, Chaim, “The Deposition of R. Gamliel, between History and Legend” [in Hebrew], Zion 64:1 (1994): 345–70); Devora Steinmetz, “Must the Partriarch Know ’Ukqtzin? The Nasi as Scholar in Babylonian Agadda,” AJS Review 23, no. 2 )1998(: 163–90 Jeffery Rubenstein, Stories of the Babylonian Talmud )Baltimore, MD: Johns Hopkins University Press, 2010(, 77–80; Simon-Shoshan, Moshe , “Creators of worlds : the deposition of R. Gamliel and the invention of Yavneh,” AJS Review 41,2 (2017) 287-313; idem, “Transmission and evolution of the story of R. Gamliel’s deposition,” in Jews and Christians in the First and Second Centuries: The Interbellum 70 – 132 CE, ed. Joshua J. Schwartz and Peter J. Tomson, CRINT 15 (Leiden: Brill, 2018).

The only entry in which Akiva’s positioning at the end can be questioned is that for the Mekhilta Ba-hodesh (and parallels), in the second column in of Table 2, where each sage speaks when his name is mentioned. It could be that he speaks last there for the significant and unique considerations that, not only does he speak longest, but also in a concluding dialogue with Eliezer, on whose account they had convened (see p. 15?? above). One could, on the other hand, suggest that Akiva’s favored speech and, by implication, predominant scholarly attainment, are coincidental indications of his political subservience to Elazar, who is name before him. See, for example, Simon-Shoshan, idem, p. 219.

60 The phrase “Shabbat shel [Elazar ben Azariah] (tSota 8.9) was adapted in bBerakhot 28a to refer to that sage’s appointed day to deliver the sermon in an arrangement worked out following the deposition and reinstatement of Gamaliel. This toseftan passage seems, then, to imply the precedence of Elazar ben Azariah. However, the parallel in Mekhilta de-Rabbi Ishmael, Bo, Pisha 16 (pp. 58-59: mi shavat sham?) does not allow for the Bavli’s construction. Sol Lieberman suggests that the Tosefta refers in a general way to which master spoke on that Shabbat, which could be on account of a number of reasons (he refers to Sifra, Metsora, [Introduction], 13 [p. 70c], and to the commentary of the Raavad there). He also notes that J. N. Epstein explained that “who’s Shabbat was it,” can ask which disciple’s turn it was to attend upon his master (Tosefta Ki-Fshutah, VIII: Nashim [Hebrew], New York: JTS, 1973, pp. 679-680; Epstein, Prolegomena ad Litteras Tannaiticas [Hebrew], Jerusalem: Magnes and Tel Aviv: Dvir, 1957, p. 427).

61 Henshke, ibid, raises the issue, but does not take this objection seriously (p. 391, n. 137). It may be that the sources portray the sages at various points in their careers. Accordingly, texts that depict different generations reclining together consider them all to have attained a certain status. Still, this not the case with Eliezer’s visit to Yose ben Peredah (see n. 65?? below).

62 Mekhilta de-Rabbi Ishmael, Yitro, Masekhta de-Amaleq 1 (p. ??)=Mekhilta de-Rabbi Shimon 18, 12 (p. ??)=Sifre Deuteronomy, Eqev 38 (p. ??), and repeated in bKiddushin 32b.

63 tBerakhot 4.15 (=bBerakhot 37a), tBetsah 2.12.

64 See n. 34 above on divergent views of whether Tarfon (or Elazar ben Azariah) is a member of Akiva’s generation or senior to him.

65 Sifre Deuteronomy, Eqev 41 (p. 85)./tt/file_convert/5e0575cb33a17772597ed1da/document.docx 5/15/23 18

A related issue, one pertinent to the Bene Beraq story is the question of whether it would be

appropriate for Eliezer, the master, to visit his disciple, especially on the latter’s turf. That is difficult to

ascertain because narratives do not often spell out matters of precedence, but there is a case where a rabbi

Eliezer went to visit etsel (on the turf of, chez) Yose ben Peredah talmido (his disciple).66 The disciple is

little known, but since this visit is also recorded in a tannaitic text,67 the visiting master is undoubtedly the

same eminent sage, Eliezer, who in our narrative visited Akiva at Bene Beraq.

Another consideration Henshke raised is that the proper way to indicate attendance at a party that

took place at Akiva’s residence, or even with him as host, would be along the lines of hayu mesubbin

etsel R. Akiva bi-Bene Beraq. The following is a paradigmatic example of that that point: אורחין שהיו Guests who had been) מסובין אצל בעל הבית וקדש עליהן היום ועקרו עם חשיכה לבית המדרש

reclining etsel [with, at the house of] the householder when Shabbat began, and then betook themselves at

nightfall to the bet midrash).68 That criticism may seem reasonable. One could build on this a case for

post-Talmudic composition: the obliviousness of the creator of the Bene Beraq narrative to such a fine

point of style could be an indication that he stood outside of the era of baraita creation. However, it is not

even followed in all tannaitic texts. For instance, while one of the examples in Table 3 (p. 25?? Below)

uses an etsel formula, the other two, one of which is in the Sifre Deuteronomy, do not specify where the

event occurred. I would furthermore stipulate that this incongruity may be explained, in part, by the

evolutionary development of our text, as it is traced in the next section.

Indeed, there is a far more telling problem with this narrative, one that marks it as

incommensurate with the ways in which sage stories, maasim and others as well, were told during the

Talmudic period. That is, the economy of expression in aggadic tales extends even to the specification, by

name, of the persons involved. This policy of narrative style accords with the rule of Chekhov’s gun,69

i.e., unless an individual character will say or do something further on in the tale, or fulfills some other

function, that character’s name is not to be proclaimed at the beginning.

The Seder of Gamaliel and the elders in Lod (tPisha 10.12)70 is a case in point. The text reads

מעשה ברבן גמליאל וזקנים

66 bEruvin 11b; ben Peredah’s status is not revealed in the original account, tEruvin 1.2.

67 tEruvin 1.2.

68 tBerakhot 5.3. This text is cited by Mandel, ibid., p. 192, but he does not explain why they would leave their host and go to the bet midrash. It is not clear that the guests were sages or their disciples; if not, and they did not expect public prayers (ibid, pp. 195-196 and n. 68), perhaps some form of instruction was expected.

69 "If in the first act you have hung a pistol on the wall, then in the following one it should be fired. Otherwise don't put it there.”

70 Discussed above, pp. 2-3??.


שהיו מסובין בבית ביתוס בן זונין בלוד והיו עסוקין בהלכות הפסח כל הלילה עד קרות

הגברהגביהו מלפניהן וניערו

והלכו להן לבית המדרש.

Gamaliel is named since he is the leader, but the elders remain anonymous, because no specific statement

or action is attributed to any one of them; the servants whose humble task is to clear away the mess are

not even denominated as a group, but merely folded by implication into the verb higbihu; ironically,

Baitos,71 the only person named besides Gamaliel, merited such public recognition because he hosted the

event.72

Only two people function in the following maʻaseh.

מעשה בר]בי[ אלעזר בן עזריה ורבי עקיבה שהיו באיןבספינה, ועשה לו רבי עקיבה סוכה בראש הספינה,

ובאת הרוח והפריחתה.73אמר לו רבי אלעזר בן עזריה: עקיבה איה סוכתך.

This incident takes place on board a ship. Other people may have been on that voyage, but they are not

mentioned because unconnected to the action. The sages’ destination, if any, is irrelevant; they just need

to be on a ship, so that the typical winds, more powerful than those experienced on land, can blow the

sukkah away.74 Both sages are named because each one has a role to play: One builds a sukkah sturdy

enough to stand up to normal winds, as halakhically required in his view, but unable to withstand the

strong winds on the Mediterranean; and the other emphasizes the irony of his situation.75

The following narrative begins with Gamaliel and the elders, also on a ship to an undetermined

destination, the purpose being to furnish a halakhically challenging setting. The narrator begins with “the

elders,” as in the Lod Seder story, but he goes on to state their names.

71 On Baitos, see Joshua Schwartz, על זונן ובנו בייתוס, Sinai 103 (1988/1989), pp. 108-122.

72 Nurit Be’eri, Exploring Ta’aniyot: Yerushalmi Tractate Ta’aniyot, -- Forming and Redacting the Traditions [in Hebrew] (Ramat Gan: Bar Ilan University), 769 [2008 or 2009]), pp. 88-91, explains that a simulated background story (סיפור דמוי-מציאות) would be constructed from borrowed and invented details and motifs, including persons and locations, into which a teaching or tradition could be incorporated; the invented narrative background with its vibrant details and emotional associations, makes it easier to remember the teaching.

73 ySukkah 2.4 (p. 52d). Told with Gamaliel and Akiva bSukah 23a.

74 Halakhic problem of where to set up a sukkah?

75 The implication is that, under the circumstances, it never was a sukkah. Gamaliel holds that a sukkah built onboard a ship must be able to withstand gusts typical of that setting, whereas Akiva holds that one need only be able to withstand winds typical of land. Akiva may have a considered position on this matter, but he no longer has a sukkah.


מעשה ברבן גמליאל והזקינים שהיו באים בספינה,ולא נמצא לולב כי אם ביד רבן גמליאל בלבד.

ונתנו רבן גמליאל מתנה לרבי יהושע,ורבי יהושע לרבי אלעזר בן עזריה,

ורבי אלעזר בן עזריה לרבי עקיבא,76וכולם יצאו ידי חובתם.

Employing a practiced narrative skill, this storyteller does not reveal their names until the character’s role

in the action is set forth.

An apparent exception, the gun that seemingly does not get fired, may be found in tBerakhot

4.18.

מעשה בארבעה זקנים שהיו יושבין בבית שער של ר'יהושע,

אלעזר בן מתיא וחנניא בן כינאי ושמעון בן עזאי ושמעון התימני, והיו עסוקין במה ששנה להם ר' עקיבא: מפני

מה זכה יהודה למלכות, מפני שהודה בתמר. הוסיפו הן מעצמן: אשר חכמים יגידו ולא כחדו מאבתם;

להם לבדם נתנה הארץ וגו' )איוב טו, יח-יט(.

That text, however, is a parenthetical insertion into the Tosefta there, and it is incomplete. In another

version, each sage does speak.77

Before concluding this section, the nature of the evidence of two phrases undocumented in

tannaitic texts will be examined. They are kol oto ha-lailah (which is amoraic), and sapper bi-yetsiat

Mitsrayim (which is unknown before the Haggadah). Henshke suggests that their lack of earlier

documentation does render them suspect. Perhaps so, but it would not be sufficient to disprove a tannaitic

provenance. The phrase sapper be-, for example, while rare in tannaitic literature, does occur in two

settings in the Tosefta, the one le-sapper be-maʻaseh ha-ari (to talk about the episode of the lion;

tBerakhot 1.11) and tYoma 2.7: le-sapper bi-genut… (to recount or discuss the disgrace of ...). The first is

actually pertinent because it involves recounting a salvation event; the second involves genut (disgrace,

shame) a rhetorical perspective to be taken up during the Seder exposition. In view of those usages, each

one itself a unicum, it would be reasonable to suggest that, if le-sapper bi-yetsiat Mitsrayim appeared in

76 Sifra Emor, parashah 12, pereq 16 (p. 102c). The identities were not revealed in the following version, which chose to enrich the narrative in other ways (tSukkah 2.11). מעשה ברבן גמליאל וזקנים שהיו באין בספינה ולא היה עמהן לולב. לקח רבן גמליאל לולב בדינר זהב. כיון שיצא

בו ידי חובתו נתנו לחבירו, וחבירו לחבירו, עד שיצאו כולן, ואחר כך החזירוהו לו.77 See Lieberman’s brief note ad locum; in his Tosefta Ki-Fshutah, Zeraim 1:69, he cites a version from a late midrash (Jellinek, Bet ha-midrash, reprint: Jerusalem: Wahrmann, 1967, 5:95) in which all do speak.

A late example features three eminences (gedole ha-medinah) who were reclining with Johanan ben Zakkai (Avot de-Rabbi Nathan B, 13, s.v., davar aher); they neither speak nor act. That text exceeds the chronological scope of this inquiry, but the publicizing of their names does have a purpose, for they are expounded upon in the ensuing account.


the Bene Beraq maʻaseh, that could indeed be a third tannaitic instance. One could surmise that it came to

be integrated into the Haggadah by the introductory units 4-6 (or 5-6), which had borrowed that phrase

from the maʻaseh (except that I observed above, and will further demonstrate below, that the linguistic

concern is irrelevant, for this expression is not original to our narrative).

The expression kol oto ha-lailah presents a more complicated problem. On the one hand, it has no

precedent in tannaitic documents (although kol ha-lailah is definitely common there).78 On the other hand,

Levi, a disciple of Judah the Prince uses it,79 as does his son, Joshua.80 But those two sages are on the

cusp, members of a liminal group, straddling between the tannaitic period and that of the Amoraim. Were

it not for the chronological indeterminacy, observations on linguistic usages like that might lend support

to an argument but would not in themselves constitute proof. A further indication, though, does negate a

hypothesis that kol oto ha-lailah could not be tannaitic: that is, a mirroring phrase concerning yom (day),

the opposite of night, viz., kol oto ha-yom, occurs several times in tannaitic documents,81 makes it easy to

imagine a tannaitic author stipulating kol oto ha-lailah in a statement or narrative.

In light of the above considerations, one must concede that, while the Bene Beraq maʻaseh

may relate an event that occurred in tannaitic times, and it may not pose some linguistic problems to a

tannaitic provenance, it definitely violates certain tannaitic literary narrative conventions. Henshke

objects to the way in which the narrator chose to identify the location of the event. He thinks that there

may be a problem with how Eliezer, the senior person, is presented, and Tarfon is completely out of

order, at the end rather than closer to Eliezer at the beginning (as in the above-mentioned story of four

sages who went to visit an ailing Eliezer, pp. 14-15?? above). Indeed, Tarfon is totally wanting in some

iterations of this incident, but the other objections may yield to further evidence. Another issue is that the

names of sages are stated even though not all of them speak or perform a specific action.

The Evolution of the Bene Beraq Seder Narrative, Stage 1: An Adaptation of tPisha 10.12

Of the problems with the Bene Beraq maʻaseh raised in Henshke and added by me, the serious,

unresolved ones are as follows. The exposition is formulated so that rabbi Akiva is a participant rather

than a host, even though he is the only one who dwells in Bene Beraq. Sages, rather than being treated as 78 According to the Bar Ilan Responsa Project database, there are 9 occurrences in the Mishnah; 23 in the Tosefta; and 29 in the Halakhic midrashim.

79 Midrash Bereshit Rabba, ed. J. Theodor and Ch. Albeck (Reprint: Jerusalem: Wahrmann, 1965), Lekh lekha 40, s.v. (17) Va-yenaga (p. 389) = Va-yera 52, s.v. (18) Ki atsar (p. 554).

80 Citing Bar Pedayah (a contemporary of Levi, his father), idem, Be-reshit 26, s.v. Tovot henah (p. 249) = Va-yera 50, s.v.(5) Va-yiqreu el (p. 522).

81 According to the Bar Ilan Responsa Project database, there are 9 occurrences in the Tosefta; one in the Mekhilta d’Rabbi Sim`on b. Jochai, ed. J. N. Epstein and E. Z. Melamed (reprint: Jerusalem: Yeshivat Shaare Rahamim and Bet Hillel, [197-?]) ad Exodus 17, 12.


a collective, are named individually, even though most do not act or speak on their own. Finally, Tarfon, a

senior figure, is either introduced last rather than at the beginning with Eliezer, or else he is completely

wanting, as in the Ginze Schechter version. Are those aberrations mere scribal oversights?

Let us take a closer look at that version.

בן עזריה ור' עקיבה אלעזר ור' יהושע ור' אליעזר ומעשה בר', ברק- שהיו מסובין בבני

, והיו מסיחים ביצי' מצ' כל אותו הלילההגיע זמן קריאת שמע של שחרית., רבותינו: עד שבאו תלמידיהם ואמרו להם

This text has only four sages. It also employs a different verbal phrase for recounting the exodus, viz.

mesihim + be… instead of mesapperim + be… It could be that the above iteration represents an early

form of this narrative, with less sages and a verb of recounting that would come to be eclipsed by another

contextual expression, mesapperim/n in later textualizations.

Pursuant to that possibility, let us consider a companion to the Ginze Schechter version, i.e., a

recension of this anecdote found in a Haggadah text appended by the copyist at the end of a manuscript of

the Siddur of Shelomoh bar Natan of Sijilmassa. That Sijilmassa-B text (Table 4, column 3) is instructive.

, ברק- שהיו מסובין בבני, בן עזריה ור' יהושע ור' עקיבא אלעזר ור' לליא. מעשה ברבן גמ1, . והיו משיחין והולכין ביציאת מצרים2. עד שעלה עמוד השחר.3'. כיון שעלה עמוד השחר,3 =4'. נכנסו תלמידיהם אצליהם,1 =5 כבר הגיע זמן קריאת שמע.:'. אמרו להם2 =6

It is arranged in a modified chiastic pattern, perhaps under the influence of the tPisha source it was

adapting. Line 1 is the exposition. That and the action, lines 2-6, revolve around the central lines 3 and 4,

which end the first scene and begin the second one respectively. Lines 1 and 5 set out the action of the

sages and of their students respectively, while units 2 and 6 report the different topics of which they

spoke.

This text is even more unique than the previous one in several noteworthy respects.

Linguistically, mesihin is used, rather than mesapperim/n, as before. Moreover, kol oto ha-lailah is

nowhere to be found. Without mentioning its nightlong duration, the extensive nature of the recounting is

conveyed by an augmented verbal construction (meshihin ve-holekhin) for a process that continued until

dawn (ad she-alah amud ha-shahar). The masters are not addressed as rabbotenu, and the specification

shel shaharit does not augment zeman qeriat shema82 (the dawning of the day makes that clear). Hence, it 82 The phrase zeman qeriat shema appears in tannaitic documents, e.g., tBerakhot 1.2, 1.4 and 3.19, but the specification “(shel) Shaharit/Arvit” does not appear earlier than Gaonic and Medieval sources. While the Sijilmassa B version could theoretically be of tannaitic provenance, this factor would suggest that all subsequent versions are post-talmudic./tt/file_convert/5e0575cb33a17772597ed1da/document.docx 5/15/23 23

is indeed ironic that, in critiquing the literary and linguistic competence of the Bene Beraq narrative in the

Haggadah, scholars were either unaware that the original form of the tale was actually different, and

lacked the linguistically problematic phrases, or they ignored the evidence. The differences, then,

illuminate some of the problems and obviate some others.

It seems clear that this supplementary version in the Shelomoh bar Natan Siddur manuscript is a

preliminary form of the anecdote. Indeed, evidently taking the Lod Seder as a structural paradigm, it is

longer than the others, told in six lines instead of four. The dawning of the day is stated twice, although

greater dramatic effect could have been achieved if instead the all-night length of this episode had been

noted. Mesihin ve-holekhin, on the other hand is a nice way to convey the long and continuous process in

which of these sages were engaged.

There are as many problems as sages. Rabbi Tarfon is wanting, which means that only four are in

attendance. (Let it be noted that this statistic comports well with our observations above that four – as

well as five -- fits within the numerical norm of named persons in a maʻaseh or other sage narrative.) The

naming of Elazar ben Azariah before Joshua is anomalous. Akiva is on the list of attendees when he ought

to be the host. But more unusual than any of those items by far is that Gamaliel is in the first position

instead of Eliezer, who is nowhere to found. Gamaliel does not appear in any of other versions of the

Seder at Bene Beraq.83

Where some have seen the toseftan and Haggadah versions of the all night Passover as mirror

image reflections of each other,84 Henshke and others have suggested that the Haggadah version is

actually an adaptation of the former, which served as its model.85 The Sijilmass-B text, preserved in an

early thirteenth-century Oriental or North African manuscript, appears to be a missing evolutionary link,

textual evidence of that process of adaptation. The Bene Beraq narrative is not a lost tannaitic tale at all; it

began, and it developed, within the context of one family of the Babylonian branch of the Haggadah as

the latter was taking form in the gaonic period.

As a way of addressing the Four Questions, the maggid section opened with a brief Avadim

hayinu declaration (Table 1 or 4, unit 1), and reminded those present that the manner in which the

Questions are answered should take four different personal identity and learning styles into account

(Table 1, unit 10; Table 4, unit 10). That baraita favored inculcating the rules of the Passover offering to

the wise son, whom it mentioned first, although it did provide for three other types. One liturgical

composer set out to counter that emphasis by interposing mBerakhot 1.5, in which Elazar ben Azariah

83 For this reason, Safrai, 1998, p. 117, concluded that Gamaliel is simply a scribal error.

84 Henshke 2016 p. 392.

85 Idem, idem, pp. 392-395.


cites Ben Zoma to instruct people to mention the Exodus at night. By way of introduction, he reworked

the maʻaseh of Gamaliel’s all-night examination of hilkhot ha-Pesah to tell of a gathering that spent the

night discussing the Exodus.

The adapter evidently decided to further differentiate the two Gamalielan Seders, by changing the

location from Lod to Bene Beraq. Perhaps the latter place was chosen simply because it was the place

where the sage named last, Akiva, dwelt. It was, accordingly, convenient to conclude the exposition of

this maʻaseh with Akiva in Bene Beraq.

With respect to that coincidence, one can suggest that it was unnecessary to specify that this was

the home of Akiva. In the three versions of a festive meal hosted by Gamaliel that will be examined in the

next section, one account specifies that sages were mesubbin etslo [at his house, or with him where he

hosted his festivity]. The other two do not.86 The location is expressed as the occasion: hayu mesubbin be-

vet mishteh beno shel rabban Gamaliel. The narrators could have formulated it, e.g., hayu mesubbin etsel

rabban Gamaliel be- mishteh beno. But they did not. Similarly, the failure to stipulate etsel rabbi Akiva

may not be an oversight, but a consequence of the retention of the formulation of the toseftan story in

switching from Lod to Bene Beraq, but with Akiva aleady stipulated as a member of the group in the final

position, immediately adjacent to the place name.

The Evolution of the Bene Beraq Seder Narrative, Stage 1 continued: Sequence of Sages

tPisha 10.12, the evolutionary germ of this narrative, does not identify Gamaliel’s collocutors,

who are simply styled zeqenim, because it did not report anything they independently said or did that

night. The Babylonian adapter, however, did want to present something that Elazar ben Azariah spoke

about at the Bene Beraq Seder as he imagined it, viz., mBerakhot 1.5, so he stipulated that sage’s

presence.

It was mentioned above that the latter, along with Joshua and Akiva, were reported to have

accompanied Gamaliel in several maasim.87 Albeit they were never reported to have been mesubbin as a

group in each other’s company, it would certainly seem appropriate to include Elazar ben Azariah in a

Seder led by Gamaliel, along with the two aforementioned colleagues.88 On the other hand, it would not

have entered the storyteller’s mind to place Tarfon at that Seder because the latter is never part of a group

86 See Table 3, below.

87 Page 14, and n. 44?? They were also together in maasim without Elazar ben Azariah, see bQiddushin 26b and bBava Metsia 11a. All four sages travel together in Derekh Erez (Rabbah=Pirqe Ben Azzai), 5.2, but only the senior figures Gamaliel and Joshua speak, the fact that this is a late text like our Bene Beraq narrative may explain this shared anomaly..

88 Discuss the halakhic, modelling function of these stories. They are not historic occurrences, but illustrative tales that act out halakhah.


accompanying Gamaliel. Rather, in their encounters, Gamaliel seems to be portrayed as challenging

Tarfon’s assertions.89

Elazar ben Azariah’s position is noteworthy. Not only does he precede Akiva, as discussed

above,90 he is named immediately following Gamaliel, i.e., he is given precedence with respect to Joshua,

who was his elder, and therefore always preceded him in a sequence of named sages. This violation of a

longstanding norm was probably done to promote the younger scholar, for he declaimed a message (unit

9/10), whereas Joshua did not say or do anything. Therefore, his presence would be signaled right after

that of the session leader; with Joshua and Akiva trailing after.91

What is important to conclude at this point is that the revisionist adapter seems to have chosen to

update the toseftan nighttime Seder to reflect an emphasis on the Exodus account. In specifying who was

present, he elected to follow a different paradigm, albeit incorrectly, but he kept to a traditional choice of

sages.

Indeed, this rabbinic-liturgical composer violated another norm when he named Joshua, the silent

sage. This positional disparity was remedied by later redactors who put Joshua where he belongs, in the

second position, followed by Elazar ben Azariah. (The later redactor was even further from the Talmudic

rules of the game in naming a silent sage, as will be explained below.) One can claim that Akiva, who

was also silent, was named last because, as presumptive host and, hence, active contributor, he would be

appropriately juxtaposed with Bene Beraq. The story could be set out in different ways.

Table 3. Three ways to begin an aggadic narrative

Mekhilta de-Rabbi Ishmael, Yitro, Amaleq 192

Sifre Deuteronomy, Eqev 3893 bKiddusin 32b

a כשעשה רבן גמליאל סעודה לחכמים, היו כל חכמי ישראל

מסובים אצלו.

וכבר היו רבי אליעזר ורבי יהושע ורבי צדוק מסובים בבית משתה בנו של רבן

גמליאל.

מעשה ברבי אליעזר ורבי יהושע ורבי צדוק שהיו מסובין בבית המשתה בנו של

רבן גמליאל.b1 עמד רבן גמליאל ושמשן. והיה רבן גמליאל עומד ומשקה עליהם. מזג רבן גמליאל את הכוס לרבי

89 He rebukes Taron in Sifre Numbers (ed. H. S. Horovitz, Lipsiae: Gustav Hock, 1917) Qorah 116, s.v. ve-atah u-vanekha (p. 133) =bPesahim 72b and disputes him in Midrash Mishle ed. Burton L. Visotzky (New York: The Jewish Theological Seminary, 1970) pp. 69-70.

90 P. 16??

91 It would have had nothing to do with the Talmudic stories wherein Elazar was chosen over Akiva and even Joshua to replace Gamaliel; otherwise, Elazar would have been introduced before Joshua in the other narratives where they appeared with Gamaliel. Henshke 2016, pp. 394-395, thinks that Elazar’s position in the center (except for the Sijilmassa B version under discussion here) was done under the influence of those stories; my view is presented inside, pp. 15-17?? and Table 2 above.

92 Pp. 195-196. There is a parallel in Mekhilta de-Rashbi, ad Exodus 18, 12 (p. 88).

93 Pp. 74-75.


אליעזר.b2 אמרו החכמים: אין אנו כדי

שישמשנו.ולא רצה לטלו,

נטלו רבי יהושע.נתן הכוס לר' אליעזר ולא נטלו,

נתנו לר' יהושע וקיבלו.c1 אמר להן ר' יהושע: הניחו לו

שישמש, שמצינו שגדול מרבןגמליאל שמש את הבריות.

אמר לו רבי אליעזר: מה זה יהושע, בדין שאנו מסובים וגמליאל ברבי עומד

ומשמשנו?

אמר לו רבי אליעזר: מה זה, יהושע, אנו יושבין ורבן גמליאל )ברבי( ]דרבי[ עומד

ומשקה עלינו?[c2] אמרו לו, אי זה זה?[c3] אמר להם: אברהם אבינו גדול

העולם, ששימש מלאכי השרת, והיה סבור בהן שהם בני אדם ערביים עובדי עבודה זרה, רבן גמליאל שישמש חכמים לומדי

תורה, על אחת כמה וכמה.c2/

[d1]

אמר להן ר' צדוק: הניחו לו שישמש, מצינו גדול מרבן גמליאל

ומאברהם ששימש את הבריות.

אמר לו רבי יהושע: הנח לו וישמש, אברהם גדול העולם שמש מלאכי שרת וכסבור שהם ערביים עובדי עבודה זרה שנאמר )בראשית יח ב( וישא עיניו וירא

והנה שלשה אנשים, והלא דברים קל וחומר, ומה אברהם גדול העולם שמש

מלאכי שרת וכסבור שהם ערביים עובדיעבודה זרה, גמליאל ברבי לא ישמשנו?

אמר ליה: מצינו גדול ממנו ששמש, )אברהם גדול ממנו ושמש( אברהם גדול הדור היה, וכתוב בו: והוא עומד עליהם! ושמא תאמרו, כמלאכי השרת נדמו לו? לא נדמו לו אלא לערביים, ואנו לא יהארבן גמליאל ברבי עומד ומשקה עלינו?

d1/ ]d2

[

אמרו לו, אי זה זה? אמר להם רבי צדוק: הנחתם כבודמקום,

ואתם עסוקים בכבוד בשר ודם?

אמר להם רבי צדוק: עד מתי אתם מניחים כבודו של מקום, ואתם עוסקים

בכבוד הבריות?

d2 / ]d3

[

אמר להם: שכינה, שבכל שעה מספיק מזון לכל באי העולם כדי צרכן ומשביע לכל חי רצון, ולא

לבני אדם הכשרים והצדיקים בלבד, אלא אף הרשעים עובדי

עבודה זרה רבן גמליאל על אחת כמה וכמה שישמש חכמים ובני

תורה.

אם מי שאמר והיה העולם משיב רוחות ומעלה עננים ומוריד גשמים ומגדל

צמחים ועורך שולחן לכל אחד ואחד,גמליאל ברבי לא ישמשנו.

הקדוש ברוך הוא משיב רוחות ומעלה נשיאים ומוריד מטר ומצמיח אדמה, ועורך

שולחן לפני כל אחד ואחד, ואנו לא יהארבן גמליאל ברבי עומד ומשקה עלינו?

The versions of this anecdote consist of an exposition (a), an action and response (b1 & 2), and two

speech acts with replies (c1 & 2, d1 & 2; the Mekhilta is set out differently from the other two, without

Eliezer, and with extra steps). What interests us is the way the actors and speakers are introduced. Since

this wedding feast was being hosted by a prominent family (the house of the nasi Gamaliel) many guests

would have been present. The Mekhilta (which just characterizes its occasion as a “feast for the sages”)

reports, right in the exposition, that “all the sages of Israel were mesubbin” with Gamaliel (etslo), but the

Sifre and the Bavli only report that Eliezer, Joshua and Tsadok were there, with the implication that

Gamaliel, the father, would be present as host. One would expect the Sifre and Bavli composers, and their

audiences, to have been aware that others would also have been present. They chose to restrict themselves

to the persons who would act or speak in the ensuing action.

This convention is followed in Late Antique rabbinic anecdotes. Like Chekhov’s gun, a character

identified by name must act or speak. There is one wrinkle in the exposition that does not name the sages /tt/file_convert/5e0575cb33a17772597ed1da/document.docx 5/15/23 27

(as with the Mekhilta above), viz., their number may be restricted to the ones who take part in the action.

For example, an exposition in Sifra Emor, parshah 12, pereq 16 (p. 102c), has Gamaliel and the elders

(zeqenim) on ship on the first day of Sukkot, with Gamaliel alone in possession of a lulav. The action has

Gamaliel gifting the lulav to Joshua, who then gifts it to Elazar ben Azariah, who does the same to Akiva.

The sequence ends with Akiva, and the auditor infers that no-one else was there, or else it would have

been reported that Akiva in turn gave the lulav to that person.94

Now, our Haggadah adapter chose to use the style of the bKiddushin 32b example: He began

maʻaseh be... and named all the sages in the exposition. However, even though the Chekhovian gun was

loaded with four bullets, only one was fired, for Elazar ben Azariah was the only one who spoke. Still,

one can count Akiva, too who, followed by the name of his town, Bene Beraq, was presumably the host;

Gamaliel, the erstwhile leader, by implication would have said or done something. That leaves Joshua

alone without a task. That fact is an indication of critical remoteness from the methods and modes of

literary creativity characteristic of the Talmudic period. The same anomaly occurs in an anecdote in

Derekh Erez Rabbah, but that it also a post-talmudic tractate.95 The conventions of one era were eclipsed

in another. That is a further sign that our story is post-talmudic.

It is indeed curious that Eliezer is missing in this early version of the adaptation, as well as

Tarfon. How they came to be in this maʻaseh, and how Gamaliel dropped out, will be accounted for as a

sequential evolutionary development, in the next section.

The Evolution of the Bene Beraq Seder Narrative, Stage 2: Eliezer and Tarfon

A later adapter decided that a Seder that emphasizes the Exodus narrative over hilkhot ha-Pesah should

have a different leader than Gamaliel (despite the fact that an adaptation of his directive in mPesahim

10.5 which, in directly addressing the child’s opening questions, is a centerpiece of the Mishnah’s Seder,

and is prominent in all later versions).96 Eliezer, an eminent sage, brother-in-law and sometime antagonist

of Gamaliel, was selected. This choice is reflected in the Ginze Schechter version, which is otherwise so

similar to Sijilmassa-B. Gamaliel’s entourage, however, was preserved; just the leader changed.

94 In the version in bSukkah 41b, Akiva returns it to Gamaliel, completing the circuit, thereby confirming that only the named sages were in that party.

95 See n. 56?? Above

96 Henshke 2016, 386, observes that some scholars (cited there, n. 125) view Gamaliel’s instruction in mPesahim 10.5 as convergent with his spending the night in discussion of hilkhot ha-Pesah in tPisha 10.11, but concludes correctly that Gamaliel’s explication of the evening menu is narration of the Exodus rather than discussion of rules. Judith Hauptman, op. cit. above, n. 18, views the two sources as representing two historical moments in the evolution of the Pasover Seder. Henshke views both as different aspects of the Passover evening program, the Mishnaic portion purposed for children, the toseftan one for adults (idem, n. 56 [pp. 54-55]).


Eliezer was not viewed in late antique rabbinic literature as a leader like Gamaliel; he does not

travel with an entourage of zeqenim. It seems that his eminence rather than his opinions got Eliezer the

appointment. He agrees with Gamaliel that the evening should focus on hilkhot ha-Pesah, although he

thinks that the discussion need only be engaged until midnight, rather than Gamaliel’s nightlong effort.97

Nor is it part of a pattern involving the sages that make up Gamaliel’s entourage and who also famously

got together to visit Eliezer when he was ailing. That source must be discounted because Tarfon, who

was not a member of a faction that accompanies Gamaliel,98 but was a member of the party visiting

Eliezer, was not included even in the Ginze Schechter party.99

As a witness to further textual autonomy from its toseftan model, the Ginze Schechter version

reflects the exchange of Eliezer for Gamaliel. All the while limiting itself to the latter’s three travelling

companions (known from elsewhere), which means that Tarfon is still wanting from the entourage. He

was only added in the subsequent Bene Beraq versions that became universal. The lateness of his

insertion is reflected in Tarfon’s position. This colleague of Joshua,100 senior to both Akiva and Elazar ben

Azariah, was simply tacked on at the end. That, of course, eclipsed the juxtaposition of Akiva with Bene

Beraq. The logic of that collocation was evidently not on the mind of the final-stage redactor who

interposed Tarfon.

The reason for his addition to the group is not clear. To the possibility that the redactor, looking

at mPesahim 10 as a prototypical Seder, hence a model for his Bene Beraq event, added Tarfon because

97 Eliezer’s view may be found in the Mekhilta de-RI, Pisha 18, s.v., Ve-hayah ki (p. 74), which is consistent with his understanding that the Paschal lamb could not be consumed past midnight (idem, idem, 6, p. 19); Gamaliel, who holds that a sacrifice that must be eaten in one day may be consumed until dawn (although it is preferable to finish them by midnight; mBerakhot 1.1), spends the whole night in discussion of hilkhot ha-Pesah (tPisha 10.10-11). Assuming that he was aware of Eliezer’s position on the appropriate nighttime activity and of the paradigmatic midnight limit, it is not clear that the aggadic redactor who substituted him for Gamaliel considered those matters relevant to his creation. Tannaim can contradict themselves or be represented as espousing contradictory positions: the first level adaptor (the Sigilmassa-B recension), who borrowed Gamaliel for the Bene Beraq maʻaseh, also created a contradictory scenario (thematizing narrative instead of rules). Even before that, within the universe of tannaitic texts, Gamaliel in tPisha 10.12 says to discuss hilkhot ha-Pesah all night long; but mPes 10.5 Gamaliel says must explicate the foods for historical/symbolic illustration (see the previous?? note). Moreover, must Bene Beraq’s supererogatory activity be constrained by the time-frame of when Passover offering must be consumed (according to Eliezer and others)? Recounting the Exodus does not map onto the phenomenon of the Paschal sacrifice and meal in the same way as the rehearsal of the hilkhot ha-Pesah functions as their analogical surrogate.

98 Cf. p. 25?? above, and n. 88??.

99 See p. 15?? above. That is the only occasion when Elazar ben Azariah is part of a group along with Akiva that is involved with Eliezer. Akiva was with Eliezer and Joshua, yHorayot 3.4 (p. 48a) = Va-yikra Rabba, Va-yiqra 5.4 (p. 110); and with Tarfon added as well, Avot de-Rabbi Nathan-A 6, s.v., Davar aher (p. 29a) = idem, Hosafah 2 to Nosah A 8, s.v., Mah hayah (p. 162).

100 Henshke says that Tarfon’s opinion is presented before Akiva when they disagree, aside from situations that have their own logic (ibid, pp. 392-393 and n. 145. Cf. 34 above?? Tarfon is presented as senior to Akiva in tZevahim 1.8; Sifra, Va-yiqra, Dibburah de-Nedavah, parashah 4.5 (p. 6b); Sifre Numbers, Be-haalotekha 75 (p. 70); yMegillah 1.10 (p. 72b) = yHorayot 3.2 (p. 47d); Kallah 1.21.


he contributed there (paragraph 6), one can object that other speakers, viz., Elazar bar Tsadoq (paragraph

3) and Yose (8) were not included. It would, rather, make more sense to look at mBerakhot chapter 1,

which is the source of Ben Azzai’s dispute with the sages cited by Elazar ben Azariah (paragraph 5), and

where Tarfon appears in paragraph 3. Others included in our maʻaseh also contribute to that chapter, viz.,

Eliezer (1, 2) and Joshua (2), i.e., everyone but Akiva, who was needed for his residence, and who was

already part of the group. Could a post-Ginze Schechter redactor have imagined the sages in mBerakhot 1

as participants in a discussion that surfaced in Elazar ben Azariah’s message, and so added Tarfon to the

Bene Beraq narrative? The fact that the houses of Shammai and Hillel, who are not represented in our

maʻaseh, appear in paragraph 3 is irrelevant, for they are too early, and merely supply background to

Tarfon’s contribution. Unfortunately, however, Gamaliel also appears there (1). Indeed, he takes part in a

discussion with his children that is thematically suggestive, for they have been partying all night, like the

Bene Beraq group, and have a question about the recitation of the Shema, albeit the evening one rather

than the morning one. Perhaps Gamaliel contributed that theme, along with Ben Zoma’s exegesis, to

make mBerakhot 1a source of inspiration for the third-stage redactor who decided to add Tarfon to

Eliezer’s Bene Beraq group. The least problematic hypothesis for his inclusion in this third stage revision,

is that its redactor was aware that Tarfon had joined a group that had paid a visit to Eliezer on his

sickbed:101 so he could, accordingly, imagine that the same individuals who had participated in that

incident might also have gotten together at Bene Beraq for a Seder. One might challenge that possibility,

as well, for Tarfon could just as well have been inserted into his proper place unless, of course, an

addition should just be interpolated at the end of the list.

Macrocosmic Issues Redux: The Evolution of the Bene Beraq Seder Narrative, Stage 2 continued:

Linguistic and Liturgical Matters

We have seen that, whatever its merits, concern over certain phrases of the Bene Beraq narrative that

seemed anomalous, e.g., kol oto ha-lailah, sapper bi-yetsiat Mitsrayim, was irrelevant. The first one does

not appear until the Ginze Schechter revision, the second only in subsequent ones.

The direction of the influence for sapper bi-yetsiat Mitsrayim was, of course, that it was

borrowed from the language in units 4, 5 and 6. They were created early on to integrate the Bene Beraq

narrative cum Elazar ben Azariah’s exposition into the Avadim hayinu sequence. When, in relation to the

key units 7 and 8 the transitionary units were developed will be discussed in the next chapter.

1. Overall development and cohesion

101 See p. 15?? Above.


The Bene Beraq and mBerakhot texts are intertwined thematically and linguistically with units 4-6, whose

authors contributed the construction, sapper bi-yetsiat Mitsrayim.102 One cannot understand them without

understanding their evolution and function. It would be incorrect to simply atomize them and treat them

as separate units that were imported from three different places. Doing so would affect how we apprehend

their meaning and purpose. This is especially important regarding text 7. Inasmuch as it is a post-talmudic

composition, it is unlikely for it to have been drawn from a hypothetical, unknown rabbinic source ; it

seems more reasonable to suppose that it was crafted to supersede the toseftan Seder story that served as

its model, with the intended function of introducing mBerakhot 1.5 (unit 8) in a way that valorized sippur

bi-yetsiat Mitsrayim. It can be best understood with reference both to its originating motivation and its

literary function. Most rabbinic mesubbin anecdotes do not relate to Passover, and mBerakhot 1.5 refers

to all the lelot (nights) of the year, but there can be no doubt about the occasion for the Bene Beraq

reclining incident because it is clear from its toseftan foil, not to mention the Haggadah context in which

it was formed, that a Passover Seder was intended. It was composed for the purpose of integrating

mBerakhot 1.5 into the Haggadah liturgy, a recontextualization that could read it as applying to the

“nights” of the Seder.103

While a date for this cannot be determined, an evolutionary position can be imagined. There was

a point during the amoraic period when an avadim hayinu pronouncement (unit 1) was recommended for

the Seder.104 It later came to be augmented in two directions, exemplified by Saadia’s Siddur and by the

Ginze Schechter Haggadah. Both of those share the baraita of the four children (Table 4, unit 10). That

means that, at a point before either units 2-3 (Saadia, Sijilmassa-A) or 4-8 (Sijilmassa-B, Ginze

Schechter) had developed, a Babylonian liturgical redactor had chosen to interject a passage similar to

that found in the Mekhilta (10) that introduced two themes, one being that four types of personalities,

learning styles or developmental stages should be taken into account in explicating the meaning and intent

of the evening ritual; and the other that, furthermore, the instruction is to be done on the Seder night, “a

time when matzah and maror are laid out before you.”

The Saadia tradition reflects further emphasis on the significance of the Exodus for

contemporaries assembled at a Seder (units 2 and 3). In an alternative augmentation, a liturgical composer 102 It is tempting to suppose that a connection with mPesahim 10.4, doresh me-Arami oved avi, is implied. However, even Mandel, op. cit., who explains that d-r-sh denotes exposition, expounding, public instruction (without reference to specific exegetical techniques, “midrash”), and cites that passage as an example, pp. 235-6, stipulates that the Mishnah requires “active recitation” of Scripture, a “content-oriented reading,” rather than narrative or discussion (sapper).

103 Not everyone accepts the notion of this redactional move, insisting, instead, that it be understood as from the perspective of mBerakhot 1. See n. 6?? above, and the chapter on ??mesihin/mesapperim?? below.

104 bPesahim 116a. Although the printed edition attributes it to Samuel, an early Babylonian Amora, other names appear in manuscript witnesses and related works, making it difficult to determine when this statement originated. Cf. n. 13?? above.


chose to reflect on avadim hayinu by preceding the baraita of the four children with the linked pair, units

7 and 8; they combined to emphasize the importance of celebrating the Exodus by articulating it (8) with

a story of a conversation that went on all night (7). This narrative looks back on the theme of avadim

hayinu (the exodus) and forward to the Mekhilta-like baraita (inculcation and instruction on it).

One may question whether this composer specifically connected units 7-8 to unit 1 by means of

the transition elements, units 4 and 6 (Sijilmassa-B, the earliest extant version,105 is our point of reference

here), or whether a later liturgist added them. It does seem clear that unit 4 was elaborated into unit 5 at a

later point (to resonate rhetorically with the sages in Bene Beraq), which replaced 4 in most versions.106

Be that as it may, units 4 (or 5) and 6 make explicit a theme that was acted out in unit 7, viz., talking

about the Exodus, while extending it to include in this practice every type of person, not just sages and

elders. Unit 6 defines it as an aspirational practice (kol ha-mesapper … hare zeh meshubbah). In

characterizing it as praiseworthy, the liturgist acknowledges that this is either an innovation or a

supererogation, or both, albeit not a requirement. While it may be possible that units 4/5-6 developed in

tandem with the inclusion of 7-8, there is problem, viz., one must account for the fact that the former

employs the verb sapper, while early versions of the unit 7 use mesihin.

105 Although it includes units 2-3 as well. Since those units are wanting, however, in the Ginze Schechter version, it seems that Sijilmassa-B contains an amalgamation of both traditions. The fact that all other witnesses to the Bene Beraq version include unit 2 from that of Saadia indicates the latter’s continuing influence on the Bene Beraq tradition as it developed.

106 The Ginze Schechter text has accumulative nature: it retained both units 4 and 5; moreover, its unit 5 consists of an amalgamation of two different elaborations.


marcus jastrow, - chagumoed.weebly.com€¦ · web viewalthough, elazar ben azariah speaks last...

Documents