Hostname: page-component-89b8bd64d-5bvrz Total loading time: 0 Render date: 2026-05-06T14:10:44.700Z Has data issue: false hasContentIssue false
  • English
  • Français

Passover and Last Supper Revisited

Published online by Cambridge University Press:  10 June 2013

Joel Marcus
Joel Marcus*
Affiliation:
Duke Divinity School, Box 90968, Durham, NC 27708-0968, USA. email: jmarcus@div.duke.edu
Rights & Permissions [Opens in a new window]

Abstract

Although Jesus' Last Supper probably took place on the night before Passover (as in John) rather than on the first night of Passover itself (as in the Synoptics), it contained elements strongly marked by the Jewish institution of the Passover seder (fixed order of service) and haggadah (ritual retelling of the exodus events). These elements were not, as some scholars of Judaism have recently argued, post-70 CE developments. Rather, evidence from Jubilees, Philo, and the NT itself indicates that seder and haggadah already existed in some form in the pre-70 period.

Keywords

Last SuppereucharistPassoversederhaggadah

Information

Type
Articles
Information
New Testament Studies , Volume 59 , Issue 3 , July 2013 , pp. 303 - 324
Copyright
Copyright © Cambridge University Press 2013 

1. Introduction

Was the Last Supper a Passover seder? Although there have always been doubters, it is safe to say that, a generation ago, the usual scholarly answer was, ‘Yes’. To be sure, the discrepancy between John and the Synoptics over the dating of the Supper was acknowledged as a problem.Footnote 1 According to Mark (14.12, 17-26), who is closely followed by Matthew (26.17) and Luke (22.7), the Supper occurred on the evening of ‘the first day of Unleavened Bread’,Footnote 2 that is, at the beginning of Nisan 15.Footnote 3 The Last Supper, therefore, took place at the beginning of Passover. In accordance with this Synoptic chronology, Jesus' death the following afternoon occurred on what was still Nisan 15, the first ‘day’ of the holiday according to the normal Jewish method of time-reckoning, in which ‘day’ begins at sunset.Footnote 4 According to John, however, Jesus died on the afternoon of the day of preparation for Passover, Erev Pesach (19.14, 31, 42), that is, Nisan 14; his Last Supper the night before, therefore, was not a Passover meal.

This discrepancy, however, was not viewed as an insurmountable problem by NT scholars affirming a Passover setting. Either they argued that the Synoptics rather than John were right on this particular,Footnote 5 or they speculated that John and the Synoptics were using different calendars,Footnote 6 or they asserted that, while John's dating might be correct, Jesus, sensing the imminence of his arrest and execution, may have modeled his last meal on the feast of deliverance he did not think he would live to celebrate.Footnote 7

But a more serious challenge to this consensus has emerged in recent years, and it has come primarily from scholars of ancient Judaism rather than NT specialists. The question these researchers have posed is: In Jesus' time, was there actually such a thing as a Passover seder? That is, was there in the early first century CE a Jewish custom of gathering on the first night of Passover at a ceremonial meal whose distinctive elements, arranged in a fixed order (the literal meaning of seder),Footnote 8 were interpreted for the edification of the participants in a ritual retelling (haggadah) that linked those elements with the exodus from Egypt?Footnote 9 And more and more of these researchers have been answering this question with a ‘no’, identifying the seder instead as essentially a post-70 CE replacement for the pre-70 tradition of Passover sacrifice, which came to an end when the Romans destroyed the Temple in which Jewish sacrifice took place.Footnote 10 And this conclusion matters for historians of early Christianity because the words of institution that Jesus speaks over the bread and wine in the Synoptics (‘This is my body… This is my blood’) are usually interpreted as his own twist on the Jewish custom of explaining the matzah and other seder elements—a connection that has helped scholars both to interpret Jesus' words and to maintain their historicity. But if there was no such Jewish custom, that whole approach falls to the ground.Footnote 11

What is the reason for these doubts about the existence of a seder rite in the pre-70 period? The central arguments are the following:

  1. 1. The foundational Pentateuchal passages dealing with the Passover festival (Exod 12–13 and Deut 16) outline neither a seder nor a haggadah, as defined above; they merely specify such things as how the sacrificial lamb should be chosen (from the sheep or the goats), how it should be cooked (roasted), and how it should be eaten (with unleavened bread and bitter herbs). No set order in the eating of these foods is prescribed, nor is it said what prayers or hymns, if any, should accompany their consumption. In other words, there is no seder in the strict sense. Neither is there a haggadah. The instructions that specify the way in which a father should reply to his son when the latter asks about the distinctive rites of the feast (Exod 12.26-27; 13.14-15; Deut 6.20-25; cf. Exod 13.8) are ad hoc; they outline the sort of thing that should be said if and when queries arise, not a fixed arrangement of ritualized questions and answers.Footnote 12

  2. 2. Later biblical and Second Temple Jewish texts dealing with Passover concentrate on the sacrifice in the Temple; they do not refer to a fixed order of foods, blessings, and hymns, or a ritual retelling of the exodus events.Footnote 13 Nor do they provide any clear evidence for a domestic celebration of Passover by Jews generally in the Second Temple period.Footnote 14 In fact, several texts militate against such a supposition, since they stipulate or suggest that the Passover sacrifice should be eaten only within the Temple courts (2 Chron 35.11-13; Jub. 49.16-21; 11QTemple 17.6-9) or the city of Jerusalem (m. Zev. 5.8).Footnote 15

  3. 3. The earliest evidence for a haggadic accompaniment to the Passover meal comes in the Mishnah (Pesaḥim 10), which was redacted in the early third century CE and basically outlines the seder as it presently exists: four cups of wine; haggadah before the meal consisting of interpretation of Passover sacrifice, matzah, and bitter herbs; recitation of the Hallel (Pss 115–118); and afikoman (post-prandial treat). But recent scholars have argued that this chapter of the Mishnah is secondary to the corresponding chapter of the Tosefta, which lacks reference to the haggadah and instead mandates a discussion of the laws (not the events) of Passover after (not before) the meal.Footnote 16 If these scholars are right, there probably was no such thing as the Passover haggadah until a very late stage in the game,Footnote 17 certainly later than the first century.Footnote 18

  4. 4. The biblical Passover meal is to be eaten in haste (בחפזון/μɛτὰ σπουδῆς; Exod 12.11).Footnote 19 According to Friedman and Hauptman, this was still the practice in late Second Temple times, as is shown by the custom of the ‘Hillel sandwich’, which was originally a device for fulfilling the biblical commandment by eating the three mandated foods all at once and thus dispatching them swiftly.Footnote 20 Hence in Hillel's time, the early first century CE, the Passover celebration had not yet developed into the sort of gracious meal, accompanied by wine and appetizers, that it would later become and that is the presupposition for the haggadah. Friedman and Hauptman trace this innovation to the Tosefta, and Hauptman views it as a post-70 adaptation to the culture of the Greco-Roman symposium.Footnote 21

2. Jubilees and Philo

Strong as this case appears to be, and supported though it is by such an impressive consensus, I remain unconvinced. It seems to me that these recent views overemphasize the biblical and rabbinic evidence and downplay or ignore evidence from the book of Jubilees, Philo, and especially the NT. The latter sources, in my opinion, point towards the emergence of a leisurely Passover meal and a domestic seder, including haggadic recital, in the pre-70 period.

The neglect of these sources distorts the results of some Passover researchers. Several of the contentions of Friedman and Hauptman, for example, are belied by evidence from the book of Jubilees. That book shows that it was not the post-70 CE redactors of the Mishnah or Tosefta who first turned the Passover celebration from a hasty repast of lamb, matzah, and bitter herbs into a ‘gracious meal’ accompanied by wine. Rather, the second-century BCE author of Jubilees is already concerned to show that the biblical regulations about eating the meal in haste applied only to the first Passover celebration, not to subsequent ones: ‘For you celebrated this festival hastily when you were leaving Egypt until the time you crossed the sea into the wilderness of Sur, because you completed it [the first Passover] on the seashore’ (Jub 49.23, emphasis and bracketed material added). The intent here seems to be to show that the note in Exod 12.11 about consuming the Passover meal in haste is not meant to apply to life in the author's present.Footnote 22 The hermeneutical strategy, therefore, is strikingly similar to that in the much later Targum Pseudo-Jonathan, which specifies that the instructions in Exod 12.11 about eating in haste apply only to ‘this time and not to future generations’.Footnote 23 Both the author of Jubilees and Pseudo-Jonathan, apparently, lived in communities in which the Passover meal was consumed in a leisurely manner. But how could they square this custom with the explicit injunction to haste in Exod 12.11? The answer both adopted was to limit the applicability of that injunction to the first Passover.Footnote 24

Thus, while Jubilees provides no evidence for a domestic celebration of Passover, and even polemicizes against it (see 49.21), it does show that, already in the second century BCE, some Jews were treating the Passover meal as a leisurely repast to be enjoyed with wine (see 49.6), contrary to the spartan regulations of Exodus.Footnote 25 And the sharp polemic of Jubilees against domestic celebration of Passover may suggest that some Jews known to the author were celebrating the feast at home.

That they were doing so a century or so later, but still before the destruction of the Temple, seems to be suggested by Philo, Spec. 2.148:

ἑκάστη δὲ οἰκία κατ' ἐκɛῖνον τὸν χρόνον σχῆμα ἱɛροῦ καὶ σɛμνότητα πɛριβέβληται, τοῦ σϕαγιασθέντος ἱɛρɛίου πρὸς τὴν ἁρμόττουσαν ɛὐωχίαν ɛὐτρɛπιζομένου.

On this day every dwelling-house is invested with the outward semblance and dignity of a temple. The victim is then slaughtered and dressed for the festal meal which befits the occasion.Footnote 26

If, as this passage seems to imply, the slaughter of the Passover sacrifice is to take place at every dwelling-house (ἑκάστη…οἰκία) in Jewry world-wide (cf. QE 1.10), we seem to be dealing with at least the rudiments of a domestic celebration of Passover.Footnote 27

We have evidence from Jubilees, therefore, that the Passover meal had become a leisurely repast by the second century BCE and from Philo that it had become (or reemerged as) a domestic celebration, at least in some circles, by the beginning of the first century CE. These are necessary conditions for the development of seder and haggadah, but Jubilees and Philo do not themselves provide unequivocal evidence for the emergence of those forms. There are, however, a couple of tantalizing hints in Philo that some form of the seder may have existed by his time. These hints are contained in two passages which, as Naomi Cohen points out, are similar in striking ways to two important sections in the Haggadah.Footnote 28

The first Philo passage, de Congressu 167, asserts that the unleavened bread of Passover, despite its biblical description as ‘the bread of affliction’ (see Deut 16.3), is not an instrument of suffering but an essential component of ‘the meal of festivity and joy’ (τὴν ἑορτῆς καὶ ɛὐϕροσύνης τράπɛζαν, my trans.).Footnote 29 This reversal of the valence of ‘bread of affliction’ is similar to that which occurs in a famous Aramaic passage in the Haggadah that also echoes Deut 16.3, since it begins, ‘This is the poor bread (הא לחמא עניא) which our ancestors ate in the land of Egypt’.Footnote 30 This passage, which is referred to as Ha Lachma after its first two words, goes on to invite the needy and hungry to come into the house where the meal is taking place and join in the paschal sacrifice by eating this ‘poor bread’. Thus, as in the Philo passage, the invitation to eat ‘the bread of affliction’ is paradoxically viewed as a cause not for sorrow but for joy. We will return to Ha Lachma below.

The second Philo passage, Quaestiones et Solutiones in Exodum 1.15, is even more striking:

Unleavened bread is (a sign) of great haste and speed, while the bitter herbs (are a sign) of the life of bitterness and struggle which they endure as slaves. That is that which is said (ῥητόν).Footnote 31 But as for the deeper meaning, this is worth noting, (namely) that that which is leavened and fermented rises, while that which is unleavened is low. Each of these is a symbol of types of soul, one being haughty and swollen with arrogance, the other being unchangeable and prudent, choosing the middle way rather than extremes because of desire and zeal for equality. But the bitter herbs are a manifestation of a psychic migration, through which one removes from passion to impassivity and from wickedness to virtue. For those who naturally and genuinely repent become bitter toward their former way of life.

Here Philo cites an interpretation of the matzah as a sign of haste, presumably that with which the Jews were forced to leave Egypt, and of the bitter herbs as a sign of their suffering in that country. The exact same connections are made in the Passover Haggadah, in a passage partly paralleled by a Mishnaic saying attributed to Rabban Gamaliel, a first-century rabbi (m. Pes. 10.5; on Gamaliel's identity, see below). Even more importantly, Philo makes it clear that the interpretations he relates are not his own invention (he goes on to give spiritual exegeses more to his liking) but belong to τὸ ῥητόν—a term that means ‘that which is said’ and could appropriately be rendered in Hebrew with haggadah.Footnote 32 As noted, this is not unequivocal evidence for the existence of the haggadah in Philo's time, but it does suggest that it may have existed by then.Footnote 33

3. The NT Evidence

Already, then, Jubilees and Philo suggest that by the early first century CE the Passover meal may have become an occasion for expounding the significance of the particular holiday foods at a leisurely repast held at home. The first unequivocal evidence for this custom, however, comes from the NT. I do not think that the seder-skeptics have fully weighed the significance of this testimony.Footnote 34

The most important datum is that, as we have already seen, all three Synoptic Gospels portray Jesus' Last Supper as a Passover meal and show him ritually distributing matzah and wine to his disciples at this meal and interpreting these elements symbolically and in sacrificial terms (‘my body [given for you]…my blood shed on behalf of many’).Footnote 35 Moreover, at least two out of the three Synoptics (three out of three if the longer reading in Luke 22.19b-20 is accepted)Footnote 36 link the ‘cup word’ with the covenant established by Moses in the exodus when they show Jesus echoing Exod 24.8, ‘Behold the blood of the covenant…’Footnote 37

As has been noted above, this Synoptic identification of the Last Supper as a Passover meal contrasts with the situation in the Gospel of John, where the Supper occurs on the night before Passover begins. Several recent seder investigators have used this discrepancy to relativize the Synoptic evidence, arguing that John is probably more accurate in dating the Supper to Erev Pesach.Footnote 38 I agree on this narrow point of chronology,Footnote 39 but in my view that does not diminish the importance of the Synoptic evidence for the question about the existence of a pre-70 CE seder. The Synoptics, after all, are strongly rooted in pre-70 traditions, as is clearly demonstrated with regard to the Last Supper story in particular by the parallel in 1 Cor 11.23-26, and Mark at least may have been composed before the destruction of the Temple—if not, shortly thereafter.Footnote 40 Even in the latter case, there was scarcely sufficient time between the destruction of the Temple in 70 and the composition of Mark perhaps a year or two later for a thorough transformation of the Passover celebration to occur. The important question for our purposes, then, is not whether or not Jesus' Last Supper actually was a Passover meal, but whether or not the Synoptic Gospels, which are rooted in pre-70 realities, portray it as such. And since the answer to that question is ‘yes’, the Synoptics provide valuable evidence for the shape of the Passover celebration before 70.Footnote 41

Moreover, it is striking that John, as noted, portrays Jesus' last meal as occurring on the night before Passover and as lacking his symbolic actions and words over the bread and wine. The Synoptics, by contrast, picture Jesus' last meal as a Passover supper, and this meal does contain those interpretative actions and words. Is this combination of Synoptic presences and Johannine absences just a coincidence? That seems unlikely; rather, it is probable that the authors of the Synoptics think that Jesus' symbolic actions and words fit into the context of a Passover meal.Footnote 42 And this makes sense in light of the important ways in which these actions dovetail with the portrayal of the Passover seder in later Jewish sources. This is not a matter of reading the evidence from these later sources back into the NT accounts but of concluding from their distinctive shared characteristics that the Passover rites depicted in these different corpora have some sort of genealogical relationship to each other.Footnote 43

4. The ‘Bread Word’ and Ha Lachma

The most important of these shared characteristics is what Jeremias refers to as Jesus' ‘altogether extraordinary manner of announcing his passion’ through ‘speaking words of interpretation over the bread and the wine’.Footnote 44 While this sort of table talk has precedents in the Greco-Roman symposium, where the foods at the feast sometimes turn into the subject of the conversation, it is unprecedented in ancient Jewish contexts—except for the Passover seder.Footnote 45 The earliest rabbinic reference to the custom is found in the Mishnaic passage to which we have already referred, Pesaḥim 10.5, in which Rabban Gamaliel designates the matzah as one of the three special foods that must be interpreted at every Passover meal. (We will return to the other two below.) This demand is repeated in the Passover Haggadah and is fulfilled there by an interpretation that treats the unleavened bread as a sign of the Israelites' need to hurry out of Egypt.Footnote 46

But it is also fulfilled near the beginning of the seder in the Ha Lachma paragraph to which reference was made earlier. For convenience of reference, I give the paragraph in full below and number its component sentences:

הא לחמא עניא די אכלו אבהתנא בארעא דמצרים (1

כל דכפין ייתי ויכל כל דצריך ייתי ויפסח (2

השתא הכא לשנה הבאה בארעא דישראל השתא עבדי לשנה הבאה בני חורין (3

  1. 1) This isFootnote 47 the poor bread that our ancestors ate in the land of Egypt.

  2. 2) All who are hungry, let them come and eat; all who are needy, let them come and partake of the Passover sacrifice.

  3. 3) This year we are here, next year we shall be (or: let us be) in the land of Israel; this year we are slaves, next year we shall be (or: let us be) free people.

The similarity in structure and meaning of ## 1 and 2 to the ‘bread word’ in the Synoptic tradition is striking. Here, for example, is Luke's version of this saying (Luke 22.19), the main elements of which are drawn from Mark and supported by the early passage 1 Cor 11.24:

καὶ λαβὼν ἄρτον ɛὐχαριστήσας ἔκλασɛν καὶ ἔδωκɛν αὐτοῖς λέγων·

τοῦτό ἐστιν τὸ σῶμά μου τὸ ὑπὲρ ὑμῶν διδόμɛνον·

τοῦτο ποιɛῖτɛ ɛἰς τὴν ἐμὴν ἀνάμνησιν.

And having taken bread and having given thanks, he broke it and gave it to them saying:

This is my body which is given for you.

Do this for the remembrance of me.

We have here, as in the Ha Lachma paragraph, an invitation to eat the matzah,Footnote 48 and this act of eating is linked with the theme of remembrance that is implied in Ha Lachma (‘that our ancestors ate in the land of Egypt’) and is so integral to the whole seder.Footnote 49 In both cases, moreover, the eating of the matzah has sacrificial overtones (‘my body given for you/partake of the Passover sacrifice’). Most importantly, Jesus' introductory formula in all versions of the saying, ‘This is my body which…’ is strikingly similar to the first words of Ha Lachma, ‘This is the poor bread which…’Footnote 50 What are we to make of these parallels?

They would be less important for our purposes—though they would still be interesting—if recent scholarship were right and Ha Lachma were a late addition to the seder service. Kulp, for example, notes that, while it ‘appears in geonic Haggadot and in most manuscripts and geniza fragments of the Haggadah…it does not appear in ancient Eretz Yisraeli Haggadot’, and Goldschmidt relates that, where the paragraph is present, the order of its sentences varies, and sometimes the first sentence (the crucial one for our purposes) is missing altogether.Footnote 51 As for Talmudic evidence, Klawans remarks that, while the Bavli discusses the biblical phrase ‘bread of affliction’ in several places (see, e.g. b. Ber. 39b; b. Pes. 36ab, 115b), it never mentions Ha Lachma,Footnote 52 and Goldschmidt observes that, while b. Taʾan. 20b offers a parallel to the second sentence of the paragraph, it does not present the invitation to the needy in the context of the seder but simply relates it to the customary charity of R. Ḥuna, a late second-century Amora.Footnote 53 There has been a recent tendency, therefore, to date Ha Lachma late; Safrai and Safrai, for example, pronounce it a product of the Babylonian Geonim, and Leonhard dates it even later, perhaps to the twelfth century CE.Footnote 54

These arguments, however, are perhaps less decisive than their framers think, given the fragmentary nature of our evidence from Jewish antiquity, in particular about folk celebrations such as the seder.Footnote 55 Moreover, the lack of attestation to Ha Lachma in rabbinic sources is counterbalanced by the parallels from other sources, which open up the possibility that some parts of it may have existed early on. We have already seen, for example, that Philo parallels the first two sentences of the paragraph by turning the biblical ‘bread of affliction’ into a paschal meal to be celebrated joyfully. There is also a noteworthy parallel between these same sentences of Ha Lachma and John 6.35—a passage that occurs, significantly enough, in a Passover setting (cf. 6.4):Footnote 56

This is the poor bread…   I am the bread of life

All who are hungry   The one who comes to me

let them come in and eat   will not go hungry

But the most compelling piece of evidence is, of course, the striking correspondence in form and meaning between Jesus' words of institution, ‘This is my body’, and the beginning of Ha Lachma, ‘This is the poor bread’. Again, it needs to be asked: Can this parallel be ascribed to chance? Such a striking correspondence in theme and wording seems unlikely to be fortuitous or an example of independent development.Footnote 57 Therefore there seems to be a prima facie case that some form of Ha Lachma, and hence of the Passover seder, already existed in the pre-70 era.Footnote 58

5. Seder as Reversal of Last Supper?

There is, however, an alternate way of explaining the parallels between the Passover seder and the Last Supper, and it has been argued with vigor in recent years by Israel Yuval. This is the theory that the seder itself is essentially a response to the Christian eucharist rather than being its source. Thus, for example, the striking parallel between Ha Lachma and Jesus' words of institution is interpreted as a polemical reversal of the latter by the former.Footnote 59 Here is Yuval's reconstruction of the way in which the seder developed in the context of what he calls ‘Jewish–Christian dialogue’ (though it would probably be truer to his theory to speak of Jewish polemic against Christianity):

During the time of the Temple the celebration of Passover included two main components, the sacrificial meal and the Hallel. For two generations after the Temple's destruction, instead of the defunct sacrifice, people generally ate a roasted kidFootnote 60 (a custom, perhaps, in distant communities before the Destruction as well) and studied the laws of sacrifice that they could no longer perform. This is the tradition described in the Tosefta's account of scholars gathering to study the laws of Passover all night long. At this stage, the Christian midrash on Exodus 12 and the paschal sacrifice emerged. In response, the Jewish Haggadah distanced itself from sacrifice and emphasized instead the duty to tell the story of the Exodus, as described in the Mishnah (M. Pes. 10). This stage was crystallized by the beginning of the second century C.E.Footnote 61

Crucial to this theory is Yuval's analysis of Gamaliel's statement about the necessity of interpreting Passover sacrifice, matzah, and bitter herbs at the seder in m. Pes. 10.5. In line with most recent seder researchers, Yuval identifies this Gamaliel as Gamaliel II, a leader of the post-70 rabbinic movement, rather than Gamaliel I, his grandfather, a pre-70 figure who is mentioned in Acts 5.34-39; 22.3.Footnote 62 The distinctive point of his analysis is to interpret Gamaliel's statement as a ‘reversed parallel’ to Christian theology, in which the pesach is Jesus (John 1.29; 1 Cor 5.7), the matzah is his body (Matt 26.26//Mark 14.22//Luke 22.19), and the bitter herbs are his suffering (Aphrahat Dem. 12.8)Footnote 63 or the punishment awaiting Israel for rejecting him (Melito Peri Pascha 93).Footnote 64 According to Yuval, then, ‘Rabban Gamaliel is demanding a declaration of loyalty to the Jewish understandings and, therefore, an implicit denial of the Christian alternative’.Footnote 65 Yuval also sees Gamaliel's ban on concluding the Passover meal with an afikoman (m. Pes. 10.8) as a response to Christian beliefs about Jesus, since Melito (Peri Pascha 66) uses the word ἀϕικόμɛνος to speak about the ‘coming’ of Jesus in incarnation and passion.Footnote 66

As this last example might suggest, Yuval tends to see parallels and polemic everywhere, and some of his arguments seem far-fetched.Footnote 67 There is, as a matter of fact, a more sensible and widely accepted explanation for the afikoman, since that term seems to be a loanword (ἐπίκωμον) that is used, along with its cognates, in Greek sources to indicate the sort of after-dinner revelry that sometimes followed Hellenistic banquets and symposia.Footnote 68 Gamaliel's dictum, ‘They do not follow the meal at which the Passover sacrifice is eaten with afikoman’, fits this context perfectly. As Baruch Bokser points out, moreover, the Mishnah did not invent the contrast between Passover feasting and the debauchery of pagan banquets. Philo, for example, warns that those at the Passover feast are not to overindulge in food and wine ‘like those in other symposia’,Footnote 69 and Josephus, in Bokser's words, ‘sets the Passover rite apart from regular banquets with the phrase “feasting alone not being permitted”’.Footnote 70 Another example comes from a NT passage that to my knowledge has not previously been mentioned in this regard, 1 Cor 5.7-8:

ἐκκαθάρατɛ τὴν παλαιὰν ζύμην, ἵνα ἦτɛ νέον ϕύραμα, καθώς ἐστɛ ἄζυμοι· καὶ γὰρ τὸ πάσχα ἡμῶν ἐτύθη Χριστός. ὥστɛ ἑορτάζωμɛν μὴ ἐν ζύμῃ παλαιᾷ μηδὲ ἐν ζύμῃ κακίας καὶ πονηρίας ἀλλ' ἐν ἀζύμοις ɛἰλικρινɛίας καὶ ἀληθɛίας.

Purge the old leaven that you may be a new lump, as you really are unleavened. For Christ, our Passover lamb, has been sacrificed. Let us, therefore, celebrate the feast, not with the old leaven, the leaven of evildoing and fornication, but with the unleavened bread of sincerity and truth.

Here we see reflected, not only the biblical ceremonies of purging the house of leaven (see Exod 12.15) and sacrificing the Passover lamb, but also a warning against letting the paschal feast become an occasion for dissipation—the same sort of reaction against Hellenistic banqueting practices that we have noticed in Philo, Josephus, and Gamaliel's dictum about the afikoman. Rather than being a post-70 CE response to Christianity, then, the afikoman seems to reflect a Jewish understanding pitting the Passover meal against Hellenistic banqueting customs, an approach that existed already in the early first century CE and is attested in an early NT text.

As for Yuval's argument that m. Pes. 10.5 is a response to Christian interpretations of the biblical elements of the Passover meal, we have seen that Philo already offers interpretations of two of the three foods mandated here, interpretations very similar to those that later appear in the Haggadah, and ascribes these interpretations to ‘the traditional exegesis’ (ῥητόν). Yuval's theory is rendered further suspect by the way in which he combines different Christian sources from widely varying times, some of them subsequent to the era of Gamaliel (Melito died around 180 and Aphrahat wrote between 336 and 345), to construct an artificial picture of a Christian understanding to which Gamaliel's statement is supposed to be a response.Footnote 71

Moreover, the movement Yuval posits from the Christians' ‘spiritual’, Christological interpretation of the Passover foods to the more literal interpretation of them in Jewish sources makes less sense than seeing the development as going in the opposite direction, from the more literal to the more spiritual. Indeed, the latter is the direction in which we can see Philo himself moving in QE 1.15, where he first cites the traditional, more literal exegesis of the bitter herbs and matzah, then develops his own spiritual exegesis, which is less tethered to the details of the biblical text. For similar reasons, it seems more likely that Ha Lachma's literalistic ‘this is the poor bread’ statement was transformed into Jesus' highly metaphorical ‘this is my body’ saying than the other way around. Furthermore, neither Ha Lachma nor Gamaliel's statement in m. Pes. 10.5 betrays any overt sign of being the sort of anti-Christian polemic that Yuval alleges. Ancient religious ideological warfare was usually not conducted so subtly, and it seems methodologically unsound to posit its existence in passages that betray no overt sign of it.Footnote 72 Yuval's theory, moreover, ignores the different liturgical contexts of the Passover seder and the Christian eucharist, the former being celebrated annually and the latter weekly. If the rabbis had intended to respond to eucharist, one might have expected this response to be incorporated into the weekly Sabbath meal rather than the annual seder.Footnote 73 A movement in the other direction, from the seder-like Last Supper to the weekly celebration of the eucharist, makes more sense, given the centrality of Jesus' death in early Christianity.

I do think that Yuval has performed a valuable service by raising the question of the function of Gamaliel's dictum in m. Pes. 10.5, but I do not think that the answer he gives is the only one possible, or the most compelling. ‘Whoever does not mention these three things at Passover has not fulfilled his obligation’ might, as Yuval posits, be a way of introducing a new religious duty. But it might also be the repetition of a traditional demand or, more likely than either, a new version of a traditional requirement. In other words, before Gamaliel's time it may have been recognized that there was an obligation to interpret the special holiday foods on the first night of Passover, but there may have been unclarity about exactly which ones needed to be interpreted, and Gamaliel's dictum may have been an attempt to end that unclarity. And this sort of new twist on a traditional custom is exactly what we see Gamaliel doing in the famous passage in b. Ber. 28b-29a in which he seeks for a way to reformulate (לתקן) one of the statutory Eighteen Benedictions, that against the heretics, to reflect the changed conditions of his own time.Footnote 74

This sort of interpretation of Gamaliel's dictum as trying to end an undesirable variety of practices also corresponds to the variation we have noted in earlier sources with respect to the foods that should be the center of attention at Passover. Jubilees mentions the sacrificial lamb and the wine, while Philo refers to and gives traditional interpretations for the matzah and bitter herbs. The Jesus of the Synoptics says words of interpretation over the matzah and the wine, whereas Gamaliel in the Mishnah specifies the Passover lamb, bitter herbs, and matzah, giving a slightly different interpretation of these elements, and in a different order, than appears in the Haggadah. The best interpretation of this variation would seem to be that, prior to Gamaliel's time, there was a Passover custom of explaining the distinctive holiday foods, but there was variation, as befits a folk ceremony, with regard to which foods needed to be explained and how. It was this variation that spurred Gamaliel to promulgate what he hoped would be an authoritative ruling.Footnote 75

6. Conclusions and Ramifications

  1. 1. Since Passover originated as a folk ceremony and probably continued to be so in later periods, the best starting point for reconstructing its shape in the Second Temple period is not priestly injunctions or the dicta of later sages but the cumulative evidence of all sources from the pre-70 period, including the NT.Footnote 76

  2. 2. As might be expected in dealing with a folk rite, those sources provide evidence for a variety of practices.

  3. 3. Common to at least several of those sources, however, is the custom of interpreting some of the special Passover foods—though the sources, again as expected, differ on exactly which foods need to be interpreted and how.

  4. 4. In the post-70 period, rabbinic sages such as Gamaliel tried to standardize these practices.

  5. 5. The sages were only partially successful in rabbinizing the seder, as can be seen, for example, in the correspondence against the Mishnah between the Haggadah and the traditional exegesis cited by Philo (see above). Even today, the Haggadah is an amalgam uniting a popular underlay (e.g. Ha Lachma, Dayyenu, Chad Gadya) with a rabbinic overlay (e.g. Gamaliel's dictum about what foods are to be interpreted, the Bnei Berak stories, the ‘I and not an angel’ midrash).

  6. 6. The variety of seder practices in the present is the latest reflection of the creative tension that has always existed within Passover between that which has been handed down (traditum) and the traditioning process that continues to introduce change (traditio), which is partly catalyzed by popular interests and pressure.Footnote 77

References

1 See, e.g., Jeremias, J., The Eucharistic Words of Jesus (Philadelphia: Fortress, 1966) 19-26Google Scholar.

2 Unless otherwise noted, all translations of NT texts are my own.

3 On the dating of the Passover, see Exod 12.6; Lev 23.5; Num 9.3, 5, 11. These texts, however, do not name the month as Nisan, which is a Babylonian term that only came into use among post-exilic Jewish writers (see Neh 2.1; Est 3.7). The earlier name for the month was Aviv (see Exod 13.4; 23.15; 34.18; Deut 16.1).

4 This despite the terminology of Mark 14.12, which uses the normal Greco-Roman sunrise-to-sunrise method of reckoning days, which was sometimes adopted by ancient Jews as well; see Marcus, J., Mark: A New Translation with Introduction and Commentary [AYB 27/27A; 2 vols.; New Haven and London: Yale University, 2000, 2009)Google Scholar, 2.944.

5 See, e.g., Jeremias, Eucharistic Words, 79-83, who ascribes John's dating to an ‘anti-passover attitude’.

6 See Jaubert, A., The Date of the Last Supper (Staten Island, NY: Alba House, 1965)Google Scholar; cf. recently Instone-Brewer, D., ‘Jesus' Last Passover—the Synoptics and John’, ET 112 (2001) 122-3Google Scholar; Instone-Brewer, Feasts and Sabbaths: Passover and Atonement (Traditions of the Rabbis from the Era of the New Testament 2A; Grand Rapids: Eerdmans, 2011) 130-5Google Scholar.

7 See recently Routledge, R., ‘Passover and Last Supper’, TynBul 53 (2002)Google Scholar 206.

8 The term itself is post-Tannaitic; see Bokser, Baruch, The Origins of the Seder: The Passover Rite and Early Rabbinic Judaism (Berkeley: University of California, 1984)Google Scholar 146.

9 On the definition of seder and haggadah, see Kulp, J., ‘The Origins of the Seder and the Haggadah’, CBR 4 (2005)Google Scholar 111. In what follows, I will use lower-case haggadah for this act of ritual retelling and upper-case Haggadah for a book that fixes this retelling in literary form.

10 See the summary by Kulp, ‘Origins’, 112: ‘Nearly all rabbinics scholars…agree that most of the elements known from the seder as described in the Mishnah are missing from descriptions in Second Temple literature… This includes the absence of a seder or a haggadah. The primal element that did exist in the Second Temple was the sacrifice of the lamb.’ Among those whom Kulp mentions as sharing this consensus are Bokser, Origins, 14-28; Safrai, S. and Safrai, Z., Haggadah of the Sages (1998; repr., Jerusalem: Carta, 2009) 9-15Google Scholar; Tabory, J., ‘Towards a History of the Paschal Meal’, Passover and Easter: Origin and History to Modern Times (ed. Bradshaw, P. F. and Hoffman, L. A.; Notre Dame: University of Notre Dame, 1999)Google Scholar 63; Hauptman, Judith, ‘How Old Is the Haggadah?’, Judaism 51 (2001) 5-18Google Scholar; Friedman, S., Tosefta Atiqta: Pesaḥ Rishon. Synoptics Parallels of Mishna and Tosefta Analyzed with a Methodological Introduction [Hebrew] (Ramat-Gan: Bar-Ilan University, 2002) 430-2Google Scholar. See also Leonhard, C., The Jewish Pesach and the Origins of the Christian Easter: Open Questions in Current Research (Studia Judaica 35; Berlin and New York: de Gruyter, 2006) 1-118CrossRefGoogle Scholar.

Within this consensus, there are distinctions. Bokser, Origins, xiii, for example, thinks that, alongside of the Temple sacrifice, which was primary, there was some sort of domestic rite in the pre-70 period, but we can know little about it. Hauptman, ‘How Old’, 9, acknowledges that, in biblical and Second Temple times, ‘people may have told the story [of the exodus] to their children’, but she thinks that we are ignorant about the form this narration took, and that whatever it was, it was far from the seder and haggadah as known today.

11 See Klawans, J., ‘Was Jesus' Last Supper a Seder?’, BRev 17, no. 5 (2001) 24-33Google Scholar, 45; cf. Leonhard, C., ‘Das alttestamentliche und das jüdische Pesachfest’, Die Osterfeier in der alten Kirche (ed. auf der Maur, H., Messner, R., and Schöpf, W. G.; Liturgica Oenipontana 2; Münster: Lit, 2003)Google Scholar 22: ‘Das gesamte Neue Testament bezeugt damit keinen Vorläufer der Pesachhaggada’.

12 As Hauptman, ‘How Old’, 9, points out, similar ad hoc questions and answers are mentioned in passages not directly connected with Passover (e.g. Exod 13.14-16; Deut 6.20-25).

13 The main texts are Ezra 6.19-22; 2 Chron 30.1-27; Jub. 49; Ezekiel the Tragedian Exagoge 167-74; Philo Spec. 2.145-149; V. Mos. 2.224-32; Josephus Bell. 6.423-27; Ant. 2.312-13; 3.248-49; 9.263-72; 10.70-72; 11.109-110; 17.213-14. They are collected and discussed in Bokser, Origins, 19-25, and Tabory, J., The Passover Ritual Throughout the Generations (Tel Aviv: Hakibbutz Hameuchad, 1996 [Hebrew]) 78-92Google Scholar.

14 Most scholars think that Passover originated as a domestic rite, or probably as two such rites (purging the house of leaven and apotropaic sacrifice), but that the majority of the biblical references reflect a later stage in which the holiday had been transformed into a pilgrimage festival centered on the Temple. See, for example, the contrast between Exod 12, which describes a domestic rite, and Deut 16.5-6, which emphasizes that the passover sacrifice may not be offered ‘within any of your towns which the Lord your God gives you; but at the place which the Lord your God will choose, to make his name dwell in it’ (rsv); cf. Moore, G. Foot, Judaism in the First Centuries of the Christian Era (New York: Schocken, 1971 [orig. 1927–30]) 40-1Google Scholar, and De Vaux, R., Ancient Israel (2 vols.; New York: McGraw-Hill, 1965)Google Scholar 2.485-93. This centralization is usually ascribed to the Deuteronomic reforms under Josiah. An awareness of the mixed signals given off by the biblical witnesses is evident in the harmonization in Jub. 49.16 (cf. v. 21): ‘It [the Passover sacrifice] is no longer to be eaten outside of the Lord's sanctuary [as implied in Exod 12] but before the Lord's sanctuary [as specified in Deut 16]’; trans. (minus bracketed material) from VanderKam, J. C., The Book of Jubilees [CSCO 511/Scriptores Aethiopici 88; Louvain: Peeters, 1989] 321-4Google Scholar; unless otherwise specified, all translations from Jubilees are from this edition). Whenever the domestic rite reemerged, therefore (if indeed it ever disappeared), it was not a new invention but the recrudescence of an ancient custom that had left its mark on an early stream of the OT.

Leonhard, Jewish Pesach, 69, however, argues that Passover never existed as a domestic rite in biblical times; the description in Exod 12 ‘is an allegorical interpretation of the liturgy at the Temple in Jerusalem…shaped as a set of fictitious rules for a primeval ritual’ (69). This seems much less likely than that the domestic rite reflected in Exod 12 was later absorbed into the Temple cult; why would anyone have shaped the Exodus text in such an inevitably misleading way?

15 See Safrai and Safrai, Haggadah, 10-11.

16 This reverses the traditional theory according to which the Tosefta, as its name implies, was a supplement to the Mishnah.

17 Hauptman, ‘How Old’, 12, acknowledges that both the Mishnah and the Tosefta contain traditions that have a pre-history, and that in individual cases Mishnaic traditions may be earlier than Toseftan ones. But the drift of her argument is that, in the most crucial ways, the Mishnaic account of Passover is secondary to the Toseftan one and that both reflect a long process of reshaping the Passover celebration after 70 CE.

18 For example, Kulp, J., The Schechter Haggadah: Art, History and Commentary (Jerusalem: The Schechter Institute of Jewish Studies, 2009)Google Scholar 10, dates the invention of the seder to the late tannaitic period (135–220 CE).

19 The first part of this verse reinforces the impression of haste: ‘In this manner you shall eat it: your loins girded, your sandals on your feet, and your staff in your hand’ (rsv).

20 The earliest evidence for this custom is in t. Pes. 2.22: הלל הזקן היה כורך שלשתן זה בזה ואוכלן: ‘Hillel the elder would wrap the three of them together and eat them’.

21 See Friedman, Tosefta Atiqta, 429, and Hauptman, ‘How Old’, 16 n. 6. On symposium elements in the seder, see Stein, S., ‘The Influence of Symposia Literature on the Literary Form of the Pesah Haggadah’, JJS 7 (1957) 13-44CrossRefGoogle Scholar, and Bokser, Origins, 62-5.

22 Segal, Contra J. B., The Hebrew Passover: From the Earliest Times to A.D. 70 (London Oriental Series 12; London/New York/Toronto: Oxford University, 1963)Google Scholar 233, who does not notice the time-limited qualification and instead asserts that according to Jub. 49.23 ‘[t]he meal must be eaten in the Temple courts, as in Deuteronomy, and with haste’; similarly, Leonhard, Jewish Pesach, 27, 238. But Jub. 49.23 clearly intends to relativize the commandment in Exod 12.11 to eat the meal בחפזון by emphasizing that that commandment applied only to the pre-sea-crossing era; after the crossing, the remaining seven days of the festival were completed on the Sinai side of the sea in the leisurely manner that thereafter prevailed at Passover celebrations. A few verses earlier, in 49.13, the author adopts a different strategy defusing בחפזון, here interpreting it as ‘carefully’ (cf. VanderKam, Jubilees, 319).

23 Tar. Ps-J. Exod 12.11: וכדא הילכתא תיכלון יתיה בזימנא דא ולא לדריא .

24 For a similar rabbinic solution to the problem, see m. Pes. 9.5, which distinguishes between ‘the Passover of Egypt’ (פסח מצרים) and ‘the Passover of generations’ (פסח דורות).

25 It is legitimate to ask how extensive the circles were that followed the injunctions laid down in Jubilees, and unfortunately there is little data to contribute to an answer, aside from the fact that fragments of the book have turned up at Qumran. But a similar question may be asked about the following that the rabbis enjoyed in the tannaitic age; see Goodman, M., State and Society in Roman Galilee, A.D. 132–212 (Oxford Centre for Postgraduate Hebrew Studies; Totowa, NJ: Rowman & Allanheld, 1983)Google Scholar and Schwartz, S., Imperialism and Jewish Society, 200 B.C.E. to 640 C.E. (Jews, Christians, and Muslims from the Ancient to the Modern World; Princeton and Oxford: Princeton University, 2001)Google Scholar.

26 LCL translation by F. H. Colson. Unless otherwise indicated, all translations of Philo are from this edition.

27 Bokser, Origins, 23 asserts that ‘Philo's presentation of the law closely follows the biblical record and does not add any local extratemple practices’, but he does not deal with the contradiction to this view presented by Spec. 2.148. Segal, Hebrew Passover, 240 and Leonhard, Jewish Pesach, 33-5 limit the applicability of Spec. 2.148 to the houses of Jerusalem, but there is no indication of such a limitation in the text. More astute is Tabory, Passover Ritual, 84, who acknowledges that Philo might be referring to what takes place in dwellings in Jerusalem but thinks it more likely, in view of the remark about ‘every house’ being converted into a temple, that he is talking about celebrants ‘who have not engaged in pilgrimage but perform the Passover sacrifice in their own homes, wherever they are’ (my trans.).

28 Cohen, N. G., Philo Judaeus: His Universe of Discourse (BEATAJ 34; Frankfurt am Main; New York: Lang, 1995) 305-10Google Scholar.

29 On τράπɛζα (lit. ‘table’) as a term for ‘meal’, see LSJ 1810 (I2).

30 Translation altered from Kulp, Schechter Haggadah, 26.

31 This passage, like most of the Questions and Answers on Genesis and Exodus, is extant only in an Armenian translation of the fifth century, which seems to have been unusually literal; see R. Marcus, ed., Philo: Supplement. Vol. 1, Questions and Answers on Genesis (LCL; London: Heinemann, 1953) ix-x. In translating the Armenian for LCL, Marcus ‘retranslated into Greek words and phrases that have philosophical or theological importance’; see Marcus, R., ‘Notes on the Armenian Text of Philo's Quaestiones in Genesin, Books I–III’, JNES 7 (1948)Google Scholar 11.

32 The Greek of this passage not being extant (see previous note), the Greek original here is somewhat conjectural, but in four of the five Armenian passages using the word čaṙ in which the Greek is extant (QG 4.172; QE 2.21, 38; Contemp. 28), the original is ῥητόν; see Marcus, R., ‘An Armenian–Greek Index to Philo's Quaestiones and De Vita Comtemplativa’, JAOS 53 (1933)Google Scholar 271. In the exception, QG 4.168, the original is διήγησις = ‘narrative’, which fits my thesis even better, since it is closer in meaning to haggadah than ῥητόν is.

Cohen, Philo Judaeus, 307-8 notes that Philo ascribes the interpretations in QE 1.15 to τὸ ῥητόν, which she translates as ‘traditional exegesis’, but does not note the similarity in meaning of this term to haggadah. In most of the other Philonic uses of (τὸ) ῥητόν (Leg. 2.19; Det. 95; Agr. 157; Ebr. 130; Sob. 65; Her. 258; Fug. 136; QG 2.5*; 4.172), the reference seems to be to the literal meaning of the scripture. But in our passage as well as Sob. 65; QG 3.23*-24; 4.168; QE 2.21, Philo uses it to refer to a scriptural hermeneutic that is somewhat imaginative, though different from the ‘spiritual’ exegesis that he embraces (asterisked passages are extant only in Armenian). R. Marcus, Philo: Supplement. 1, ix, notes that Philo's ῥητόν ‘corresponds to the “literal” or “historical” interpretation of the Church Fathers and to the pešaṭ of the Rabbis’.

33 Hauptman, ‘How Old’, 17 n. 26 criticizes Cohen for asserting that ‘the basic rubrics of the text of the haggadah’ were already current and traditional in Philo's day (Cohen, Philo Judaeus, 313): ‘What she has shown, to my mind, is that Philo knew one of the developing midrashim on the three Passover foods’. The issue between Cohen and Hauptman, then, is whether Philo knew this midrash as a customary part of the Passover service or apart from it. Since Hauptman thinks that that service did not exist in Philo's time, she cannot allow the former possibility.

34 Safrai and Safrai, Haggadah, 9-15, in their section on ‘The Pesah Holiday During the Second Temple Period’, ignore the NT entirely. As we shall see below, other seder investigators use the Johannine evidence to relativize the testimony of the Synoptics about the seder-like features of the Last Supper, but I do not think they have drawn the right conclusions from this discrepancy. Yuval, I. J., ‘Easter and Passover as Early Jewish–Christian Dialogue’, Passover and Easter: Origin and History to Modern Times (ed. Bradshaw and Hoffman) 98-124Google Scholar does take the evidence of the Gospels and other early Christian literature seriously, and criticizes other Jewish scholars for not doing so (103), but his usage of this material does not seem to me to be compelling; see below, Section 5.

35 ‘Given for you… Do this in my remembrance’ is present only in Luke 22.19 among the Synoptics, but it is paralleled in 1 Cor 11.24.

36 On the textual question, see Bovon, F., Luke 3: A Commentary on the Gospel of Luke 19.28–24.53 (Hermeneia; Minneapolis: Fortress, 2012) 154-6Google Scholar, who deems the problem insoluble.

37 Leonhard, Jewish Pesach, 33, introduces a red herring when he impugns the Synoptic testimony by pointing out that it does not align with Exod 12: there is no mention of bitter herbs, a wanderer's dress, or eating hastily, and ‘Jesus and the apostles do not even remotely resemble a “family” or a “house”’. But strict adherence to the pattern prescribed in Exod 12 is not required for a meal to be considered a Passover repast; as we have seen above, both the author of Jubilees (49.23) and the rabbis recognized that their Passover meals differed from the hurried ‘Passover of Egypt’ (Leonhard mentions the Jubilees text twice [pp. 27 and 238] but without engaging this aspect of it). As for Jesus and his disciples not constituting a ‘family’ or a ‘house’, this objection ignores the widespread early Christian image of Jesus and his disciples as an eschatological ‘family’ (see, e.g., Mark 3.31-35 and cf. Marcus, Mark, 1.277). Josephus, moreover, speaks of the Jews eating the Passover sacrifice in ‘fraternities’ (ϕρατρίας; Bell. 6.423; Ant. 2.312; 3.248), a term that seems to transcend family groupings (see LSJ 1953). Although Josephus wrote in the post-70 CE era, the Judaism he described was usually that of the pre-70 period.

38 See, e.g., Bokser, Origins, 29-30; Klawans, ‘Last Supper’, 29.

39 The basic arguments are: (a) There would have been theological pressure to transform the Last Supper into a Passover meal because of the primitive Christian theologoumenon identifying Jesus as the Passover lamb (see 1 Cor 5.7; John 1.29, 36; 19.29, 36). (b) The legal activity that the Gospels attribute to the Jewish and Roman authorities is implausible on the first day of Passover, as is the travel implied by Mark 15.21//Luke 23.26. (c) Mark 14.1-2//Matt 26.1-3 and Mark 15.42 go against Mark's Tendenz by preserving hints that Jesus was crucified on the day before Passover began; see Theissen, G., The Gospels in Context: Social and Political History in the Synoptic Tradition (Minneapolis: Fortress, 1991) 166-7Google Scholar; Marcus, Mark, 2.932, 1041, 1070. The main counter-argument is that John's dating, too, may be theologically driven, since, as has just been seen, he thinks of Jesus as the Passover lamb, and so ex hypothese he has him executed at about the hour that the Passover lambs were slain, i.e. noon on Erev Pesach (see John 19.14; Philo Spec. 2.145; m. Pes. 5.3; cf. Jeremias, Eucharistic Words, 82). I do not find this argument convincing, however, because (a) John does not explicitly say that Jesus died as the lambs were being slaughtered. (b) Awareness of the timing of the custom is unlikely in readers whose knowledge of Judaism is so deficient that they need to be told, for example, that Passover is ‘a feast of the Jews’ (John 6.4). (c) Other evidence seems to favor a midafternoon or later slaughter of the sacrificial animals, so the posited synchronicity does not work (see Exod 12.6; Jub. 49.12; Josephus Bell. 6.423; Philo QE 1.11; m. Pes. 5.1; cf. Brown, R. E., The Death of the Messiah: From Gethsemane to the Grave: A Commentary on the Passion Narratives in the Four Gospels [2 vols.; ABRL; New York: Doubleday, 1994]Google Scholar 1.847, 2.1371-2). Contra Instone-Brewer, Feasts and Sabbaths: Passover and Atonement, 120, who implausibly thinks that the Synoptics, which have Jesus die at 3 pm (Mark 15.34), intend to link his death with the start of the Passover offerings—even though the Synoptics date Jesus' death to the first day of Passover, not its eve.

40 See Marcus, Mark, 1.37-39.

41 Contra Kulp, Schechter Haggadah, 237, who dismisses the Synoptic evidence by saying, ‘While these words are attributed to Jesus, whether or not Jesus actually uttered them is debatable and ultimately unknowable. All we know is that they are attributed to him by the authors of the Gospels, and therefore existed (in written form) by the time the Gospels were written. Scholars generally assume that the three synoptic gospels…were written in the decades following the destruction of the Temple, around the same time that Rabban Gamaliel lived, and before the redaction of the Mishnah.’ This is inaccurate, at least as far as the dating of Mark is concerned, and it glosses over the Synoptics' rootedness in pre-70 traditions.

42 Klawans, ‘Last Supper’, 30-1 points out that bread and wine are the basic elements of any formal Jewish meal, not just the Passover seder. He recognizes, however, that what is distinctive about the seder and about the Synoptic Last Supper is that words of interpretation, not just blessings, are spoken over the bread and wine.

While wine is not specified as part of the Passover meal in the foundational texts in Exod 12–13 and Deut 16, and hence Gamaliel does not in m. Pes. 10.5 identify it as one of the seder elements that needs to be interpreted, it does form part of the Passover meal already in Jub. 49.6. The Mishnah and Tosefta specify that four cups of it must be drunk, and some later rabbinic authorities and the Passover Haggadah offer symbolic interpretations of it; see Jeremias, Eucharistic Words, 58-9; Bokser, Origins, index s.v. ‘wine’.

43 The method of using distinctive characteristics to trace genealogical relationships was of course pioneered by Charles Darwin in On the Origin of Species (orig. 1859) and has subsequently been transferred to many other realms, including textual criticism; see S.C. Carlson, ‘The Text of Galatians and its History’ (PhD diss., Duke University, 2012).

44 Jeremias, Eucharistic Words, 56. For a recognition of the importance of this parallel, even though he ends up disagreeing with Jeremias, see Klawans, ‘Last Supper’, 30-1. Jeremias, Eucharistic Words, 41-62 mentions thirteen other parallels between the Synoptic Last Supper and the Passover seder, the most important of which are reclining at the meal (Mark 14.18 pars.) and singing a hymn (Mark 14.26 pars.), which Jeremias assumes to be one of the Hallel psalms. Both of these details make sense in a Passover context, but neither is the sort of distinctive parallel that the interpretation of the matzah is, since the Gospels do not specify that the hymn was a psalm, and in the Gospel tradition Jesus also reclines at meals that are not Passover seders (see Mark 2.15; 14.3; Luke 7.37).

45 See Stein, ‘Influence’, esp. 18-23, 32-3.

46 ‘This matzah which we eat, what is it for? It is because the dough which our ancestors prepared did not have sufficient time to rise before the King, King of all kings, the Holy One, Blesssed be He, was revealed to them and redeemed them.’ The passage goes on to cite Exod 12.39 as a prooftext. Trans. from Kulp, Schechter Haggadah, 56.

47 For the translation הא as ‘this is’, see Glatzer, N. H., The Passover Haggadah with English Translation, Introduction and Commentary (New York: Schocken, 1979)Google Scholar 21, and Kulp, Schechter Haggadah, 27. Jastrow, M., A Dictionary of the Targumim, the Talmud Babli and Yerushalmi, and the Midrashic Literature (New York: Judaica, 1982Google Scholar [orig. 1886–1903]) 328, translates הא as ‘behold’ and Sokoloff, M., A Dictionary of Jewish Palestinian Aramaic of the Talmudic and Geonic Periods: Second Edition (Dictionaries of Talmud, Midrash and Targum II/Publications of the Comprehensive Aramaic Lexicon Project; Ramat-Gan: Bar Ilan University; Baltimore and London: The Johns Hopkins University, 2002)Google Scholar 158, renders it as ‘behold, here is’. The Hebrew term behind הא in Ha Lachma is הנה, which is sometimes rendered in ancient Greek versions and modern English translations as ‘This is’ or ‘This was’. For example, Jesus' words of institution over the cup in Mark 14.24//Matt 26.28, which repeat the τοῦτό ἐστιν formula from the words over the bread, are an echo of Exod 24.8, which uses הנה, rendered in the Targums on this verse as הא; cf. Heb 9.20, which renders Exod 24.8 as τοῦτο τὸ αἷμα τῆς διαθήκης. ‘This is’, then, is an acceptable English translation for הנה in Exod 24.8, and it is so rendered in njps; cf. Clines, David, ed., The Dictionary of Classical Hebrew (8 vols.; Sheffield: Sheffield Academic, 1993–)Google Scholar 2.574, 577 [1 and 1i] which translates הנה in some circumstances as ‘there is’ and ‘he/she/it is’.

48 As Yuval, ‘Easter and Passover’, 105-6, notes, the correspondence is even closer in the Matthean form of the saying (Matt 26.26), in which Jesus explicitly says λάβɛτɛ ϕάγɛτɛ (‘Take, eat’).

49 Epitomized above all in the statement, בכל דור ודור חיב אדם לראות את עצמו כאלו הוא יצא ממצרים (‘In every generation, each person must regard himself as if he himself had come out of Egypt’, my trans.). This statement is not found in better Mishnah manuscripts but is present in the extant Haggadot, though the prooftexts that follow it vary; see Safrai and Safrai, Haggadah, 210-11; Kulp, Schechter Haggadah, 237-8.

50 This similarity was already highlighted by J. Scaliger in his De Emendatione Temporum in 1583; see Grafton, A., Joseph Scaliger: A Study in the History of Classical Scholarship (2 vols.; Oxford-Warburg Studies; Oxford and New York: Clarendon/Oxford University, 1983–93)Google Scholar 2.317; cf. Yuval, ‘Easter and Passover’, 120 n. 33.

51 See Goldschmidt, D., The Passover Haggadah: Its Sources and History (Jerusalem: Bialik Institute, 1960 [Hebrew]) 7-9Google Scholar. He lists the version of Saadia Gaon, some of the Geniza fragments, and MS. Cambridge 145 as missing the first sentence of Ha Lachma.

52 Private communication, 17 May 2004.

53 Cf. Goldschmidt, Haggadah, 8; Kulp, Schechter Haggadah, 191. On Huna's dates, see Marcus, J., ‘A Jewish–Christian ‘Amidah?’, Early Christianity 3 (2012) 5-6CrossRefGoogle Scholar.

54 Safrai and Safrai, Haggadah, 179, offer no evidence for their specific dating, beyond saying that the Geonim ‘wrote many declarations and prayers in Aramaic’. But Aramaic, of course, was in use before the Geonic period as well, and there is nothing distinctively Geonic about the Aramaic of Ha Lachma. Leonhard, C., ‘Die Pesachhaggada als Spiegel religiöser Konflikte’, Kontinuität und Unterbrechung: Gottesdienst und Gebet in Judentum und Christentum (ed. Gerhards, Albert and Wahle, Stephan; Studien zu Judentum und Christentum; Paderborn: Schöningh, 2005) 162-3Google Scholar, thinks that Ha Lachma is a response to the Christian eucharistic theology of the late Middle Ages, specifically, perhaps, to the practice of elevating the host. On this interpretation of Ha Lachma as anti-Christian polemic, which develops further the approach of Yuval, see the next section.

55 Cf. Goldschmidt, Haggadah, 7, who, although dating Ha Lachma late, says that the origin of its components is ‘in the customs of the people rather than in the dicta of sages’. Similarly, Lawrence A. Hoffman, ‘A Symbol of Salvation in the Passover Seder’, Passover and Easter: The Symbolic Structuring of Sacred Seasons (ed. Bradshaw and Hoffman) 120 observes the absence of Ha Lachma from tannaitic and amoraic sources but nevertheless inclines towards a first-century composition for it, noting, ‘Students of ritual have long been aware that the actual extent of available custom is not represented in the Mishnah, say, or even the Tosefta or the Yerushalmi. These books represent only an isolated segment of contemporary usage.’ Cf. Kulp, Schechter Haggadah, 25, who mentions a fact that in my view offers a striking proof that folk customs can remain literarily invisible for a long time. A Barcelona Haggadah from around 1350 (BL Ms. Add 14761, fol. 28b) contains an illustration depicting the Sephardic custom of lifting the seder plate—four hundred years before this practice is first mentioned in writing!

56 On the Passover symbolism in John 6, see Borgen, P., Bread from Heaven: An Exegetical Study of the Concept of Manna in the Gospel of John and the Writings of Philo (NovTSup; Leiden: Brill, 1965)Google Scholar. I am grateful to Dale Allison for pointing out the parallel between Ha Lachma and John 6.35.

57 Contra Kulp, ‘Origins’, 121, and Kulp, Schechter Haggadah, 12, 191, 237, who ascribes the similarities to independent development, since both early Christians and post-70 CE Jews were influenced by Greco-Roman symposium customs.

58 If some form of Ha Lachma did exist in the pre-70 era, it was probably one that lacked the third sentence, since all three pre-70 parallels (Philo's de Cong. 167, the ‘bread word’ in the Synoptics, and John 6.35) link up with the first two sentences of Ha Lachma but not with the third, which on other grounds also seems to have a different origin. Unlike the first two sentences, which are in Aramaic, sentence #3 is a mixture of Hebrew and Aramaic. Also, unlike the first two sentences, #3 presupposes a Diaspora setting.

59 This opinion is apparently shared by Hoffman, ‘Symbol’, who speaks of ‘the use of bread as a symbol in both the Lord's Supper and in the early seder’ as ‘two sides of the same coin’ (123) and of Ha Lachma as ‘an obvious Jewish parallel to the institution of the Lord's Supper’ (124). This language might suggest concurrent independent developments, but since Hoffman dates Ha Lachma to just after the destruction of the Temple (122) and Jesus' ‘bread word’ to the pre-destruction era (115), the implication would seem to be that Ha Lachma is modeled on the Last Supper saying.

60 This assertion, however, is controversial; see the discussion in Bokser, Origins, 89-94, 101-6.

61 Yuval, ‘Easter and Passover’, 117-18.

62 Klawans, ‘Last Supper’, 32, says that ‘virtually all scholars working today’ believe that the Gamaliel of m. Pes. 10.5 is Gamaliel II. The argument in favor of this identification draws on t. Pes. 10.12, which shows Gamaliel II occupied with other sages in exposition of Passover laws. Moreover, as Klawans points out in a private communication, almost all the other sages mentioned in m. Pesaḥim 10 are post-70 figures. The chief difficulty with this theory is the necessity of interpreting פסח = ‘passover offering’ in m. Pes. 10.5 as ‘a nonsacrificial animal roasted to resemble the passover sacrifice’, but this sort of transformation of the term did occur in the post-70 period; cf. Alon, G., The Jews in their Land in the Talmudic Age (70–640 C.E.) (Cambridge, MA and London: Harvard University, 1989) 261-5Google Scholar; Bokser, Origins, 42, 89-90, 101-6. In favor of Gamaliel I is Casey, M., Aramaic Sources of Mark's Gospel (SNTSMS 102; Cambridge/New York/Melbourne: Cambridge University, 1998)Google Scholar 238, who, however, relies on assertion (‘This must surely be Gamaliel I…’). If this were right, Yuval's case would be impossible, since Gamaliel's formulation could not be construed as a response to the changed post-70 situation. Gamaliel I, moreover, unlike Gamaliel II, is portrayed as sympathetic to Christians in Acts 5.34-39.

63 But this passage from Aphrahat does not identify the bitter herbs with Christ's suffering, as Yuval asserts, nor is that interpretation given by Rouwhorst, G. A. M., Les hymnes pascales d'Ephrem de Nisibe: analyse théologique et recherche sur l'évolution de la fête pascale chrétienne à Nisibe et à Edesse et dans quelques Églises voisines au quatrième siècle (2 vols.; Supplements to Vigiliae Christianae 7; Leiden and New York: Brill, 1989)CrossRefGoogle Scholar 1.146, whom Yuval cites for support on this point. Ephrem (d. 373), however, does link the bitter herbs with Jesus' passion (Virg. 8.9; cf. I. Lizorkin, ‘Aphrahat's Demonstrations: A Conversation with the Jews of Mesopotamia’ [PhD diss., Stellenbosch University, 2009] 164).

64 Versions of this theory were already argued by Sulzbach, A., ‘Die drei Worte des Seder-Abends’, Jeschurun 4 (1917) 216-19Google Scholar; Fisher, S. Y., ‘Sheloshah Devarim’, Hatsofeh Lechokhmat Yisrael 10 (1926) 238-40Google Scholar; and Goldschmidt, Haggadah, 52; cf. Yuval, ‘Easter and Passover’, 121 n. 38.

65 There is some ambiguity in Yuval's phrasing here. When he says that Gamaliel was ‘demanding a declaration of loyalty to the Jewish understandings’, that seems to imply that such ‘Jewish understandings’ already existed. But how does that cohere with the previous paragraph, which describes Gamaliel's dictum as ‘a reversed parallel’ of the Christian interpretation?

66 Yuval, ‘Easter and Passover’, 106-7.

67 See the criticism of Yuval's ‘polemicamania’ in Kulp, ‘Origins’, 124. To give my own examples: Yuval, ‘Easter and Passover’, 102, claims that the Bnei Berak account, in which Elazar ben Azaryah speaks of finding a scriptural justification for ‘mentioning the exodus from Egypt at night’, is designed to ‘establish the new edict’ to recite the Passover story during the holiday. But it is easier to understand this passage as providing a prooftext for a customary practice than as establishing a new one. Contrary to Yuval's assertion, there is no evidence of anti-Christian polemic in the discussion of ‘all the days of your life’ later in the Bnei Berak account or in the depiction of the wicked son in the ‘four sons’ midrash. Nor is it obvious that ‘the end of Dayyenu, “[God] built us a Temple to atone for all our sins”, is a kind of afterthought, indicating that Temple offerings atone, in contrast to the Christian claim that atonement comes through the crucifixion of Jesus’ (105); on most of these criticisms, cf. Leonhard, ‘Pesachhaggada’, 146-8, 155-60, 165-6.

On the other hand, it is possible that the ‘I and not an angel’ midrash and the absence of Moses' name in the Haggadah are directed at Christian conceptions of Jesus as an angelic figure and a second Moses (Yuval, ‘Easter and Passover’, 109-10 and 122 n. 49; cf. Leonhard, ‘Pesachhaggada’, 165), though it is also possible, as recently suggested by Henshke, D., ‘“The Lord Brought Us Forth from Egypt”: On the Absence of Moses in the Passover Haggadah’, AJSR 31 (2007) 61-73CrossRefGoogle Scholar, that these circumstances are a response to Jewish revolutionary activism associated with the exodus typology. But even if Yuval is right about there being anti-Christian polemic at some points within the Haggadah, that does not necessarily mean that the Haggadah as a whole is a polemic against Christianity; cf. Leonhard, ‘Pesachhaggada’, 166-7.

68 See, e.g., Polybius Hist. 26.1.4; Plutarch Is. Os. 357 F6; Alcibiades 32.2; De tuenda sanitate praecepta 128D9; Athenaeus Deipnosophistae 5.8 (180A); Diogenes Laertius Vitae 7.13. For this interpretation, see Lieberman, S., Hayerushalmi Kiphshuto: A Commentary (Jerusalem: Darom, 1934)Google Scholar 521; Bokser, Origins, 65-6; Tabory, Passover Ritual, 65-6; Safrai and Safrai, Haggadah, 45; Hoffman, ‘Symbol’, 112-13; Kulp, ‘Origins’, 122.

69 Spec. 2.148: οὐχ ὡς ɛἰς τὰ ἄλλα συμπόσια χαριούμɛνοι γαστρὶ δι' οἴνου καὶ ἐδɛσμάτων.

70 μόνον γὰρ οὐκ ἔξɛστιν δαίνυσθαι, Bell. 6.423; cf. Bokser, Origins, 65-6. Also skeptical about Yuval's discernment of anti-Christian polemic in the afikoman is Leonhard, ‘Pesachhaggada’, 147-8.

71 Cf. Leonhard, ‘Pesachhaggada’, 164.

72 Cf. Barclay, J. M. G., ‘Mirror-Reading a Polemical Letter: Galatians as a Test Case’, JSNT 31 (1987) 73-93Google Scholar. Similarly, Leonhard, ‘Pesachhaggada’, 168, suggests that we should find specific signs of polemic, such as an antithetical formulation (‘not X, but Y’, as in the ‘not by the hand of an angel’ midrash), before attributing a polemical intention to a text.

73 I owe this point to the anonymous NTS reviewer, to whom I am also indebted for several other valuable corrections and suggestions.

74 On this interpretation of לתקן in b. Ber. 28b-29a, see Marcus, J., ‘Birkat Ha-Minim Revisited’, NTS 55 (2009)CrossRefGoogle Scholar 540. This passage is also important for Yuval's case, since it shows Gamaliel commissioning a liturgical edict against the Christians (and on this interpretation of the target of Birkat Ha-Minim I follow the previously cited article in agreeing with Yuval against recent revisionists). But that does not mean that all of Gamaliel's liturgical reforms had the same purpose. Birkat Ha-Minim is obviously polemical, whereas m. Pes. 10.5 lacks a polemical tone.

75 Whether it would have functioned as authoritative outside of the rabbinic circles to which Gamaliel belonged, however, is a good question that has been raised by recent scholarship on the role of the rabbis in the early post-destruction era; see above, n. 25.

76 Cf. the distinction between ‘folk religion’ and ‘book religion’ and between ‘official’ and ‘popular’ religion in Stone, M. E., Ancient Judaism: New Visions and Views (Grand Rapids: Eerdmans, 2011)Google Scholar 3, 9.

77 Recent examples include feminist seders and ‘peace’ seders. There is even a Haggadah for Jews and Buddhists (ed. Pearce-Glassheim, E.; Mill Valley, CA: Modern Haggadah Distribution, 2006)Google Scholar! For the traditum/traditio distinction, see Fishbane, M., Biblical Interpretation in Ancient Israel (Oxford: Clarendon, 1985) 6-19Google Scholar.