The Legal Idiosyncrasy of Rabbi Yehoshua's View

The debate between Rabban Gamliel and Rabbi Yehoshua regarding the obligation to recite the Amidah in the evening, which serves as the tinder that sparked the revolt against the former, does not bear any of the normal markers of tannaitic sources in the Bavli (other than the names of the disputants). That is, the debate is not preceded with an introductory term such as תנו רבנן (our rabbis taught) or תניא (it is taught). Instead, the debate appears in a memra of the first-generation Babylonian sage Shmuel. According to Shmuel, Rabban Gamliel ruled that one is required to recite this evening prayer, while Rabbi Yehoshua claimed that it was optional. Indeed, tannaitic and early amoraic teachings render the tannaitic origins of the debate—or to be more precise, the lenient view of Rabbi Yehoshua—suspicious.

To begin with, nothing in tannaitic corpora suggests such a view. M. Berakhot 4:1, presenting the end times for the various daily prayers, states that “the prayer of the evening—it has no fixed time” (אין לה קבע).Footnote ² Although the Bavli will interpret this admittedly unusual phrase to mean that the author of the statement maintains the position that Ma‘ariv is optional, in its context the phrase more likely refers to its relatively loose time requirements.Footnote ³ The Tosefta gives the same impression, using the identical language of “it has no fixed time,” alongside competing views arguing for a more specific time for the evening Amidah (עם נעילת שערים). To my knowledge, tannaitic midrash also contains no suggestion that this evening prayer is optional.

To the contrary, a number of tannaitic and early amoraic sources imply the precise opposite, namely, the obligatory nature of the evening prayer. For example, T. Berakhot 3:6, in explaining the origins of thrice-daily prayer, relates that Psalms 55:18 (“Evening, morning, and noon, I complain and moan, and [God] hears my voice”) teaches that one may not simply recite the Amidah three times in a row at some point during the day; rather, these prayers must be spaced throughout the day. No hint is offered here that the evening recitation is in some way less significant than the other two: “Evening—this is the Amidah of the evening; morning—this is the Amidah of the morning; noontime—this is the Amidah of the afternoon.”

A similar sense arises from T. Berakhot 3:7, which discusses the “brief prayer” (תפילה קצרה) to be recited in times of danger. In this context, Rabbi Elazar b. Ẓadok relates that his father would make use of a modified shortened version of the Amidah on Friday nights. Louis Ginzberg claimed that this demonstrated an early tradition that did not include the evening prayer, reading Ẓadok's action as a pious stringency: although nothing was required to be said at night, this sage voluntarily took on a prayer practice specifically for Friday evenings.Footnote ⁴ A more straightforward reading of this testimony, however, is that though normally Rabbi Elazar b. Ẓadok's father recited the evening prayer in its entirety, on Friday nights specifically he would use an abbreviated version. Context suggests such a reading: Rabbi Elazar b. Ẓadok's statement appears in a discussion of shortened prayers to be recited in times of danger and what the precise content of such shortened prayers ought to be.

Most striking, as already noted in several cases by medieval commentators, are the many places where a failure to recite the evening Amidah properly means that one must repeat it, which indicates its obligatory status. We find such a case already in the Tosefta: “If one did not mention ‘the powers of the rains’ in [the blessing] ‘revivification of the dead,’ or if one did not request the rains in the blessing of the years, this person is returned [i.e., must repeat the Amidah]; if one did not say the Separation in [the blessing] ‘who graciously gives knowledge,’ one says it over the cup. If one did not say it over the cup, this person is returned” (T. Berakhot 3:9). The “Separation” referred to here is the insertion for the Amidah at the end of Shabbat, that is, an evening Amidah; yet the reason that one need not repeat in this case has nothing to do with the optional status of Ma‘ariv, but rather the additional opportunity that one will have because of the ritual of reciting such a blessing over a glass of wine. Where one did not engage in the latter, one must indeed repeat the evening Amidah.

Even more examples of this phenomenon appear in the Talmuds, both in otherwise unattested baraitot and in early amoraic statements. Most generally, at B. Berakhot 26a, the early Palestinian Amora Rabbi YoḥananFootnote ⁵ states that “if one erred and did not pray in the evening, such a person should pray twice in the morning,” clearly assuming that the failure to recite the evening Amidah is an omission of an obligation that must be remedied.Footnote ⁶ Later on the same page, we find a baraita interpreting Ecclesiastes 1:15: “‘That which is crooked cannot be made straight.’ This refers to one who missed the evening recitation of the Shema, or the morning recitation of the Shema, or the Amidah of the evening or the Amidah of the morning.”Footnote ⁷ And in the following sugya (B. Berakhot 26b), a baraita states that one who fails to recite the afternoon Amidah on Friday or Saturday afternoon should say it twice that evening.Footnote ⁸ While these sources disagree regarding the ability to make up in part for missed prayer (the baraita on 26a seemingly rejecting the possibility, and both the baraita on 26b and Rabbi Yoḥanan on the previous page allowing for it), they are unanimous in their treatment of Ma‘ariv as no different from the other two daily prayers.

Similarly, two traditions about the addition to the Amidah for Rosh Ḥodesh clearly imply the obligatory nature of the evening prayer. At B. Berakhot 30b, a baraita states that one who forgets this insertion need not repeat the Amidah, because this person will be able to say it in the morning prayer.Footnote ⁹ A statement attributed to Rav shortly thereafter comes to the same conclusion, but with different reasoning: because the rabbinic court would sanctify the new moon only during the day, an omitted Rosh Ḥodesh insertion in the evening does not demand a repeated performance of the prayer. As with the baraita, however, the obligation in Ma‘ariv is taken for granted.Footnote ¹⁰

Rav in particular seems to bear no doubts about the status of the evening Amidah. Both Talmuds report a view, attributed to Rav, that reciting the Ne‘ilah prayer exempts one from needing to recite Ma‘ariv (Y. Berakhot 4:1 [7c]; B. Yoma 87b). A number of authorities in the Yerushalmi disagree with this view in ways that highlight the seriousness with which they took the obligation in the evening prayer. For example, Rabbi Yonah challenges Rabbi Abba, one of the tradents for Rav's view here, asking how a prayer composed of a mere seven blessings (i.e., Ne‘ilah) can fulfill the obligation of the eighteen blessings of Ma‘ariv. Several figures (Rabbi Abba b. Memel, Rabbi Yehoshua b. Levi, Rabbi Yose b. Bun) relate a view that Ne‘ilah in fact does not fulfill one's obligation to recite the evening prayer. But even the lenient view of Rav shares the assumption that the evening prayer is an obligation that must be dispensed, whether or not the infrequently recited Ne‘ilah counts to fulfill that obligation.

Tannaitic texts, as well as positions attributed to both tannaitic and early amoraic authorities in both Talmuds, thus speak unequivocally of an obligation to recite the evening prayer. One leniency regarding Ma‘ariv, found in the Yerushalmi, is striking, however. Those from the “house” of Rabbi Yannai state that once a person gets into bed (עלה אדם על מיטתו) having not yet recited the evening Amidah, he need not be harassed into reciting it (אין מטריחין אותו לירד) (Y. Berakhot 4:1 [7c]).Footnote ¹¹ In one sense, this ruling distinguishes the evening prayer from the other two recitations of the Amidah; we do not find any leniency for those prayers that where it is inconvenient for someone to get out of bed such a person is exempt. At the same time, this leniency regarding Ma‘ariv—the one daily prayer when one might reasonably already be in bed—is functionally equivalent to the statement, found in M. Shabbat 1:2, that one who has already sat down before a barber “close to Minḥah” need not interrupt the grooming in order to recite the afternoon prayer. Yet no one would suggest that this leniency implies that the afternoon prayer is optional!Footnote ¹² The specific exemption for one who is already in bed highlights the challenges peculiar to mandatory evening prayer; such attention to its particularities does not suggest that this prayer is not optional. Indeed, were it merely a personal practice, even one that were encouraged, there would be no need to release someone already in bed from needing to get up to pray!

Rulings attributed to tannaitic and early amoraic times thus reveal no general tendency to treat Ma‘ariv as fundamentally different from the other two daily prayers. Some leniencies vis-à-vis Ma‘ariv may exist (e.g., it may be fulfilled through the recitation of Ne‘ilah, one need not get out of bed to recite it), but that it is obligatory goes unquestioned in these sources. But before moving on to the Babylonian sugya in which Rabbi Yehoshua's leniency with regard to the evening service appears, it is worth considering one baraita in the Bavli that some commentators have argued suggests a less rigid approach to Ma‘ariv. At B. Berakhot 26b, in the midst of the well-known sugya asking whether thrice-daily prayer has its roots in the daily offerings in the temple or rather the practices of the patriarchs, a baraita cited in support of the former view connects the timing of each prayer to its parallel in the temple. Thus, the morning Amidah is timed to be in accord with the rules regulating the morning daily offering, while the afternoon Amidah takes as the source of its rules those determining the time of the afternoon offering. Regarding the evening Amidah, the baraita states: “And why did they say that the evening Amidah has no fixed time [אין לה קבע]? Because the limbs and fats that were not consumed by evening were offered throughout the night.” The comparison of the evening prayer to the remainders of the day's offering, combined with the Bavli's interpretation of the phrase אין לה קבע to mean “optional” rather than “it has no fixed time,” led some traditional commentators to understand this as explaining the optional nature of Ma‘ariv. That is, since there might sometimes be such leftovers, while on other days in the temple there may not have been, the evening prayer took on a less obligatory status.Footnote ¹³

The baraita, however, implies no such thing. First of all, the baraita finds a parallel at T. Berakhot 3:1. The Tosefta, however, includes only the sections dealing with the morning and the afternoon prayers; the toseftan authors/editors do not connect Ma‘ariv to cultic practice, thus suggesting that this part of the Bavli's baraita may be a later emendation to an earlier version of the baraita.Footnote ¹⁴

Indeed, a parallel to the baraita appears as well in the Yerushalmi (4:1 [7b]), where the text states that the sages “did not find something to hang [Ma‘ariv] on, and they taught it plain [setam],” connecting this explicitly to the language of אין לה קבע in the Mishnah. The connection to the leftover limbs and fats is then made by the Amora Rabbi Tanḥuma. The Palestinian version of this tradition thus makes it likely that the connection of the evening prayer to an aspect of the cult was a post-tannaitic development. But in its Palestinian context, the initially assumed lack of connection to an aspect of the sacrificial system (“they did not find something to hang it on”) actually explains the meaning of the troubling phrase אין לה קבע—Ma‘ariv has nothing on which to be hung, that is, nothing to which it can be attached/fixed. This interpretation has nothing to do with the obligatory or optional nature of the prayer.

Even the formulation that we find in the Bavli does not make any connection to questions of obligation. Rather, the point of the comparison to leftover parts from the sacrificial animals is that they “were offered throughout the night,” which would answer well the question implied by the contextual meaning of the phrase אין לה קבע: Why may one recite the evening prayer all night long? Because those parts burned “throughout the night.” No mention is made of the inconsistent nature of these leftover parts, offered some nights but not on others, such that this would be responding to a question about one's fundamental obligation. Rather, the editors of this baraita as it appears in the Bavli—whoever they were—still understood the phrase אין לה קבע in accordance with its tannaitic meaning, namely, “it has no fixed time,” and did not view it as implying anything about the evening prayer's being optional.Footnote ¹⁵

A survey of sources relating to the evening Amidah in Mishnah, Tosefta, and both the Palestinian and Babylonian Talmuds reveals the idiosyncrasy of the view attributed to Rabbi Yehoshua, namely, that Ma‘ariv is optional. When early sources address the evening prayer, they take for granted its mandatory nature and, for the most part, treat it as thoroughly equivalent to Shaḥarit and Minḥah. To be sure, inconsistencies in the way the evening prayer is treated by comparison to the morning and afternoon prayers appear in amoraic statements in both the Bavli and the Yerushalmi, but in none of them do we find an implication that Ma‘ariv is therefore optional. Importantly, all of these rulings are attributed to tannaitic sources, Rabbi Yoḥanan, or Rav. Although Rav is of course an important Babylonian Amora, rabbinic sources portray him as having roots in rabbinic Palestine, a student of Rabbi. It is striking, for example, that Shmuel, a more unambiguously Babylonian personality, does not figure in any of these sources. But it is to his tradition that we now turn.

The Debate of Rabban Gamliel and Rabbi Yehoshua

As noted above, in the Bavli, the debate regarding Ma‘ariv, though ascribed to tannaitic sages, bears none of the normal markers of tannaitic texts.Footnote ¹⁶ Rather, the debate appears in the mouth of Shmuel:

The Bavli posits the first-generation Babylonian Amora Shmuel as the source for the tannaitic debate. However, the sugya is then followed by the famous story of the deposition of Rabban Gamliel, which includes the debate of Rabbi Yehoshua and Rabban Gamliel and is indeed introduced by a tannaitic marker:

It thus seems that though the only legal context in which the debate appears is a memra, we indeed have a claimed tannaitic attestation in the Bavli, that is, the story based on the debate. What is more, the baraita telling this story finds a close parallel in the Yerushalmi (Y. Berakhot 4:1 [7c–d] = Y. Ta‘anit 4:1 [67d]), including the debate with the same tannaitic figures. In the Yerushalmi, the story does not bear a specific marker of tannaitic origins, but that may be simply because it appears there as a continuation of what came before, namely, a presentation of the debate independent of the story.Footnote ²³ There, the debate is indeed marked as tannaitic: “Rabbi Yaakov b. Aḥa said: It was taught there [תניי תמן]: The evening prayer, what is its status? Rabban Gamliel says ‘obligatory,’ but Rabbi Yehoshua says ‘optional.’”

It is worth summarizing the somewhat confusing presentation of the tannaitic debate about Ma‘ariv in the two Talmuds. In each Talmud we find two appearances, close to each other, of the debate. The first appearance is a legal tradition, presented as an independent statement, attesting to tannaitic debate about the status of Ma‘ariv. The second appears as a recapitulation of this tradition, but now found as a narrative component of a larger story. In the Bavli, the independent statement of the debate is attributed to the Amora Shmuel, but a baraita tells the story of the debate and its consequences. The Yerushalmi, by contrast, provides no indications as to the origins of the story, but does mark the debate when presented independently as a tannaitic teaching; it is likely that the lack of attribution for the story results simply from presenting it as a continuation of the debate. Thus, despite the lack of attestation of a lenient view regarding the evening Amidah in tannaitic corpora, as well as the many tannaitic and early amoraic views that contradict the existence of such a view, we do find a claim, in both Talmuds, that this tradition is indeed tannaitic.

Leaving the question of the dating of this tradition for the moment, it is important first to note the somewhat unusual attribution that the Yerushalmi uses for ascribing tannaitic authority to the debate, namely, “it was taught there” (תניי תמן). In his dictionary of technical terms in the Yerushalmi, Leib Moscovitz describes this phrase thus: “It seems [נראה] that ‘it was taught there’ presents Babylonian baraitot that were transmitted by Amoraim.”Footnote ²⁴ If correct, then the Yerushalmi, despite preserving the debate as a tannaitic tradition, affirms the sense we get from the Bavli's placement of the debate in the mouth of Shmuel, namely, that the notion of a disagreement regarding the obligation to recite the evening prayer has its origins in rabbinic Babylonia.

Moscovitz is careful to express some doubt about his description of this technical term. He notes: “I say ‘it seems’ because there are no clear proofs from the sugyot that one should explain the term in this way.”Footnote ²⁵ This particular example, however, supports Moscovitz's tentative conclusion; as I argued above, sources found in Palestinian corpora, as well as those found in the Bavli attributed to Palestinian sources, all lend the impression that an evening recitation of the Amidah is required in the same way as the morning and afternoon performances. Only this passage, which in the Yerushalmi is attributed to a tannaitic teaching taught “there,” that is, in Babylonia, suggests that Ma‘ariv is optional. Furthermore, Rabbi Yaakov b. Aḥa, the third-generation Palestinian sage who transmits the teaching in the Yerushalmi, describes himself elsewhere in the Yerushalmi as having arrived from Babylonia.Footnote ²⁶ A full consideration of all the appearances of the term “it was taught there” in the Yerushalmi is a desideratum and may well weigh in on the question of early Babylonian traditions.Footnote ²⁷ In this case, it nonetheless seems likely that the attribution to Rabbi Yehoshua of a view that the evening prayer is optional was recognized in amoraic times as a uniquely Babylonian tradition.Footnote ²⁸

A Babylonian Tannaitic Tradition?

The notion that Rabbi Yehoshua believes the evening prayer to be optional is distinctively tied up with Babylonian tradition. But at what point did this Babylonian tradition develop? Is this an example of an early, perhaps even tannaitic Babylonian teaching? In a number of recent studies, Barak Cohen has argued that many traditions that appear to be examples of this phenomenon are in fact simply reworkings of earlier Palestinian teachings.Footnote ²⁹ Although the work of the previous sections provides a counterexample of a tradition, claimed to be tannaitic and attributed to Shmuel, that is indeed most likely Babylonian in its origins, no compelling evidence exists to suggest that this Babylonian tradition predates Shmuel. Rather, the idea that the evening prayer is optional, contradicted by so much early material in both Talmuds, achieves some sort of wider significance only in later strata of the Bavli. This uneven distribution of the idea makes it unlikely to be truly tannaitic, and thus supports Cohen's more general argument against seeing in passages such as this one examples of early Babylonian traditions.

Before considering the other contexts in which the view appears I must first return to the locus classicus itself, the Bavli passage treated above. Recall that following Shmuel's presentation of the tannaitic debate between Rabbi Yehoshua and Rabban Gamliel, he himself rules in accordance with the stringent view (“And the law follows the one who says obligatory”),Footnote ³⁰ while Rav supports the lenient view. Such is the reading of all of the manuscripts as well as the Soncino printing. However, the Vilna printing reads as follows: “For Rav Yehudah said that Shmuel said: The evening prayer—Rabban Gamliel says: obligatory, Rabbi Yehoshua says: optional. Abbaye said: The law follows the one who says obligatory. And Rava said: The law follows the one who says optional.” Shmuel's memra is identical but for its ending, at which point his ruling in accordance with Rabban Gamliel (found in the manuscripts) is absent in the Vilna printing, as is the competing view attributed to Rav. In their stead, we find the fourth-generation Amoraim Abbaye and Rava, the former ruling stringently (like Shmuel in the manuscripts) and the latter (like Rav) ruling leniently.

One tempting interpretation of this phenomenon would be to view the reading of the Vilna printing as a mistake. Given the unanimity of the manuscripts, one would already be predisposed towards skepticism regarding the inclusion of Abbaye and Rava in the Vilna printing. Furthermore, though the difference between the two appears large, one can see how an overthinking scribe might have misread or miswritten “Rav” as “Rava,” at which point Abbaye would become the obvious disputant.Footnote ³¹ However, the reading of the Vilna printing may find some support in medieval commentators. Although most commentators clearly have a version of the passage in line with that found in the manuscripts, the Rif (Isaac ben Jacob Alfasi ha-Cohen, ca. 1013–1103) reads here a debate between Rava and Abbaye.Footnote ³² There may thus be an early version of the passage that indeed attributes the debate to these later amoraic sages. I will return to this point below.

In any event, the view appears in other contexts only in anonymous portions of the Bavli. But before turning to those, the Yerushalmi's presentation of the debate must be examined more carefully. Recall that the Babylonian baraita appears, in the Yerushalmi, in the mouth of Rabbi Yaakov b. Aḥa, a third-generation sage. The only other Amora in the Yerushalmi who relates to this debate is Rabbi Ḥanina,Footnote ³³ whose statement immediately follows Rabbi Yaakov b. Aḥa's Babylonian teaching. Rabbi Ḥanina connects the debate about Ma‘ariv's obligatory status to the debate, referenced above, about the relationship between the evening prayer and the recitation of Ne‘ilah. Rabbi Ḥanina asserts that those who exempt a person from reciting the evening prayer on a day when they have recited Ne‘ilah (i.e., Rav, though Rabbi Ḥanina does not identify this view explicitly with his name) do so because they fundamentally view Ma‘ariv as optional, and those who do not exempt such a person from Ma‘ariv are in accord with the position that it is obligatory.Footnote ³⁴ Taken at face value, Rabbi Ḥanina's conflation of these two debates renders the view that the evening prayer is optional absurd, since one need not wonder whether some other prayer will fulfill an obligation when this activity is not even obligatory. Perhaps Rabbi Ḥanina understands the view of “optional” in a less extreme way, that is, “optional” in this context simply means “less obligatory,” a significant reining in of the lenient view.Footnote ³⁵ If so, given that this is the only engagement with the position in the Yerushalmi, it may even suggest a broader Palestinian resistance to the straightforward meaning of the lenient position, though the dearth of Palestinian materials engaging the debate makes such speculation impossible to prove or disprove.

More important here is the identity of this Rabbi Ḥanina. Louis Ginzberg points out that this figure cannot be the “usual” Rabbi Ḥanina b. Ḥama, since that figure was roughly contemporaneous with Rav, on whose opinion he is commenting.Footnote ³⁶ Rather, he must be a later figure, though precisely how late cannot be determined.

The ambiguous dating of Rabbi Ḥanina's engagement with the claimed tannaitic debate is doubly significant. First, it is distinctly possible that the attribution of leniency to Rav in the Bavli is itself a reformulation of Rabbi Ḥanina's point. That is, Babylonian editors, aware of a tradition that linked the lenient view regarding Ma‘ariv's obligatory status to the view that claimed Ne‘ilah fulfilled one's obligation in Ma‘ariv could have gone the next step, formulating a memra that attributes ruling like Rabbi Yehoshua to Rav.Footnote ³⁷ A legal narrative of this sort goes a long way towards explaining the contradiction between Rav's ruling like Rabbi Yehoshua at B. Berakhot 27b and his statement that Ne‘ilah fulfills one's obligation in Ma‘ariv at B. Yoma 87b (see further on this below), since the very attribution to him of such leniency was itself the result of Babylonian reworking of Rabbi Ḥanina's more modest claim.

The difficulty in dating Rabbi Ḥanina's use of the Babylonian “baraita” is important as well because it is the only example in either Talmud of an Amora making use of this allegedly tannaitic debate.Footnote ³⁸ The debate appears in three locations in the Bavli, and two of them are clearly produced by the anonymous voice of the text. I begin with the most difficult example, since it is the only one in which it appears that an Amora references the debate; I will argue that this is likely a later interpolation, however:

Ze‘iri (not Abbaye, following the vast majority of the textual witnesses) here refers to the debate, stating that according to the view that Ma‘ariv is optional, once one has removed one's belt, this person no longer need be bothered to recite the evening prayer. An anonymous voice responds, picking up on the implication that the other view (that is, Rabban Gamliel's) would indeed hold that someone who had loosened a belt would have to return to a more dignified state and pray before eating dinner. Thus, we appear to have here a case of a later Amora making use of and building on the debate, attributed to R. Yehoshua and Rabban Gamliel, that appeared in the mouth of Shmuel.

I think it likely, however, that the reference to the evening prayer being optional is a later addition to Ze‘iri's statement. First of all, note that the statement is not necessary for Ze‘iri to make a meaningful claim. To begin with, the phrase is unnecessary; were we to remove the words “according to the view that says the evening prayer is optional,” Ze‘iri would still be saying something noteworthy, namely, that Babylonians, because of their clothing customs involving belts, do not need to worry about the evening prayer once they have removed their distinctive belt. Of course, the fact that we can remove this phrase does not mean that we should, but such extraneous phrases are often indicators of later editorial work, and in this case, it means that such editorial intrusion is at least plausible, if not yet probable.

More important here is the relative incoherence that the phrase produces. As the anonymous voice here objects, to suggest that the leniency raised by Ze‘iri applies only if one follows the view that Ma‘ariv is optional makes little sense, since the Mishnah has stated that one need not interrupt a meal in order to recite the afternoon prayer, and Rabbi Ḥanina has commented that the undoing of one's belt indicates the beginning of a meal. If undoing one's belt essentially exempts a person from reciting the afternoon prayer, then why should the situation be any different with regard to the evening prayer, irrespective of its status as obligatory or not? The sugya tries to resolve this in customary Bavli fashion, but a simpler explanation exists: Ze‘iri was indeed making a general statement about the evening prayer and the sartorial loosening that one might engage in at the end of the day at home.

Such an explanation makes Ze‘iri's original statement and the memra attributed to his teacher Rabbi Ḥanina legally identical (the releasing of one's belt exempts a person from needing to recite a prayer),Footnote ⁴⁶ though a number of important differences remain: Ze‘iri makes what appears to be an independent statement, whereas Rabbi Ḥanina comments on the mishnah; Ze‘iri is discussing the evening prayer, whereas the mishnah on which Rabbi Ḥanina comments relates to the afternoon prayer. Perhaps Ze‘iri's statement is simply an example of a student expanding on and extending his teacher's idea; what Rabbi Ḥanina said regarding Minḥah, his student Ze‘iri applies to Ma‘ariv. Alternatively (and I think more likely), the very attribution of such an interpretation of the mishnah to Rabbi Ḥanina could have resulted from Ze‘iri's statement. The latter is an Aramaic statement that references a distinctively Babylonian phenomenon (the wearing of a belt). We can see how a later editor might have sought to make the statement more general and authoritative by transforming it into a pithy, Hebrew memra that functions to explain a mishnah. The teacher-student relationship and the greater prominence of Rabbi Ḥanina would explain the attribution of the newly formed memra to the teacher rather than the student. Either way, reading Ze‘iri as making a general claim about Ma‘ariv, rather than one dependent on Rabbi Yehoshua's view that the evening prayer is optional, removes the problems that the anonymous sugya raises. The anonymous Babylonian editors, in the interest of creating such a sugya, could have interpolated into Ze‘iri's statement the phrase connecting it to the debate about the evening prayer's status, thus creating the opening for their discussion.Footnote ⁴⁷

That the phrase referencing the debate about the status of Ma‘ariv is a later addition is further buttressed by the two other sugyot that make use of that debate. In both, the phrase occurs only in the anonymous stratum of the Talmud. At B. Berakhot 4b, the debate appears in the anonymous layer of a sugya commenting on a baraita:

Here, the anonymous voice makes use of the debate about the obligatory status of the evening prayer to explain an unusual locution in the baraita. The extreme formulation that someone who transgresses any rabbinic detail merits the death penalty, the anonymous voice here argues, serves to emphasize that Ma‘ariv is indeed mandatory, despite the existence of a lenient view. Without making broad assumptions about the dating of the anonymous layer of the Bavli, we cannot say with any confidence when the anonymous editor made use of the debate between Rabbi Yehoshua and Rabban Gamliel. Nonetheless, the usage quite clearly is not ascribed to any Amora.

The debate appears as well at B. Yoma 87b. Here we return again to the debate about the relationship between Ma‘ariv and Ne‘ilah:

Once again, the anonymous voice invokes the debate about the status of Ma‘ariv—this time, specifically Rav's having ruled in accordance with the lenient view—to engage with an earlier source, in this case Rav's statement that recitation of the Ne‘ilah prayer fulfills one's obligation in the evening prayer. Once again, no later Amora cites the debate—either in its tannaitic guise (i.e., the debate between R. Yehoshua and Rabban Gamliel) or its early amoraic form (Rav and Shmuel). Also important, this anonymous Babylonian voice makes clear that “optional” in this debate truly means optional—unlike the Yerushalmi's Rabbi Ḥanina, who equated the view of Rav regarding Ne‘ilah with the lenient view regarding Ma‘ariv generally, the anonymous voice here views these positions as being in tension with each other precisely because it assumes a powerfully lenient position regarding the evening prayer.

Perhaps significant as well is the fact that in one manuscript of the passage from Yoma, the anonymous voice refers not to Rav as the amoraic sponsor of the view that the evening prayer is optional, but rather to Rav Papa. I noted above that a very small minority of textual witnesses to B. Berakhot 27b present the middle-generation amoraic sages Rava and Abbaye, rather than Rav and Shmuel, as disagreeing about which tannaitic view to follow. Given Rav Papa's close relationship to Rava, perhaps this reading of B. Yoma 87b represents another example of a later dating of the debate, not to Rav and Shmuel, but rather to middle-generation Amoraim (Rava, Abbaye, Rav Papa). Taking this version of the debate more seriously leads to the following summary:

1. 1. A debate about Ma‘ariv attributed to the tannaitic sages R. Yehoshua and Rabban Gamliel appears in no tannaitic corpora, but in both Talmuds.

2. 2. In the Yerushalmi, this tradition is referred to as a Babylonian tannaitic teaching. In the Bavli, it appears in a memra of the first-generation Babylonian sage Shmuel with no markers of tannaitic origins. (Though the story based on it and which follows it in the Bavli is indeed marked by the term תנו רבנן.)

3. 3. Babylonian Amoraim also debate the issue, though there is some ambiguity as to whether these are early or middle-generation figures.

4. 4. The debate is referred to in three other locations in the Bavli (nowhere else in the Yerushalmi). In two of those cases, it appears clearly in the anonymous layer of the sugya, while in the third, though it is couched as part of an Amora's statement, it is likely a later, anonymous interpolation.

If one were to take the attributions of the debate at B. Berakhot 27b seriously as locating the debate in tannaitic times and then being rearticulated by first-generation Amoraim Rav and Shmuel, it is striking that nowhere in the Bavli do we find later Amoraim referencing it. The case of Ma‘ariv as an optional prayer thus may offer partial support for Cohen's finding that traditions asserted in rabbinic texts to be early (i.e., tannaitic) and Babylonian are in fact not evidence for an independent Babylonian tannaitic tradition. Assuming that the attributions to Shmuel and Rav in most of the witnesses of the Bavli are correct, the assignation of a lenient view to Rabbi Yehoshua dates back to the earliest stratum of amoraic Babylonian authorities, but there is no reliable evidence that it stretches back earlier than that. What is more, the lack of later amoraic sages commenting on this tradition, as well as the possible ambiguity about whether early-generation Amoraim (Rav and Shmuel) or middle-generation figures (Rava, Abbaye, and Rav Papa) articulated it, hints that the origins of this tradition were even later.

However, unlike the examples that Cohen has considered, which he argued all had their origins in Palestinian teachings, this tradition does appear to be decidedly Babylonian, both in its origins and in its association in the eyes of early Amoraim in both Babylonia and Palestine.Footnote ⁵³ Perhaps further consideration of תניי תמן passages in the Yerushalmi will shed light on whether this finding is generalizable.

Why Babylonia?

Until now I have argued that the idea that the evening prayer is optional has its roots in amoraic Babylonia and was influential primarily, if not uniquely, there. Though Rabbi Yehoshua's opinion that Ma‘ariv is not obligatory appears once in the Palestinian Talmud, it appears to have been fairly inconsequential in amoraic Palestine. Indeed, the lone Palestinian sage to relate to the debate in the Yerushalmi, Rabbi Ḥanina, apparently interpreted it to mean that “optional” does not truly mean obligation-free, since he aligned this position with the view that recitation of Ne‘ilah exempts one from reciting Ma‘ariv. By contrast, Rabbi Yehoshua's position appears three additional times in the Bavli. Significantly as well, when the Bavli presents the locus classicus for the debate at B. Berakhot 27b, it connects Rabbi Yehoshua's view to the mishnaic phrase אין לה קבע, thus implying that the Mishnah's editing has preferred his position. In effect, then, Palestinian rabbinic culture assumed a thrice-daily prayer practice, reporting but marginalizing as foreign a position that would exempt a Jewish adult from reciting the evening prayer; in amoraic Babylonia, the lenient view was at minimum a viable option, and seems even to have been preferred.Footnote ⁵⁴ Are there reasons why a notion of twice-daily prayer should develop and hold greater sway specifically in rabbinic Babylonia?

Though our data are too sparse to be particularly definitive, external factors are relatively unlikely to be at work in this case. Zoroastrians participate in neither twice- nor thrice-daily prayer, but rather have five daily prayer services.Footnote ⁵⁵ What is more, the evidence from early Christian communities is decidedly ambiguous in its bearing on this question. The Didache mandates recitation of the Lord's Prayer three times a day (8:3).Footnote ⁵⁶ Even leaving aside the complicated questions of how to treat this possibly first-century work of ambiguous geographical origins and its significance for different Christian communities in antiquity, the Didache's call to thrice-daily prayer could easily lead one to competing narratives. Would Babylonian rabbis have abandoned their own tradition of statutory prayer three times a day so as to differentiate themselves from local Christians? Or would they be more likely to embrace this practice, feeling emboldened by similar practices among their neighbors? Either is conceivable, and thus the possible influence of local Christians, while plausible, remains a thoroughly ambiguous datum. What is more, a number of early Christian works, including the Didascalia apostolorum, another work with connections to ethnically Jewish Christians, mandate prayer seven times a day!

A more promising route for thinking about the Babylonian move towards a twice-daily model of prayer is considering it in the internal context of Israelite and Jewish models of statutory prayer. The Hebrew Bible presents at least two different paradigms for daily worship.Footnote ⁵⁷ The cultic service requires twice-a-day offerings, morning and afternoon (Numbers 28:4, 8). Similarly, Deuteronomy 6:4–9 implores the reader to “speak” some set of words when lying down at night and rising up in the morning, a set of verses that, though understood variously in Second Temple times, by the earliest rabbinic texts already refers to a liturgical recitation twice daily.Footnote ⁵⁸ Both of these produce a twice-a-day model of prayer.

At the same time, at least two biblical texts seem to allude to thrice-daily prayer. I discussed above T. Berakhot 3:6, which cites Psalms 55:18 (“Evening, morning, and noon, I complain and moan, and [God] hears my voice”) as grounding thrice-daily prayer. Though the verse likely is not discussing regular prayer—“complaining” and “moaning” sounds more like spontaneous communication with the divine, and “evening, morning, and noon” may well be a turn of phrase intended to mean “all day long”Footnote ⁵⁹—a reader with a thrice-daily prayer life could easily retroject their own practice back onto the verse. And indeed, Daniel 6:11 features Daniel going to pray three times a day in accordance with his custom, an unambiguous reference to daily prayer, and one that fits well with a more literal reading of Psalms 55:18.Footnote ⁶⁰

These passages led Michael Weitzman to speculate about the concurrent development of a thrice-daily prayer cult developed by “Levites in the provinces” alongside the twice-a-day practice of the temple.Footnote ⁶¹ We need not accept the precise details of his suggestion to see how biblical readers could derive two different schema for daily prayer, one with three prayers a day, and the other with two. We can also see how twice-daily prayer would be more closely associated with the temple and cultic practice, while the model of praying three times a day would have fewer cultic associations.Footnote ⁶²

The tannaitic paradigm of prayer three times a day can be read as an expansion of that biblical model—a move already made by T. Berakhot 3:6, which cites not only the verse from Psalms, but also Daniel 6:11, and which appears as well in both the Palestinian and Babylonian Talmuds.Footnote ⁶³ The more cultic, twice-daily model found its own inheritors in the texts of the Qumran community, which mandate prayer morning and evening.Footnote ⁶⁴

Weitzman, assuming that separation from the cult would also mean a greater interiorization of religious expression, suggested that the thrice-daily model of statutory prayer would have been stronger in the Diaspora and with groups less tied generally to cultic sacrifice.Footnote ⁶⁵ Indeed, the Qumran community, for all of its critiques of the central Jerusalem temple, reflects a prayer practice and discourse that keeps an ideal temple service at the core.Footnote ⁶⁶ And yet, in the context of rabbinic literature, it is precisely the diasporan community of Babylonia that pushes in the direction of a “return” to a cultic model of prayer. Perhaps this turn reflects a greater antiquarianism and desire to reflect temple practices more closely on the part of Babylonian Amoraim.Footnote ⁶⁷

The historical reconstruction offered here rules out another narrative for describing the origins of the debate about the status of Ma‘ariv. Ginzberg suggested that “those who said that the evening prayer is optional … were seizing on an ancestral custom; just as in temple times they established for them as an obligation only two prayers, parallel to the two regular offerings, so too after its destruction they did not see a reason to add another obligatory prayer.”Footnote ⁶⁸ However, I have argued here that leniency regarding the evening prayer was a Babylonian innovation, rather than a doggedly persistent “ancestral custom.” What is more, given my description of this debate as reflecting a Babylonian-Palestinian divide in the amoraic period, the reference in the book of Daniel to thrice-daily prayer specifically in Babylonia argues against viewing Ma‘ariv-as-optional as the preexisting practice of this community. Our only evidence for some kind of pretalmudic “ancestral” practice in Babylonia describes a thrice-daily prayer model. Rather, the possibility of eliminating Ma‘ariv and moving towards semidaily prayer, though nostalgic for the temple, was inventive rather than ancestral.

* * * * *

The debate of Rabbi Yehoshua and Rabban Gamliel about Ma‘ariv is likely one of the best-known legal disputes in rabbinic literature. Yet, the existence in tannaitic Palestine of a view such as that attributed to Rabbi Yehoshua—that the evening prayer should be only optional—is exceedingly unlikely. The opinion finds no expression in the Mishnah, Tosefta, or tannaitic midrashim, and these tannaitic corpora in fact generally reveal an assumption of a thrice-daily prayer practice. Views attributed to Rav and Rabbi Yoḥanan in the Talmuds are consistent with what we find in those tannaitic works as well.

The origins of this tradition are reflected as well in the technical presentation of the debate in the two Talmuds. Although the Babylonian Talmud presents the story in which the debate features as a baraita, introduced with the phrase תנו רבנן, it appears in the Bavli as an independent legal statement attributed to the first-generation Babylonian Amora Shmuel. Similarly, the Yerushalmi marks this tradition as distinctively Babylonian, with the Babylonian-affiliated sage Rabbi Yaakov b. Aḥa introducing it with the formula תניי תמן and thus attributing it to the rabbinic community of Babylonia. In this case, these attributions appear to be significant, testifying meaningfully to the Babylonian origins of the debate between Rabbi Yehoshua and Rabban Gamliel.

Although most witnesses to the Bavli present the first-generation sages Rav and Shmuel as already debating which tannaitic view to follow, there is some reason to be skeptical of this particular attribution, especially in light of the contradiction between Rav's opinion in favor of Rabbi Yehoshua and his many positions, in both Talmuds, that assume an obligation in Ma‘ariv. In any event, in the Bavli, the debate is engaged with and used in other contexts only by the anonymous voice. All of this makes the possibility that this debate represents a pre-amoraic Babylonian tradition relatively unlikely.

Finally, although I am unable to identify the factors that made a twice-a-day prayer regime more appealing in rabbinic Babylonia than in Palestine, external factors are likely not helpful as an explanation. Rather, the relationship between these two different models of daily prayer to the temple service is a more fruitful avenue for considering these legal developments; the Babylonian interest in twice-daily prayer likely reflects a growing interest in connecting prayer to memories of the cult. Further exploration of this particularly Babylonian interest may well provide the necessary context for understanding the creation of the view attributed to Rabbi Yehoshua.