The Josephan Influence

Almost every scholar who has commented on the issue considers the rabbinic story heavily indebted to Josephus.Footnote ⁹ Daniel Schwartz has aptly summarized both the connections to, and discrepancies from, the Josephan account. He notes:

In other words, together with direct parallels, some elements of the talmudic story, though they do not correspond directly to Josephus's account, seem to be based on other passages in Josephus, which contribute to the claim that the rabbinic story has a prominent Josephan influence.

At the same time, Schwartz duly notes the contradictions with Josephus's account:

However, I think that these efforts by Schwartz and others to find direct parallels to Josephus, what Schwartz calls the “correct” lines of the story, can be questioned. Schwartz claims, for example, that Herod “did marry a Hasmonean princess and she did die in confrontation with him.” But in the rabbinic story he does not marry her, and she dies by suicide, not because Herod ordered her death. That Herod “did rise to power by replacing” the Hasmoneans, as Schwartz puts it, does not exactly correspond to the massacre recounted in the Talmud. Schwartz's additional connections to other Josephan passages are likewise not probative: The wife of Pheroras, according to Josephus, did not in fact die, but survived her jump from the roof.Footnote ¹² In any case, suicide by jumping off the roof is a common Bavli narrative motif, attested of a Roman officer in B. Ta‘anit 29b, the laundryman who failed to call on R. Yehudah Ha-Nasi on the day of his death in B. Ketubot 103b, the heretic (min) who challenges R. Yehudah Ha-Nasi in B. Ḥullin 87a, and the mother and father of a slain child in B. Ḥullin 94a. So there is little reason to see that Josephan passage about the wife of Pheroras as the source of the maiden's suicide, which is a stock motif available in other Bavli anecdotes.

Eli Yassif, adopting a more folkloristic perspective, mentions the Josephan connection in a looser manner:

It is easier to accept this perspective, which places less emphasis on specific passages and more weight on the general contours of the story derived from Josephus among other historians, narrative elements which probably circulated and metamorphosed in oral form for centuries thereafter. Yassif is on target in characterizing the “historical” dimension of the Bavli account as a “grain of truth,” (with an emphasis on the “grain”). The addition of “folk narrative motifs” seems indisputable, and the rabbinocentric tendencies are indeed typical. Yet once we concede that the Josephan substratum is minimal and deviations from Josephus are prominent, the question arises as to other sources of the Bavli tradition.

A. A. Halevi has noted that the Herod story appears only in the Bavli and nowhere in Palestinian sources.Footnote ¹⁴ And J. N. Epstein suggested that the custom of preserving corpses in honey is Babylonian, citing Herodotus's observation that the Babylonians bury their dead in honey.Footnote ¹⁵ Herodotus also mentions a Persian custom of preserving bodies in wax, claiming that, “the Persians cover corpses in wax before burying them in the ground.”Footnote ¹⁶ The sources of the story, then, should perhaps be sought among Sasanian traditions in addition to Josephus.Footnote ¹⁷

Sasanian Sources

Traditions of the rise of Ardashir, founder of the Sasanian dynasty, which may have influenced the Bavli story of Herod, have been preserved in a number of Sasanian sources. These include: (1) The Karnamag i Ardashir i Pabagan, (The Book of the Deeds of Ardashir Son of Pabag [Papak, Babag]), a Pahlavi text probably composed in the sixth century CE on the basis of earlier traditions.Footnote ¹⁸ Like other Pahlavi texts, the extant version postdates the Muslim conquest. Scholars have characterized this account as a “romance” and as “popular.”Footnote ¹⁹ (2) The Xwaday-namag (Khvaday-namag) (The Books of Kings), an official court history of the Sasanian dynasty probably assembled from earlier materials during the reign of Xusro (Kosrow) I (Anushirwan; 531–579), and revised and updated by his successors.Footnote ²⁰ The original Xwaday-namag itself is not extant, but it was used by Muslim historians and authors, including Tabari, in his History,Footnote ²¹ and Firdowsi, in his Shahnameh.Footnote ²² (3) Other Sasanian traditions found in Byzantine and Armenian authors.

A brief summary of the beginning of the account in Karnamag i Ardashir i Pabagan is as follows:

Pabag, a nobleman and the governor of Fars, has auspicious dreams about his shepherd Sasan, a descendant of the old Persian kings,Footnote ²³ and gives Sasan his daughter in marriage. She gives birth to a son, Ardashir, whom Pabag adopts and educates. King Ardawan (Ardawan IV / Artabanus IV) hears of the young Ardashir's exquisite talents and equestrian ability, and summons Ardashir to his court to join his retinue. During a hunting expedition Ardashir quarrels with Ardawan's son, when the prince claims to have shot an antelope that Ardashir in fact killed. Ardawan sides with his son and reprimands Ardashir, and then relegates him to the stables to take care of the horses. This naturally embitters Ardashir and kindles resentment toward the royal family.

Some time thereafter, Ardawan consults with his astrologers and soothsayers about what they see in the stars. He receives an ominous response:

A favorite servant girl of the king, referred to as a “certain maiden” (kanīzag-ēw) overhears the astrologers' talk. She had become enamored of Ardashir and fallen in love with him. She informs Ardashir of the astrologers' words. The two take a pair of fine horses and flee.

Ardawan discovers his slave girl is missing, hears of their flight, pursues them, but cannot overtake them, and returns home after three days. Ardashir gathers allies, raises an army, and fights a series of battles, eventually killing Ardawan and other members of the royal family, and marrying his daughter. After making more alliances and fighting additional battles he becomes King of Kings, inaugurating the Sasanian dynasty.

One can immediately recognize the beginning of the Herod story in this Sasanian text, which, after transforming and condensing from written to oral form, could briefly be summarized along the lines of the Bavli story as: “Ardashir was a servant of the House of Ardawan. One day he heard of a voice saying that whoever rebels now against his masters will succeed. He arose and killed many of his masters, and spared one maiden.”Footnote ²⁶

In other words, I suggest that the rabbis understood the transition from the Hasmonean to Herodian dynasties through the prism of the transition from the Parthian (Arsacid) to Sasanian dynasties. They identified Herod with Ardashir, and constructed a story of Herod's usurpation and rise to power on the basis of Ardashir's usurpation and rise to power as recounted in Sasanian sources.

Here are the particulars of the two accounts in more detail.

(1) Herod was a Servant of the Hasmonean House: It is clear that Ardashir can be characterized as a “servant of Ardawan” or “a servant of the House of Ardawan,” exactly as the Bavli states of Herod vis-à-vis the Hasmoneans. This can be said fairly of Ardashir's initial position in the king's retinue, where he is a vassal or attendant. His subsequent demotion to the stables is a clear indication of servitude. Geoffrey Herman has shown, based in part on talmudic evidence, that the stable master was near the bottom of the hierarchy of positions in the royal administration, and evidently considered among the baser forms of service.Footnote ²⁷ Moreover, Ardashir understands the astrologers' prognostication of a “servant who will escape from his master” to be applicable to himself. The version of this scene in the Shahnameh depicts Ardashir's relegation to servitude in the stables more explicitly: Ardawan commands Ardashir: “Go! Look after my Arab horses and keep your place in the stables. Be a groom in the stalls and assistant to any man at any task. Ardashir, his eyes running tears, thus came to be the groom in the stalls and assistant to any man at any task.”Footnote ²⁸ And after Ardashir writes to his father of the unhappy turn of events, Pabag writes back in rebuke: “Why did you thrust yourself ahead of his sons? You are his attendant, not his kinsman.”

Apart from this account of the Karnamag that claims that Ardashir descended from the old Persian kings and that his grandfather Pabag was an aristocrat, there seems to have existed a different “popular” tradition of Ardashir's origins preserved in Byzantine sources. According to this tradition Ardashir emerged from society's lower classes. Thus the seventh-century historian and poet George Pisida (a.k.a. George the Pisidian), writes in his poem, “The Heracliados,” about the Roman emperor Heraclius's campaign against the Persians in 622:

Similarly, the sixth-century Byzantine historian Agathias Scholasticus, in his work “On the Sasanians,” writes of Ardashir as “at first obscure and undistinguished.”Footnote ³⁰ According to Arthur Christensen, this popular version stems from the tendency of the folk to see rulers emerge from lower-class or common origins.Footnote ³¹

Thus the description of Herod as a servant/slave (‘avda) of the Hasmonean dynasty closely parallels the traditions of Ardashir as a servant of Ardawan and the Parthian House, whether we follow the version of the Karnamag, where Ardashir was an aristocratic vassal relegated to servitude in the stables, or the popular version attested among Byzantine authors, where Ardashir was originally a slave. The description also dovetails to some extent both with rabbinic traditions concerning the genealogical status of Herod's descendants and with Josephus's account that Herod's father Antipas was a Hasmonean vassal, although Josephus himself apparently considered Herod to be fully Jewish.Footnote ³²

(2) The voice Herod Hears that Inspires him to Rebel: The Vilna printing of the Talmud reads that Herod heard a bat kala (bat kol). However, the evidence from the manuscripts and early printings is divided:

We thus have two main families: MS Escorial together with the Pesaro and Venice printings, where Herod hears “a certain man” or “them,” and MSS Munich, Florence, and Paris, where Herod hears a kala or bat kala (Paris). The Vatican manuscript and the Vilna printing offer a conflation of the two (in MS Vatican a scribe has crossed out ha-hu gavra and inserted a superlinear bat).

Of the readings kala and bat kala, it seems that kala is original, as one can easily see how kala would have been changed by scribes to the more familiar bat kala, but not why bat kala would be altered to the vague and indefinite kala (and it is unlikely that the word bat was accidentally omitted.Footnote ³³) Moreover, a bat kala proclaiming such a general and indefinite opportunity (“Every slave who rebels now …”) is unusual, if not unattested, as bat kols in the Bavli typically have a more specific target or audience. The closest parallel I can find is the bat kol heard after R. Yehudah Ha-Nasi's death, “Whoever was present … is designated for the world to come.” But even this bat kol applies only to those present, not to anyone who takes the specified action.Footnote ³⁴ Furthermore, if this is a bat kol, we have to ask why a bat kol, which generally discloses elements of the divine will as a sort of prophecy, would make such an announcement: how does the rogue Herod's rebellion against the Hasmoneans accord with the divine will? (a problem which bothered Naḥmanides and other medieval commentators).Footnote ³⁵ If the reading kala is original, however, we are left with the questions: what kind of kala—voice, or rumor, or report—is this, why would Herod trust it, and why would it prove true?Footnote ³⁶ The reading of MS Escorial and the first printings, “he heard them say” or “he heard a certain man say” presents the same difficulty: why listen to this talk, and why did it in fact come true? However, if the Bavli storytellers here mean to emulate the Sasanian account, then these problems disappear: Herod hears the voice or sayings of astrologers and soothsayers analogous to Ardashir. Or, to put it in slightly different words: the rabbinic storytellers are attempting to mirror the Sasanian story where the astrologers’ kol, based on the configuration of the stars, inspires the usurper to rebel.

This conclusion is bolstered by the reading found in Haggadot ha-talmud and a manuscript of the Yalkut Shim‘oni, which read: “He heard the Chaldeans (=astrologers) saying” (שמע לכלדאי דקאמרי).Footnote ³⁷ Now it is always difficult to figure out the provenance of the readings found in Haggadot ha-talmud and their relationship to the manuscripts of the Bavli, although some recent work suggests that this compilation preserves early and original readings.Footnote ³⁸ If this reading is original, we could conjecture a development along the lines of:

The initial kalda'i (Chaldean / ל]כלדאי]) simply dropped out or was intentionally removed because of unease with astrology (although in general the Bavli places a great deal of stock in astrology, as did the Iranians).Footnote ³⁹ This was corrected by the addition of the object of the verb he “heard”: the simple inhu (them) in Escorial, le-ha-hu gavra (a certain man) in Venice and Pesaro printings, and the generic kala (a voice) in Munich and Florence. The reading kala was changed to the more familiar bat kala, in Paris. We find a conflation of the readings bat kala and ha-hu gavra in the Vatican manuscript and the Vilna printing, the traces of which one can still see in Vatican where ha-hu gavra is crossed out in place of kala, and the bat added superlinearly.Footnote ⁴⁰

But this is only a conjecture, owing to the uncertainty regarding the provenance of these versions, which are not found in the Talmud manuscripts themselves.Footnote ⁴¹ It is also possible that the reading kalda'i (Chaldean) simply reflects a different oral version deriving from early times. And perhaps we should also consider the possibility that kala de-’a(mar) (a voice that said), was a corruption of kalda'i during the period of oral transmission, as the pronunciation of the ק and כ were no longer distinguishable. But whatever the source of the reading in Haggadot ha-talmud and the Yalkut manuscript, the parallel to the Sasanian tradition is stunning. We can safely say that this reading accurately reflects the background of the narrative: Herod heard the astrologers disclose that the time is auspicious for any slave who rebels against his masters to succeed. As in the Sasanian accounts, these astrologers have read the stars accurately.

(3) Herod killing the entire Hasmonean House: The Bavli account states that, “He [Herod] arose and killed all of his masters. He left [only] that maiden.” In the Karnamag Ardashir does not kill all of the House of Ardawan. He kills Ardawan and marries his daughter, but he imprisons two of Ardawan's sons while two others escape and subsequently write to their sister, urging her to poison Ardashir.Footnote ⁴² However, in Tabari's version, perhaps deriving from the Xwaday-namag, as opposed to this more romantic version of the Karnamag, Ardashir indeed kills off all of Ardawan's kin, sparing only one maiden:

Thus different accounts circulated as to the fate of Ardawan's extended family. Tabari offers the neat and tidy story of complete annihilation (with the exception of one daughter), which is of course a standard plot of folk narratives. The Bavli's story of Herod coheres with this version.

(4) Herod and the Hasmonean Maiden: Here the two accounts deviate, although there is enough in common to argue for the rabbinic story's dependence on the Sasanian tradition. In the Karnamag it is the maiden who falls in love with Ardashir and who hears the astrologers. Thus the rabbinic story would be more parallel if it read: “One day a certain maiden set her eyes on him. She heard a voice saying …” However, Ardashir returns the maiden's love (and promises to make her queen) and hears of the astrologers' report from her, so the rabbinic account is not too far off.

A second divergence is of course the fact that Ardashir marries Ardawan's daughter—or in other versions of the tradition, Ardawan's niece or another female relative—and she eventually gives birth to Shapur, the next emperor, whereas the Hasmonean maiden commits suicide, and Herod must content himself with a sticky corpse.Footnote ⁴⁴ This difference is a product of the divergent—indeed opposite—purposes of the storytellers. The authors of the Karnamag and Xwaday-namag write to glorify and legitimate the Sasanian dynasty. Hence the marriage to the former king's daughter and offspring with both Sasanian and Arsacid blood. Firdowsi, in the Shahnameh version, hits us over the head with this idea: “Then Sabak came before Ardashir and said, ‘Wise king, demand Ardavan's daughter in marriage: she has splendor and beauty and the dignity of station. The crown will be in your hands.’”Footnote ⁴⁵ The rabbinic storytellers, by contrast, mean to dishonor Herod, delegitimize his claims to the throne, and expose his attempt to mask his lower-class origins. Therefore they relate that Herod failed to marry a Hasmonean descendant—though Herod tried to create the impression that he did so—and remained a slave in the final reckoning, as the concluding Roman message makes abundantly clear.

It should also be noted that the mention of Herod preserving the corpse in honey for seven years has a type of parallel in the Sasanian story, though again employed for the opposite purpose.Footnote ⁴⁶ In the Sasanian stories, Ardashir orders his minister to kill Ardawan's daughter, whom he had married (in the Karnamag and Shahnameh, because she tried to poison him; in Tabari, because of the oath mentioned above). She informs the minister that she is pregnant, so he sequesters her instead, and she soon gives birth to Shapur. The minister keeps her hidden for seven years, until Ardashir regrets having no son and heir, at which point the minister produces the wife and son. Moreover, this assiduous minister takes care to castrate himself at the outset, placing his severed genitalia in a box and having Ardashir set his seal upon it, in order to remove himself from suspicion. In Tabari's version the minister hides the woman in a cellar, telling Ardashir: “I have consigned her to the bowels of the earth.”Footnote ⁴⁷ When he then produces her and has Ardashir open the box with his genitalia, he explains, “We consigned her to the bowels of the earth, as our monarch had commanded us, and we exonerated ourselves in his sight from any guilt (i.e. by self-mutilation), lest any calumniator find any means to forge lies against her.”

One can make a case that the account of Herod and the maiden's corpse comprises a type of inversion of the Sasanian narrative. Ardashir marries and sires a son with a daughter (or other female relative) of the former king; Herod fails to marry or produce offspring with a Hasmonean maiden. Ardashir commands the daughter to be killed, yet she survives; in the rabbinic story she dies, though Herod tries to create the false impression that she survives. The minister hides her “in the bowels of the earth” for seven years to create the impression that a living woman is a corpse; Herod preserves her in honey for seven years to create the impression that a corpse is a living woman. The minister castrates himself to prevent calumnies and lies that the son is not of Ardashir's (and Ardawan's) line. The maiden kills herself for exactly the opposite purpose: to guarantee that claims of Hasmonean descent among Herod's progeny will be recognized as lies. In short, the rabbinic storytellers employ this narrative element of the Hasmonean maiden to further their agenda: to illustrate that Herod, the slave and usurper, never became a legitimate king, neither he nor his descendants. The Sasanian storytellers employ the same narrative element in line with their purposes: to demonstrate that Ardashir, the usurper, was indeed the legitimate king, both he and his descendants, who in fact descend from the Parthian kings too.Footnote ⁴⁸

(5) Herod Building the Temple: One purpose of the Bavli story, as noted, is to explain how the wicked Herod nevertheless performed the great mitzvah of rebuilding the glorious temple, and apparently did an excellent job of it. Temple building provides another parallel to the Karnamag, which portrays Ardashir as building fire-temples throughout the lands he conquered. Indeed, Sasanian royal ideology in general portrayed the Sasanian kings as the great defenders of Zoroastrianism, as religious monarchs who restored the true faith to the Iranian lands after the heterodox ways of the Arsacids.Footnote ⁴⁹ Thus in the Karnamag, after making an alliance with a local potentate named Banug, Ardashir “expressed gratitude to the Yazads, and there founded a district named Buxt-ardasir, and ordered to found ten Victorious Fire(-temples) over the sea (coast).”Footnote ⁵⁰ Again, after the battle with, and defeat of, Ardawan, “He [Ardashir] founded a city called Ardasir-xarrah, dug a large lake and brought out of it four water-courses, there established a Fire(-temple), excavated a huge mountain and made a canal, founded many villages, landed estates and inhabited places, and ordered to establish there many Victorious fires.”Footnote ⁵¹ And we find other such passages.Footnote ⁵² The rabbinic storytellers of course told not of the multiplicity of temples of the Sasanian account but of the single Jerusalem temple. This tradition of a king, founder of a new dynasty, who dedicated resources and energy to building temples, may have been one reason the rabbinic storytellers associated Herod with Ardashir and viewed Herod's rise to power on the basis of Ardashir's ascent.

Apart from temples, Ardashir is portrayed as a great builder in general, especially the builder of cities, as the quotes above attest, mentioning that Ardashir founded cities and villages, landed estates, watercourses, and a canal. Tabari lists eight large cities that Ardashir built, and mentions others too.Footnote ⁵³ The talmudic story does not feature Herod's impressive building campaigns, apart from the temple. But to the extent the rabbis preserved some historical memory of Herod's prodigious building, they had another reason to identify him with Ardashir, and hence to model the story of Herod upon Ardashir's rise to power.

(6) The Romans' Rebuke: The talmudic story concludes with the Romans rebuking Herod for building the temple without permission and reminding him, in insulting fashion, that he remains a slave. An interesting parallel appears in Tabari's account, where Ardawan, after hearing of Ardashir's rebellion and building activities, sends him a letter of insult:

As in the rabbinic story, the rebuke is sent by emissary. Here too the incumbent power rebukes the usurper for building without permission and insists that such enterprises require prior authorization. Here too the rebuke focuses on the usurper's inferior lineage, that Ardashir “presumed beyond your rank in society.” And here too there is an almost poetic insult with a repetition: “O Kurd brought up amongst the tents of the Kurds,” echoing, “You are not a king [rekha], nor the son of a king.” There is a difference, of course, in that the talmudic story places the rebuke in the mouths of the Romans, whereas the Sasanian account attributes it to the soon-to-be-deposed king. But this too is a function of the different agendas of the storytellers. In the Sasanian tradition, Ardawan naturally rebukes the usurper, but Ardashir responds in kind, with insults of his own and a defense of his kingship. The rabbinic storytellers, on the other hand, wish to underscore Herod's failure and unworthiness, hence they conclude their account with the Romans putting Herod back in his true place, that is, a slave.

(7) Herod, Bava b. Buta, and Herod's “Remedy,” Takkanah: One unit of the Herod story that finds no direct parallel in the Sasanian accounts of Ardashir is the massacre of the sages and encounter with Bava b. Buta. This part of the story is probably best explained, along the lines of Eli Yassif's observations quoted above, as the sages anachronistically reading themselves back into history and, as leaders and representatives of the people, a logical target of a usurper.Footnote ⁵⁵ However, there may be some connection to the Sasanian accounts, though I concede that claims of influence here are far more speculative than the points noted above.

The motif of the usurping king's regret and even repentance for killing appears in the Karnamag. The author relates that once when Ardashir was hunting he saw a male antelope save his female partner from Ardashir's arrows at the cost of his life, and then the female do the same to save her young. Ardashir weeps upon seeing this, moved by the animals' compassion, and explains that he is reminded “again (of) that wife, and the child that was, innocent, in the mother's womb; and I am thoughtful and mournful of killing them, that it can be a grievous sin even concerning the soul.”Footnote ⁵⁶ The chief of the magi asks Ardashir why he weeps and whether anything can be done as a remedy (čārag), and then observes that “Ardasir has come to repentance (pasemanih) of that act” and informs him that he in fact spared the maiden and hid her away, because she was pregnant, as discussed above. Upon being presented with his son, Shapur, Ardashir “offered much thanks unto Ohrmazd the Lord” and then “ordered to set up on the same place a city called Valas-sabuhr, and established ten Victorious Fires there, and sent immense property and wealth to the House of the sacred fire.”Footnote ⁵⁷ Here we have the king expressing regret for killing, mention of a remedy by the chief magus, and the king respond by building fire temples and dedicating immense wealth to the temple. Clearly there are important differences, as Ardashir regrets ordering the killing of his pregnant wife and fetus, which was in fact never carried out, whereas Herod regrets killing the sages, which he perpetrated. Yet there are several salient points in common: the chief magus and leading sage, Ardashir's regret for ordering his wife be killed and Herod stating he should not have killed the sages (this may be called “regret” even though the word itself is not used)Footnote ⁵⁸; mention of a remedy (takkanah, čārag); and then the building and enriching of a temple/temples.Footnote ⁵⁹

In addition, the Karnamag tells of a second massacre ordered by Ardashir.Footnote ⁶⁰ Some time after killing Ardawan, Ardashir becomes frustrated at not being able to subdue all of Iran and secure his rule. He consults with an Indian magician or soothsayer (ked i hindigan), who informs him that he must intermarry with the family of Mihrag in order to establish complete rule over Iran. This infuriates Ardashir, who orders the entire family of Mihrag to be killed. However, a wet nurse saves a three-year-old girl, placing her in foster care. Eventually Shapur marries her and she gives birth to Hormazd, and in this way Ardashir and his descendants are able to conquer all of Iran and become the Shahen-Shah, the King of Kings.

Thus the Sasanian traditions tell of a double massacre: Ardashir massacres the House of Ardawan and the House of Mihrag, and in both cases only one daughter survives. Similarly, the Bavli story tells of a double massacre, of the Hasmoneans and the sages, and in both cases only one individual survives. There is also the Indian sage consulted by Ardashir, perhaps parallel to Bava b. Buta, with whom Herod consults.

These points are more speculative, as there is much less of a direct parallel. Rather there are a few common elements configured differently in the two accounts. But these are at least as compelling as the indirect parallels to Josephus, such as the wife of Pheroras jumping off a roof, which some scholars point to as evidence of Josephan influence.