PARALLEL HISTORICAL TRADITION IN JOSEPHUS
AND RABBINIC LITERATURE
SHAYE J.D. COHEN
The value of rabbinic literature for the study of the history of
the second temple period is the subject of ongoing debate. The
rabbinic evidence about the period between Alexander the Great and 70
CE is of two sorts, explicit and implicit. The former category
consists of sayings attributed to, and stories about, figures who
lived during the period. Implicit evidence is provided by anonymous
texts, whose origins may ascend to the Maccabean or pre-Maccabean
periods, and by material attributed to the masters of Yavneh, which
may allow deductions about the state of Jewish law and society in the
immediately preceding century. In this essay I am interested not in
sayings and legal rulings but in the people to whom rabbinic tradition
ascribes these sayings and rulings. I am interested not in the
implicit evidence but in those texts which explicitly describe the
conditions of the second temple period.
Much of the rabbinic evidence is unparalleled in other sources.
Shimon ben Shetah and Yohanan ben Zakkai loom large in rabbinic
tradition, but are completely unknown outside of it, as are most of
the figures listed in the chain of tradition in Abot 1. Similarly^ the
Houses of Hillel and Shammai are unknown to non-rabbinic sources.
Many alleged events of the period are known only to rab
tradition. For example, only tSota 13.6 p. 232L (and parallels)
knows that Simon the Righteous heard a voice in the temple announcing
the death of QSGLGS, presumably Caius Caligula, and only mSota 7.8
describes the moving scene which occurred one Sukkot when the
assembled multitudes assured King Agrippa that he was their brother
and therefore entitled to be king. The omission of these two stories
from Josephus is remarkable, since elsewhere he tells stories of the
same type. John Hyrcanus heard a voice in the temple when his sons
1. They are mentioned in various Christian texts of the fourth century
and later, but these need not be discussed here.
2. In this essay, I generally give only one reference for each
rabbinic text; the parallels can be traced easily.
Shaye J.D. Cohen
fought Antiochus Cyzicenus (AJ 13.282-283 // tSota 13.3 pp. 231-232L).
The lineage of the Hasmonean and Herodian kings was not spotless, and
once a Jew even attacked the Jewishness of Agrippa I (AJ 19.332-334).
But Josephus omits the two analogous incidents known from rabbinic
tradition. (In fact, in Josephus' view of things the story involving
Agrippa is an impossibility, because according to AJ 4.209 it is the
high priest, not the king, who recites the reading of the Torah every
seventh year). The origin and value of all this uniquely rabbinic
material are problematic.
A good part of the rabbinic evidence, however, is paralleled by
other sources. That the Maccabees buried the relics of the old altar
is known from both mMiddot 1.6 and 1 Maccabees 4.46. The martyrdom of
a woman and her seven sons (2 Maccabees 7) is told in several
different rabbinic versions. The execution of Jesus is described in
bSanhedrin 43a. Many of the stories in the rabbinic Alexander cycle
(notably bTamid end) are paralleled in the Greek Alexander romances.
But the single source which provides more parallels to rabbinic
tradition than any other is the work of Josephus. In his Essai sur
1'histoire et la q^oqraphie de la Palestine d'aprfes les thalmuds et
les autres sources rabbiniques (Paris 1867), Joseph Derenbourg
assembled many of the rabbinic texts which parallel and supplement the
narrative of Josephus. Following Derenbourg's lead some scholars have
compiled anthologies to show the rabbinic view of second temple
history, while others have attempted to determine the
interrelationship and relative historicity of the parallel versions of
some of the stories.
I have been working for some time on a literary study of the
entire corpus of paralleled material. The questions I hope to answer
are historiographical rather than historical. First, which source
preserves the data in more "original" form? Second, how have Josephus
and the rabbis modified the data which reached thefo? What are their
biases, goals, and methods? Third, what sources did the rabbis use
for the history of the second temple period? Did they read Josephus,
either directly or indirectly, or did they use Josephus1 sources? If
the latter, do the rabbinic parallels help identify the nature of
these sources? Although my work is far from complete, I should like
to present in preliminary form some general conclusions.
My answer to the first question (which source preserves the data
in more "original" form?) needs to be stated at the outset. The War
and the Antiquities were completed over a hundred years before the
redaction of the earliest rabbinic text, the Mishnah. Some of the
content of the Mishnah, the Tosepta, etc. undoubtedly derives from the
second temple period, but the current literary form of all of these
works is not older than the third century CE. Therefore I assume that,
unless there is a clear reason to argue the contrary, the Josephan
traditions are older and more "original" (I do not say more reliable)
than the rabbinic. As far as I have been able to determine in not a
Parallel Tradition in Josephus and Rabbinic Literature
single case is there a compelling reason to argue the contrary; at
best the rabbinic accounts sometimes appear to be coeval with the
Josephan, but never prior. If this is correct, Josephus provides a
"control" for the study of the rabbinic texts, and that control should
facilitate the solution of my second and third sets of questions. The
study of the rabbinic parallels to Josephus is primarily a study
rabbinic historiography.
The Definition of Parallels
The paralleled traditions cover a great range of topics and
themes. The people of the second temple period known by name to both
Josephus and the rabbis include various gentile rulers, Jewish rulers,
high priests of the Roman period, rabbinic (or proto-rabbinic)
figures, revolutionaries involved in the war against the Romans, and
assorted others. Sometimes the image projected by these people in the
two traditions varies considerably, as I shall discuss below. The
events which are recounted in both traditions include many examples of
the interaction of the Jews with their gentile rulers, miracles
associated with the temple, the deeds and misdeeds of the priests and
high priests, and the role of the Pharisees/rabbis in society.
When is a parallel a "parallel"? The relationship of the
rabbinic to the Josephan tradition ranges from the nearly identical to
the very dissimilar. In order to simplify matters I present four
types. (1) In some cases the rabbinic and the Josephan versions are
so close to each other as to be practically identical. Some examples
of this type are the rabbinic and Josephan accounts of the voice heard
by John Hyrcanus (Yohanan the High Priest in rabbinic parlance) in the
temple (mentioned above), of the miraculous suspension of rain during
the daytime when the Herodian temple was being built (AJ 15.425 //
bTaanit 22b-23a), and of the pollution suffered by various high
priests on the evening of the day of Atonement (tYoma).
(2) In most cases, however, the identification is more
problematic because, although the traditions clearly involve the same
event or person, the two versions differ in so many points that the
3. I exclude the following material from this study: 1. Scriptural
exegesis (i.e. rabbinic parallels to AJ 1.1-11.296) and historical
geography. 2. "Archaeological" matters, notably the layout and
procedures of the Temple, and the constitution and prerogatives of the
Sanhedrin. These are enormously complex areas which are best avoided
here, unless both Josephus and rabbinic tradition share an anecdote or
some specific claim. 3. Those stories which appear for the first time
in the Hebrew scholia to Megillat Taanit: see the scholia for 7
Kislev, 28 Tebet, 2 Shebat, 22 Shebat. Megillat Taanit itself is
problematic and obscure, and is best treated separately.
Shaye J.D. Cohen
differences as well as the similarities have to be explained. The
major examples this type are the rabbinic and Josephan accounts
meeting between Alexander the Great and the high priest ^'AJ 11.3
// bYoma 69a), of the breach between John Hyrcanus (Yannai the K
rabbinic parlance) and the Pharisees (AJ 13.288-298 // bQiddusin
of the hostile reception which Alexander Jannaeus received when
officiating in the temple (AJ 13.372 // mSuka 4.9), of the trial of
Herod before the sanhedrin (BJ 1.210-215 and AJ 14.168-184 //
bSanhedrin 19a-b), of Herod's murder of Mariamme (BJ 1.432-444 and AJ
15.62-87 and 202-242 // bBaba Batra 3b-4a), of the conversion to
Judaism of the royal house of Adiabene (AJ 20.34-53 // Genesis Rabbah
46:9), and of the census of the people by Cestius Gallus at Passover
time (BJ 6.423 // tPesahim 4:15). In all these cases the rabbinic and
the Josephan accounts differ in numerous details large and small, but
the identity of the stories is clear nevertheless.
(3) In some cases the Josephan and rabbinic stories share the
same structure but otherwise are so different that the similarity is
not readily apparent. Here are three examples, each of which requires
a separate study, (a) According to both Josephus and the rabbis a
schismatic temple was founded as the result of a feud between two
bothers who shared the high priesthood. At issue was the liaison of
one of the brothers with a woman. Upon being ousted from Jerusalem,
the brother founded a temple for the service either of the God of
Israel or of some other deity. In Josephus the temple is that of
Samaria (AJ 11.302-303, 306-312, 321-325); in rabbinic tradition it is
that of Heliopolis (bMenahot 109b and yYoma 6.3 43c-d). (b) According
to both Josephan and rabbinic accounts a man named Zadoq, in
association with a colleague, broke away from the Pharisees to
establish a new and intrusive hairesis (sect or philosophical
school). In Josephus Zadoq the Pharisee, together with Judas the
Gaulanite (or Galilean), founded the fourth philosophy, a distinct and
illegitimate school (AJ 18.4-10 and 23-25). In the Fathers according
to Rabbi Nathan (A 5 and B 10, p. 13b ed. Schlechter) Zadoq and Boethus
rebel against Antigonus of Sokho (a link in the chain of tradition)
and establish the Sadducees and Boethusians. (c) The third example of
this type of parallel is the well known story of the prophecy of
Josephus himself and Rabban Yohanan ben Zakkai before Vespasian. In
both accounts the hero of the story escapes from his dangerous
situation by a strategem, prophetically hails Vespasian as king and
the destroyer of the temple, and is rewarded by the Romans for his
efforts.
(4) In some cases an isolated motif is shared by two stories
which otherwise seem unrelated. In AJ the Jews debate the Samaritans
before Ptolemy Philometor (AJ 13.74-79); in rabbinic tradition the
Jews debate various nations before Alexander the Great (bSanhedrin
91a). Josephus retails the anti-Jewish charge that Antiochus IV found
10
Parallel Tradition in Josephus and Rabbinic Literature
an ass' head in the temple (Against Apion 2.80);^ in rabbinic midrash
the Ammonites and Moabites find the cherubim in the temple and accuse
the Israelites of idolatry (Lamentations Rabbati petihta 9 p. 4b
Buber). In order to disqualify him from the priesthood, Antigonus
cuts or bites or bites off Hyrcanus' ears (AJ 14.365-366 and BJ
1.269-270); in order to disqualify him from the priesthood Rabban
Yohanan ben Zakkai maims the ear of a Sadducean high priest (tPara
3.7-8 p. 632Z).
Rabbinic Historiography
Parallels of the second, third, and fourth types illustrate the
methods used by the rabbis in preserving and transmitting historical
material. The general pattern which emerges matches almost exactly
the portrait of "popular historiography" drawn by Hippolyte Delehaye
in his The Legends of the Saints (English translation 1962). There is
a fundamental lack of concern for chronological exactitude. Stories
and sayings cluster around a small number of archetypal heroes who
absorb the material of lesser figures. In the sixteenth century
Azariah de Rossi recognized the typological nature of the rabbinic
character of Simon the Righteous. Other figures of this type are
Yannai the King, Yohanan the High Priest, Shimon ben Shetah, and
Agrippa the King. The Rabbis do not know who any of these figures
were or when they lived; the Josephan parallels show that rabbinic
tradition has bestowed upon each of them events which originally were
performed over various periods by various people. Even Yohanan ben
Zakkai has absorbed stories originally told about others, notably
Antigonus (the contemporary of Herod) and Josephus. These techniques
resemble not only those of Christian hagiography but also those of
rabbinic midrash. The fondness of rabbinic midrash for archetypes,
type stories, and clustering has been well analyzed by Izhaq Heinemann
in his Darke HaAqqadah (1954). Just as the midrash declares that
Ibzan and Boaz are one and the same, that Cyrus, Darius, and
Artaxerxes are one and the same, the rabbis declare that Yannai the
King and Yohanan the High Priest are one and the same (bBerakot 29
Even when they do not explicitly state that they are creating
composite figures, they are doing so.
The paralleled traditions about both the people and the events
of the second temple period demonstrate also the "rabbinization" of
Jewish history. Simon the Righteous appears in Josephus as a high
priest, not as a link in the chain of Pharisaic or rabbinic
4. This charge is known from various sources; see M. Stern, Greek and
Latin Authors on Jews and Judaism (1974) 1.97-98.
5. Jacob Neusner, The Rabbinic Traditions about the Pharisees before
70 (Leiden, 1971).
11
Shaye J.D. Cohen
tradition. Other priests are similarly assimilated by rabbinic
historiography (Joshua ben Gamala, Hananya the seqan of the pr
and Ishmael ben Phiabi). Josephus says that Menahem was a leader of
the Sicarii in the fall of 66, that Simon ben Gamaliel was'a leader of
the revolutionaries in the spring of 67, and that Zechariah ben
Amphicallaeus (Abqulos, in rabbinic parlance) was one of the leaders
of the Zealots a year later, but in rabbinic tradition all three
figures are "rabbis," none is a revolutionary. If the bene Bethera,
who are associated with Hillel in bPesahim 66a are to be identified
with the military colony of Babylonians established by Herod in the
village of Bathyra (AJ 17.26), we observe the same phenomenon. In
rabbinic tradition high priests, revolutionaries, and mercenaries
become scholars. Once again, rabbinic midrash, which frequently
rabbinizes Biblical heroes (e.g., Moses "our Rabbi"), parallels this
process.
The rabbinization of events is just as clear. Biblical
quotations and other rabbinic touches are added to a story's nar
framework. Even more important are the substantive changes which are
introduced to make a story accord with rabbinic ideology, in this case
to make it part of pro-rabbinic anti-Sadducean polemic. In Josephus
the masses pelt Jannaeus with their citrons because they hate him and
his rule; in the Tosepta they pelt an anonymous Boethusian high priest
because he refused to perform a rabbinic ceremony. Thus interpreted,
the story becomes part of a cycle designed to show that the Sadducees,
although they disagree with the rabbis, must nevertheless obey them
(yYoma 1.5 39a). A second story in the cycle is enriched by a motif
derived from another context entirely (Yohanan ben Zakkai slits an ear
of the high priest). These rabbinic storieg are more interested in
speaking about rabbis than about Pharisees. Josephus, of course, is
unaware of this distinction and claims that the Sadducees must obey
the Pharisees. For Josephus the adherents of the Fourth Philosophy
(Sicarii) are the illegitimate deviants from Pharisaism; for the
rabbis the Sadducees and Boethusians have this distinction. In
similar fashion, rabbinic tradition takes the sanhedrin, an
inter-denominational council chaired by the high priest, and converts
it into a rabbinic council chaired by two rabbinic figures. The
history of the second temple period is thereby made useful for the
rabbinic memory.
The Sources of Rabbinic Tradition
One topic which awaits detailed investigation is the nat
6. On this distinction see S.J.D. Cohen, "The Significance of Yavneh:
Pharisees, Rabbis, and the End of Jewish Sectarianism," HUCA 55 (1984)
27-53.
12
Parallel Tradition in Josephus and Rabbinic Literature
the source(s) used by both Josephus and the rabbis.^ In many cases in
AJ the paralleled story seems to be an addition from some other source
into the basic narrative (for example, the break between Hyrcanus and
the Pharisees; Jannaeus is pelted with citrons; Honi the rain maker,
AJ 14.22 // mTaanit 3.8; the siege of Jerusalem by Hyrcanus II, AJ
14.25-28 // bSota 49b, bMenahdt 64b, and bBaba Qamma 82a). On the
rabbinic side some of the parallel narratives were originally written
for a purpose other than the one for which they are adduced in their
present literary contexts. The story about Simon the Righteous is
cited by the Talmud in a discussion about the permissibility for a
high priest to wear his priestly vestments outside the temple. The
long narrative about the break between Yannai and the Pharisees is
cited by the Talmud in a discussion about pedigrees and genealogies.
The narrative about the siege of Aristobulus II and his supporters in
Jerusalem by Hyrcanus II is set in three different halakic contexts by
the Bavli. These are the three longest parallels between Josephus and
rabbinic tradition. All three are found in the Bavli, not the
Yerushalmi, and all three seem to be cited by the rabbis from some
written source, a sort of l)r-Yosippon, which was unaware of the uses
to which the stories would be applied in their rabbinic settings. Two
of the three (Hyrcanus/Yannai and the Pharisees; siege of Jerusalem)
seem to be accretions to Josephus' source, implying that he had access
to this history as well.
In the third case, however (the high priest and Alexander the
Great), thre author of the story in bYoma 69a has used Josephus.
Elsewhere I tried to demonstrate that the account in AJ 11 is a
combination of three strands which originally were separ
Josephus, not his source, is responsible for combining them. If this
is correct, the rabbinic account must be dependent upon Josephus
because it too combines two of the three strands (a story about the
Samaritans and Alexander, and a story about a high priest and
Alexander).
In at least one other instance rabbinic dependence upon Josephus
is demonstrable. In tPara 9.2 p. 637Z, in a discussion of the streams
which are permitted to provide water for the red heifer ceremony, some
disputants of R. Judah remarked "All the waters of creation failed
during the war; Siloam (was so dry that) an ant could walk in it."
This statement seems to be dependent upon Josephus' claim in BJ
5.409-410, in a speech which he delivered before the walls of
Jerusalem, that "Siloam and all the springs outside the town were
7. The speculations of D. R. Schwartz in JQR 72 (1982) 241-268, esp.
262-268, do not convince me.
8. "Alexander the Great and Jaddus the High Priest according to
Josephus," AJSreview 7-8 (1982-83) 41-68.
13
Shaye J.D. Cohen
failing (before Titus arrived), insomuch that water was sold by the
amphora, whereas now they flow ... freely for your enemies ..." This
proves, according to Josephus the orator and author, that God has
abandoned the Jews and is supporting the Romans. But the facts of his
argument are false. Nowhere in BJ does Josephus say that the besieged
were afflicted by thirst. Hunger, yes, thirst, no. Dio Cassius says
explicitly that during the siege of Jerusalem the Romans were harassed
by a shortage of water while the Jews suffered no l^ck. Many ancient
sources refer to Jerusalem's abundant water supply. Josephus,
therefore, is lying for the sake of his theological argument, but his
argument reappears in tPara: all the streams, especially Siloam,
failed during the war. The passage must be dependent on Josephus.
In several other cases the Josephan narrative must be earlier
than the ra^inic (notably the break between Hyrcanus/Yannai and th
Pharisees), but dependence by the rabbis on Josephus is not
demonstrable. More significant, however, is the fact that'in not a
single case is the rabbinic version earlier than the Josephan.
Conclusion
The major problem in the study of rabbinic historiography is the
lack of a control on the rabbinic texts. The Josephan parallels
provide such a control, and illuminate the ways in which the rabbis
molded the traditions which they received. Future research will have
to distinguish clearly the diverse strata and genres within rabbinic
tradition. We must distinguish tannaitic from amoraic, Palestinian
from Babylonian, midrash from Talmud, and, perhaps, every individual
document from every other, if we we are to get a clearer understanding
of the rabbinic interpretation of the past.
9. Dio Cassius 66.4.5 = Stern #430; for other passages see Stern's
commentary on # 41 (Timochares). For the archaeological data see J.
Wilkinson, "Ancient Jerusalem: Its Water Supply," PEQ 106 (1974)
33-51.
10. S.J.D. Cohen, Josephus in Galilee and Rome (Leiden 1979) 254-255.
The only other possibility is that Josephus and the Tosefta are
independently following a standard motif. Cf. Psalms of Solomon
17.21. The phrase "the waters of Siloam and the waters of creation"
might be a Palestinian expression; see yTaanit 2.1 65a (the parallel
in the Bavli 16a has "all the waters in the world").
11. Neusner, Traditions 1.175, anticipated by Derenbourg 80 n. 1.
14