The Tragedies of Ezekiel

Thomas D. Kohn

V
ERY LITTLE G REEK TRAGEDY remains from the Hellenistic
period. We do however have substantial fragments
from a playwright named Ezekiel.1 This paper examines
those fragments in the context of modern criticism and of the
ancient authors who preserved them, and suggests that instead
of deriving from a single play, they are from a tetralogy.
Eusebius, in the Praeparatio Evangelica, preserves a lengthy
quotation from Alexander Polyhistor, polÊnouw Ãn ka‹ polu-
mayØw énÆr (a “thoughtful and much learned man”), author of
Per‹ ÉIouda¤vn. 2 Among the authors Polyhistor discusses is
Ezekiel, ı t«n tragƒdi«n poihtÆw (“the maker of tragedies,”
9.28.1). Praep.Evang. 9.28–29 contains excerpts from and
discussion of a dramatic work called the Exagoge, which retells
the biblical story of the Jews’ Exodus from Egypt. Polyhistor
preserves 269 lines of iambic trimeter; some of these lines are
also quoted by Clement of Alexandria (Strom. 1.23.155–156),
who refers to Ezekiel more specifically as ı t«n ÉIoudaÛk«n
tragƒdi«n poihtÆw (“the maker of Jewish tragedies”). Nothing
is known definitely about the playwright except for his name.
But the common assumption, due to both his name and his

1 The fragments have been recently gathered, edited, and extensively com-
mented upon by Howard Jacobson, The Exagoge of Ezekiel (Cambridge [Mass.]
1983), and Carl Holladay, Fragments from Hellenistic Jewish Authors II
(Atlanta 1989): cited hereafter by authors’ names. References are to the text and
line numbers of Jacobson.
2 Praep.Evang. 9.17.1; the quotation from Polyhistor continues through
9.39.5. I refer to the edition of Karl Mras (Berlin 1954).

Greek, Roman, and Byzantine Studies 43 (2002/3) 5–12

© 2003 GRBS
6 THE TRAGEDIES OF EZEKIEL

subject matter, is that Ezekiel was a Jew.3 Common opinion also

holds that he could not practically have lived anywhere but in
Alexandria. 4 Ezekiel makes use of the language of the Sep-
tuagint, providing a terminus post quem of the mid-third century
B . C . E .; and Alexander Polyhistor flourished in the mid-first
century B.C .E. Within those roughly 200 years, there has been
great scholarly disagreement as to when Ezekiel lived, with Hol-
laday concluding that “the question of dating Ezekiel is perhaps
best left open and general.”5
Concerning the text itself, scholars have been able to mold
these fragments into a “traditional” five-act structure.6 This can
be done with little difficulty; however, it leaves little room for
adequate development of plot, or for suitable choral passages,
especially since most reconstructors assume the portrayal of at
least one confrontation between Moses and Pharaoh.7 But the
feature of such a reconstruction which has proved most striking
to critics is Ezekiel’s apparent violation of the so-called Unities
of Time and Place.8
Aristotle, in the Poetics, says that tragedies seek to take place
within a single twenty-four-hour period (1449b12–16). Al-
though there was no hard and fast rule that the action of a play

3 Jacobson (5) and Holladay (301–303) provide the prevailing arguments,

and other alternatives.
4 Holladay 312–313.

5 Jacobson 5–13, Holladay 308–12 (at 311).

6 The five-act structure is a “tradition” going at least as far back as Horace

at Ars Poetica 189–190. C. O. Brink, Horace on Poetry I (Cambridge 1963) 114,
posits that five acts, while not Aristotelian, “are however likely to be Hel-
lenistic.” And Niall Rudd, Horace: Epistles Book II and Epistle to the Pisones,
(Cambridge 1989) on Ars P. 188–190, states that “one assumes [Horace] was
using a Hellenistic scholar who attempted to divide tragedy into five parts or
acts.” See Jacobson 33ff and Holladay 306–308 on how this five-act structure
has been applied to Ezekiel; see also the recent article by Aleksandra Kleczar,
“The Exagoge of Ezekiel: Analysis of the Dramatical Structure of the Play,”
Eos 87 (2000) 113–118.
7 See Jacobson 35.

8 Notably Otto Zwierlein, Die Rezitationsdramen Senecas (Meisenheim am

Glan 1966) 138–146.
 THOMAS D. KOHN 7

must occur in a single place and in a single day, most of the

surviving Attic tragedies do at least approach these guidelines.
Certainly, even those plays which violate this “rule” never cover
so great a span as does the Exagoge. A conservative estimate,
omitting any presumed meetings between Moses and Pharaoh,
is that there were at least four different settings: before the tents
of Raguel, on Mount Horeb, in the palace of Pharaoh, and some-
where in the desert. And a vast amount of time must have
elapsed between the scenes, with the whole action spanning at
least several months, if not years. Admittedly, this departure
from fifth-century practice may simply represent the natural evo-
lution of the genre. Perhaps by the Hellenistic period tragedians
were attempting works of epic scope. Ezekiel, however, breaks
a much more important Unity: the Unity of Action.
Aristotle is fairly firm that a tragedy must have a single,
unified plot.9 Admittedly, the Attic tragedians do not always
adhere to this rule, either; but, as with the other two Unities, no
one violates it to so great an extent as Ezekiel. It is hard to
imagine how the extant fragments of the Exagoge could possibly
be conceived of as a single plot. But these fragments, I suggest,
do not necessarily come from a single play. Polyhistor and
Clement do refer to Ezekiel’s work as “the Exagoge.” 10 But
rather than being the name of one tragedy, it could refer to a
trilogy or tetralogy. Trilogy names appear in Aristophanes, and
the scholia make it clear that the Alexandrian scholars under-
stood that he was referring to a set of plays.11 So giving three or
four plays a collective title was a known practice in the
Hellenistic period. Furthermore, Ezekiel’s use of language from
the Attic tragedians shows that he had studied them, and
therefore would have been familiar with the practice of writing

9 Among other places, Poetics 1450b23ff.

10 At Praep.Evang. 9.29.4, 9.29.12, 9.29.14, 9.29.15, and Strom. 1.155.
11 ÉOr°steia in Frogs (1124), LukoÊrgeia in Thesmophoriazusae (135), with
scholia.
8 THE TRAGEDIES OF EZEKIEL

three or four plays intended for successive performance, pos-

sibly covering a single story or theme, and all grouped together
under a single title. Finally, composing a tetralogy rather than a
single play would be in keeping with what Jacobson (12)
describes as the “‘archaizing’ element” found in Ezekiel’s work.
In order to obviate the seeming violations of the Unities, I
would suggest the following division of the fragments of Ezekiel
into plays. The first play would begin with the extant intro-
duction, a Euripidean-style prologue delivered by Moses,12 in
which he describes briefly the origin of Jews in Egypt and their
subsequent oppression, and then his rescue as a baby by the
daughter of the king (lines 1–31). Moses goes on to recount his
upbringing, his slaying of an Egyptian, and his witnessing a fight
between two men, which led to his exile from Egypt (32–58).
This is followed by his meeting with the daughters of Raguel, a
name equivalent to the biblical “Jethro” (59). In the next sur-
viving passage, Sepphora describes to Moses both the land of
Libya and its ruler, her father (60–65). According to Polyhistor,
Ezekiel next spoke of the watering of the animals and the
marriage of Moses and Sepphora.13 Then two lines of dialogue
are quoted, in which Sepphora discloses to Chum, a figure not
from the Bible, her betrothal to Moses (66–67). There has been
much scholarly speculation as to who this man may be.
Jacobson surveys the opinions, and concludes that Chum was a
previous suitor who was displaced by Moses.14 If this is so, the
main action of this first play revolves around a foreign prince
(Moses) who is betrothed to a local princess (Sepphora), in-
curring the wrath of an earlier lover (Chum). This is a common
theme in Greek and Roman myth—Medea and Jason, Lavinia

12 See Jacobson 69–70 on the Euripidean qualities of the opening speech.

13 Praep.Evang. 9.28. See Holladay 360 and nn.63–68 on Polyhistor’s com-
ments.
14 Howard Jacobson, “The Identity and Role of Chum in Ezekiel's Exagoge,”
Hebrew University Studies in Literature 9 (1981) 139–146.
 THOMAS D. KOHN 9

and Aeneas, etc. A betrothal with its ensuing complications

features in some form in a number of Attic tragedies;15 and so
including such a theme in the Exagoge would be in keeping with
Greek tragic tradition. Furthermore, Chum’s reaction would
logically lead to what Aristotle says is the best kind of tragic
plot, disaster among friends, e.g. the killing of a f¤low, whether
accomplished or averted.16 Clearly, an entire play would be
necessary to develop adequately the implications of the
marriage.
At this point in the Praep.Evang. (9.29.1–3) Polyhistor pauses
in his discussion of Ezekiel to compare the tragedian’s version
of the story with that of Demetrius (FGrHist 722). If Polyhistor
were quoting from a single tragedy, it seems odd and arbitrary
for him to stop here. But if the dialogue between Sepphora and
Chum marks the end of Polyhistor’s discussion of the first play,
this is a very convenient and sensible place for Polyhistor to
pause. Moreover, Clement does not quote from Ezekiel after this
point, making it likely that he had access only to the first play.
The second play would portray how Moses became the leader
of the Jews. It opens with the extant discussion between Moses
and Raguel, who is now his father-in-law. If Polyhistor pre-
served a single play, an unreasonable span of time must be
represented as elapsing after the previous scene with Sepphora
and Chum; but if this is a new play, any amount of time can
comfortably be allowed since the earlier talk of the impending
nuptials. Moses and Raguel now discuss Moses’ prophetic
dream, in which he is adored by stars while sitting upon a great
throne. Raguel interprets this as a sign of Moses’ impending
greatness (68–89). Then, presumably after a choral interlude,
come the surviving fragments in which Moses on Mt Horeb

15 E.g., Euripides’ Medea, Aeschylus’ Suppliant Women , and even Euripides’

Iphigenia at Aulis.
16 Poetics 1453b14ff. See Elizabeth Belfiore, Murder Among Friends (Oxford
2000) 3–20.
10 THE TRAGEDIES OF EZEKIEL

receives his charge from God. This section portrays the incident
of the burning bush, with the voice of God sending Moses to
Egypt to free his people (90–119); it also shows God teaching
Moses how to change his staff into a serpent and back again,
and how to turn his hand white by placing it in his tunic
(120–131). After this, God foretells the plagues He will visit
upon Egypt and then sets down some instructions for the
celebration of Passover (132–174). Next we can postulate
another choral passage during which Moses descends from the
mountain, after which is the extant passage in which Moses
repeats God’s instructions to Raguel and his fellow shepherds
(175–192), and he or they depart for Egypt. The prominence of
prophetic dreams in Greek tragedy has been well noted.1 7
Furthermore, in this play the Hebrew god plays as active a role
in directing the path of his follower as, for example, Apollo
does in Euripides’ Ion or Dionysus in the Bacchae.
At this point in the Praef.Evang. there is no intervening discus-
sion of Demetrius or any other version. But there is an intriguing
oddity. After line 192, Polyhistor refers for the only time to t“
drãmati t“ §pigrafom°nƒ ÉEjagvgÆ (“the play called the
Exagoge”: 9.29.14). All his other references are either to t ª
tragƒd¤& (9.28.3) or simply to tª ÉEjagvgª (9.29.4, 12, 15).
Thus, I suggest that the third play, which in fact dealt with the
actual Exodus, was called Exagoge, and that this name, for
some reason, also was applied, confusingly, to the tetralogy as
a whole.18 It does appear that Polyhistor is interested in making
a distinction between the preceding lines and the following ones.
The logical assumption is that they come from different plays.
The third play would be similar in structure to Aeschylus’
17 On the common use of dreams in Greek tragedy see Holladay 437 and Erich
Gruen, Heritage and Hellenism (Berkeley 1998) 130–131.
18 Clement also refers to his source as t“ §pigrafom°nƒ drãmati ÉEjagvgÆ
(Strom. 1.155). But since he provides lines only from the first play it is likely
that his copy of the work contained only the first play, labeled with the
tetralogy name, which he sensibly mistook for the name of the first tragedy.
 THOMAS D. KOHN 11

Persae.19 Accompanied by a lamenting Chorus of Egyptians, the

queen of Egypt would recount the meetings between Moses and
Pharaoh, the plagues, the decision to let the Jews go free, and
the subsequent decision to get them back. Then the Egyptian
soldier enters and delivers the extant fragment, reporting the
disaster at the Red Sea (193–242). Finally, the queen would be
left to mourn and lament, just as Atossa wept over the defeat
of Xerxes. One would be justified in imagining that Ezekiel
followed Aeschylus’ lead and let the ghost of the dead Pharaoh
make an appearance.
At this point, Polyhistor again pauses to discuss Demetrius
(Praep.Evang. 9.29.15). As before, this interruption would seem
intrusive and arbitrary, unless the soldier’s speech that precedes
and the scout’s speech that follows derive from different plays.
The final fragments, in which a Hebrew scout tells Moses
about the discovery of an oasis and a phoenix (243–269),
would then belong to a fourth drama, occupying the satyr play’s
position. Ezekiel was familiar enough with the works of
Euripides to know that the fourth play did not literally have to
involve satyrs, 20 but it should be different in tone from the
previous three. The relief of the Jews, having escaped from bond-
age and arrived safely at an oasis, together with the strange
account of the mystical bird, would suitably lighten the mood.21

19 Several commentators on the Exagoge have noted similarities between this

scene and the Persae: Holladay 494ff; Jacobson 136–140; Bruno Snell, “Eze-
chiels Moses-Drama,” AuA 13 (1967) 150–164; J. Wieneke, Ezechielis Iudaei
poetae Alexandrini fabulae quae inscribitura ÉEjagvgÆ fragmenta (Munich
1931) 93–94.
20 Ezekiel's familiarity with Euripides and other classical Greek authors is
evident in the frequent references to them in his verse; such references are noted
by Jacobson and Holladay (see especially 303), where they occur.
21 Although the fragment does not use the word “phoenix,” nor contain the
most obvious attributes of the phoenix (its miraculous death by flames and
rebirth from its own ashes, or its set cycle of recurrence), the physical de-
scription of the bird, its gait, and its being the “king of birds” is enough to
confirm Polyhistor’s identification (1.29.15). See R. van den Broek, The Myth of
the Phoenix according to Classical and Early Christian Traditions (Leiden
(1972).
12 THE TRAGEDIES OF EZEKIEL

This division into four plays fulfills a number of functions.

First, it explains why Polyhistor and Clement refer to Ezekiel as
the “maker of [plural] tragedies.” Second, it divides the story of
the Exodus into discrete actions. Third, it alleviates the
objectionable violations of Time and Space. Each play can now
take place within sunrise and sunset of a single day; and only
the second play requires any change of setting, from the foot of
Mt Horeb to the top and back down again. Fourth, it allows for
the leisurely development of stories, characters, and ideas, as
well as giving room for choral passages, none of which would fit
comfortably in the rapid shifts across times and places required
if all the fragments came from one play. Finally, it fits with the
way Alexander Polyhistor presents the material, interrupting his
discussion of the fragments at places which coincide with
logical breaks between plays. Since he explicitly names the third
play Exodus, we can guess that the first play might be named
The Daughters of Raguel and the second The Shepherds of Libya,
after their respective Choruses, while the fourth could be The
Phoenix, The Oasis, or even The Scouts. But whatever their
names, it is sensible to think that the fragments of Ezekiel’s
work come not from a single play, but from a tetralogy.22

November, 2002 Classics Department

University of Mississippi
University, MS 38677
[email protected]

22 Versions of this paper were presented at “Crossing the Stages: The

Production, Performance and Reception of Ancient Theater,” University of
Saskatchewan, 23 October 1997, and at the APA Annual Meeting,
Philadelphia, 5 January 2002. Thanks to those who heard these presentations
and made comments. Special thanks also to Professors Erich Gruen, Nita
Krevans, Robert Sonkowsky, and Michael Gleason for reading previous
drafts. Finally, extra special thanks to the anonymous reader at GRBS for
making invaluable suggestions.