The Two-Act Structure: A Narrative Device in Akkadian Epics

In: Journal of Ancient Near Eastern Religions
Author: Sophus Helle1
View More View Less
  • 1 Aarhus University, Department of Comparative Literature, Aarhus, Denmark
Full Access


Most Akkadian epics are organized according to the same structure: the narrative arc is divided into two acts, of which the second mirrors and expands the first. The structure has already been observed in Atra-hasis, Enuma Elish, Gilgamesh, and Etana, but the recurrence of the pattern has not previously been noted. The essay explores the widespread application, individual adaptations, and literary significance of this device, noting its presence in nine cuneiform compositions.


Most Akkadian epics are organized according to the same structure: the narrative arc is divided into two acts, of which the second mirrors and expands the first. The structure has already been observed in Atra-hasis, Enuma Elish, Gilgamesh, and Etana, but the recurrence of the pattern has not previously been noted. The essay explores the widespread application, individual adaptations, and literary significance of this device, noting its presence in nine cuneiform compositions.

1 Introduction1

Akkadian literature is a treasure trove of stories about the gods, goddesses, and mythical figures of ancient Iraq. But when studying myths and religious narratives that survive in the form of literary “epics,” one has to reckon with the process by which the stories of gods and heroes were reworked to fit the demands of literary expectation.2 This process of creative adaptation transformed the mythical material into the format of narrative poetry, replete with the templates and clichés that appealed to ancient audiences. This article seeks to explore one such literary template, that is, one narrative structure that determined how Babylonian religious narratives were ultimately told.

I will argue that the narratives of most Akkadian epics are structured according to a specific pattern. Epics such as Atra-hasis, Enuma Elish, Gilgamesh, and Etana are divided into two acts, of which the second mirrors and extends the first. Atra-hasis, for example, consists of a first act that describes the creation of humanity, and then of a longer act that describes humanity’s near-total destruction. Likewise, in Enuma Elish, the god Ea’s battle against Apsû is repeated and enlarged in his son’s Marduk’s battle against Tiamat. The epics can thus be divided into two mirroring parts, and I have assigned the term “acts” to these parts because I view them as closely connected but also self-standing narrative entities. I propose that a similar division into mirroring acts can be found in at least four and up to six Akkadian epics, with Anzu being the clearest exception.

Intriguingly, the expansion of the second act is consistent with other known aspects of Akkadian poetry. In his doctoral dissertation from 1916, Hans Ehelolf proposed what is known as the Akkadian Wortfolgeprinzip, according to which words are grouped together in order of length.3 Ehelolf began by examining lists of individual words – both in lexical lists, incantations, and royal inscriptions – finding that the shortest words consistently appear first. He then broadened his investigation to look also at syntactical sequences in which it is not the length of the single words that determines its position but that of the larger grammatical units. More broadly still, Niek Veldhuis claimed that “[g]enerally, in Akkadian poetic language, the second of a pair will be longer than the first.”4 This principle is far more readily apparent at the level of stanzas, such as in word pairs or verse couplets. But, perhaps, the proposed pattern may be merely one more instance of the same principle, just working on a much larger compositional scale.

Such a pattern, recurring as it does across multiple works of literature from different periods, is best understood as a literary device: a structure that is employed, consciously or otherwise, in a number of compositions, gaining new meaning in each of them. By viewing this recurrent structure as a narrative device, I am taking it to be neither universal nor unique to the Akkadian epics, but merely prevalent among them. The pattern does not recur in every Akkadian epic (as mentioned, Anzu is a clear exception), nor is it restricted to Akkadian epics only. In short, the stories of gods, such as Ea and Marduk, or heroes, such as Gilgamesh and Etana, were told employing a range of artistic devices, and this article proposes that one of these devices was the bisection of the plot and the application of a generalized Wortfolgeprinzip to the resultant narrative couplet.

1.1 Overview

In seeking to establish a recurrent pattern across works of literature, one is faced with the question of whether the proposed pattern is really there, or whether it is simply an effect of the common human tendency towards spurious pattern recognition and confirmation bias.5 The epics discussed here are complex compositions, whose narratives consist of innumerable episodes, characters, and interlocking structures. To see one particular pattern recurring in each of them might well be an instance of what is known as pareidolia: the seeing of patterns where none exists.

As a safeguard against this tendency, if only a partial one, I begin my analysis with four epics for which the structure under discussion is well-known or has at least already been noted by other Assyriologists. For these epics, I confine myself to pointing out that the narrative structures noted by William Moran in Atra-hasis, A. Leo Oppenheim in Enuma Elish, Herman Vanstiphout in Gilgamesh, and Michael Haul in Etana are not isolated instances, but rather share a number of specific similarities. Only then do I go on to discuss two narratives for which this pattern has not previously been noted, namely Nergal and Ereshkigal and Erra and Ishum. By basing my suggestion on the arguments of established scholars, I hope to lessen the subjective bias.

As a further safeguard against bias, I begin by establishing three specific criteria for identifying the proposed device. For each epic, I seek to establish the following:

  1. That its narrative can be divided into two distinct parts, which I refer to as “acts.” The two need not be completely unconnected (elements from one act often carry over to the other), but they should each contain a separate conflict, climax, and resolution.

  2. That the second act expands the first. This can be achieved in different ways: in some epics, the second act is simply longer than the first; in others, the expansion is not quantitative, but qualitative – the scope of the story is expanded.

  3. That there is some form of mirroring between the two acts. In short, the two acts are not disjointed stories that have been arbitrarily combined to form an epic but are bound together by a narrative relation that makes the whole more than the sum of its parts. I note three kinds of relations between the acts: a contrast, where aspects of the first act are reversed in the second; a parallelism, where the two acts follow the same narrative sequence; and some combination of the two.

Finally, besides the six epics that I believe meet these criteria, I discuss two, Anzu and Adapa, that appear not to do so. Through such counterexamples, I hope to show that the criteria are not so general as to include all works of literature, or even all Babylonian epics, making the similarities between the other epics more significant.

2 Previously Noted Instances

In the following section I present four epics for which this structure has already been noted in the secondary literature: Atra-hasis, Enuma Elish, Gilgamesh, and Etana.

2.1 Atra-hasis

The Old Babylonian Version of Atra-hasis is perhaps the clearest example of the proposed “two-act structure.”6 Its division into two parts was made particularly clear in William Moran’s 1987 study of the epic, in which he argued that Atra-hasis should be seen as “two stories, two myths. The first, the Igigu-myth, is both a story in itself as well as the preface to the second, the Deluge-myth.”7 This separation has become a commonly accepted feature of the epic. Dahlia Shehata, for example, introduces the text as follows: “Die altbabylonische Erzählung von der Sintflut kann in zwei Großteile gegliedert werden: Menschenschöpfung und Sintflutgeschichte.”8 Likewise Helge Kvanvig writes of Atra-hasis that “[a]s a literary composition the poem clearly falls into two parts,” while Claus Wilcke states that “[d]er Dichter des Atram-hasīs-Epos koppelt die Sinflutgeschichte an einen zweiten Mythos, den von der Erschaffung der Mennschheit.”9

Each of the parts constitutes an independent mythical tradition. The creation of humanity has strong parallels in Enuma Elish, which does not include an account of the Flood.10 Conversely, the story of the Flood is also found in Gilgamesh, which does not include the creation of humanity (though it does allude to it, through the creation of Enkidu). There is some inconsistency as to where exactly the dividing line between the acts is drawn,11 but one way or the other, it remains clear that Atra-hasis consists of two recognizably separate parts. Just as clear is the expansion of the second act, which is longer than the first. Depending on where the line is drawn, the first act consists of either 320 or 352 lines and the second of either 893 or 925. Whatever the case, the second act is slightly less than three times the length of the first.

What then of the mirroring between them? In Atra-hasis the relation between the acts is most often described as a contrast. Simply put, the first act concerns the creation of humanity and the second its all but total destruction. Kvanvig further points to a more general contrast between creation and destruction:

What the gods destroyed in the flood was not only the human race, but their creation of the earth. The “end” therefore stands in contrast, not only to the creation of humans, but to the initial work carried out by the gods. The flood story is thus in contrast with the first part of the poem. Creation and flood is [sic] placed as the ultimate opposition…. The flood stands as an antithesis to the creation recorded at the beginning of the poem.12

The contrast between the creation of humanity and its subsequent destruction in the Flood has been noted for its Biblical parallels: Hans-Peter Müller, for example, has drawn attention to the “Komplementarität von Menschenschöpfung und Flut,” as this complementarity is at work in both Genesis and Atra-hasis.13 Müller views the stories of creation and Flood as “Mythos” and “Anti-Mythos,” respectively, conveying two contradictory messages to the human audience:

[W]ährend der Schöpfungsmythos das Dasein des Menschen in seiner Welt garantieren will, indem er aufzeigt, warum das, was da ist, da sein darf, will der Anti-Mythos von der Flut hinwegbannen, was dieses Dasein bedroht und seine Rechtfertigung in Frage stellt.14

In Müller’s summary, the contrast between the acts is therefore not only one of creation versus destruction; more fundamentally, it concerns the justification of human existence that is first established and then withdrawn.

Further support for the antithetical structure of the story can be found in the ironic repetition of lines from the first act in the second, where they take on a significance completely contrary to their original meaning.15 When Belet-ili is called to create humanity in the culmination of the first act, the other gods say to her: “You are the birth-goddess, creator of humanity!” But in the culmination of the second act, they say: “You are the birth-goddess, creator of destinies!”16 As noted by Lambert, the second address is to be interpreted as a call for Belet-ili to impose natural death on the previously long-lived humans, as the word “destiny” (šīmtu) is a common Akkadian euphemism for death. The two addresses to Belet-ili thus employ the exact same phrase when asking her to create human life and human death, respectively.17

To summarize, Atra-hasis is clearly divided into two acts, with the second act just as clearly expanding the first, as it is almost three times as long. Further, there is a central contrast between the theme of creation in the first act and the theme of destruction in the second.

2.2 Enuma Elish

Enuma Elish is a complex composition, consisting of a narrative section, a hymnic enumeration of the names of Marduk, and a short epilogue. It is the narrative section that I examine here, as it can be further subdivided into two parts: first the creation of the gods and Ea’s battle against Apsû, then Marduk’s birth and his battle against Tiamat. This subdivision, which Vanstiphout calls “the ‘double’ nature of the narrative,” was first recognized by Oppenheim:

From the literary point of view, it should be noted that the Ea story as told in Enûma Eliš in its condensed form is basilically [sic] a doublette of the Marduk story. It is used, however, very skillfully by the poet as an overture which touches briefly upon the main “Leitmotifs” and thus instills curiosity, and prepares the listener for what is to come.18

Likewise, Dina Katz argued that the “two sections of the text are structurally and thematically parallel,” pointing out that both narrative sequences end with the same line: “After he had bound and slain his enemies …” The line refers first to Ea’s killing of Apsû and then to Marduk’s killing of Tiamat.19 The parallelism between the stories has recently been explored in further depth by Gösta Gabriel.20 As analyzed by him, both acts consist of five parts: (1) a story of divine origin, (2) the disturbance of primordial beings and the plotting of deicide, (3) the countermeasures of the Anshar-gods, (4) the creation of (a part of) the world, and finally (5) the construction of a divine dwelling. The strict parallelism between the two sequences is made clear by a set of lexical equivalences between each part:

  1. The creation of the gods parallels the birth of Marduk, both as to the verb employed (the N-stem of banû) and the place they occur (within the Apsû).21

  2. The disturbances of the primordial beings revolve around a contrast between ṣalālu, “to sleep” (I 23, 108, 116, 119), and dalāḫu, “to disturb” (I 38, 40, 50, 116, 122).

  3. Both combats end with the same line, as noted above: “After he had bound and slain his enemies …” (I 73 and IV 123).

  4. Ea’s creation of the Apsû and Marduk’s creation of the universe employ the verb kânu, “to make firm” (I 71 and IV 144, V 8, 62, 66).

  5. Ea’s construction of a home inside Apsû and Marduk’s construction of Babylon also share a number of lexical parallels. Both Babylon and Ea’s home are referred to as that god’s “dwelling” and “chamber” (šubtu in lines I 71 and V 122, kummu in lines I 75 and V 124). And both acts of construction employ the same verb: ušaršid-ma and lušaršid-ma, “he founded” and “I will found” (I 77 and V 123). In fact, the epic specifically refers to the Esagila, Marduk’s temple in Babylon, as the meḫret Apsî, the “replica of the Apsû” (VI 62), clearly identifying the two dwellings as each other’s mirror image.

Further, both acts end with a specific kind of naming. As Ea builds his home, he names that part of the world “Apsû,” to which is added the brief comment, “whose shrines he revealed.”22 As noted by Jean-Marie Durand, the comment is a midrashic interpretation of the word “Apsû,” written ZU.AB, according to the techniques of Babylonian hermeneutics: ZU is read as “to know,” here in the sense “reveal,” while AB is read as eš3, “sanctuary.”23 Gabriel links Ea’s naming of Apsû to Marduk’s naming of Babylon (V 129). But the naming of Apsû finds a much larger parallel in the climatic list of Marduk’s names and the associated epithets, each of which, according to a Babylonian commentary text, can be read as an interpretation of the apposite name following the same hermeneutical techniques by which “Apsû” was linked with the recognition of sanctuaries.24

The two acts thus run closely parallel. While Gabriel’s study finds also a number of important differences between them, I would note that in case after case, those differences amount to a narrative expansion in the second act.25 For example, Gabriel notes that the “größte Differenz zwischen dem ersten und zweiten Konflikt ist die Schilderung der umfangreichen Rüstungsmaßnahmen Tiāmtus,” that is, the long description of the army amassed by Tiamat.26 This is an expansion, not only in the line count, but in scale and narrative tension: the threat posed by Tiamat is much greater than that posed by Apsû. Likewise, Gabriel remarks that: “Bei der Gegenaktion der Anšar-Götter weicht der zweite Konflikt am umfangreichsten von dem ersten Konflikt ab.” Again, the difference between the acts is one of expansion. Finally, Gabriel notes a contrast between Ea’s creation of the “ersten definierten Weltteil,” namely the Apsû, and Marduk’s “allumfassende raum-zeitliche Ordnung.”27 Once more, the difference constitutes an enlargement of the first act: from the creation of one defined part of the world to a universal definition of cosmic order.

Further, Gabriel points out that this parallelism is only one among many narrative structures at work in this composition, including, for example, a number of overlapping circular structures that he detects in the epic.28 However, this does not take away from the present argument: while the narrative parallelism in Enuma Elish can be identified as an instance of a recurrent literary device, this is not the only structure at work in the epic. All the epics examined here are complex, multilayered compositions that employ many other patterns beside the two-act structure. Further, those structures may interact with one another. According to Gabriel, the narrative parallelism in Enuma Elish is joined by a “linear structure,” namely the exaltation of Marduk over the course of the epic. As he notes: “In die Parallelstruktur wird so nun mit Marduks Erhöhung ein neuer, linearer Erzählstrang eingefügt, der in die parallele Erzählung über Konflikt und Schöpfung eingeschrieben wird.”29 The introduction of this new theme in the second act would seem to deviate from the two-act structure, as the parallelism is now no longer exact. Yet, as Gabriel also makes clear, this exaltation is primarily expressed through an expansion of themes from the first act, as with the contrast between Ea’s limited creation of a part of the world and Marduk’s establishment of universal order. The linear theme of Marduk’s rise to power does not counteract the two-act structure but is rather reinforced by it.

The narrative section of Enuma Elish can thus be divided into two acts, and it is clear that there is a close parallelism between them. It is also clear that the second act expands the first, being about ten times longer.30 But whereas the relation between acts in Atra-hasis takes the form of a contrast between creation and destruction, in Enuma Elish one sees rather a strict parallelism of narrative sequences and lexical equivalences.

2.3 Gilgamesh

The division into two acts is perhaps less readily apparent in the case of Gilgamesh than in Atra-hasis or Enuma Elish (in the following, I limit my discussion to the SB version of Gilgamesh, since the plots of the OB and the MB versions are still difficult to reconstruct). But the case for such a division has been made by a number of scholars, most notably Andrew George, who argues that the narrative’s “division into eleven tablets is itself symmetrical,”31 with the first part describing the two great successes of Gilgamesh (the killing of Humbaba and the Bull of Heaven) and the second his two great failures (falling short of immortality and losing the Plant of Youth). Likewise, Vanstiphout’s article on the narrative structures of Gilgamesh includes a section on what he terms the “symmetry” of the epic, in which he argues that the story can be divided into five parts:32

  • Part A (tablets I and II) concerns Gilgamesh’s encounter with Enkidu.

  • Part B (tablets III through V) tells of the heroes’ journey to the Cedar Forest.

  • Part C (tablet VI) narrates the episode of Ishtar and the Bull of Heaven.

  • Part D (tablets VII and VIII) tells of Enkidu’s death.

  • Part E (tablets IX through XI) concerns Gilgamesh’s journey to Uta-napishti.

According to Vanstiphout, part C is the pivotal episode around which the parallelism of the epic turns. Taking tablet VI as the center of the story, “the parallelism or symmetry in the formal distribution between a ‘first part’ [A + B] and a ‘second part’ [D + E] becomes manifest.”33 The story of Gilgamesh losing Enkidu thus mirrors the story of Gilgamesh meeting Enkidu, while Gilgamesh’s journey to Uta-napishti mirrors his journey to Humbaba. Similar arguments have been made by other scholars. Hope Nash Wolff, for example, notes that the epic “separates into two parts, each more concerned with one of the heroes,” in that the first act centers on Enkidu and the second on Gilgamesh. Laura Feldt and Ulla Koch argue that the epic “may be divided into two sections, namely before and after the death of Enkidu,” and further that these “two parts may be seen as structurally parallel.”34

In fact, it may even be the case that the prologue of the epic explicitly refers to the mirroring of Gilgamesh’s two journeys – to Humbaba and Uta-napishti, respectively. The Ugaritic version of the prologue notes that Gilgamesh “explored both daises.”35 The word here translated as “both,” mitḫāriš, derives from mitḫāru, which has a broad range of meanings that includes “corresponding” and “symmetrical.” The text thus states that Gilgamesh travelled to matching daises, parakkī, a metonym that is used to describe seats of power, which I identify as the dwellings of Humbaba and Uta-napishti. In his edition of the Ugaritic text, George doubts that these dwellings can truly be considered places of political power, as thus identifiable as parakkī.36 But in fact, both figures can plausibly be identified as kings or at least the rulers of the Cedar Forest and Shuruppak, respectively.37 If this reading of the line holds up, it would be a particularly explicit reference to the mirroring structure of the narrative in the very first stanza of the epic.

Keith Dickson has also drawn attention to another symmetry in the epic, namely that the epic “unfolds within the space between two trees.”38 In the first act the two heroes travel west until they reach a mythical forest, the Forest of Cedars; in the second act Gilgamesh, now alone, travels east and again reaches a mythical forest, the Forest of Jeweled Trees. But Dickson adds a caveat to this argument, as the symmetry is not quite exact: “The fact is that Gilgamesh, after emerging from the tunnel of the sun, enters the garden of jeweled trees and then passes beyond it.”39 Dickson therefore argues that the episode of Uta-napishti, which takes place after Gilgamesh’s arrival at the Jeweled Trees, is a later addition to the narrative. But this need not be so, for the story as preserved fits perfectly with the two-act structure.40 Gilgamesh’s eastward journey in the second act mirrors and expands his westward journey in the first act: he reaches but then goes beyond a mythical forest, where his first journey had ended.

The expansion one finds in Gilgamesh is not, as is in Atra-hasis and Enuma Elish, a quantitative extension of the line count. In fact, the two acts are almost exactly the same length: five tablets, separated by the pivotal tablet VI. The expansion is rather a qualitative extension. The cosmological scope of the epic is expanded in the second act as Gilgamesh goes beyond the Jeweled Trees and travels to world’s end. Likewise, Vanstiphout argues that beyond the symmetry he describes in Gilgamesh there “is also a manifest expansion or enlargement of narrative styles, though not in any quantitative sense.”41 Vanstiphout speaks of an accumulation of disappointments, in that Gilgamesh’s failures in the second act grow progressively more intense: he loses first his friend, then the dream of eternal life, and finally the concrete youth of his body. Vanstiphout also notes the increased formal complexity of the second journey, which consists of a sequence of small, independent episodes, as compared to the straightforward narrative of the first journey.

The Gilgamesh epic may thus be divided into two parts, the second of which expands the first, though in a qualitative rather than quantitative sense. What, then, of the mirroring between the acts? I have already noted some instances of symmetry – meeting and losing Enkidu, traveling to the Forest of Cedars and to the Jeweled Trees – but the mirroring is particularly clear at the thematic level. Giuseppe Furlani opposed what he saw as an exaggerated focus on the theme of death and immortality in the secondary literature on Gilgamesh, emphasizing instead the theme of friendship.42 One may argue that each of the two themes dominates one act of the epic. The first dwells on the love between Gilgamesh and Enkidu and the second on Gilgamesh’s loss of Enkidu. Friendship and grief are presented as each other’s emotional mirror image: Enkidu becomes “like a husband” in the first act, while Gilgamesh veils his corpse “like a bride” in the second, connecting love and death in a neat reversal.43

This distinction between a first act focused on love and success and a second act focused on loss and tragedy is confirmed by the way the epic marks its own boundaries. As has often been noted in the secondary literature, the Walls of Uruk, which are described in both the prologue and the final lines of the epic, serve to delimit and encircle the text, opening and closing the story.44 Beside these two matching scenes, the walls make only a single other appearance in the preserved text. When the heroes have defeated the Bull of Heaven, Ishtar goes up on the Walls of Uruk and cries: “Woe to Gilgamesh!” (VI 151–53). All too soon, Ishtar’s outcry becomes a reality, as woe indeed befalls Gilgamesh in the form of Enkidu’s death. Besides the outer limits of the text, the Walls of Uruk thus also mark the internal limit between the two acts and the emotional transition from glory to tragedy.

2.4 Etana

If the division into two acts is less than obvious in Gilgamesh, it is if anything too obvious in Etana. The epic begins with a mythological prologue describing the invention of kingship, and the remainder of the narrative consists of two clearly separate stories: the first is about the betrayed friendship between a snake and an eagle, and the second tells of how Etana flew to the heavens on the back of that eagle. As summarized by Geoffrey S. Kirk, Etana “consists of two lightly linked stages and a prelude”; likewise, Lorenzo Verderame states that “è possibile ricostruire due unità narrative che corrono parallele.”45 That the epic consists of two parts thus seems incontestable.

It is likewise undeniable that the second act is the more prominent of the two, though it is unclear whether it is also the longest. Etana is preserved in three recensions – OB, MA, and SB – all of which are fragmentary and differ substantially from one another. In the following, I concentrate on the SB version, which is the best preserved, but even for this version one cannot establish a line count for the two acts; there is even debate about the number of tablets it comprised.46 Yet despite such uncertainty, scholars assume that the second act takes narrative precedence: as its modern name indicates, the epic is generally considered to be the story of Etana’s flight to heaven, onto which the story of the eagle and the snake was grafted.47

While the division into acts and the prominence of the second are generally accepted, the relation between them is an issue of considerable doubt. The two seem rather freely strung together; and as noted by the editor of the epic, Michael Haul, the looseness of this connection is generally considered to be a problem:

Ein wesentliches und bis heute ungelöstes Problem in der Interpretation des Etana-Epos stellt seine zweigliedrige Gestalt dar. […] Die missliche Lage des Adlers liefert die Begründung dafür, warum der Adler nach seiner Rettung durch Etana zum Himmel emporträgt. Sie verbindet so beide Erzählungen miteinander. Ansonsten jedoch scheinen sie gänzlich unmotiviert nebeneinander zu stehen. Oder entsprach ihrer Zusammenfügung eine tiefere mythische Bedeutung, die uns entgangen ist?48

Numerous suggestions have been made as to the content of this “deeper mythical significance,” but none has so far gained much traction among Assyriologists.49 In light of the proposed two-act structure, two points from the scholarly discussion about Etana’s coherence bear mention: the contrast between up and down, and the parallelism between two scenes of friendship. The relations between the acts of Atra-hasis, Enuma Elish, and Gilgamesh lead one to expect some combination of contrasts and parallels in Etana as well, and that is exactly what one finds.

The most obvious contrast is up versus down. The first act ends with the snake casting the eagle down into a ditch, while the second ends with the eagle carrying Etana far up into the heavens. Not only do the two actions constitute a reversal of direction, they also exhibit a qualitative extension of scale, not unlike the one observed in Gilgamesh: the heavens to which the eagle flies are far higher than the ditch is deep. Kirk elaborates on this contrast, arguing that the eagle and the snake live at the “polar extremes” of the tree – its crown and roots, respectively. This opposition is developed further in the second act, as the eagle is thrown down and then “rises so high in the air, even above the abode of the sky-god, that he begins to crash earthward once more.”50 According to Kirk, this sequence of reversals has an important symbolic layer. The eagle’s betrayal of the snake, “itself a reversal of Shamash’s rule of law, has resulted in a complete reversal of the natural relations between eagles and snakes.”51 This imbalance is rectified when the eagle first rises up into the heavens and then back to earth.

However, there is also a fundamental parallelism at work between the acts. Haul builds on Kirk’s argument, proposing a more general cosmological coherence between the stories:

Es ergibt sich die spiegelbildliche Verschränkung beider Erzählungen: Etana behebt die kosmische Zerrüttung, indem er den Adler rettet und ihm die Rückkehr an den ihm zugewiesenen Ort in der Baumkrone (bzw. in den Himmel) ermöglicht, während der Adler die Zerrüttung der irdischen Ordnung des Königtums behebt, indem er Etana zu einem Nachkommen verhilft.52

The two acts thus mirror each other through their respective restorations of order. But there is also another parallel between them, namely, the theme of friendship: both acts concern the relation between the eagle and an ally. Jan Dietrich has recently noted the similarity between the eagle’s address to the snake and its address to Etana. To the snake, it exclaims: “Come, let us make a friendship! Let us be friends, you and I!” And to Etana, it says: “My friend! Let us be friends, you and I!”53 It is worth considering whether this parallel means that the two acts had similar outcomes as well. Does the eagle betray Etana as it had betrayed the snake? Or does the epic rather hinge on a contrast between its betrayal in the first act and its reformed loyalty in the second? Until the ending of the epic is recovered, there is no way to know for sure.

3 Additional Instances

Having established that similar structures have been observed in Atra-hasis, Enuma Elish, Gilgamesh, and Etana, I now move on to two epics in which the structure has not previously been noted: Nergal and Ereshkigal and Erra and Ishum. It must be said from the outset that both are very difficult texts, and their fragmentary states of preservation mean that the plots cannot be restored with certainty. Additional complications, such as the difference between the preserved versions of Nergal and Ereshkigal and the innovative style of Erra and Ishum, present further problems still. However, I argue that both narratives revolve around a pair of pivotal events, and that those events may be related to one another according to the logic of the two-act structure.

3.1 Nergal and Ereshkigal

One difficulty in analyzing Nergal and Ereshkigal is the significant difference between the two versions in which it is preserved: the MB version from Amarna and the SB version from Sultantepe and Uruk. The versions deviate at a number of crucial junctures. Here I deal only with the SB version, leaving aside the question of whether the Amarna version also employed the two-act structure. Further, I follow the reconstruction of the fragmentary plot presented by Benjamin Foster.54

The division of the narrative into two acts is most clearly apparent in Nergal’s two descents to the netherworld. The story is structured around a polarity between the heavens and the netherworld, and the journeys back and forth between these places, first by the messengers Kakka and Namtar and then by Nergal himself, form the bulk of the narrative. As the text makes clear, crossing back and forth between the heavens and the netherworld is a momentous decision for all but the messenger gods: Ereshkigal is not allowed to leave her realm, and Nergal risks becoming trapped there just by entering it. As such, his descents are moments of crucial importance in the cosmology of the text.

There is also a clear parallelism between Nergal’s two descents. Upon arriving to Ereshkigal’s realm, he is said both times to “enter her wide courtyard,” whereupon the two deities “embrace one another and go passionately to the bedroom.”55 The narrative of Nergal and Ereshkigal thus revolves around two strikingly similar scenes, allowing us to divide it into two halves. In the first act, Ereshkigal’s messenger Namtar goes up to the heavens and is not properly greeted by Nergal, who is therefore forced to descend to the netherworld to make amends. Here, he makes love with Ereshkigal, but then slips away. In the second act, Namtar again ascends to the heavens looking for the disguised Nergal. Again, Nergal is somehow forced to descend into the netherworld and again makes love with Ereshkigal. As in Enuma Elish, the two acts mirror each other through an exact repetition of the narrative sequence.

But there is also a crucial difference between the acts. In the first, the two gods make love for six days, but then Nergal slips away while Ereshkigal sleeps. In the second, they make love for seven days, and it seems that by staying a full week in the netherworld, Nergal is somehow bound to it, becoming Ereshkigal’s husband. The episode of passionate sex is thus repeated but also extended from six to seven days and its outcome is thereby changed. Again, as in Enuma Elish, this “repetition with a difference” is highly significant. While the extension of the second act in Enuma Elish signaled Marduk’s elevation to king of the gods, the extension of the sexual encounter in Nergal and Ereshkigal seals Nergal’s fate as king of the netherworld. If the epic can be said to employ the two-act structure, then it is a highly creative use of the device.

Another difference between the acts involves Ereshkigal’s summons. Ereshkigal twice demands that the heavenly gods send Nergal down to her but does so for different reasons in each instance. Here, however, one must contend with challenges stemming from the divergent versions, since Ereshkigal’s first summon is not preserved in the SB manuscripts. If one cautiously assumes that this episode can be restored from the Amarna version (though note Reiner 1985: 58), one finds a neat reversal. Affronted at Nergal’s lack of respect, Ereshkigal first demands: “Bring him before me! Let me kill him!” Then, infatuated with him after their all but weeklong lovemaking, she exclaims: “Send me [that god]! Let him be my husband, let him spend the night with me!”56 Though the wordings of the summons are similar, the intentions behind them could not be more different. Jean Bottéro has suggested that the whole epic revolves around a pun: Ereshkigal summons Nergal to put him to death (mūtu) but receives him as a husband (mutu).57

The similarity between the summons may further strengthen the suggestion that the narrative mirroring of love and death is at the heart of the epic. Though it is a difficult composition in many ways, Nergal and Ereshkigal may tentatively be read as being structured around a parallelism between two sequences: Ereshkigal summons Nergal, he descends to the netherworld, and the gods make love. The relation between the acts amounts to a “repetition with a difference,” as Ereshkigal intends first to kill him and then to marry Nergal and as their sexual encounter is extended from six to seven days.

3.2 Erra and Ishum

Erra and Ishum also presents a number of difficulties, most notably because of an innovative style that sets it apart from almost all other Akkadian literature.58 One key problem in establishing the plot of Erra is the relation between the episode of Marduk’s statue being renewed and Erra’s subsequent rampage. It has often assumed that Erra tricks Marduk into abandoning his throne so as to usurp the kingship and unleash chaos in his absence – the need for Marduk to refurbish his statue being nothing more than a ruse invented by Erra. However, this direct causal connection between the two episodes has been called into question on the basis of l. II 49, which reads “[…] he established his residence.”59 According to Foster, this refers to Marduk returning to his throne after his statue was refurbished, and so in his translation, Marduk is said to have “reoccupied his dwelling.”60 However, if Marduk does indeed return to the throne after his statue is renewed and before Erra runs amok, the latter cannot be a direct consequence of Marduk’s absence. Foster therefore summarizes the plot as follows:

Erra does such an excellent job of guarding in Marduk’s absence that the universe continues to operate as it should. Erra is praised by the gods and is expected to go quietly home when Marduk returns. Resentful and disappointed that there has been no violence, even though this is the result of his own valor, Erra, resolves to go on a rampage, alleging that he has not been given sufficient respect.61

George likewise presents a carefully hedged reading:

Marduk consents to Erra’s plan but only when Erra promises to maintain order in his absence. When Marduk returns Erra takes offence, either at some slight or because he has been tricked out of the opportunity to use his power.62

But if Erra does not usurp power in Marduk’s absence, how then are we to understand the relation between the episodes? What is the purpose of the first episode if it is not a pretext for Erra’s rampage? I would suggest that they can be read as separate disturbances of order and, hence, as two contrasting acts. After a prologue introducing Erra’s lust for war, the epic would consist of one act in which Marduk refurbishes his statue and then returns to his throne and a second act in which Erra goes on a rampage against the entire world before finally being soothed by his counsellor Ishum. Crucially, both acts concern a deviation from cosmic order. Marduk warns that if he leaves the throne, evil winds will rise and light will become dark, and Erra assures him that he himself will guard against the forces of chaos that will inevitably attempt to seize their chance in Marduk’s absence. As for Erra’s rampage, there is no question that it can be considered a disturbance of order: he overturns all rules and destroys all that there is to destroy.

Machinist has argued that Erra and Ishum fundamentally concerns the relation between order and disorder, rest and violence.63 This interplay is only made clearer by my suggestion, according to which the epic consists of two such cycles of order and disorder, violence and relaxation. Further, George has argued that the epic creates a distinction between necessary and unnecessary violence: this is likewise supported by my division of the story into two acts, each of which is concerned with a very different kind of violence.64 Both episodes involve a cleansing destruction and an eventual renewal. When Erra convinces Marduk to leave his throne, he promises that he will keep the order “while you enter that house, fire cleanses your apparel and you return to your place.”65 But the sequence he describes is in fact a miniature version of his own rampage: interruption of order, cleansing fire, and return to order. When Erra does calm down, he commands the restoration of Babylon’s might and blesses all who respect his sovereignty.

Viewing both episodes as interruptions of order also highlights the difference between them. Thanks to Erra, the disturbance occasioned by Marduk’s absence is kept in check, while the disturbance he himself brings about is unboundedly chaotic. The destruction and restoration of Marduk’s apparel is a strictly local event, confined to the ritual house where the apparel can safely be burned and remade. Indeed, the entrance to the house is fiercely guarded by Erra himself. But when Erra goes on his killing spree, the destruction he causes takes place on a truly global scale. Not only is the second act longer than the first,66 it also expands the cycle of destruction and renewal from a single house to the entire world.

Erra and Ishum is a challenging text, and my suggestion that it employs the two-act structure is a very tentative one. But I believe that the two-act structure would at least explain why the story of Marduk’s statue is included in the epic. Following the logic of this device, it serves as a miniature mirror image to Erra’s rampage, highlighting through contrast the latter’s chaotic and global nature.

4 Summary of the Narrative Device

I have proposed that the narratives of the six epics analyzed above share a set of characteristics that can be gathered under the label of “the two-act structure.” In the table below, I have summarized the main findings for each epic, showing that they all fulfil the three criteria set out in the introduction: a division into two acts, an expansion of the second act, and a mirroring between the acts. But even if this interpretation is correct, a number of questions remain. Why employ this narrative device? What literary effects are achieved by the division of the story into two parts, and how do the patterns of mirroring and expansion fit into each composition?

One effect of the two-act structure is that the first act foreshadows the second, presenting a miniature version of the drama that will then be played out on a larger scale. This is especially the case for Enuma Elish, in which the episode of Ea’s battle against Apsû “instills curiosity, and prepares the listener for what is to come.”67 This form of foreshadowing necessitates a dynamic expansion of the second act – otherwise it would be merely a repetition of the same narrative sequence. But as a miniature version, the first act lays out a simple pattern for the second act to embellish and complicate. This device – a simplified version of the story nested within the story itself – is reminiscent of the mise en abyme from later traditions, such as the “plays within a play” found in Shakespeare’s Hamlet and Midsummer Night’s Dream.

But if the relation between acts is one of contrast rather than repetition, the structure produces an altogether different effect. Here one sees what Jonathan Culler terms a “pattern of aletheic reversal,” whereby the contrast serves to expose a fundamental truth.68 The reversal between acts functions not only as a narrative element, but also as a revelatory statement about a contrast inhering in the world itself. The clearest example is once more Atra-hasis, whose acts contain opposite statements on the human condition. The first explains the necessity of human life, the second the necessity of human death. Those antithetical propositions illuminate each other to form a single “dramatic whole.”69 The reversals found in Gilgamesh could likewise be described as aletheic. The first act explores the wonders of friendship while the second turns to its unavoidably tragic consequences. Gilgamesh loves and loses; and through the reversals that link the acts together the epic confronts us with a fuller picture of the double-edged nature of human attachment.

Further, the simple opposition between acts comes to generate its own narrative impulses, which can be further elaborated and explored. This is certainly the case in Erra and Ishum, if the story has been analyzed correctly above. The small-scale burning and renewal of Marduk’s apparel is mirrored and massively expanded in Erra’s rampage. This second destruction is described in agonizing, horrific detail, extrapolating the original contrast between control and chaos ad infinitum, growing into a complex tale that far outpaces any symmetry.

Finally, in most of the epics one sees not only a parallelism or a contrast, but a combination of the two. Their interaction is best illustrated by Nergal and Ereshkigal, where reversal and repetition are cleverly combined to give the exact same sequence of events a completely different background and emotional tenor. Nergal is called first to his death and then to his wedding; he first slips away and is then bound to the netherworld. Contrast and parallelism are folded into each other to develop the same basic structure in opposite directions. Likewise, the expansion of the second act is not merely a formal element devoid of meaning but is often used for literary effect. In Enuma Elish, for example, the expansion highlights Marduk’s ascendency over the other gods. Marduk’s exploits thus not only repeat those of his father Ea but also surpass them at every turn. Marduk’s foe, Tiamat, is necessarily also an unspeakably greater threat than that faced by Ea, and whereas Ea’s creation was limited to a single part of the world, Marduk’s creation encompasses the entire universe. In short, the pattern of expansion is not applied unthinkingly, but is made to fit the specific purposes of each composition.

5 Exceptions and Uncertain Cases

I now turn to three compositions that do not employ the two-act structure (Anzu, Adapa, and Ishtar’s Descent) and three compositions that do seem to employ it, but which do not belong to the genre of Akkadian epics (Inana’s Descent, The Poor Man of Nippur, and The Cuthean Legend of Naram-Sin). Taking the two-act structure as a narrative device means that it is not a universal aspect of Akkadian literature, but rather a recurrent element that could either be employed or left out. The exceptions discussed below thus do not constitute a counterargument to my proposal; rather, they highlight the fact that the narrative device was just that, a dominant but not necessary pattern. Further, the fact that the three criteria set out above are not fulfilled by stories such as Anzu and Adapa shows that these criteria are sufficiently rigid: if all stories could fulfil them, the pattern would not be significant. Conversely, I do not hold the narrative device to be unique to the Akkadian epics either; the Sumerian version of Inana’s Descent certainly seems to fulfil the three criteria I have set out.70 A special case is The Poor Man of Nippur, which clearly abides by the two-act structure but does not treat mythical matters, meaning that it can be read as a satirical reuse of a pattern otherwise associated with “high” epic narratives. Likewise, the Cuthean Legend of Naram-Sin cannot be considered an epic in its form or a myth in its content, but again its use of the two-act structure is striking and innovative.

5.1 Anzu

It has often been noted that Enuma Elish incorporates and reworks elements from Ninurta’s battle in Anzu.71 But in Enuma Elish, this battle is further provided with a miniature doublet, and one finds no such doublet in Anzu, leading one to wonder whether this story can also be said to employ the two-act structure.

In her edition of Anzu, Marianne Vogelzang divides the story into not two but three parts: the Theft of, the Battles over, and the Retrieval of the Tablet of Destinies. Each of these parts is structured around an opposition: Anzu against Enlil, Ninurta against Anzu, and Ninurta against Enlil. Vogelzang pays close attention to the ways in which symmetry and repetition are used to structure the plot of the epic, but the parallels she finds do not divide the story neatly into two acts; rather, they are echoes running across all three parts of the story.72

At first sight, it would seem as if Anzu does not employ the two-act structure. But one might point to a two-fold division between the main conflicts of the epic: first Anzu holds the Tablet of Destinies, then Ninurta holds the Tablet of Destinies. Each represents an independent conflict and resolution – and thus possibly, by the definition set out in the introduction, its own act. One might further claim that these “acts” mirror each other in that they both deal with the same problem: the Tablet of Destinies is being held hostage. However, this reading would soon run into problems. First, it is not certain that Ninurta does indeed refuse to return the Tablet after conquering it from Anzu. The text is fragmentary at this point, and the interpretation of the conflict relies on a few broken sentences.73 At any rate, if indeed l. III 22’ can be restored as Ninurta’s claim that “I will not [return] the Tablet of Destinies,” such a “rebellion” hardly amounts to an equivalent of Anzu’s defiance. If anything, it is a much shorter and less dramatic echo of the epic’s core conflict, Ninurta’s battle with Anzu.74 The narrative structure thus seems to be the exact opposite of the expected expansion in the second act, so that even if the epic can be divided into two mirroring acts, it would still fall short of the second criterion set out in the introduction. One way or the other, Anzu cannot be said to employ the two-act structure.

5.2 Adapa

Adapa likewise cannot easily be divided into two acts. Drawing a comparison between the stories of Adapa and Etana, Kirk notes that the “Etana myth has a two-stage structure that the Adapa tale could not easily emulate.”75 If the two-stage structure mentioned by Kirk can be equated with the narrative device described here, one might conclude that Adapa does not employ it. One possibility for dividing Adapa into two acts would be to take the story of Adapa’s encounter with the South Wind as a separate first act, making the story of Adapa’s ascent to heaven a second act, and indeed, Michalowski refers to Adapa’s breaking of the South Wind’s wing as the “first climax” of the story.76 Since I argued in the introduction that what I consider to be “acts” are distinguished by having independent climaxes, Michalowski’s comment raises the possibility that this episode could conceivably be a separate act. In his study of Adapa, Shlomo Izreʾel emphasizes that, between his encounter with the South Wind and his ascent to the heavens, Adapa lies in the sea suspended between life and death.77 This “in-between” state could be understood as an uneasy balance at a narrative juncture, halfway between stories. There is also no question that, if the episode of the South Wind is taken as a separate section, then the second “act” would be much longer.

However, to see this initial episode as an act of its own would rather stretch the definition of the term. The episode is so simple and short that it does not truly bear comparison to the acts of, say, Gilgamesh and Atra-hasis. Among the stories discussed above, the least complex of the episodes I considered to be an independent “act” is the refurbishing of Marduk’s statue in Erra and Ishum, and even this episode accounts for some 150 lines – about a fifth of the epic.78 While the fragmentary state of Adapa makes it impossible to establish a comparable line count, there is little doubt that the episode of the South Wind is much shorter than that, even in relative terms.

Further, there is no obvious mirroring between the two episodes. Adapa’s encounter with the South Wind and his ascent to heaven are causally connected, as he goes up to Anu to atone for his misdeed. But the two episodes do not form a parallel or a contrast, at least none that is as clear to my mind as those noted for the other epics. Izreʾel’s study describes a number of structural relations and reversals, thoroughly combing the short tale for binary oppositions, but his study does not result in anything that can be considered a structural contrast between the two episodes – only mirrorings that run across the entire text.79 Either way, Adapa does not meet the criteria of the two-act structure in any obvious way.

5.3 Ishtar’s and Inana’s Descent

The name of Ishtar’s Descent is, strictly speaking, only appropriate for the first half of this story, since Ishtar’s descent to the netherworld is followed by her ascent back to the land of the living (that is, if the laconic and somewhat cryptic ending of the Akkadian version is read through the lens of the Sumerian version, Inana’s Descent). The story of Ishtar’s Descent is clearly divided into two parts, the first of which runs from l. 1 to 75 and the second from l. 76 to 139, and the transition between the two is a clear-cut shift in narrative focus, as the story moves abruptly from Ishtar’s imprisonment in the netherworld to her servant Papsukkal’s despair on earth. Also clear is the contrast between the two halves: Ishtar moves downward in the first act and upward in the second. This is underscored by her passage through the seven gates of the netherworld: Ishtar is forced to deposit an item of clothing at each gate on the way down, which is returned to her at each gate on the way up. The items of clothing are of course returned in the reverse order as Ishtar moves backwards through the gates, yielding a neat set of reversals: up/down, giving/receiving, there/back.80

However, this does not mean that Ishtar’s Descent fits easily into the two-act structure. Most importantly, there seems to be no element of expansion. There is certainly no quantitative expansion of the line count, as the second act is slightly shorter than the first. But there is no obvious qualitative expansion either. Since Ishtar returns to her original state at the end of the story, her downward and upward movements are the same length. One could argue that her return from the land of the dead is a more significant feat than her imprisonment there (simply put, resurrection is trickier than death), but Ishtar’s attempt at a coup of the netherworld in the first act is surely also an exceptional occurrence.

Further, the identification of the two-act structure in Ishtar’s Descent is more complicated than in the previous cases, because it is not originally an Akkadian narrative, but a Sumerian story that was reworked into Akkadian. The question of whether Ishtar’s Descent employs the two-act structure therefore raises the problem of the whether this structure also existed in Sumerian literature. The Sumerian version of the story, Inana’s Descent, does in fact evince the expansion of the second act that is lacking in its Akkadian counterpart. The division into two parts and the reversal of direction is much the same, but the Sumerian version adds an episode after Inana’s return from the netherworld that is only briefly summarized in the Akkadian one. Inana has to provide a substitute to take her place in the netherworld and refuses to give her servants Ninshubur, Shara, or Lulal to the netherworld demons, but instead allows them to seize her lover, Dumuzi. Through the addition of this episode, the second act is made longer than the first (172 and 238 lines, respectively). This is intriguing, but the larger question of whether the two-act structured also existed in Sumerian literature would require a much broader investigation.

5.4 The Poor Man of Nippur

Whereas the two-act structure is at best ambiguous and probably absent from the stories of Anzu, Adapa, and Ishtar’s Descent, the story of The Poor Man of Nippur is an exceptional case for a different reason. It is exceptional because it so clearly displays the two-act structure and directly comments upon it, but without belonging to the genre of myths and epics with which I have associated the narrative device so far. All the stories treated concern gods, goddesses, heroes, and mythical figures, but with The Poor Man, we turn to much more prosaic matters.

It is the story of Gimil-Ninurta, who is abused by the mayor of Nippur and then avenges himself by thrashing the mayor three times. Jerrold Cooper divides the story into five parts: (I) an introduction, (II) the poor man’s first contact with the mayor, and (IIIV) his three acts of revenge. Cooper goes on to note that the introduction stands apart from the rest of the tale, as the four episodes of contact between Gimil-Ninurta and the mayor share a number of structural similarities. However, episodes I and II are also strongly linked, leading Cooper to view the latter as a “pivotal episode,” whose role in the story “is to transform the initial situation of episode I into the tale of vengeance of episodes IIIV.”81 The story’s two-act structure is thus readily apparent:

  • There is a clear division into two acts, each with their own narrative sequence: in Cooper’s division, episodes III and IIIV.

  • The second act extends the first, as Gimil-Ninurta’s revenges triple the mayor’s abuse.

  • The two parts mirror each other, as crime is met with punishment.

In fact, Gimil-Ninurta puts this structure clearly when he says to the mayor: “For the one burden you’ve laid on me, I’ll pay you back three to one!”82 His acts of revenge clearly follow the logic of the two-act structure, as one event is counterpoised with an opposite three times its size. Just as in Nergal and Ereshkigal, the two-act structure is not just a formal aspect of composition but is also used creatively as part of the plot itself. If one accepts that the narrative device proposed in this essay was a prevalent element in Akkadian epics, Gimil-Ninurta’s tale may be read as an ironic twist on a standardized aspect of “high” literature, which is here put to more humorous uses. The Poor Man of Nippur may be a purposefully slapstick version of the two-act structure, satirizing the narrative structure of epic tales.

5.5 The Cuthean Legend of Naram-Sin

The final story to be treated here is The Cuthean Legend of Naram-Sin, which tells of the tragic figure of the Old Akkadian king Naram-Sin, who is punished for defying the will of the gods by an invasion of monstrous enemies that overrun his land.83 Like The Poor Man of Nippur, the story clearly employs the two-act structure but does not neatly fit into the genre of Akkadian epics.

At the very least, the generic classification of the story is ambiguous. On the one hand, it shares a number of literary features with texts such as Gilgamesh, notably its presentation as the autobiography of a tragic king, whose tale the reader is invited to find on a stela lying in a tablet box.84 On the other hand, the Cuthean Legend differs from the epic in that it tells of a king who modern historians be identified as a real figure, and so cannot be properly said to belong to the same genre of mythical stories as Enuma Elish or Atra-hasis. In fact, the Cuthean Legend may be said to have an actual historical background in the royal inscriptions of Naram-Sin, where the king boasts of crushing a wide-spread revolt against his reign with nine victories in a single year.85 Over time, this account was reversed, exaggerated, and fictionalized, as Naram-Sin was instead said to have lost nine battles in a single year and the enemy he was said to have fought became increasingly overwhelming and supernatural. As Foster puts it, “[i]n treatments of the theme in the late second and first millennia BCE, the story became ever more fantastic and Naram-Sin ever more chastened and humble, to the extent that he counsels prudence and pacifism rather than military derring-do.”86

I would further argue that as the story of Naram-Sin’s battle became less historically accurate and ever more a literary drama, it was also made to follow the two-act structure.87 The mirroring structure of the story consists in a contrast between Enmerkar and Naram-Sin. Both kings face a terrible invasion, both use divination to ask the gods for counsel and assistance, and both ultimately meet a tragic fate. Enmerkar, however, manages to defeat the enemy, but does not write down the story of his struggles (l. 28–30), leaving Naram-Sin unprepared when he then faces the same challenge. By contrast, in the ending of the Cuthean Legend, the defeated Naram-Sin does set down an account of his struggles in the form of the very text that we have been reading, so as to instruct future kings that they may learn from his sad example (l. 149–51). The contrast between the two rulers and the ending of their otherwise similar stories is clearly expressed: Enmerkar “did not write on a stele and leave it behind for me,” while Naram-Sin “wrote a stele for you,” that is, the future ruler reading the story.88

The expansion of the second act with respect to the first is likewise unmistakable, as Enmerkar’s story is told in 36 lines and Naram-Sin’s in 150. The expansion is also qualitative in that the enemy faced by Enmerkar is only referred to in passing while the monstrous horde faced by Naram-Sin is described in horrifying detail, much like the extended description of Tiamat’s army in the second half of Enuma Elish. Intriguingly, the contrast between the two acts is here used, as in Nergal and Ereshkigal, not only as a matter of literary form in the telling of the narrative, but also as a crucial plot device within the progression of the narrative itself. The contrast between the king who did not write down his story and the king who did is ultimately what explains the very existence of the text itself, according to its own logic: the story was composed precisely because of the mirroring relation between its two acts.

The Legend is a particularly interesting example of the two-act structure because it serves as a case study of how originally historical, non-literary material was reshaped over time and eventually came to assume the narrative form that I have argued was so distinctive for Akkadian epics. It thus gives us a glimpse of how literary demands affected the way in which stories that may originally have circulated in other formats – such as historical artefacts or oral tales – came to be arranged and restructured when they were written down as literature, which in many cases is the only medium in which they have survived.

6 Conclusion

As they were reworked into a deliberate and contrived artistic form, religious narratives from the cuneiform world were made to fit a set of literary expectations. In this article, I have argued that one of those literary expectations was that the stories be arranged into two mirroring acts, with the second being the larger of the two. While neither unique nor universal, this structure does seem particularly prevalent among the Akkadian epics, as it is found in up to nine of the twelve cases examined here. There are substantial variations in how the epics employ the pattern. In some, the expansion of the second act is an enlargement of the verse count; in others, it is a qualitative extension of scope. In some, the relation between the acts is one of parallelism; in others, one of contrast. But there are also important similarities in the structure of these stories, which I believe justify viewing this pattern as a recurrent narrative device.

This finding has further implications for the study of ancient Near Eastern myth. In some cases, most notably Atra-hasis, the pattern was apparently achieved by combining two separate myths to form the bipartite structure of the epic. In others, such as Enuma Elish, an already existing mythical template was provided with a miniature doublet. The proposed pattern thus helps to explain why myths were reshaped the way they were.

Finally, I have noted in passing some possible parallels to other literary and religious traditions. A brief comparison with Inana’s Descent raised the question of whether the same pattern is also to be found in Sumerian literature, a question that cannot be answered within the limits of this study. Müller’s interpretation of the narrative contrast in Atra-hasis drew a parallel to a similar structure in the Biblical narrative of creation and the Flood, a parallel that this essay has also left unexplored. In short, the proposed structure opens the way for further investigations into the religious narratives of the ancient Near East.


  • Al-Rawi, Farouk N.H., and Jeremy A. Black. 1989. “The Second Tablet of ‘Išum and Erra.’ Iraq 51: 111122.

  • Annus, Amar. 2001. The Standard Babylonian Epic of Anzu. SAACT 3. Helsinki: Helsinki University Press.

  • Borger, Rykle. 1979. Babylonisch-Assyrische Lesestücke. 2nd edition. 2 volumes. AnOr 54. Rome: PBI.

  • Bottéro, Jean. 1977. “Les noms de Marduk, l’écriture er la ‘logique’ en Mésopotamie ancienne.” In Ancient Near Eastern Studies in Memory of Jacob Joel Finkelstein, ed. Maria de Jong Ellis, 527. MCAAS 19. Hamden: Archon Books.

    • Search Google Scholar
    • Export Citation
  • Bottéro, Jean. 1987. Mésopotamie: L’écriture, la raison et les dieux. Paris: Éditions Gallimard.

  • Cagni, Luigi. 1969. L’epopea di Erra. StSem 34. Rome: Istituto di Studi del Vicino Oriente.

  • Cooper, Jerrold S. 1975. “Structure, Humor, and Satire in the Poor Man of Nippur.” JCS 27: 163174.

  • Culler, Jonathan. 2001. The Pursuits of Signs: Semiotics, Literature, Deconstruction. Augmented edition. Ithaca: Cornell University Press.

    • Search Google Scholar
    • Export Citation
  • Davila, James R. 1995. “The Flood Hero as King and Priest.” JNES 54: 199214.

  • Dickson, Keith. 2007. “The Jeweled Trees: Alterity in Gilgamesh.” Comp. Lit. 59: 193208.

  • Dietrich, Jan. 2014. “Von der Freundschaft im alten Testament und alten Orient.” WdO 44: 3756.

  • Durand, Jean-Marie. 1994. “100. Enûma Eliš I 76.” NABU 1994: 91.

  • Ehelolf, Hans. 1968 [1916]. Ein Wortfolgeprinzip im Assyrisch-Babylonischen. Reprint. LSS 6. Leipzig: Zentralantiquariat der DDR.

  • Feldt, Laura, and Susanne Ulla Koch. 2011. “A Life’s Journey – Reflections on Death in the Gilgamesh Epic.” In Akkade Is King: A Collection of Papers by Friends and Colleagues Presented to Aage Westenholz on the Occasion of His 70th Birthday, 15th of May 2009, ed. Goiko Barjamovic et al., 111126. PIHANS 118. Leiden: NINO.

    • Search Google Scholar
    • Export Citation
  • Foster, Benjamin R. 2005. Before the Muses: An Anthology of Akkadian Literature. 3rd edition. Bethesda: CDL.

  • Foster, Benjamin R. 2007. Akkadian Literature of the Late Period. GMTR 2. Münster: Ugarit-Verlag.

  • Foster, Benjamin R. 2016. The Age of Agade: Inventing Empire in Ancient Mesopotamia. London: Routledge.

  • Furlani, Giuseppe. 1946. “L’Epopea di Gilgames come Inno all’Amicizia.” Belfagor 1: 577589.

  • Gabriel, Gösta. 2014. Enūma eliš – Weg zu einer globalen Weltordnung: Pragmatik, Struktur und Semantik des babylonischen “Lieds auf Marduk.” ORA 12. Tübingen: Mohr Siebeck.

    • Search Google Scholar
    • Export Citation
  • George, Andrew R. 2003. The Babylonian Gilgamesh Epic: Introduction, Critical Edition, and Cuneiform Texts. 2 volumes. Oxford: Oxford University Press.

    • Search Google Scholar
    • Export Citation
  • George, Andrew R. 2007. “The Gilgameš Epic at Ugarit.” AuOr 25: 237254.

  • George, Andrew R. 2013. “The Poem of Erra and Ishum: A Babylonian Poet’s View of War.” In Warfare and Poetry in the Middle East, ed. Hugh Kennedy, 3971. London: I.B. Tauris.

    • Search Google Scholar
    • Export Citation
  • George, Andrew R., and Farouk N.H. Al-Rawi. 2014. “Back to the Cedar Forest: The Beginning and End of Tablet V of the Standard Babylonian Epic of Gilgameš.” JCS 66: 6990.

    • Search Google Scholar
    • Export Citation
  • Gurney, Oliver R. 1956. “The Sultantepe Tablets. V. The Tale of the Poor Man of Nippur.” AS 6: 145164.

  • Haul, Michael. 2000. Das Etana-Epos: Ein Mythos von der Himmelfahrt des Königs von Kiš. GAAL 1. Göttingen: Seminar für Keilschriftforschung.

    • Search Google Scholar
    • Export Citation
  • Helle, Sophus. 2015. “95. Contrast through ironic self-citation in Atra-ḫasīs.” NABU 2015: 158160.

  • Helle, Sophus. 2016. “Emotions in Gilgamesh: Desire, Grief, and Identity in the Standard Babylonian Epic of Gilgamesh.” Unpublished MA-thesis, Copenhagen: University of Copenhagen.

  • Izreʾel, Shlomo. 2001. Adapa and the South Wind: Language Has the Power of Life and Death. MC 10. Winona Lake: Eisenbrauns.

  • Katz, Dina. 2011. “Reconstructing Babylon: Recycling Mythological Traditions Toward a New Theology.” In Babylon: Wissenskultur in Orient Und Okzident, ed. Eva Cancik-Kirschbaum, Margarete van Ess, and Joachim Marzahn, 123134. Topoi 1. Berlin: De Gruyter.

    • Search Google Scholar
    • Export Citation
  • Kilmer, Ann D. 2006. “Visualizing Text: Schematic Patterns in Akkadian Poetry.” In If a Man Builds a Joyful House: Assyriological Studies in Honor of Erle Verdun Leichty, ed. Ann K. Guinan et al., 209221. CM 31. Leiden: Brill.

    • Search Google Scholar
    • Export Citation
  • Kinnier Wilson, James V. 1985. The Legend of Etana: A New Edition. Warminster: Aris and Philips.

  • Kirk, Geoffrey S. 1970. Myth: Its Meaning and Functions in Ancient and Other Cultures. Cambridge: Cambridge University Press.

  • Kramer, Samuel Noah. 1980. “Sumerian Literature and the British Museum: The Promise of the Future.” PAPS 124: 295312.

  • Kvanvig, Helge S. 2011. Primeval History: Babylonian, Biblical, and Enochic: An Intertextual Reading. JSJSup 149. Leiden: Brill.

  • Lambert, Wilfred G. 1980. “The Theology of Death.” In Death in Mesopotamia, ed. Bendt Alster, 5366. CRRAI 26. Copenhagen: Akademisk Forlag.

    • Search Google Scholar
    • Export Citation
  • Lambert, Wilfred G. 2013. Babylonian Creation Myths. MC 16. Winona Lake: Eisenbrauns.

  • Lambert, Wilfred G., and Alan R. Millard. 1969. Atra-ḫasīs: The Babylonian Story of the Flood. Winona Lake: Eisenbrauns.

  • Machinist, Peter. 1983. “Rest and Violence in the Poem of Erra.” JAOS 103: 221226.

  • Machinist, Peter. 2005. “Order and Disorder: Some Mesopotamian Reflections.” In Genesis and Regeneration: Essays on Conceptions of Origins, ed. Shaul Shaked, 3161. Jerusalem: Israel Academy of Sciences and Humanities.

    • Search Google Scholar
    • Export Citation
  • Michalowski, Piotr. 1980. “Adapa and the Ritual Process.” RO 41: 7782.

  • Michalowski, Piotr. 1999. “Commemoration, Writing, and Genre in Ancient Mesopotamia.” In The Limits of Historiography: Genre and Narrative in Ancient Historical Texts, ed. Christina S. Kraus, 6990. Leiden: Brill.

    • Search Google Scholar
    • Export Citation
  • Moran, William. 1987. “Some Considerations of Form and Interpretation in Atrahasis.” . In Language, Literature and History: Philological and Historical Studies Presented to Erica Reiner, ed. Francesca Rochberg-Halton, 245255. AOS 67. New Haven: AOS.

    • Search Google Scholar
    • Export Citation
  • Müller, Hans-Peter. 1985. “Das Motiv für die Sintflut: Die hermeneutische Funktion des Mythos und seiner Analyse.” ZAW 97: 295316.

    • Search Google Scholar
    • Export Citation
  • Novotny, Jamie. 2001. The Standard Babylonian Etana Epic. SAACT 2. Helsinki: Neo-Assyrian Text Corpus Project.

  • Oppenheim, A. Leo. 1977. Ancient Mesopotamia: Portrait of a Dead Civilization, ed. Erica Reiner. Revised edition. Chicago: University of Chicago Press.

    • Search Google Scholar
    • Export Citation
  • Pettinato, Giovanni. 2000. Nergal ed Ereskigal, il poema assiro-babilonese degli inferi. Rome: Accademia Nazionale dei Lincei.

  • Ponchia, Simonetta, and Mikko Luukko. 2013. The Standard Babylonian Myth of Nergal and Ereškigal. SAACT 8. Helsinki: Neo-Assyrian Text Corpus Project.

    • Search Google Scholar
    • Export Citation
  • Reiner, Erica. 1985. Your Thwarts in Pieces, Your Mooring Ropes Cut: Poetry from Babylonia and Assyria. Ann Arbor: Horace H. Rackham School of Graduate Studies at the University of Michigan.

    • Search Google Scholar
    • Export Citation
  • Seri, Andrea. 2014. “Borrowings to Create Anew: Intertextuality in the Babylonian Poem of ‘Creation’ (Enūma eliš).” JAOS 134: 89106.

    • Search Google Scholar
    • Export Citation
  • Shehata, Dahlia. 2001. Annotierte Bibliographie zum altbabylonischen Atramḫasīs- Mythos: inūma ilū awīlum. GAAL 3. Göttingen: Seminar für Keilschriftforschung.

    • Search Google Scholar
    • Export Citation
  • Sladek, William R. 1974. “Inanna’s Descent to the Netherworld.” PhD diss., John Hopkins University.

  • Soden, Wolfram von. 1994. “Der altbabylonische Atramchasis-Mythos.” In Mythen und Epen II/4, ed. Karl Hecker et al., 612645. TUAT 3. Gütersloh: Gütersloher Verlaghaus.

    • Search Google Scholar
    • Export Citation
  • Vanstiphout, Herman L.J. 1990. “The Craftmanship of Sîn-leqi-unninnī.” OLP 21: 4579.

  • Vanstiphout, Herman L.J. 1992. “Enuma Elish as a Systematic Creed: An Essay.” OLP 23: 3761.

  • Veldhuis, Niek. 1999. “The Poetry of Magic.” In Mesopotamian Magic: Textual, Historical and Interpretative Perspectives, ed. Tzvi Abusch and Karel van der Toorn, 3548. AMD 1. Groningen: Styx.

    • Search Google Scholar
    • Export Citation
  • Verderame, Lorenzo. 2016. Letterature dell’antica Mesopotamia. Milan: Mondadori.

  • Vogelzang, Marianne. 1988. “Bin Šar Dadmē: Edition and Analysis of the Akkadian Anzu Poem.” PhD diss., Rikjsuniversiteit Groningen.

  • Walls, Neal H. 2001. Desire, Discord, and Death: Approaches to Ancient Near Eastern Myth. ASOR Books 8. Boston: ASOR.

  • Westenholz, Joan Goodnick. 1997. Legends of the Kings of Akkade. MC 7. Winona Lake: Eisenbrauns.

  • Wilcke, Claus. 1999. “Weltundtergang als Anfang.” In Weltende: Beiträge zur Kultur- und Religionswissenschaft, ed. Adam Jones, 63112. Wiesbaden: Harrassowitz.

    • Search Google Scholar
    • Export Citation
  • Winitzer, Abraham. 2019. “History as Scholarship in the Early Babylonian Divination Literature (Part I).” KASKAL 16: 259277.

  • Wisnom, Selena. 2019. Weapons of Words: Intertextual Competition in Babylonian Poetry. CHANE 106. Leiden: Brill.

  • Wolff, Hope Nash. 1969. “Gilgamesh, Enkidu, and the Heroic Life.” JAOS 89: 392398.

  • Wolff, Hope Nash. 1973. “Review of Lambert, W.G. and Millard, A.R., Atra-ḫasīs: The Babylonian Story of the Flood.” JAOS 93: 7576.

    • Search Google Scholar
    • Export Citation
  • Zgoll, Annette. “monumentum aere perennius – Maerring und Ringkomposition im Gilgameš-Epos.” In Von Göttern und Menschen: Beiträge zu Literatur und Geschichte des Alten Orients: Festschrift für Brigitte Groneberg, ed. Dahlia Shehata, Frauke Weiershäuser, and Kamran V. Zand, 443470. CM 41. Leiden: Brill.

    • Search Google Scholar
    • Export Citation

The article was first presented at a meeting of the Copenhagen Near Eastern Studies group, and then as a poster at the 62nd Rencontre Assyriologique Internationale. I thank the audiences at both events for their valuable discussions of my proposal, and in particular Ann Guinan for a stimulating discussion. I would like to thank also the anonymous reviewer and journal editor for their suggestions.


In the following, I understand “epic” in the maximalist sense of the word, as an extended narrative poem set in the distant past that concerns the feats of gods or extraordinary humans.


Ehelolf (1968 [1916]).


Veldhuis (1999, 40).


I thank Ann Guinan for bringing this point to my attention.


The text is edited in Lambert and Millard (1969) and Shehata (2001). Not enough is preserved of the later version of the epic to reconstruct its narrative structure.


Moran (1987, 245).


Shehata (2001, 4).


Kvanvig (2011, 19) and Wilcke (1999, 70).


In both stories, a rebellious god is sacrificed and his blood used to create humanity so as to relieve the gods of their labor.


Lambert and Millard (1969, 9) and von Soden (1994, 615) bisect the story at l. I 352, while Moran (1987) places the dividing line before the fragmentary section beginning with l. I 320 – see also Kvanvig (2011, 19–21).


Kvanvig (2011, 26–7).


Müller (1985, 295).




Wilcke (1999, 98), and Helle (2015) with further examples.


attī-ma šassūru bāniat awīlūti, I 194; and [attī-ma ša]ssūru bāniat šīmāti, III vi 47 (Shehata 2001, 60 and 161).


Lambert (1980, 57–8). See also Wolfe (1973).


Vanstiphout (1992, 47) and Oppenheim (1977, 214). The text is edited in Lambert (2013).


ištu lemnēšu ikmû isādu, I 73 = IV 123, (Lambert 2013, 54–5, 92–3). See Katz (2011, 129). For further references to previous studies of this structure, see Gabriel (2014, 183–4).


Gabriel (2014, 182–97).


ibbanû-ma ilū qerebšun, I 9, “the gods were created within them,” meaning within Apsû and Tiamat; and ina qereb Apsî ibbani Marduk, I 81, “Marduk was created within the Apsû” (Lambert 2013, 50–1, 54–5).


uʾaddû ešrēti, I 76, Lambert (2013: 54–5).


Durand (1994).


See Gabriel (2014, 192) and Bottéro (1977).


See also Katz (2011, 130): “The motifs that governed the story of Ea are at work again in relation to Marduk, but the extent of the narrative is different.”


Gabriel (2014, 187).


Gabriel (2014, 189–91).


Gabriel (2014, 197–218).


Gabriel (2014, 189).


According to Gabriel (2014, 183), the acts run from l. I 7–78 and l. I 79–VI 69, while according to Katz (2011, 130), they run from l. I 29–72 and l. I 110–IV 122, since she includes only the episodes of conflict and not those of creation. This yields acts of 72 and 757 lines, and 44 and 475 lines, respectively.


George (2003, 48).


Vanstiphout (1990, 40–53). The text is edited in George (2003); see also George and al-Rawi (2014) for the new manuscript of tablet V.


Vanstiphout (1990, 49).


Wolff (1969, 392), Feldt and Koch (2011, 112).


iḫīṭ-ma mitḫāriš parakkī (George 2007, 239).


George (2007, 246).


For Humbaba as king, see George and al-Rawi (2014, 74); for Uta-napishti as king, see Davila (1995, 206). Note also that the Cedar Forest is explicitly identified as a parakku in l. V 7. Further, both figures are described as “the one of whom people talk”: respectively Humbaba (ša iqabbû lūmur, OB version, l. II 182, George 2003, 202–3) and Uta-napishti (ša idabbubūš lūmur, SB version, X 250, George 2003, 692–3).


Dickson (2007, 193).


Dickson (2007, 195, emphasis in the original).


One may note here the distinction made in the introduction between the narrative content of myths and the narrative form of epics: it is possible that a strict symmetry between the two forests was a part of an original myth about Gilgamesh and that, in the making of the epic, this was reworked to fit the expected structure of Akkadian epics.


Vanstiphout (1990, 53, emphasis in the original).


Furlani (1946).


Respectively, kīma muti, OB version, l. II 111, and kīma kallāti, SB version l. VIII 59 (George 2003, 176–7 and 654–55). See also Vanstiphout (1990, 50n.20) and Helle (2016).


See e.g. Zgoll (2010).


Kirk (1970, 125) and Verderame (2016, 65). The text is edited in Kinnier Wilson (1985) and Haul (2000).


Or even more – see e.g. Novotny (2001, xii–xiv).


See Haul (2000, 64n.244 and 245) and Kirk (1970, 126).


Haul (2000, 249).


See Haul’s discussion of previous proposals (2000, 64–74).


Kirk (1970, 128).


Kirk (1970, 129).


Haul (2000, 73, emphasis added).


Dietrich (2014, 41): [al]ka nīnū-ma rūʾ[ūta i nīpuš] / [] itbārū an[āku u atta], SB version l. II 8–9; and ibrī lū itbārānu anā u atta, OB version l. I vi 6’ (Haul 2000, 170–1 and 114–5).


Foster (2005, 506–9). See also the summaries in Walls (2001, ch. 3) and Reiner (1985, 50–6). Both versions are edited in Pettinato (2000).


īrum-ma ana palkî kisallīša, l. iii 48’ = vi 28’; and innadrū-ma aḫāmiš kilallān / ana mayyāli šitmuriš īterb[ū-ma], l. iv 9’–10’ = vi 34’–35’ (Pettinato 2000, 88–9 and 106–7, 92–3 and 106–7, respectively).


ana mu[ḫḫīya] šūbilānišum?-ma ludūkšu, Amarna version l. 27; and [ilu šâšu šu]purannašū-ma lū ḫāmerī libīt ittīya, SB version l. v 6’ (Pettinato 2000, 60–1 and 98–9, see also Ponchia and Luukko 2013, xciv).


Bottéro (1987, 443), see also the discussion in Walls (2001, 128, and passim).


See Foster (2007, 106–9). The text is edited in Cagni (1969), see also the edition of tablet II in al-Rawi and Black (1989).


[…] irtame šubassu, l. II B 34 according to the numbering in Cagni (1969, 84–5).


Foster (2005, 892, emphasis added).


Foster (2007, 65).


George (2013, 54).


Machinist (1983; 2005).


George (2013).


adi atta ana bīti šâšu terrubū-ma Girru ṣubatka ubbabū-ma taturru ašrukka, I 181, (Cagni 1969, 78–9, translation from Foster 2005, 889–90).


I would follow Foster (2005, 894–5), in seeing a crucial shift in the narrative in pericope C2, l. 36’. This would yield a first act consisting of tablet I and a substantial part of tablet II, and a second act consisting of the rest of tablet II and tablets III through V, about twice as long as the first.


Oppenheim (1977, 214).


Culler (2001, 69).


See Wolff (1973).


Nor is the device necessarily restricted to cuneiform literature. Given the numerous parallels between A New Hope and The Force Awakens (such as the mirroring between the Death Star and the much larger, but narratively identical Starkiller Base) one might claim that the Star Wars franchise employs the same device. After all, just like Enuma Elish, the Star Wars films begin high in the heavens, a long, long time ago….


See e.g. Machinist (2005), Seri (2014), and most recently Wisnom (2019).


Vogelzang (1988, 133–4, 145, and 202–24).


See e.g. Annus (2001, xiii).


See e.g. the description in Machinist (2005, 38, emphasis added), according to whom Ninurta is reluctant to return the Tablet, “recalling the much greater resistance to the assembly that Anzu had offered.”


Kirk (1970, 130–1). The text is edited in Izreʾel (2001).


Michalowski (1980, 79).


Izreʾel (2001, 141–2).


Taking the episode of Marduk’s statue as spanning from l. I 124 to at least l. II 79, and following Cagni’s (1969, 26) estimate that the epic originally consisted of c. 750 lines.


Izreʾel (2001, 141–6).


For further examples of symmetry in Ishtar’s Descent, see Kilmer (2006, 214). The Akkadian version is edited in Borger (1979, 86–91). The main edition of the Sumerian version remains Sladek (1974), though see also Kramer (1980, 299–310).


Cooper (1975, 163–4, 167). The text is edited in Gurney (1956).


aššu ištēt biltu ša tēmedanni ša ištēn 3 rībētu arâbka, l. 67–68 (Gurney 1956, 152–3).


The three versions of the text are edited in Westenholz (1997, 263–8).


Respectively, the Gilgamesh I 24–28, and the Cuthean Legend l. 1–3 and 149–155. For the relation between the two texts, see e.g. Michalowski (1999, 80–84).


See Winitzer (2019, 267–8) with references to previous literature.


Foster (2016, 270).


The Cuthean Legend exists in three versions, an OB, MB, and NA version. The division into two mirroring acts is clearly present already in the Middle Babylonian version, but this version is too fragmentary to allow a full study. The Old Babylonian version lacks the beginning of the text and so one cannot know for certain whether it followed the same structure as the later versions.


Respectively, ina narê ul išṭur, l. 29, and narâ ašṭurka, l. 151 (Westenholz 1997, 306 and 326).

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 165 105 0
Full Text Views 81 72 5
PDF Views & Downloads 144 124 10