Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

The three movements of the Iliad
Heiden, Bruce
Greek, Roman and Byzantine Studies; Spring 1996; 37, 1; ProQuest
pg. 5

The Three Movements of the Iliad 

Bruce Heiden 

T HE ILIAD FOR US is a text to be read; for its composer, his 
audiences, and several generations of audiences after them, 
it was a live vocal performance. Scholars of Greek epic 

and related genres have become increasingly sensitive to the 
losses in affect and significance that occur when such a per­
formance survives only in the form of its 'libretto'.! But for 
students of Homer the desire to recapture the power of the per­
formed Iliad confronts the silence of the historical record: the 
first traces of performances of Homer date from the mid-sixth 
century, perhaps more than a century after the epic's compo­
sition.2 Denys Page probably spoke for many in declining to 

speculate about how the questions concerning Homeric per­
formance might be answered. 3 Others have been less cautious. 
The issue of the performance structure of the Iliad has proven 
an especially fertile ground of speculation, because the evident 

1 For recent theoretical discussion of these losses and efforts to recuperate 
them, see J. M. Foley, The Singer of Tales in Performance (Bloomington 1995) 
1-59. 

2 The view that our Iliad is the result of a major revision that occurred in 
sixth-century Athens has gained adherents during the past two decades: M. S. 
Jensen, The Homeric Question and the Ora/-Formulaic Theory (Copenhagen 
1980); K. STANLEY, The Shield of Homer (Princeton 1993: hereafter 'Stanley') 
279-93; R. Seaford, Reciprocity and Ritual (Oxford 1994) 144-90; C. Nagy, 
"Homeric Questions," TAPA 122 (1992) 52, describes the "textualization" of 
Homeric epic as a centuries-long process that entered a "definitive" stage in 
mid-sixth century Athens. For recent arguments against a late sixth-century 
Athenian recension, see R. Janko, The Iliad: A Commentary IV (Cambridge 
1992) 29-32; M. L. West, "The Date of the Iliad, " MusHelv 52 (1995) 203-19, 
proposes 688 as a terminus post quem for composition; E. F. Cook, The 
Odyssey in Athens (Ithaca 1995) 128-70, argues that the Odyssey ·acquired 
its present form in the region of the Saronic Gulf during the seventh and sixth 
centuries B.C.E." The argument presented in this paper does not depend upon 
the date of the Iliad, but assumes that the composer established a relatively 
fixed text for performance. A. Parry (YCS 20 (1966] 175-216) compellingly 
explained why the question "Have we Homer's Iliad?" must be answered 
·yes." 

J D. Page, The Homeric Odyssey (Oxford 1955) 75 n.8. 

5 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

6 THE THREE MOVEMENTS OF THE ILIAD 

coherent design of the poem indicates that its composer in­
tended it to make a profound impression as a whole upon its 
audiences. Yet the immense length of the Iliad renders con­
tinuous performance a physical impossibility. Scholars have 
therefore tried to imagine a performance structure that might 
accommodate the necessary pauses but preserve unity of im­
pact. This paper will review the proposals that have addressed 
this issue, evaluate their plausibility, and offer a new contribu­
tion to the discussion. 

I 

Three theories of Homeric performance structure may be 
dealt with briefly; these lack plausibility because they under­
estimate either the aesthetic importance of the epic's coherence 
or the requirement that its outstanding features be accessible in 
performance. According to A. B. Lord, Homer, like a South 
Slavic guslar, could have simply stopped whenever he got 
tired;4 this overly close identification of Homer with South 
Slavic singers has been criticized by A. Parry (supra n.2: 212) 
among others. Lord later suggested that the Iliad may have 
achieved its monumentality through the influence of the scribe 
who took it down from its oral composer-in-performance, 
who would never have intended it to be performed as a whole 
at all, at least not after the performance in which it was dictated. 5 

As poems until the late fifth century in Greece were heard and 
not read, Lord's theory would mean that Homer, or rather 
Homer's scribe, had knowingly created a poem whose most 
extraordinary effects he alone could appreciate: not likely, 
especially since Lord does not even suggest that an eighth cen­
tury scribe could have had any poetic taste independent of the 
oral masters. Notopoulos argued that Homer's epics were 
designed for performance in short segments, probably corre­
sponding to the 'books' transmitted in the manuscripts, over a 
period of several weeks at the private banquets of aristocrats. 6 A 
performance so structured could not have made a coherent 

4 A. B. Lord, ·Homer and Huso I: The Singer'S Rests in Greek and South­
slavic Heroic Song," TAPA 67 (1936) 10lr13. 

5 A. B. Lord, The Singer of Tales (Cambridge 1960) 149-57. Lord's scenario 
has recently been adopted by B. B. Powell, Homer and the Origin of the 
Greek Alphabet (Cambridge 1991) 228ff. 

6 J. A. Notopoulos, ·Studies in Early Greek Oral Poetry," HSCP 68 (1964) 
1-18. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 7 

impact upon audiences. It is altogether possible that such per­
formances, or performances even less coherent, did take place; 
but it is unlikely that the composer of the Iliad never intended 
his epic to be performed in a more coherent format. 

More plausible is the view that the composer of the Iliad in­
tended it to be performed in marathon sessions over a brief 
span of days, probably at a pan hellenic religious festival. Only in 
such a performance could the epic have been appreciated as a 
whole.7 Wade-Gery's suggestion of a three-day division was 
inspired by the analogy of the competition in tragedy at the 
City Dionysia in Athens, which occupied three days, as well as 
by the three "movement" division of the Iliad previously pro­
posed by Sheppard without reference to performance. 8 The 
obvious basis for Sheppard's scheme is the flanking of the third 
day of battle (Books 11-18) by the many narrated days that pre­
cede and follow it. This three-part scheme was adopted with 
slight modification by Wade-Gery and continues to be influen­
tial in the recent new proposal of Taplin. 9 

The idea of a three-part performance does not necessarily 
entail the conclusion that the divisions in the performance were 
exactly determined by the composer in advance. The points at 
which a day's performance ended and the next day's began may 
have been variable. But this does not seem likely. Beginnings, 
pauses, and resumptions of performance can be moments of 
great emphasis, and one can hardly imagine that so skilful a 
conductor of his audience's feelings as Homer would have left 
the divisions in the performance of his epic to chance or failed 
to exploit their potential. Thus adherents of the three-part per­
formance structure have maintained that the Iliad was designed 
to be performed with two specific, major internal pauses and 

7 Cf G. S. K1RK, The Iliad: A Commentary I: Books 1--4 (Cambridge 1985: 
hereafter 'Kirk') 12f. 

8 H. T. WADE-GERY, The Poet of the Iliad (Cambridge 1952: hereafter 
'Wade-Gery') 13-16; J. T. Sheppard, The Pattern of the Iliad (London 1922); 
G. Murray, The Rise of the Greek Epic· (Oxford 1934) 187-92, had earlier 
suggested a panhellenic festival as the occasion for which the Iliad might have 
been composed; M. Breal, Pour mieux connaitre Homere (Paris n.d.) 43ff, had 
linked composition and performance to the establishment of sacred athletic 
contests; before Sheppard and likewise without reference to performance, E. 
Drerup, Das funfte Buch der !lias (Paderborn 1913) 364 with 0.2, 428, 
suggested a three-part division of the whole epic. 

9 O. T APLlN, Homeric Soundings (Oxford 1992: hereafter 'Taplin') 11-31. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

8 THE THREE MOVEMENTS OF THE ILIAD 

that the location of these pauses can be precisely identified In 

the text.10 

Five different schemes for dividing the Iliad into three 
movements for performance have been proposed. They are 
summarized in the following table: 

II III 

Wade-Gery 1-9 10-18.353 18.354-24 

Davison 1-8 9-16 17-24 

Schadewaldt 1-9 11-18 19-24 

Taplin 1-9 11-18.353 18.354-24 

Stanley 1-7 8-17 18-24 

These diverse schemes are based upon several aesthetic criteria. 
Davison's criterion is length: dividing the poem into six four­
book segments yielded units of approximately equal length that 
might have been convenient for recitation by a team of 

10 Wade-Gery 13-16; J. A. Davison, "Thucydides, Homer, and the 'Achaean 
Wall'," GRBS 6 (1965) 23ff; W. Schadewaldt, Der Aufbau der I/ias (Frankfurt 
a.M. 1975) 24; Taplin 11-31; and Stanley 261-68. Davison's proposal receives 
acceptance and further argumentation from A. Thornton, Homer's Iliad: Its 
Composition and the Motif of Supplication (=Hypomnemata 81 [Gottingen 
1984]) 46-63. I do not include in this discussion the tripartite divisions 
proposed by C. M. Bowra, in A. J. B. Wace and F. H. Stubbings, edd., A 
Companion to Homer (London 1962) 43; Drerup (supra n.8); Sheppard (supra 
n.8); B. Andreae and H. Flashar, ·Strukturaequivalenzen zwischen den 
homerischen Epen und der friihgriechischen Vasenkunst," Poetica 9 (1977) 
217-64; and M. Silk, Homer: The Iliad (Cambridge 1987) 37, because these 
scholars do not claim that their schemes of division shaped performances. 
Sheppard divided the Iliad into three "movements" consisting of Books 1-10, 
11-18, and 19-24; Andreae and Flashar divide it into three sections consisting 
of Books 1-7, 8-18, and 19-24; Silk prefers 1-9, 10-17, and 18-24. C. H. Whit­
man, Homer and the Heroic Tradition (Cambridge 1958) 329 n.46, suggests 
that Sheppard's scheme was plausible for performance. Kirk (45f) endorses a 
rough three-movement division, without specific breaks and not tied to per­
formance. Bowra's division, consisting of Books 2-8 (preceded by Book 1 as 
prologue), 9-15, and 16-22 (followed by Books 23 and 24 as epilogue) is the 
closest to what I shall suggest in the second part of this paper. For discussion 
of proposals to divide the Iliad into three movements for performance, see N. 
J. Richardson, The Iliad: A Commentary VI: Books 21-24 (Cambridge 1993) 
2ff. Stanley (265) suggests that his long middle movement might have been 
divided into two to create a four-day performance. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 9 

rhapsodes. 11 For Stanley the criterion is formal: his umts of 
performance derive from and are identical to units that he 
argues are constituted through patterns of verbal, narrative, and 
thematic repetition in the text. Schadewaldt and Taplin base 
their performance divisions upon the narrative divisions of the 
text. As Taplin explains (18f, 26), 

Once the handling of narrative-time is set out ... one feature 
stands out as by far the most remarkable: the huge central day 
lasts all the way from book 11 to book 18 .... It is held to­
gether as a narrative unit by Zeus' promise of success to Hek­
tor until the sun goes down .... From the appreciation of the 
massive structural unity of the central day follows the observa­
tion that the poem as a whole is constructed in three parts, or 
three movements .... My theory is, simply, that the three part 
structure matches and arises from Homer's own performance; 
and that the two clusters of forward-markers to 'tomorrow' 
in narrative-time also serve as forward-markers in performance­
time. 12 

For Wade-Gery (16), the first to propose a three-movement 
performance, the explicit criterion was thematic, but his words 
indicate that he was thinking of narrative divisions as well: 

The first performance ends with the fruitless embassy: the 
second, when the death of Patroklos has had its full reverbera­
tion: the third, with Hektor's ransoming. In the poem, thus 
articulated, the main theme is now seen clearly to be Achilles: 
at the end of each day's performance we have reached (and 
fully explored) a vital stage of his experience. 

The common respect for prominent narrative divisions in deter­
mining the performance divisions may explain why the 
schemes of Wade-Gery, Schadewaldt, and Taplin, though not 
identical, are quite similar to one another. 

None of the above schemes of performance division has won 
acceptance from a consensus of scholars-no doubt largely due 
to the aforementioned skepticism regarding any proposals for 

11 Kirk (44-47) skeptically discusses four-book divisions but without refer­
ence to performance. 

12 For Schadewaldt (supra n.lO: 24) night in the narrative signifies the end of 
a day of performance: "Zweimal in dem Gedicht schliesst mit einer Nacht ... 
ein Stuck der Handlung .... Danach mag die I1ias an drei Festtagen vorge­
tragen worden sein .... Wenn ein Vortragstag zu Ende ging, war auch im 
Gedicht Abend geworden; wenn man morgens wieder begann, stieg auch in 
dem Gedicht wieder die Morgenrote auf." 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

10 THE THREE MOVEMENTS OF THE ILIAD 

fixed performance divisions. But among those who do feel that 
the investigation of plausible performance divisions is worth­
while, the lack of consensus would seem to indicate unease 
with the aesthetic and thematic implications of the divisions pro­
posed. Perhaps the problem lies with the criteria used by the 
authors of these proposals to derive and support them. For 
none of these scholars have based their schemes on aesthetic 
principles reculiar to situations of performance. The narrative 
and forma divisions identified by Taplin, Schadewaldt, Wade­
Gery, and Stanley, if for the sake of argument we grant their 
validity, can be found in the text regardless of how the text is 
performed or whether it is performed at all. Davison's divisions 
are indeed based upon considerations of performance, but only 
upon its mechanics (assignment of segments to rhapsodes, a 
matter about which we actually have no evidence whatsoever). 
Taplin alone has observed that the situation of performance 
makes special aesthetic demands. He does this when he claims 
that the performance divisions he has proposed would have 
stimulated audiences to return for the second and third days of 
recitation: «Anyone who had heard the Iliad up to book 9 or up 
to book 18 would be very reluctant to stop with the closure at 
those junctures."13 Nevertheless, Taplin has not undertaken a 
systematic analysis of beginnings and endings of performances, 
and he assumes that divisions in the performance coincide with 
those in the narrative. 14 

Once we turn full attention to the relationship between narra­
tive divisions and performance divisions in performances with 
which we are familiar, it rapidly becomes apparent that although 
these two types of divisons may coincide, they need not do so, 
and that they are often found in tension with one another. The 
examples of motion picture serials and television soap operas 

!3 Taplin 26; Stanley (263) also briefly considers how his proposal would 
flow as a performance. 

14 In the discussion that follows the terms "performance" and "narrative" 
roughly correspond to the terms" recit" and" histoire" in the fundamental 
narratological studies of G. Genette. Narratologists in fields other than 
classics usually translate these terms into English differently; I have main­
tained usage of "performance" and "narrative" in the senses familiar to 

classicists from existing treatments of Homer, especially that of Taplin. In 
narratology "histoire" (often translated by "story") denotes the actual or 
fictitious e'lJents that a story is about, "recit" (often translated as "narrative" or 
"discourse") the telling that signifies those events. See Genette, Narrative Dis­
course, tr. J. E. Lewin (Ithaca 1980) 27, and S. Richardson, The Homeric 
Narrator (Nashville 1990) 3, 209 nA. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 11 

readily come to mind. In these dramatic formats a virtually un­
breakable rule dictates that performances should end at a point 
that interrupts the movement of the narrative toward a climax, 
and begin again as the anticipated crisis is approached. For a 
paradigm closer to Homer, we can turn to Athenian tragedy, 
where choral interludes frequently interrrupt the action at sus­
penseful points, such as Soph. 0 T 1086, when Jocasta has 
rushed into the palace and the arrival of the shepherd is awaited, 
and Eur. Med. 1250, when Medea rushes into the palace to kill 
her children. Similar phenomena may be found in the text of 
the Iliad itself. The ancient scholiasts often observed Homer's 
skill in creating an effect of suspense;lS such effects need not 
coincide with performance divisions, but they may at least 
indicate that pauses in performance could also have been 
suspenseful rather than c1osural. 

Yet two performance divisions in the Iliad can be determined 
with near certainty: the beginning of the first day's performance 
and the end of the last. 16 Homer does not begin the Iliad with 
the Wedding of Peleus and Thetis, the birth of Achilles, the 
Judgment of Paris, the abduction of Helen, or the beginning of 
the Trojan War. As everyone knows, Homer plunges in medias 
res, in the tenth year of the war. Nor does he begin at sunrise, 
or offer any indication of the time of day. Homer begins the nar­
rative of the Iliad at a moment of crisis, the quarrel between 
Achilles and Agamemnon (1.M), from which he backtracks to 
an equally critical moment, Agamemnon's rejection of Apollo's 
priest Chryses (1.11f), and then finally to his extended narra­
tion of Chryses' supplication of Agamemnon. 17 The abruptness 
of this beginning becomes the more apparent when we com­
pare the way Achilles begins his account for Thetis of the 
quarrel with Agamemnon (1.366-69): 18 

15 For examples and discussion see N. J. Richardson, -Literary Criticism in 
the Exegetical Scholia to the Iliad, " CQ N.S. 30 (1980) 269f. 

16 Performances of Homer sometimes began with hymns to the gods, and 
may have been followed by recitations from the epic cycle. Such extensions of 
the performance, if they were actually contemporary with the composition of 
the Iliad, need not have obscured the points at which the performance of the 
Iliad itself began or ended. 

17 On -epic regression" see T. Krischer, Formate Kon'lJentionen der homer­
ischen Epik (=Zetemata 56 [Miinchen 1971]) 136-40. 

18 j. Griffin and M. Hammond, -Critical Appreciations VI: Homer, Iliad 
1.1-52," GaR 29 (1982) 128£, and 1. j. F. de jong, -Iliad 1.366-392: A Mirror 
Story," Arethusa 18 (1985) 13-16, have many illuminating observations about 
the differences between Achilles' narration and that of the Homeric narrator. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

12 THE THREE MOVEMENTS OF THE ILIAD 

c;>xoj.l.e9' it; 81l~T1v, iepTtv 1tOAlV 'Huirov<><;, 
-rilv oe ol£1tpo.9oj.l.EV 'tE lC<X1. "lY0j.l.£V £V9o.0E 1to.v't<X. 
lC<X1. 'ta j.l.tv elS Oo.oooV'to j.l.e'ta O<p{OlV utEt; 'AX<XlWV, 
£lC 0' EMv 'A'tpeion XPUOTJ{o<x lC<xUmo.pnov.19 

Achilles begins at the beginning, the sack of Thebe, because 
otherwise Thetis might not understand why Chryses sup­
plicated Agamemnon in the first place. But Homer, more 
interested in arresting his audience's attention, begins his rer­
formance at a critical moment well past the starting point 0 the 
story, and fills in the omitted narrative information later. 

Likewise the Iliad does not conclude with the death of 
Achilles or the fall of Troy, although its narrative emphatically 
points ahead to these events. Nor docs it end at sundown, bed­
time, or any other point in time designated as a termination of 
movement in the narrative. On the contrary, the Iliad ends 
some time after sunrise (24.784-89): 

£VVllj.l.<XP j.l.ev 'to{ YE UY{VEOV a01tE'tov UAT1V' 
',\ 'I' " ~,~ , " , A ' , <X!\,!\, O'tE on oelC<X'tTI E<o<xvn <O<XEqt~VPO'tO<; nco<;. 

lC<X1. 'to't' ap' £~ECPEPOV 9p<xouv "ElC'top<x Oo.ICpU XEOV'tEt;, 
tv oe 1tUpU lmo.'tU VEICPOV 9 EO<XV , £V 0' E~<XAOV 1tUp. 
n~<><; 0' UP\'YEvEl<X <oo.vu pOOoOUIC'tUA<><; UW<; 
'tl;j.l.0t; ap' Uj.l.<pl. 1tUPTtV lCAU'tOU "EIC'topot; aypE'to A.<X0t;.20 

Homer does not allow even the cremation of Hector to appear 
as a point of rest in the narrative, although it is the point at 
which his performance stops. He reminds us not only that the 
battle will resume, but that it may resume at any time, despite 
the promise of Achilles to Priam (24.654-70) that the period of 
mourning for Hector and the resumption of the battle would 
be marked off from one another by a firm division (24.799f): 

19 

20 

-We went to Thebe, the holy city of Eetion, and we sacked it and 
brought all the booty here. The sons of the Achaeans fairly 
distributed it among themselves, selecting Chryseis, whose 
cheeks are beautiful, for the son of Atreus." 
-For nine days [the Trojans] gathered a huge amount of 
firewood. But when dawn, which brings light to mortals, 
appeared on the tenth day, with tears streaming down their cheeks 
they carried out Hector, the brave. They set his body 
on the top of the pyre, and set fire to it. And when the rosy 
fingers of early-born dawn appeared, the people gathered 
around the pyre of Hector, the famous." 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 

ptllCPa of: 01111' £x£av' 1t£pl of: O"lco1tol £'la'to mxv't!1, 
11" 1tplv ecpoPIl1l8£l£v il)1CV1llll0£~ 'AXalOUl 

13 

The Trojans' anxiety lest the fighting exceed the temporal frame 
assigned to it seems to project within the narrative Homer's 
own practice of running the story over the frame of the per­
formance. 

The examples of the only points in the Iliad where perfor­
mances are certain to have begun or ended suggest that in this 
work Homer may have preferred a tension between the bound­
aries of the performance and those of the narrative. 22 Of course 
this does not prove that the internal divisions of the three-day 
performance must have displayed the same tensions, but it 
strongly undercuts the assumption of Taplin and Schadewaldt 
that the breaks between the days of performance should have 
coincided with breaks in the narrative. On the contrary, it sug­
gests that it would be worthwhile to seek plausible breaks in 
performance that interrupt the flow of the narrative and pro­
duce a sense of openness at both beginnings and endings. 

Stanley's theory of performance division, like that of Taplin 
and Schadewaldt, assumes that openings and closures in per­
formance would coincide with textual openings and closures of 
another type: patterns of repeti tion. Of course any validity Stan­
ley's proposal may have must depend upon the validity of the 

21 ~[The Trojans] built the burial mound in haste. And sentries 
were posted on all sides, lest the well-greaved Achaeans attack 

too soon." 
22 Examination of the beginning and end of the Odyssey reveals a some­

what different tendency. The Odyssey of course does not begin with the sack 
of Troy, although that is where Odysseus begins his own narrative of his 
wanderings (,It..l69£v, 9.39). Homer's Muse begins ·somewhere within the 
narrative" (tWV CxIlOe£V, 1.10). The moment chosen is somewhere between 
Poseidon's departure for the Ethiopians and his return, in the window of 
opportunity when Athena and Zeus can plan the homecoming of Odysseus 
without interference. But because at this time the situations of Odysseus on 
Ogygia and Penelope in Ithaca have remained essentially unchanged for a 
long while, the beginning of the Odyssey is considerably less abrupt than that 
of the Iliad. The ending of the Odyssey has rather a stronger sense of closure 
than that of the Iliad, although it stops short of the end to Odysseus' 
wanderings and of his death, which were foretold by Teiresias (11.119-37). 
The termination in fighting at the end of the Odyssey occurs only because 
Athena intervenes to change the direction of the narrative: Odysseus and 
Telemachus would have killed all the allies of the suitors, if Athena had not 
shouted out and held back the whole host (24.528ff). Thus at the end of the 
Odyssey coincidence between narrative and performance closure is achieved 
by placing within the narrative a goddess who violently interrupts its forward 
movement, a dea ex machina as it were. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

14 THE THREE MOVEMENTS OF THE ILIAD 

formal patterns he claims to have found, a matter which at this 
time remains controversial. But even if Stanley's formal scheme 
of the Iliad is correct, it does not go without saying that Ho­
mer's scheme of performance would have harmonized with 
this scheme rather than conflicted with it. It is true that formal 
links between Books 1 and 24 coincide with the beginning and 
end of the whole epic performance at these points. But the 
strong sense of closure thus achieved might have been per­
missible or desirable precisely because Book 24 ends the last 
day of performance, when the sense of continued forward 
momentum could be diminished. Such a sense of closure might 
not have been as desirable at the end of performance day 1 or 
day 2. Stanley himself has argued (260f) that c10sural elements 
are sometimes postponed beyond the book-end where regular 
patterning would place them. By the same logic it should at least 
be possible that larger design patterns as well might not exactly 
coincide with larger performance movements. 

There are additional reasons to be skeptical of Stanley' scheme 
as well as the others. One of these concerns the lengths of the 
performances yielded by the various schemes of division. The 
following table shows the number of lines in each performance 
unit that has been proposed: 23 

I II III 

Wade-Gery 5,691 5,984 4,018 

Davison 4,978 5,583 5,132 

Schadewaldt 5,691 5,669 3,754 

Taplin 5,691 5,405 4,018 

Stanley 4,413 6,909 4,371 

A glance at the table reveals that in each scheme the third day of 
performance is much shorter than the second; Davison's is the 
only scheme in which the third day of performance is not the 
shortest. Our sense of Homeric aesthetics, however, should 
lead us to expect that Homer's last day of performance would 
have been his longest, for Homer characteristically creates em­
phasis in his compositions by expanding the dimensions of 
important segments. Thus although many a warrior's death in 
the Iliad is prepared for, described, and reacted to in only a few 

23 Line totals are based on the standard enumeration (i.e ., including brac­
keted lines), except that Schadcwaldt and Taplin omit Book 10. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 15 

lines, that of Hector is developed to the scale of an entire 
book.24 As the conclusion of the epic is its most emphatic point, 
here too one finds Homer at his most expansive; Book 24, one 
of the longer books at 804 lines, is entirely devoted to develop­
ment of a single event, the redemption of Hector's body. Seven 
books have over 800 lines, but only two such books follow one 
another in sequence: Books 23 and 24.25 

We may also note that in classical poetics and rhetoric it is 
often (though by no means invariably) the case that in a series of 
three members the last is expanded to greater length than the 
first two and thus emphasized. In Homeric versification one 
observes this in the important verse-type that Kirk calls (20f) 
the "rising threefolder." Homer's use of such verse at positions 
of great emphasis (the ends of speeches, for example) indicates 
his conscious valuation of the three-part crescendo as an 
emphatic rhythmical cadence. A similar phenomenon in larger 
units of composition may be observed in Odysseus' narration 
of his adventures at Od. 9, 10, and 12. In each of these books 
Odysseus narrates three adventures, the first two of ap­
proximately equal length, the third much longer. 26 In Book 9 he 
relates his encounter with the Ciconians (22 lines), then the 
Lotus-Eaters (22 lines), and finally the island of the Cyclopes 
(461 lines). In Book 10 he narrates his encounter with Aeolus 
(79 lines), then the Laestrygonians (52 lines), and finally Circe 
(441 lines). In Book 12, after Circe warns him of the challenges 
to follow, he first encounters the Sirens (36 lines), next Scylla 
and Charybdis (59 lines), and finally he comes to Thrinacia, the 
island of the Cattle of the Sun (159 lines). The pattern is varied 
by a short coda when Odysseus, having lost his ship, is carried 
on a piece of debris back to Charybdis (21 lines). Homer's 
tendency to create emphasis through a rhythm of three mem-

24 On emphasis through expansion see M. W. Edwards, Homer, Poet of the 
Iliad (Baltimore 1987) 74; R. P. Martin, The Language of Heroes (Ithaca 1989) 
206-19, analyzes expansion of formulaic expressions in the great speech of 
Achilles to the Embassy. 

2S Many scholars doubt that the twenty-four -books" of the Iliad in the 
manuscripts correspond to units of the original composition. I retain the tradi­
tional designation of -books" for the sake of clarity and convenience. One 
could equally well say that the episodes of the funeral of Patroclus and the 
redemption of Hector are amonf the most amply expanded in the poem. For 
a recent defense of the traditiona division into books, see Stanley 249-61. 

26 C. Rothe, Die Odyssee als Dichtung und ihr Verhiiltnis zur !lias (Pader­
born 1914) 72; for a recent discussion and bibliography see G. W. Most, -The 
Structure and Function of Odysseus' Apologoi," TAPA 119 (1989) 21. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

16 THE THREE MOVEMENTS OF THE ILIAD 

bers, of which the last is the longest, therefore suggests that in a 
three-movement performance the last movement might well 
have been longer than either of the first two. We must feel 
uneasy with the short final movement implied by the perfor­
mance schemes of Wade-Gery, Davison, Schadewaldt, Taplin, 
and Stanley. 27 

Thematic significance is a final criterion that focuses unease 
with the proposed divisions. When we remember that in a 
three-day performance each day's recitation would itself have 
been monumental, it would seem desirable that the audience's 
thoughts and feelings be oriented toward the epic's most 
important themes at each beginning and end, lest they be lost 
from sight amid the excitement of the narrative and the 
mayhem of the battlefield. Although the proposed divisions do 
sometimes begin or (more often) end a day's performance on a 
thematically significant note, no coherent patterns of emphasis 
emerge. Even if we concede that Wade-Gery's endings at 9 and 
18.353 focus upon the theme of Achilles, I fail to see that this or 
any other particular themes are emphasized in common by the 
episodes he proposes to start the second and third days of 
recitation, the Doloneia and the Shield of Achilles, in neither of 
which Achilles appears at all. Beginning the second day with 
Book 11 (the aristeia of Agamemnon, Patroclus' meeting with 
Nestor) offers little improvement in this respect. Stanley's 
endings at Book 7 (single combat of Ajax and Hector, building 
of Achaean trench and wall) and Book 17 (battle over Patroclus' 
corpse) likewise do not seem to offer audiences a coherent 
thematic orientation toward the massive poetic creation they 
were hearing; Stanley's beginnings at Books 8 and 18 are no 
better in this regard. Davison's scheme of eight book divisions 
yields an emphatic opening for the second day in Book 9, but 
this seems thematically uncoordinated with Book 17, which 
opens his third day. Of course no law of composition would 
have required Homer to emphasize important themes in the 
manner suggested. Nevertheless, the thematic incoherence of 
the beginnings and ends proposed cannot inspire confidence 
that they are Homer's. 

27 Cf Richardson (supra n.10: 3): ·One might object that divis ion into 
books 1 to 9, 11 to 18, and 19 to 24 would make the last movement noticeably 
shorter than the other two, but that is not necessarily a drawback." The short 
final movement clearly makes him uneasy. I would agree, however, that it is 
not necessarily a drawback. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 17 

II 

Perhaps the problem can be reformulated. If we adopt the 
hypothesis that the Iliad was performed over three days, then 
each day's performance had a beginning and an end. Let us call 
the three beginnings B., B 2, and B) and the three endings E., E 2, 
and E). As each B must immediately follow the preceding E, the 
three hypothetical performances of the Iliad can be schema­
tized as follows: 

Bl -------------------------------- El 

II B2 (=E1 + 1) ---------------------E2 

III B3 (=E2 + 1) --------------------- E3 

Fortunately, two of these elements are already known, Bl and 
E), the points at which the entire epic begins and ends, Books 1 
and 24. Pcrhaps the unknowns B2 and B) can be extrapolated 
from the characteristics of the known B h and the unknowns E 1 

and E2 from the characteristics of the known E). Of course the 
composer may not have felt that he wanted to begin his second 
performance in a manner similar to that with which he began 
his first; he may have preferred to follow a principle of poikilid. 
("variation "). Poikilia, howevcr, is not an especially evident fea­
ture of Homeric composition, which tends to repeat a limited 
repertoire of typical narrative situations with fairly subtle varia­
tions occurring within the basic formats.28 So it is worthwhile to 
assume experimentally that the composer employed a typical 
format for beginning each rerformance and likewise a typical 
format for each ending. I he did, then perhaps the distin­
guishing features of each format may be discovered in the one 
beginning and one ending that we are sure of. 

The results of this procedure can be considered more plau­
sible than the previous proposals only if the following condi­
tions are met. (1) The characteristics used to identify Books 1 
and 24 as potentially typical of Iliadic beginnings and ends must 
be very prominent and distinguish these books from both one 

28 On typical motifs in the Iliad see G. S. Kirk, The Iliad: A Commentary II: 
Books 5-8 (Cambridge 1990) 15-27; M. W. Edwards, The Iliad: A Commen­
tary V: Books 17-20 (Cambridge 1991) 11-15, 19-23, 24-41. Of course Homer 
employs typical expressions for beginning speeches (e.g. 'tov l)' 0:7ta~Elp6~EVO<; 
7tPOOt<j)l1 "pdcov 'Aya~t~vcov 'Vel sim.) and ending them (e.g. co<; E<pa't'/<p<loav, 
,,'tA.), as well as for beginning and ending scenes; on the latter see Stanley's 
discussion (249-61) of the formal characteristics of the traditional book­
divisions. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

18 THE THREE MOVEMENTS OF THE ILIAD 

another and almost all other books of the Iliad beyond any 
reasonable doubt. 29 The characteristics should not therefore be 
formal elements frequently repeated in early Greek epic com­
position, such as typical scenes and formulae, for these would 
not be peculiar to any given episodes and thus could not 
characterize them. Nor should they be details of debatable 
significance. Plausibly defining characteristics must be unusual, 
dominant features of the narrative that would be obvious to 
anyone who was following the story. (2) The number of books 
with which Book 1 shares its distinguishing characteristics must 
equal exactly the number of books with which Book 24 shares 
its distinguishing characteristics. Otherwise the procedure will 
yield an unequal number of beginnings and ends, which is 
impossible. (3) The beginnings extrapolated from Book 1 must 
interlock perfectly with the endings extrapolated from Book 
24-ie., B2 must equal E t + 1, etc. 

The single most prominent characteristic of Book 1, I find, is 
Achilles' decision to withdraw from battle, followed by Zeus' 
promise to bring suffering upon the Greeks, thus setting the 
entire narrative on its course. All other events in the book 
provide a setting for Achilleus' withdrawal. As the composer 
began the first day of performance with a crisis in which 
Achilles' decision determines the forward movement of the 
plot, the hypothesis is that the second and third days of 
performance began likewise. Of the twenty-four books, 
Achilles plays a major role in only ten: 1, 9, 16, 18-24. Achilles' 
absence from Books 2-8 is determined by his withdrawal in 
Book 1. His virtually complete absence from Books 10-15 is 
determined by his rejection of the embassy in Book 9. His 
frequent, almost constant involvement in Books 18-24 is 
determined by the death of Patroclus in Book 16, which results 
from Achilles' decision, narrated in that book, to send his friend 
into battle in the first place. Achilles' announcement in Book 18 
that he will return to battle to take vengeance on Hector, and 

29 Again, I use the term -books" for the sake of convenience. It is clear that 
the epic begins and ends with coherent episodes devoted respectively to 
Achilles' angry withdrawal from the battle and his redemption of Hector's 
corpse. The dimensions of these episodes are identical to those of Books 1 and 
24 In the manuscripts. Other coherent episodes, also identical to transmitted 
-books," can also be identified in the epic, e.g., the embassy to Achilles 
(=Book 9) and the exploits of Patroclus (=Book 16). For the purposes of this 
discussion it is necessary to assume only that each day's performance began 
and ended with a coherent episode, by analogy with the coherent episodes 
with which the epic as a whole begins and ends. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 19 

his reconciliation with Agamemnon in Book 19, are not 
decisions of the same magnitude of effect as that taken in Book 
16. In fact they are not decisons at all; once Achilles learns that 
Patroclus has been killed he feels compelled to return to the 
fighting. Thus the Iliad contains exactly three books in which 
Achilles makes a decision that determines the subsequent 
course of the narrative, Books 1, 9, and 16.30 As Book 1 began 
the first day's performance, the process of analogy suggests that 
Book 9 began the second day's performance and Book 16 the 
third day's.31 

The most prominent characteristic of Book 24, I find, is Zeus' 
decision to proclaim a decree that confronts and overcomes the 
wishes of Hera and, more importantly, those of Achilles; in so 
doing, Zeus determines the close of the narrative by arranging 
that the body of Hector should be returned to his people for 
burial, and he arranges the exchange of the body for ransom. It 
is true, of course, that Book 24 is the occasion of the most 
profound moments of human pathos and insight in the epict, 
the speeches of Priam and Achilles to one another and the 
lamentations of Andromache and Helen over the corpse of 
Hector. Nevertheless, these insights and the speeches that 
manifest them only occur because Zeus has created the 
conditions in which they can occur. There is no end in sight to 
Achilles' savagery when the Olympians debate the fate of 
Hector's corpse at the beginning of the book; that is why Zeus 
must intervene by sending Thetis to convey his will to her son. 
Moreover, Priam can appear in Achilles' lodge to deliver the 
ransom and receive the body only because Zeus has dispatched 
Iris to persuade him to go, has sent an omen to assure him that 
he will be safe, and provided Hermes to serve as his escort. 
Thus at the conclusion of the Iliad, Zeus appears as the force 
powerful enough to blunt the will of even the most powerful 
mortal, and as the force that furnishes mortals with the 
opportunity for their most profound emotions and insights. 32 

30 Cf Janko (supra n.2: J11): "The poem's structure is based on advances 
and retardations: books 1, 9, 16, 22, and 24 are the main advances .... " 

31 Cf Bowra in Wace and Stubbings (supra n.l0: 43): Book 9 marks the 
transition to the second section of the epic and Book 16 the beginning of the 
third. 

J2 In this regard we want to remember that the suggested three-day 
performances may well have occurred at festivals sacred to the Olympians, 
and that Homer and his audiences thought of the Muses, Zeus' daughters, as 
indispensable to the existence of the poem. On the Muses see A. Ford, Homer: 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

20 THE THREE MOVEMENTS OF THE I LlAD 

Homer displays the power of Zeus in almost every book of 
the Iliad. Nevertheless, two points far exceed all others in the 
extent of power displayed: the two speeches in which Zeus 
foretells to Hera and to Homer's audience the chief events that 
he has decreed for the future course of the Trojan War and of 
the entire narrative to follow, i.e, the .1lOC; llouAll. The power of 
these speeches is tremendous, because Zeus' word once 
spoken is irrevocable. In telling Hera what is to happen, Zeus' 
very act causes the events to happen, and Homer's audience 
hears his very words re-enacted by the rhapsode. And Zeus 
speaks these words without hesitation, deliberation, or such mis­
givings as he expresses before the deaths of Sarpedon and 
Hector. roc; yap 8ta<pu'tov fa'tl. The critical provision that 
Patroclus must die is indicated only in a prepositional phrase 
and merely by the participle 8uvoV'toc;, placed with subtle 
emphasis at the end of the hexameter (8.476). Such is the 
efficacy of Zeus' speech. 

Achilles never learns of Zeus' plan. Nevertheless, as soon as 
Zeus proclaims to Hera that the Greek cause cannot prevail 
until Achilles returns to battle after the death of Patroclus, the 
audience instantly recognizes that Zeus' plan will destroy 
Achilles' most earnest hopes even as it fulfils the 'letter' of the 
agreement solemnized in Book 1. In the speeches proclaiming 
the Plan of Zeus, Homer first reveals the gap between what 
Achilles has proposed and what Zeus will dispose. Thus these 
speeches, like the events of Book 24, display the power of Zeus 
confronting and overwhelming that of Achilles. 

The first of these speeches occurs in Book 8, the second in 
Book 15. 33 Therefore the process of analogy with Book 24 sug­
gests that Book 8 concluded the first day of performance and 

The Poetry of the Past (Ithaca 1992) passim. For discussion of the implications 
of the sacred occasion and divine inspiration of Homer's performances see B. 
Heiden, ·The Ordeals of Homeric Song, H Arethusa, forthcoming. 

J) These 'books', like Books I, 9, 16, and 24, are coherent episodes of narra­
tive. 'Book 8' begins when Zeus convenes the immortals on Olympus and 
forbids anyone from interfering with his plan for the battle. Zeus' resistance 
to interference from Hera remains the theme until a new one is introduced 
with Agamemnon's distress and the embassy to which it leads (,Book 9'). 
'Book IS' begins when Zeus awakens from sleep to reassert his control over 
the battlefield. His supervision remains the theme until a new one is 
introduced when Patroclus arrives at the lodging of Achilles ('Book 16'). Thus 
'Book 8' and 'Book 15' may be treated as coherent bodies of narrative that 
correspond with the coherent episode 'Book 24'. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

BRUCE HEIDEN 21 

Book 15 the second day.34 This would give the following 
scheme of divisions: 

II 

III 

1 ------ 8 

9 ------ 15 

16 ---- 24 

It is readily apparent from the table that the process of extrapo­
lation yields only two plausible beginnings and two plausible 
ends and that these interlock perfectly. Each day's performance 
would therefore have begun with a dramatic assertion of Achil­
les' will and ended with a dramatic counterassertion of Zeus'. 
Thus the proposal satisfies each of the three conditions we 
established for plausibility. 

This division is extremely coherent thematically; the progres­
sion of the epic as a whole, and that of each of its three move­
ments, would recapitulate that of the proem, where the state­
ment of Achilles' wrath and its consequences (1.1-5) is abruptly 
concluded by the assertion that the will of Zeus was fulfilled 
(~tO~ ()' rt£A£l£tO ~OUA~. 1.5). But it also provides performances 
of aesthetically pleasing proportions (as determined by analogy 
with Homeric practice elsewhere) and at least moderately sus­
penseful performance breaks. This proposal yields perfor­
mances of the following lengths: 3s 

II 

III 

1 ------ 8 

9 ------ 15 

16 ---- 24 

4,978 lines 

4,716 lines 

5,999 lines 

The third day's performance has the greater length desirable for 
a bravura climax. Performance would halt on the first day with 
the Trojans restlessly encamped on the plain and anticipating 
victory when battle resumes. Performance on the second day 
would begin immediately amidst the busy night that follows the 
second day of battle, at a critical juncture, just as it did on the 
first. The conclusion of performance on the second day would 

J4 Cf. Bowra (supra n.)l), who says nothing explicit about the ends of sec­
tions, but whose view that Books 9 and 16 began the second and third 
sections respectively implies that Books 8 and 15 ended the first and second 
sections. 

JS Line totals are based on the standard enumeration and assume the 
authenticity of Book 10. Exclusion of Book 10, however, would not affect my 
argument, which aims only at determining points at which the performances 
may be plausibly supposed to have begun and ended. 



Reproduced with permission of the copyright owner.  Further reproduction prohibited without permission.

22 THE THREE MOVEMENTS OF THE ILIAD 

interrupt the narrative of the third day of battle just as the 
critical juncture, the firing of the Achaean ships, is about to be 
reached, with Patroclus left hurrying midway between the 
lodging of Eurypylos and that of Achilles (15.401-05). The third 
day of performance would begin immediately, as the first two 
had, at a moment of dramatic crisis, as Patroclus arrives at 
Achilles' lodge. Not only would audiences return for another 
day of performance after these breaks, they would be in their 
places for the very first lines of recitation. 

The hypothesis that the characteristics of internal points of 
pause and resumption could be extrapolated from the charac­
teristics of the whole epic's beginning and end, has uncovered a 
surprisingly clear and consistent articulation of episodes in 
which Achilles asserts his will with those in which Zeus asserts 
his. There is nothing implausible about a performance of the 
Iliad with major pauses after Books 8 and 15. Imagining such a 
performance enhances our sense of the epic's structure, drama, 
and significance. We shall never know for sure how the com­
poser of the Iliad intended his masterpiece to be performed. 
But if we have found a unique way to structure performance of 
the Iliad that truly does it justice, we may at least conjecture that 
Homer could have had the idea first. J6 

THE OHIO STATE UNIVERSITY 

July, 1996 

36 My thanks to Professors Mark Edwards, David Konstan, Joseph Reed, 
Thomas G. Rosenmeyer, and the anonymous referee for GRBS for their 
helpful comments on earlier versions of this article.