Oral Tradition, 9/2 (1994): 371-389

Homer’s Style:
Nonformulaic Features of an Oral Aesthetic
Joseph Russo

From Oral to Aural and Back Again

In the Homeric epics we have a text created within a highly traditional

diction, a special poetic language, for performance before a large public
situated entirely within that tradition. We do not have poetic language in our
modern sense, that carefully honed personal and private idiom meant for the
eye and (to a lesser extent) the ear of a small number of connoisseurs.
Therefore those who make up Homer’s modern audience need to know if
there is a certain ideal way to hear, or read, and respond to certain stylistic
habits of his that our experience of modern literature has not prepared us to
understand very well. That is Question One, and the important one to
answer if we are interested in experiencing Homeric poetry in its full
complexity and idiomatic richness.
Are the Iliad and Odyssey genuine oral compositions? That is
Question Two, which I believe it is not, and may never be, possible to
answer with absolute certainty. For all the disagreement and verbal combat
over this issue—from Parry’s earliest critics in the1930’s to the pages of the
New York Review of Books from March 5 to June 25, 1992—the fact is that
recovering the exact genesis or technique of Homer’s composition will
always be beyond us. Therefore knowing exactly how he composed, just
how much of his verse came from improvisation while performing and how
much from prior memorization, and whether the newly available skill of
writing was used to any degree, should be less important to us than
appreciation of the distinctive and sometimes almost odd rhetoric found
throughout his poetry, and of an underlying aesthetic that can make sense of
both the distinctiveness and the oddness. Almost twenty years ago, at a
comparatists’ conference on Oral Literature and the Formula (Stolz and
Shannon 1976), I suggested we shift from emphasis on oral to aural style
372 JOSEPH RUSSO

in an attempt to pursue the aesthetics of this style rather than its genesis.
While it is still theoretically possible to doubt that Homer is an oral poet,1 it
remains beyond doubt that he is “aural” in that he composes in a style
guided by the ear and meant to be heard, a style that pleases through verbal
play based on an aesthetic of repetition and variation, and of relaxed fullness
of expression wherever the context allows it. And yet while not discarding
that emphasis on the style itself, I now believe it is fruitful to return to the
issue of orality in connection with some salient but non-formulaic features of
this style, seeking to understand all of them as counterpart phenomena to
formulas per se, and some of them as most likely generated by composition
in the act of performance as described by Albert Lord in The Singer of Tales
(1960).
It is significant and perhaps surprising that none of these features of
an oral-derived style has to do with the employment of formulas as such.
For decades the definition and analysis of the formula dominated the
argument over Homer’s orality, but the presence or density of formulas in a
text has proved ultimately to be an insufficient basis for arguments in favor
of an oral Homer.2 At this point in the history of Homeric scholarship, our
understanding of Homer’s technique may be best served by describing
certain favorite devices or tropes and explaining their shared aesthetic.3

1
The safest position is to describe the Homeric texts as “oral-derived.” See Foley
1990:5-8 and passim; 1991:22.
2
Smith (1977) offers a classic example of a traditional epic text that is formulaic
but not orally (re)composed in the act of performance. For the difficulties in using
formula density to prove orality, see Hainsworth 1964, Russo 1976. For the balance
between formulaic and nonformulaic language and Homer’s freedom to use both, see the
important study of Finkelberg 1989.
3
Of course various studies of this kind have been done before. Edwards (1966)
sharpens our awareness of Homer’s style by presenting a survey of characteristic devices of
word (primarily adjective) position, enjambement, and sentence structure as these are
related to colon structure. His overall emphasis is on the many devices of linkage, and to
the limited extent that his study is aesthetic as well as descriptive, he does well to
emphasize “the peculiar smoothness in the progression of thought in Homeric verse” (148),
which is also my concern. Occasionally his aesthetic judgement lapses into apology for a
mere “filler” that “pads out the verse,” a “meaningless grammatical link,” and the like (see,
e.g., 144-47). Yet these stylistic features embody perfectly the principle of “epic fullness,”
a term coined by Bassett (1926:134). In an earlier study of devices of linkage between
successive speeches, Bassett (1920) illuminated a related aspect of the Homeric aesthetic,
 HOMER’S STYLE 373

Common Tropes of Extension

The bulk of my paper will be devoted to the description and

explication of certain rhetorical tropes that give Homeric style its peculiar
flavor, an archaic taste for redundancy and familiarity discreetly seasoned
with variation and ornamentation. When, following Parry’s epoch-making
insight, we sought the key to Homeric oral style in the use of formulas, our
concern was to examine style in order to demonstrate the poet’s technique
for producing verses rapidly in the act of performance. In moving from an
emphasis on the generation of language to an emphasis on the aesthetic
presentation of language, I am not abandoning my belief that Homer’s style
is either oral or orally derived, but moving the focus of investigation to a
related question. Why is Homer’s style is so uniquely pleasing, and how
may the sources of its charm reside in a variety of rhetorical features distinct
from formularity but related to it through a shared aesthetic?
It is interesting to note that scholarship on Homeric language and
compositional technique has often called attention to features that are the
opposite of charming and pleasing. Homer’s awkward moments and
inconsistencies have more recently been regarded benignly as natural
products of oral genesis (Janko 1990; Willcock 1977; Gunn 1970, based on
Lord’s prior demonstrations of composition by theme). But earlier they were
viewed as compositional gaucheries that would have been avoided by a
writing poet who composed more carefully (Combellack 1965), and still
earlier as clear evidence of scribal miscopying or imperfect conflation of
multiply authored sections (see almost any page of the editions of Leaf
1900-02, Von der Mühll 1946). I refer to such small-scale features as
redundancy, confused syntax and bad grammar, anacoluthon, traditional
phraseology awkwardly transferred to new contexts, verses out of place
(because of the performer’s memory lapse or the copyist’s oversight?),
awkward or abrupt transitions, and so on. And on the larger scale of theme

the “principle of continuity,” which he pointed out was already well understood a century
ago by scholars like Bougot (Etude sur l’IIiade d’Homère, 1888) and Zielinski (Die
Behandlung gleichzeitigen Ereignisse im antiken Epos, 1901), with their principles of
“affinity” and “continuous narrative.” My study differs from these predecessors in its
focus on a range of phenomena perhaps too diverse to have been accorded equally serious
attention in previous discussions of Homeric style, and in its attempt to describe these
seemingly unrelated phenomena as all emanating from the epic impulse toward
expansiveness, which is at the heart of the oral aesthetic.
374 JOSEPH RUSSO

and plot, comparable phenomena would be the various inconsistencies—

from unfulfilled predictions and unreconciled alternatives to outright
contradictions—too well known to need repeating here. It is indeed a
curious truth that the strongest evidence for Homer as an orally composing
poet comes from the existence of these stylistic and narrative infelicities,
which seem to suggest not that our text is inartistically composed or the
product of layers of authorship, but rather that it is the transcription of a live
performance (Janko 1990:328). We shall return to a detailed consideration
of some of these “negative” features.
We shall begin, however, with those more “positive” qualities named
above, features of style that are both orally (or aurally) inspired and
aesthetically pleasing and successful as narrative devices. Consider three
phenomena actively used in the construction of phrases and sentences,
which I shall call appositional, explanatory, and metonymic extension. I
suggest that the basic epic trope, what we might call the master trope of
traditional epic phrase-making, can be conceived in its simplest essence as
Item Plus. I am referring to the wide-ranging impulse toward repetition and
expansion that earlier scholarship has identified under a variety of names
referring to different but often related phenomena: the “traditional epithet,”
“hendiadys,” the “adding-on style,” levxi~ eijrhmevnh, “parataxis,” and so
forth, as well as Bassett’s principles of “continuity” and “epic fullness”
mentioned above (note 3). My own terminology attempts to identify a single
aesthetic impulse that issues forth in three varieties of rhetorical expansion.
In plain English, appositional extension means item + slightly different
aspect of the same, explanatory extension means item + aspect that
significantly widens its reference or image, and metonymic extension means
item + expansion that serves as a natural bridge to the next (closely related)
idea. It is my contention that underlying the various stylistic tropes and the
principles named variously by past scholarship as “affinity,” “continuity,”
and “progression,” there is one major unifying impulse that shows itself in
variety of ways. This is the fundamental impulse toward repetition and
fullness.4

4
What I call appositional extension is essentially the phenomenon well characterized
by Monro 1901 in his note on 15.175, the phrase genehv te tovko~ te: “the kind of
hendiadys formed by two nearly synonymous words,” and he compares kradivh qumov~ te,
u{bri~ te bivh te, ajneivreai hjde; metalla`/~ and similar phrases, adding “The two
meanings are fused, as it were, into a single more complete conception.” While this is true
enough, my point is to emphasize the same reality from the opposite side, finding
 HOMER’S STYLE 375

Let me illustrate this feature of style with an extended passage that

renders a dialogue between a Homeric hero and a lesser goddess. At Odyssey
4.363-905 Menelaos is telling Odysseus’ son Telemachus about what he
learned from the sea nymph Eidothea. The exchange of sentences and ideas
between the hero and the goddess offers no purple patch of rhetoric, no
specially climactic exchange of speeches; rather it is typical epic diction at
its most representative. Note the many ways in which a word or idea is
either repeated or extended, and how certain extensions are tightly bound to
the next idea. I have underlined appositional extensions with a solid line,
explanatory extensions with a broken line, and metonymic extensions with a
dotted line.
And then all provisions would have perished, and the strengths of men,
-----------------
if someone of the gods hadn’t grieved for me and pitied me,
---------------------
the daughter of strong Proteus the old man of the sea, 365
- - - - - - - - - - - - - - - - - - - - - ----------------------------------
Eidothea; it was her spirit that I especially stirred,
-------------- - - - - - - - - - - - - - - - - - - - - - - - - - - -
who met me as I was wandering alone away from my comrades
-----------------
because they were constantly roaming the island and fishing

with bent hooks, and famine was wearing away their bellies—
- - - - - - - - - - - - ……………………………………………
but she standing near me spoke a speech and addressed me: 370
----------------------------
You are a fool, stranger, excessively so, and a slack-wit,
--------------------------
or do you willingly dally, and take pleasure in suffering grief?
………………………… -------------------------
How long now you are held on this island, and there is no means
----------------
you are able to find,…………………………………………
and your comrades’ heart is shrunken.
-------------
So she spoke, and then I in answer replied to her: 375
---------------------

significance in the fact that an essentially single conception is commonly expressed in

twofold fashion, in obedience to the epic inclination to fullness and redundancy.
5
In all Homeric citations I use Arabic numerals for books of the Odyssey and
Roman numerals for the Iliad. All translations are my own.
376 JOSEPH RUSSO

I shall speak out to you, for all that you are a goddess,
------------------------------------
that it is no way willingly I am held here, but ………………….
rather I must have

given offense to the gods, they who keep wide heaven.

………………………… --------------------------------------------
But you now tell to me—the gods are aware of everything—
-----------------------
who of immortals fetters me and binds me from my passage, 380
----------------------- - - - - - - - - - - - -
and the homecoming, ………………………………………how I will make it over the fishy sea?
---------------
So I spoke and she answered at once, bright among goddesses:
-----------------
Now indeed O stranger will I speak to you without guile.
------------------------------ - - - - - - - - -
A certain one frequents these parts, the unfailing old man of the sea,
---------------------------------------------------
immortal Proteus the Aigyptian, …………………… the one who knows 385
---------------------------------------------------
the ocean’s every depth, Poseidon’s underling.
……………………….. ---------------
They say he is my father and that he gave birth to me.
---------------------------------------------
If somehow you might be able to lie in ambush and to seize him,
-------------------------
he would be able to tell you the way and the measures of passage
--------------------
and the homecoming, …………………………………………
how you will make it over the fishy sea. 390
---------------

In this passage we can see the three categories of extension deployed

as natural and graceful functions of epic poetic diction, serving the crucial
principal of linkage. The exchange between hero and goddess is unhurried
and easygoing in its natural-seeming inclination to say things twice (and
occasionally thrice), as if to clarify the presentation of most ideas by re-
presenting them in other wording or under another aspect. Some of my
distinctions may be disputed—it may not always be clear whether an
extension adds a merely appositional or a more explanatory aspect to a basic
idea, or whether it moves primarily towards explanation or metonymic
connection—yet I remain satisfied that however we draw the distinctions,
the overall effect is one of a heavy reliance on doubling and metonymic
connection to give the idiom its characteristic epic grace and fullness. The
norm is to present ideas and persons with a slight redundancy, and to avoid
 HOMER’S STYLE 377

the abrupt or unexpected. It is against this normative background that the

surprising phrase or thought, when introduced, will strike us with special
force.6

More Specialized Tropes of Extension

More specialized tropes of extension may be added here by adducing

two characteristic phenomena of Homer’s style, single word appositional
doubling, or glossing, and hysteron-proteron, or prothysteron.
Single word apposition is commonly used throughout the epics to
restate the identity of a noun in terms usually more specific (a distinct trope
from the more commonly cited doublets based on synonym or hediadys, cf.
note 4 above). A complete inventory would be impossible, but
representative examples follow. It is apparent that around a centrally
important noun like “man,” ajnhvr, epic diction has accumulated something
very analogous to a formular system. Since the epic world is less a woman’s
world, a smaller system exists for the noun gunhv.7

men heroes ajndrw`n hJrwvwn Od. 4x, ajndravsi hJrwevssi xiii.346

men shepherds a[ndre~ te nomh`e~
men hunters a[ndre~ ejpakth`re~
men spearmen a[ndrw`n aijcmhtavwn Il. 2x
men shieldbearers a[ndrw`n ajspistavwn
man bronzesmith ajnh;r calkeuv~
bronzesmith men calkh`e~. . . a[ndre~
carpenter men tevktone~ a[ndre~
leader men hJghvtore~ a[ndre~
cowherd men bouvkoloi a[ndre~
suppliant . . man iJkevtew . . . ajndrov~ Il. 2x
king man basilh`i . . . ajndri;

6
The Iliad and Odyssey prologues offer good illustration of the opposite pole of
Homeric style. They show minimal repetition and extension; instead, the sentences move
swiftly and almost confusingly in their rapid changeover of subjects and swift piling up of
causes and effects. The poet sets out to excite and attract his audience by putting off his
redundant, extended, and relaxed manner and making them pay heightened attention to his
words.
7
Book and line citations are given only when significant distinctions might be made
between the two epics. The interested reader may consult the Concordances or Ibycus.
378 JOSEPH RUSSO

doctor man ijhtro;~ . . . ajnhvr

human man brotw`/ ajndri;
man hunter ajndro;~ qhrhth`ro~
Sintian men Sivntie~ a[ndre~
Arkadian men [Arkade~ a[ndre~
Thracian men qrh/`ke~ . . . a[ndre~, qrh/kw`n a[ndrw`n Il. 3x
Koan men Kovwn . . . ajndrw`n
Phoenician men Foivnike~ . . . a[ndre~
woman household-manager gunh; tamivh
woman house-mistress gunh; devspoina
woman corn-grinding slave gunh; . . . ajletriv~
woman female-day-laborer gunh; cernh`ti~
person(s) wayfarer(s) a[nqrwpo~ oJdivth~, xvi.263, oJditavwn
ajnqrwvpwn 13.123
ox bull bou`~ . . . tau`ro~
eagle bird aijeto;~ o[rni~
birds. . . vultures o{rnisin . . . aijgupivoisin
pig boar sui; kaprivw/, sui; . . . kaprivw/ Il. 2x
falcon hawk i[rhx kivrko~ 13.86-87

Since the doubling most often consists of a more narrowly descriptive

noun added to a more generic noun, this phrase habit has something in
common with the familiar noun-epithet combination so fundamental to epic
diction. Both figures embody Item Plus in the form of noun + descriptive
word. The similarity is all the more pronounced when the epithet of the
combination—Parry’s “traditional epithet”— is not an adjective but a noun,
a not uncommon occurrence, typically when nomina agentium or
patronymics are used as modifiers (e.g. Kronivdh~ Zeuv~, nefelhgerevta
Zeuv~, iJppovta Nevstwr, Phlhiadevw jAcilh`o~, Laertiadevw jOdussh`o~,
jAgamevmnono~ jAtreivdao, Kullhvnio~ ajrgeifovnth~, and so on). What I am
proposing, then, is that the trope long familar to us as the epic “noun-epithet
combination” be classified as one of several “common tropes of extension”
as described above, the epithet bestowing an extension that is always
appositional and is explanatory to the extent that it enlarges the idea or
image.8
8
After creating my threefold classification of tropes of extension and expansion, I
realized that O’Nolan (1978) had anticipated me in one point. In his “Doublets in the
Odyssey” he defines doublet as “a combination of two terms that are to all intents
synonymous” and equates it with the hendiadys noted by Monro (above, note 4). He then
explicitly connects the doublet to the noun-epithet formula as being “twin tool[s]. . . of the
 HOMER’S STYLE 379

With hysteron-proteron we encounter a more curious type of

expression through extension. This is the extension in reverse causal or
chronological order of an action through its immediate coupling with
another act that must have preceded it. Some of the results would be
humorous in their illogic or the impossible images they conjure up, were this
trope not such a fundamental piece of archaic Greek language and thought
that all scholars of ancient Greek—and we assume a fortiori all Greek
authors—habitually accept it as normal. What strikes a modern reader as out
of sequence according to strict literal or logical criteria, is for an ancient
Greek quite proper: the resultant state or final action is named first, because
it is nearer to hand, and then that which preceded and/or caused it is named
second. A far from exhaustive Homeric sampling, largely Odyssean, is as
follows. (For the Iliadic examples I am indebted to Macleod 1982.)

4.49-50 When the servants bathed them and rubbed them with oil,
then they put cloaks around them and tunics

4.207-8 . . . a man to whom Zeus

gives prosperity in his marrying and his being born

4.723 . . . of all women who were reared with me and born with me

5.264 . . . having clothed him with fragrant clothing and having bathed him

10.352-53 One of them spread fine cloths on the armchairs,

purple ones over the top, and underneath she spread linens

12.134 And their lady mother rearing them and giving them birth

13.274 I asked them to drop me at Pylos and to take me on board

14.200-1 . . . and many other

sons grew up in the palace and were born there

14.279 He [the king] saved me and took pity on me

15.188 there they slept the night and he set guest-gifts for them

16.340-41 . . . when he gave the message

epic composer’s craft,” both being “formulae of the style of heroic narrative” (22).
380 JOSEPH RUSSO

he went to the pigs, and left the yard and the palace

l6.173-74 first she put a well-washed cloak and a tunic

around his torso

xxi.537 they opened the gates and pushed aside the bolts

xxiv.206 for if he is going to capture you and see you with his eyes

xxiv.346 [Hermes, dispatched by Zeus to aid Priam]

arrived quickly at Troy and at the Hellespont

Some of these cases admittedly give less awkwardness than others. In

15.188 and 16.340-41, for example, we may feel the effect only of a mild
afterthought supplementing the main activity as if offering a piece of
background information. In most instances we feel the reversed order more
forcefully. And yet the recurrence of some of these formulations shows that
some prothysteronic expressions tended toward formular status. Forms of
the verb for being born (givgnomai) are regularly localized at the end of its
verse, and the act of putting on a heavier outer covering and a lighter under
covering—whether on couches or men’s bodies—seems to be formulaic in a
sequence that must begin with the second, outer layer and then pass to the
inner. The act of bathing seems, in comparison, less formulaically fixed in
second position. In 5.264 the bathing of Odysseus illogically follows the
maids’ clothing him, but in 4.49-50 the first sequence of actions presents
bathing followed, properly, by rubbing with oil, but then presents the
servants putting cloaks on the guests before putting on the underlying tunics.
Most prothystera can be absorbed easily enough by a reading or
listening audience, apparently because the two acts often form a closely
connected sequence in behavior that is highly familiar and to some extent
ritualized (bathing, clothing, and hospitality are often elaborated in the
familiar “typical scenes” well documented by Arend 1933). In effect they
merge in the mind as if a single two-part activity. The least familiar and
truly unique prothysteron, however, that of 13.274, occurs in a lying tale of
Odysseus that is furthest from a typical epic description, and may be an ad
hoc creation improvised (by the poet as well as by his character!) in
performance. The awkwardness is remarkable: not only does the
description lack the cushion of familiarity, but the two actions of dropping
 HOMER’S STYLE 381

off and picking up a passenger are too formally opposed to permit easy
merger into a single two-part activity.

Ambiguous syntax, bad grammar, and anacoluthon

Some of Homer’s rhetorical features can be seen as more distinctively

“negative” in the resistance to lucidity they offer the hearer or reader.
Usually on a larger scale than the tropes of extension, but far less frequent,
we find perplexing instances of ambiguous or incorrect syntax or grammar,
sometimes in the form specifically called anacoluthon, where a construction
is begun but dropped in favor of a new construction, and sometimes resumed
after an interval in an inappropriate or awkward form. From the perspective
we have established, we may understand anacoluthon as the unraveling of
tightly organized syntax in deference to the impulse to take in new matter
too quickly, an aspect of what we have called epic expansiveness or fullness.
A simple example is found in the simile comparing the Myrmidons to wasps
in Iliad xvi.264-65:9

They, having a brave heart,

each one flies forward and defends his own children

The Greek begins with a plural participle for “having,” then surprisingly
switches to the singular “each one” and a singular verb for “flies.” The
apposition of plural and singular may be felt as slightly awkward or strained,
but is not unduly distracting. But more severe cases can be cited.
At xvi.317-22 Nestor’s two sons, Antilochus and Thrasymedes, each
slay a Trojan opponent in battle. The poet begins with the plural subject,
“sons,” as if intending to describe the success of each in turn, “the one did
this, the other did that.” But it looks as if an afterthought takes over once

9
I am indebted to Janko 1992 for singling out the anacolutha of xvi in his
exemplary Commentary. Throughout his volume he shows a strong interest in oral
phenomena, but I should warn the reader that he uses “oral apposition” in a sense
different from mine, as equivalent to anacoluthon such as at xvi.317-22 (1992:359; and at
353 he calls anacolouthon “oral syntax”). Some of the anacolutha of xvi are also
discussed by Chantraine 1953:15-16, who notes that in xvi.265 the word pa`~, “each
one,” eases the transition; and he adduces 9.462-63 as showing a “similar freedom of
movement” to 12.73ff.
382 JOSEPH RUSSO

the first slaying is underway: the interesting and unexpected intervention of

the brother of the just slain Trojan, who lunges into the scene to take
revenge. The syntax is admittedly ragged and the sequence jerky, as
Atymnion’s brother Maris attacks Antilochus and is intercepted by
Thrasymedes who dispatches him. The literal translation is:

And the sons of Nestor—one wounded Atymnion with a sharp spear,

Antilochus, and drove the bronze point through his flank,
and he fell forward. But Maris from close up with his spear
lunged at Antilochus in a rage over his brother
standing in front of the corpse; but godlike Thrasymedes
got him first before he could wound, and did not miss him. . . .

The anacoluthon need not be felt as a blemish, since the slightly

disorganized and abrupt movement of the description nicely captures—
perhaps mimics—the presence of confusion and the unexpected on the
battlefield. There are, however, still stronger anacolutha to cite.
Consider xvi.401 ff., where Patroklos’ stabbing of Thestor is a
complete syntactic mess, hard to follow with the eye or the ear. I translate
literally, keeping close to the Greek word order, to capture the full confusion
of the original:

He fell with a thud. But he [Patroklos] to Thestor, Enops’ son,

in his second onrush, he sat huddled back in his well-polished chariot,
he had lost all his nerve and the reins had slipped
from his hands—him did he come up close to and stab with a spear.

The impulse to expansiveness here seems to have overrun all the boundaries
of clear syntactic organization. It is from a passage like this that we can
make the strongest case for our text as the record of a live performance, for
what writing poet, with the capacity to review critically what he had created,
would have let such language stand?
Homer’s lengthiest anacoluthic fault is probably at Od. 12.73ff., the
very long delay of 28 verses before the second member of a pair is named.
It will take a close look at the Greek to appreciate the full awkwardness here.
Verse 73 begins with “The two peaks, the one reaches the wide sky,” oiJ de;
duvw skovpeloi oJ me;n oujrano;n eujru;n iJkavnei. Both “peaks” and “the
one” are in the nominative case, a forgiveable instance of “oral syntax,”
where a writer of correct Greek prose would have put the plural in the
 HOMER’S STYLE 383

genitive case and said “Of two peaks, the one. . . .” The serious
awkwardness does not lie here, however, but in the way the second peak
seems for twenty-eight verses to have been forgotten, although “the one”
included the untranslatable particle me;n that implies that a second parallel
and related item will soon follow. But the poet instead enlarges upon details
surrounding the first peak, then upon other details to which he rambles rather
loosely. When, at verse 101, he finally comes around to mentioning the
second peak, he has of course lost any notion that it was to be in the
nominative, part of the nominative apposition used in 73. “The other peak
you will see,” Odysseus is told, to;n d j e{teron skovpelon o[yei, with the
peak now in the accusative case. Homer does retain a thin fiber of
connection by using the particle dev answering to me;n 28 lines earlier—
probably the longest deferment of dev after me;n in Greek poetry and perhaps
in any surviving Greek text!
Anacoluthon of this breadth certainly suggests a poet composing and
expanding his description in the act of performance, and the combined
presence of all the anacolutha throughout the Homeric text certainly must
count as likely markers of an oral style.10 By this I mean a style that may be
based in part on memorization and more or less faithful reproduction of
some passages, but which in passages like those we have examined is very
likely obedient to the vagaries of momentary inspiration and ad hoc creation
of verses.11
There are many more instances of careless or incorrect grammatical
constructions we could cite. The famous description in xxiv.527-28 of Zeus’
two jars containing good and evil gifts for humankind is marred by a
construction so unclear that Pindar apparently mistook Homer’s meaning

10
Chantraine’s (1953:12-21) chapter II, “La construction appositionnelle et la
syntaxe d’accord,” although not presented as an argument for oral composition, implicitly
makes a good case for Homer’s style as oral. Chantraine was one of the early scholars to
endorse Parry’s work, and his discussion of Homeric syntax seems often to assume “oral
syntax.”
11
See Willcock 1977 and 1978 for an excellent discussion of ad hoc language in
the Iliad, with the assumption Homer is an oral poet.
384 JOSEPH RUSSO

and understood that there were three jars.12 And this is only one of many
cases where we must abandon strictly grammatical readings of the text and
construe ad sensum rather than ad litteram, as the commentaries repeatedly
remind us.13
If we see anacoluthic and grammatically weak construction as signs of
oral composition on the restricted level of phrase, verse, and passage, we
can go further and add another kind of non-sequitur on the level of thematic
construction. Homer’s composition by theme and scene-type has been well
documented and hardly needs further illustration. But there exist passages
where the thematic sequence seems to have been tampered with. Although
such inconsistencies have traditionally been condemned by modern scholars
as textual corruptions (and were suspected or athetized by the Alexandrian
editors), some recent studies have shown how these incongruities may well
be inevitable in the oral recomposing of traditional material. Gunn (1970)
has two good examples of this phenomenon. He analyzes Od. 5.85-96 and
l5.130-50 and makes a good case for the text as the transcript of a
performance in which the singer has momentarily skipped a beat, moving
either too soon or too late to the appropriate verse, and thereby slightly
muddling a sequence that should have been perfectly clear. Other examples
of such orally derived awkwardness are given by Willcock (1978) and Russo
(1987). Such “mistakes” must naturally seem more grievous to the eye, on
the printed page, than to the ear of those immersed in the flow of a live
performance. One is tempted to imagine what cannot be historically
confirmed for Homer—but is in fact experienced by a modern audience in a
live performance of drama, music, or song—namely that the audience
instinctively compensates for the mistake, be it omission or intrusion, by
supplying what is needed or subtracting what is inappropriate, and soon has
forgotten that anything was amiss.
If we are fully aware of these and similar oral stylistic features of our

12
Pyth. 3.80-82. I owe this observation to Macleod 1982:133. The literal
rendering of Homer’s Greek actually does suggest three jars: “A pair of jars lies at Zeus’
threshold, of the evil gifts that he bestows, and another one of good.” It is the force of a
long tradition of construing the passage ad sensum that gives the generally accepted
picture of two jars.
13
The General Index in vol. II of Leaf 1900-02:658, for example, has an entry
“Constructio ad sensum,” listing xi.690, xiii.564, xvi.281, xvii.756, xviii.515, 525, and
xxii.84.
 HOMER’S STYLE 385

text, what does it mean for our editing of such texts? Regarding such
features as repetition, apposition, expansion, and other forms of the Item
Plus trope, we need do nothing more than relax and enjoy a slower tempo for
unfolding the segments of a story. But regarding those features of style that
have often been criticized as awkward and mistaken use of language, we
must give up the centuries-old project of emending them to make more
correct Greek or more consistent expression. Richard Janko (1990) has
recently suggested that we must always have an eye on the apparatus, wary
of modern editors’ tendency to normalize differences or awkwardnesses that
we now recognize as likely to be orally generated. He goes on to remind us
that we still possess no Homeric texts edited with a full post-Parry mentality.
Janko’s article is an excellent starting point for any future editor of a Greek
oral epic text. He gathers several paradigmatic examples where, for two
similar passages, the manuscripts offer variation in one word and the
editors—both the ancient Alexandrians or the modern Oxford scholar—have
ignored this difference and made the passages entirely uniform.

Conclusion

My arguments in the course of this paper may have some aura of

paradox about them. I have claimed that stylistic redundancy bordering on
“padding” is really an aesthetic plus (no conscious pun was intended here),
and that anacoluthon and other inconsistencies of sequence are acceptable
and in fact the natural products of a great poet’s technique. There is an
entire realm of comment left untouched here, the extensive comment that
could be made on Homer’s very obvious excellence as a wordsmith and a
story-teller, which I and others have offered elsewhere.14 An unfortunate
tradition continues in Homeric criticism whereby opponents of the oral
school of criticism imagine themselves as defenders of the poet’s

14
See my recent commentary on Od. 17-20 (Russo et al. 1992), as well as my
previous publications. Contributions to the appreciation of Homer’s uniqueness and
creativity with both diction and theme, written under the influence of Parry and Lord, are
too numerous to list here, but include Hainsworth 1964, 1970; Russo 1968; Segal 1970;
Nagler 1974; Willcock 1977, 1978; Edwards 1980; Sacks 1987; Finkelberg 1989; Martin
1989; Foley 1987, 1990, 1991. See also Holoka 1973 for an excellent survey of
“Homeric Originality” that includes many orally oriented studies.
386 JOSEPH RUSSO

individuality and creativity, and see the followers of that school as

promoting a Homer who is little more than a mechanical reproducer of
traditional fixed phrases. Such objections to the oral theory began with
some of the earliest critics of Parry’s work, who at least had some justice
when they disputed Parry’s exaggerated claim that Homeric verse is
virtually all formulaic, a claim long abandoned by Parry’s successors. But
anti-oralist critics continue to publish earnest defenses of Homer’s capacity
for unique language, subtle allusion, successful plot construction, brilliant
handling of character, imagery, and so on.15 I must point out that Homerists
in the Parry-Lord tradition have not overlooked excellences of this sort, as
the bibliographical record shows. But the oralist perspective has also been
able to make unique contributions to criticism beyond the more obvious
kinds of observation that show Homer to be a fine poet in some of the same
ways that literate poets are fine poets. The originality of the best criticism in
the Parry-Lord tradition lies in its development of a new aesthetic, one that
finds a particular kind of beauty in features of style, construction,
allusiveness, and referentiality16 that would not strike the reader as obvious
virtues in contemporary literature. This paper will be successful to the
extent that it has unfolded a few of these virtues, and taken steps toward
integrating them into a larger vision of Homeric style.

Haverford College

15
A good case in point is Richardson 1987. He shows that hapax legomena have
a greater place in Homeric diction than one would expect if one believed (but who
nowadays does?) Parry’s claim that epic verse is almost entirely formulaic, and that these
unique elements contribute moments of great poetry. This is a good observation, but
there is no reason why it cannot be harmonized with the belief in a great oral poet capable
of fine poetic effect both by staying within his traditional diction and occasionally
reaching beyond it for a new word, phrase, or idea. It is unfortunate that Richardson
concludes his insightful study with an uninformed swipe at oralist critics, whom he
crudely caricatures as believing in a poet “tied to the apron-strings of his tradition” and
simply “reshuffling the index cards.” Such criticism can only stem from limited
acquaintance with the best scholarship in the oralist tradition.
16
I use the term in Foley’s (1991:38-60) technical sense, which opens up a new
perception of the great “depth” behind some of Homer’s traditional language.
 HOMER’S STYLE 387

References

Arend 1933 W. Arend. Die typischen Szenen bei Homer. Problemata, 7.

Berlin: Weidmann.

Bassett 1920 Samuel H. Bassett. “USTERON PROTERON

OMHRIKWS (Cicero, Att. 1, 16, 1).” Harvard Studies in
Classical Philology, 31:39-53.

Bassett 1926 ______. “The So-Called Emphatic Position of the Runover

Word in the Homeric Hexameter.” Transactions of the
American Philological Association, 57:116-48.

Chantraine 1953 Pierre Chantraine. Grammaire Homérique. Tome II,

Syntaxe. Paris: Klincksieck.

Combellack 1965 Frederick M. Combellack. “Some Formulary Illogicalities

in Homer.” Transactions of the American Philological
Association, 96:41-56.

Edwards 1966 Mark W. Edwards. “Some Features of Homeric

Craftsmanship.” Transactions of the American
Philological Association, 97:116-79.

Edwards 1980 ______. “Convention and Individuality in Iliad 1.”

Harvard Studies in Classical Philology, 84:1-28.

Finkelberg 1989 Margalit Finkelberg. “Formulaic and Nonformulaic

Elements in Homer.” Classical Philology, 84:179-97.

Foley 1987 John Miles Foley. “Reading the Oral Traditional Text:
Aesthetics of Creation and Response.” In Comparative
Research on Oral Traditions. Ed. by J. M. Foley.
Columbus, OH: Slavica. pp. 185-212.

Foley 1990 ______. Traditional Oral Epic: The Odyssey, Beowulf, and
the Serbo-Croatian Return Song. Berkeley: University of
California Press.

Foley 1991 ______. Immanent Art: From Structure to Meaning in

Traditional Oral Epic. Bloomington: Indiana University
Press.
388 JOSEPH RUSSO

Gunn 1970 David M. Gunn. “Narrative Inconsistency and the Oral

Dictated Text.” American Journal of Philology, 91:192-
203.

Hainsworth 1964 J.B. Hainsworth. “Structure and Content in Epic Formulae.

The Question of the Unique Expression.” Classical
Quarterly, 14:155-64.

Hainsworth 1970 ______. “The Criticism of an Oral Homer.” Journal of

Hellenic Studies, 90:90-98.

Holoka 1973 James P. Holoka. “Homeric Originality: A Survey.”

Classical World, 66:257-93.

Janko 1990 Richard Janko. “The Iliad and its Editors: Dictation and
Redaction.” Classical Antiquity, 9:326-34.

Janko 1992 ______. The Iliad: A Commentary, Vol. V: Books 17-20.

Cambridge: Cambridge University Press.

Leaf 1900-02 Walter Leaf, ed. The Iliad. London: Macmillan. 2 vols.

Lord 1960 Albert B. Lord. The Singer of Tales. Cambridge, MA:

Harvard University Press.

Macleod 1982 C. W. Macleod, ed. Homer. Iliad. Book xxiv. Cambridge:

Cambridge University Press.

Martin 1989 Richard P. Martin. The Language of Heroes: Speech and

Performance in the Iliad. Ithaca: Cornell University Press.

Monro 1901 D.B. Monro, ed. Homer’s Odyssey, Books xiii-xxiv.

Oxford: Clarendon Press.

Nagler 1974 Michael N. Nagler. Spontaneity and Tradition: A Study in

the Oral Art of Homer. Berkeley and Los Angeles:
University of California Press.

O’Nolan 1978 Kevin O’Nolan. “Doublets in the Odyssey.” Classical

Quarterly, 28:23-37.

Richardson 1987 N. J. Richardson. “The Individuality of Homer’s

Language.” In Homer: Beyond Oral Poetry. Ed. by J. M.
Bremer, I. J. F. De Jong, and J. Kalff. B. R. Grüner:
Amsterdam. pp. 165-84.
 HOMER’S STYLE 389

Russo 1968 Joseph Russo. “Homer Against his Tradition.” Arion,

7:275-95.

Russo 1976 ______. “Is ‘Oral’ or ‘Aural’ Composition the Cause of

Homer’s Formulaic Style?” In Oral Poetry and the
Formula. Ed. by B. A. Stolz and R. S. Shannon. Ann
Arbor, MI: Center for the Coordination of Ancient and
Modern Studies. pp. 31-54.

Russo 1987 ______. “Oral Style as Performance Style in Homer’s

Odyssey.” In Comparative Research on Oral Traditions.
Ed. by J. M. Foley. Columbus, OH: Slavica. pp. 549-65.

Russo et al. 1992 ______, Manuel Fernandez-Galiano, and Alfred Heubeck.

A Commentary on Homer’s Odyssey, Vol. III, Books xvii-
xxiv. Oxford: Clarendon Press.

Sacks 1987 Richard Sacks. The Traditional Phrase in Homer. Leiden:

E. J. Brill.

Segal 1970 Charles P. Segal. “Andromache’s Anagnorisis: Formulaic

Artistry in Iliad 22.437-476.” Harvard Studies in
Classical Philology, 75:33-57.

Smith 1977 John D. Smith. “The Singer or the Song? A Reassessment

of Lord’s ‘Oral Theory’.” Man, n.s. 12:41-53.

Stolz and Shannon 1976 B. A. Stolz and R. S. Shannon, eds. Oral Literature and
the Formula. Ann Arbor, MI: Center for the Coordination
of Ancient and Modern Studies.

Von der Mühll 1946 Peter Von der Mühll, ed. Odyssea. Basel: Helveticae.

Willcock 1977 Malcolm M. Willcock. “Ad Hoc Invention in the Iliad.”

Harvard Studies in Classical Philology, 81:41-53.

Willcock 1978 “Homer, the Individual Poet.” Liverpool Classical

Monthly, 3:11-18.