Does Hector’s Helmet Flash?

The Fate of the Fixed Epithet in the Modern English Homer

Oral Tradition, 33/1 (2019):89-114

The most important question raised by studies in oral tradition is “So what?”

—John Miles Foley, “Oral Tradition and Its Implications”

With the publication of his fine press Odyssey in view, the great American printer and typographer Bruce Rogers wrote in April 1931 to his translator, T. E. Lawrence, with a request for material to thicken a prospectus. Lawrence duly sent along notes “copied from the back of the book,” which became the edition’s “Translator’s Note” (quoted in Knox 1991:xiv). Among many notable pronouncements, Lawrence offered his deductions concerning the shadowy historical figure lurking behind the text (1932:vii):

In four years of living with the novel I have tried to deduce the author from his self-betrayal in the work. I found a bookworm, no longer young, living from home, a main-lander, city-bred and domestic. Married but not exclusively, a dog-lover, often hungry and thirsty, dark-haired. Fond of poetry, a great if uncritical reader of the Iliad, with limited sensuous range but an exact eyesight which gave him all his pictures. A lover of old bric-a-brac, though as muddled an antiquary as Walter Scott. … Very bookish, this house-bred man. His work smells of a literary coterie, of a writing tradition.

Here Lawrence ventures a solution to the ancient mystery of the Homeric epics’ origins, what has come to be known in the modern period as “the Homeric Question,” though, as Gregory Nagy has argued, we are wiser to speak of Homeric questions—“in the plural”—since there have long been many questions embedded in the Homeric Question, and each generation seems to add to the pile (Nagy 1996:1).

Lawrence’s “Homeric answers” contrast the usual declarations on these matters as a realist portrait does a silhouette. Where most attempts are filled with disclaimers, Lawrence, with knowing cheek, provides an account full of colorful details. For present purposes, I want to highlight Lawrence’s emphasis on the ancient writer’s bookishness. Lawrence’s Homer inhabits a world overflowing with papyrus. He was surrounded by books enough to breed “bookworms” and fellow readers enough to form a “literary coterie.” While a participant in the “epic tradition,” Lawrence’s Homer was also an innovator—the first novelist of the European tradition. The


Odyssey thus appeared to Lawrence the work of not just a single but a singular author, a literate of literates. And this notion of the author framed, in turn, Lawrence’s labors: to translate the Odyssey was, on this account, to carry over a written document from an ancient literary culture to a modern one.

I begin with Lawrence’s “bookish” deductions because they represent, in the context of English translations of the epics at least, the high-water mark of the notion of Homer as a writer thoroughly at home in the culture of letters. Only one year after the Rogers-Lawrence Odyssey was published, 1933, Milman Parry began fieldwork among folk singers in the Balkans—the guslari—that would render such a portrait of the artist as a bookworm obsolete and raise, in turn, new questions about the task of the Homeric translator. Parry’s contact with the illiterate guslari offered auricular evidence of what his prior work on the text of Homer had led him to conclude: that ancient singers composed lengthy narrative poems in the act of performance, thanks, in particular, to inherited units of utterance—metrical hand-me-downs, if you will. His example par excellence was the body of epithets applied to gods, humans, animals, and recurring events such as “rosy-fingered” (attached to Dawn), “swift-footed” (Achilles), and “resourceful” (Odysseus). In the guslari, in other words, Parry found a tantalizing vision of the inner workings of poetic oral traditions—whether living or long dead. From this encounter, he was able to imagine bards in archaic Greece improvising an epic on the spot—no writing required.

Parry’s “oral Homer” has passed, in turn, through what John Miles Foley well describes as the “predictable and natural life-cycle” of such a revolutionary idea (1997:147). A seemingly inevitable pushback arose in subsequent years in the equally uncertain name of whoever finally did write the poems down—whether Homer himself or some scribe (the “adapter,” whom Barry Powell ventures to name as Palamedes), perhaps shortly following the development of the Greek alphabet in the eighth century BCE (Powell 2014b:14). Yet no matter how savvy about the new technologies of writing scholars now believe Homer to be, the epics’ author(s) can no longer answer to Lawrence’s altogether-literary description. As Bernard Knox observed, it no longer seems possible “that anyone at such an early stage of Greek literacy could have been a bookworm, a member of a literary coterie, or an inheritor of a literary tradition” (1991:xiii). Scholars now often characterize Homer as the “heir” of an oral tradition whose imprint on the written epics remains palpable. The poet is now often imagined as standing between, rather than squarely in, the worlds of orality and literacy.1 

As Parry himself realized, his theories have implications beyond the parlor-debates of classicists; they also matter to those who would transmit the Homeric epics into modern languages, translators. For this group, what to make of Parry’s ideas (and the “literate” Homer pushback) isn’t simply a theoretical problem—it’s also a practical one. Translators wrestle with not only the question of what the texts of the Homeric epics are—to what degree they are literature, to what extent “orature.” Translators also must render ancient poetic techniques in a

1 In How to Read an Oral Poem, for example, Foley places the Homeric epics in his third category of oral poetry: Voices from the Past. He describes its membership criteria as follows (2002:46): “[I]t offers a slot for those oral poetic traditions that time has eclipsed and which we can consult only in textual form. … Any given poem’s original composition may have been oral or written; in many cases we can’t tell whether the document we hold in our hands is a direct transcription of an oral performance or an artifact that some generations of editing and recopying removed from performance. The particular version that survives may even have been composed as a text, written down by a poet adhering to the rules of Oral Performance.”


manner pleasing to modern readers. Within the English line of translators that concerns us here, the first to address Parry’s ideas as a set of practical concerns for translation appears to be Robert Fitzgerald, whose Odyssey debuted in 1961 and Iliad in 1974. “The problem [for the Homeric translator],” he observed in an interview that will be discussed below, “is to bring a work of art in [an oral] medium into another medium formed on different principles and heard and understood in a different way” (1985:109). Another translator whose work we will consider, Robert Fagles, contended that Parry’s work faces translators with a “Homeric Question” of their own: how to convey the poem’s oral dimensions in the translator’s medium of writing? (1990b:ix). Not every translator would agree with Fitzgerald’s or Fagles’ assessments exactly, mind you. Their remarks are useful, though, in displaying the translator’s altered situation in the wake of Parry’s work.

There are many materials that we might use to gauge Parry’s effect on the modern English Homer. One could, for example, track the fates of ritualized and linguistically repetitive episodes, the so-called “type scenes,” such as when heroes don armor for battle or hosts sit down to feast with guests. One could look at lines that operate according to discernible patterns or that are even repeated verbatim, such as those that introduce speeches. In the present case, we will focus on an even a smaller, though no less pervasive, sign of the oral tradition’s influence—the aforementioned epithets.

In concentrating on epithets as a problem of translation, we in fact follow Parry’s own lead. In his groundbreaking thesis at the University of Paris, L’épithète traditionnelle dans Homère (1928), he raised the question of what translators are to make of what he called the “fixed” epithet, which he characterized as traditional noun-modifier combinations adapted to— or, perhaps better said, productive of—the dactylic hexameter line. Parry contended that such “fixed” epithets cannot be accurately translated for a modern audience, an argument that we will trace in detail below.2 That Parry believed them inherently troublesome for translation is reason enough perhaps to make a review of how translators have tackled them. But they are also significant, I argue, because they exemplify the Translator’s Homeric Question. To translate the epithets is to contend with not only an ancient poetic practice but also the gap between the conditions in which the epics emerged—an oral culture—and the one in which they will be consumed—a literate one.

The epithets have additional interest here given that our translations belong to the “English Homer” tradition. They have, in fact, often been perceived as posing a difficult pill for modern taste, and translators have tinkered with them from the beginning. The Jacobean courtier George Chapman, the first Englishman to translate the Homeric epics in full from the Greek, dropped epithets, made substitutions, and varied expressions seemingly at whim. Consider the fate of the epithet on which I will be focusing here, koruthaiolos (κορυθαίολος), which has been traditionally translated as “of the flashing helm” (or some variation such as “shining” or “shimmering”). Chapman renders it, by turns, “warlike,” “helm-decked,” “fair-helmed,” “helm-

2 The reader should note that in my subsequent discussions of Parry’s thesis, I will be using the 1971 translation by Adam Parry. As paper copies of the thesis can be difficult to come by, readers wishing to consult the French text alongside my remarks below may wish to use the digital version available on the website of the Center for Hellenic Studies ( 1928).


moved,” and “helm-graced” (Miola 2017:8). Moreover, in his Odyssey, Chapman conjures “Homer-like epithets” nowhere to be found in the Greek—including “wave-beat-smooth” and “golden-rod-sustaining” (Gordon 2016:30). Many of the most venerable “English Homers” fiddle with the epithets to some degree. Alexander Pope did it early in the eighteenth century, as did William Cowper at that century’s end. In the demonstration passages that he included in the still-influential On Translating Homer lectures, the Victorian Matthew Arnold did, too. Writing to Rogers in 1930, Lawrence claimed that he “slaughtered [epithets] freely. From now on we will put in only enough to remind ourselves of a bad business” (1930a:49). A few months later, however, Lawrence would admit that “Some of them have value. The author wrote them deliberately, as part of the epic tradition, and the text loses if they all disappear. Loses dignity, I should say” (1930b:54-55).

The present piece is, then, in truth a study of two traditions in translation. First, I examine how modern translators have practiced their craft in light of their post-Parry understandings of the oral tradition’s relation to the written texts of the epics. Second, I consider the English Homer as itself a tradition. Now, translators of the Iliad and the Odyssey are obviously members of a tradition insofar as all are engaged in the same long-running process of transmission that includes Chapman, Pope, and Lawrence. All have composed texts designed to make ancient poems available to modern audiences. Yet in calling the English Homer a tradition, I also mean to observe the lines of influence and resistance (which is a kind of negative influence) among the translations themselves. No translation stands alone. (Even Chapman’s work responded to the earlier efforts of those who attempted to make an English Homer out of French translations.) In some ways, moreover, the translations considered here constitute a tradition analogous to the oral one envisaged by Parry.3 The notion of oral performance has led a number of modern translators to conceive of themselves as English “performers” of the Greek text rather than its straightforward transmitters. The analogy of a performative tradition helps us to conceive of the ways that translations are shaped by their predecessors—and often judged by readers against the background of previous translations—just as oral performances in antiquity were informed by what came before, even while (deliberately or not) introducing variation and innovation.4 

We begin below with Parry’s reflections on the challenge of translating fixed epithets, which I balance with an influential dissenting opinion about the epithets’ functions, that of Parry’s son, Adam. I then examine how five modern Homer translators—Fitzgerald, Robert Fagles, Stanley Lombardo, Stephen Mitchell, and Barry Powell—have framed their efforts in relation to the oral tradition on the one hand and the modern reader on the other. We will also consider how their strategies build on or turn away from those of their predecessors. To illustrate their approaches in action, I focus on the aforementioned koruthaiolos, which appears on thirty-

3 I am grateful to one of the anonymous reviewers for crystalizing this aspect of the study. Subsequent sentences draw heavily on the reviewer’s comments.

4 Also germane here is Foley’s argument in Homer’s Traditional Art that “a tradition is always evolving within certain rules or boundaries, always proving a somewhat different ‘thing’ from one performance to another and from one practitioner to another” (1999:xii). Foley memorably characterizes oral traditions as “work[ing] like language, only more so” (for example, 2002:127). The latter point helps to frame the problem that the epithets pose for translators: how to translate not just the words’ semantic meanings but their meanings within the traditional “language” of the epics.


eight occasions in the Iliad, attached to Hector in all but one. (Ares is its other one-time bearer). This is not, let us recognize at the outset, an accidental choice: koruthaiolos often appears in the translators’ own discussions of their approaches. The tradition has marked it as a representative phrase.

There are, of course, other translations that we might consider. Emily Wilson, the most recent English translator of the Odyssey, has rightly observed that we are in a “bull market” for Homer translation—no less than twenty new English Homers having been published in the last two decades (2017a).5 My selections have been made with an eye toward revealing the variety of strategies that translators have employed, including omitting epithets (on occasion or in full) and altering their phrasing (on occasion or at nearly every appearance). This study is thus in no small way a catalogue of tactics.6 It is not my ambition, let me emphasize, to advocate for one strategy over the others, though I do at times observe some of their narrative implications. This study, then, is best understood as a diagnostic exercise—the examination of how the recovery of one Homeric tradition, the ancient oral tradition, has impacted the ongoing development of another, that of English translation.

Two Parrys

Parry, once again, must be credited with first recognizing that his conception of the fixed epithet had implications for the task of the translator. His exploration of the issue in L’épithète traditionnelle dans Homère thus represents the necessary starting point for our reflections, for whether or not translators reach the same conclusions, Parry’s thought now represents the jumping-off point for scholars and translators alike. Yet Parry obviously hasn’t had the final say on the fixed epithet or, more broadly, the oral tradition’s mark on the texts of Homer. In light of subsequent studies, Parry’s characterization of the epithets now seems monolithic and inflexible—too fixed. To understand the epithets’ fate in modern English translations, then, we need to consider an alternative assessment. As noted above, I concentrate on one influential response, that of Parry’s son, Adam. Specifically, we will review A. Parry’s less “rigid” (his phrase) notion of the fixed epithet laid out in his “Language and Characterization in Homer” (1972).

To begin with the father: Milman Parry raises the matter of translation in the last section of L’épithète traditionnelle’s fourth chapter, which he titled: “Can the Fixed Epithet Be Translated?” Before diving into his thinking in this section, we need to acknowledge the particular understanding of the “fixed” epithet that Parry develops in the preceding pages of the thesis. Now, the idea that the epithets were “traditional” epic material long preceded Parry. (Lawrence, too, as we have seen, conceived of them as elements of the “epic tradition.”) Parry’s breakthrough was to situate the epithets within a historical process of refinement: he argued that they had been gradually selected over several generations of a poetic tradition—which he would

5 For reflection on the recent contributions of three women—Wilson, Caroline  Alexander, and Alice Oswald—to the stack of new English Homers, see my essay “On Women Englishing Homer” (Gibson 2019).

6 I am grateful to the other anonymous reviewer for the suitably Homeric characterization of this piece as a “catalogue.”


within a few years pronounce an oral tradition—for their usefulness in generating lines of dactylic hexameter.7 Already in the nineteenth century, at least one English translator, Frances Newman (Arnold’s combatant in the early 1860s), had observed that Homer chose some epithets for “their convenience to his metre” (1861:76). In 1928, Parry argued that this was half-right: the epithets were indeed chosen for metrical convenience—but the sifting was done by the tradition, over generations, rather than a single author. The epithets must, he argued, be understood above all as a “functional” element of the epics, equipment for archaic song-building.

In Chapter 4, “The Meaning of the Epithet in Epic Poetry,” Parry took the bold next step in light of this analysis, declaring that the epithets were not “meaningful” in the standard literary sense. Moderns, he observed, assume that words find their places in poetic lines in order to contribute something to the meaning of the passage at hand. The writer, literates assume, has a point to make, a scene to build, a feeling to express, and he or she chooses each and every word to that immediate end. Fixed epithets, according to Parry, didn’t work like that in archaic Greece. They are “invariably used without relevance to the immediate action whatever it may be” (italics mine; 1987:118).8 Fixed epithets are, Parry declared, “ornamental” epithets. They serve the needs of the metrical line, we might say, rather than the sentence or scene at hand.

They are not altogether meaningless, however. Parry grants that fixed epithets have an effect, but it must be registered over the course of the epic or epic poetry as a whole rather than in individual instances: “For [Homer] and his audience alike, the fixed epithet did not so much adorn a single line or even a single poem, as it did the entirety of heroic song” (137). Of their general significance, he observes, for example, that the epithets play an important role in enumerating the qualities of the heroes, whether in general (via a frequently applied epithet such as “godlike”) or in regard to an individual character (“fleet-footed Achilles”). On any one given occasion, though, the fixed epithet has no bearing on what’s transpiring in the narrative: it is meaningless in context. They may make the music move, but they never touch the story. Parry grants that some epithets in Homer are locally significant, dubbing these “particularized” epithets, yet he states in no uncertain terms that this small class does not include the thick stock of fixed epithets. Fixed epithets are always ornamental.

In the original context, then, the epithets were experienced on profoundly different terms than moderns ordinarily approach the language of literature. Parry, as we have seen, understood the epithets as common goods in respect to both the bard and his audience. For the former they were the traditional tools of the trade. For the latter they were the familiar strains of heroic song. In so many ways, Parry imagined these phrases as being in the cultural air—the bards breathing them out, their auditors in. And they were like the air in being absolutely necessary to the poetic enterprise and yet remaining imperceptible unless absent: “The audience would have been infinitely surprised if a bard had left them out; his always putting them in hardly drew their

7 On the shift in Parry’s account of the epithets, John Miles Foley writes in his Traditional Oral Epic: “[I]t was not until his ‘Studies’ I (1930) and II (1932) that Parry first broached the possibility that his earlier demonstration of the  traditional  character  of  Homer’s  epics  must also  mean  that  they  were  composed  orally” (1993:122). I follow the general consensus, in turn, in viewing Parry’s expeditions to the Balkans as enriching, even solidifying, what Parry had begun to work out on paper a few years earlier.

8 Hereafter, all quotations from Adam Parry’s translation of L’épithète traditionnelle, published as The Traditional Epithet in Homer.


attention” (137). Parry memorably distills the point thusly: “Homer’s listeners demanded epithets and paid them no attention.”

In the aforementioned concluding section of the fourth chapter—which he, again, titled “Can the Fixed Epithet Be Translated?”—Parry despairs about the possibility that the unique nature of the fixed epithet can be adequately rendered in translation. The trouble isn’t that no English counterparts can be found for Greek words. Instead, the central problem of translation here is cognitive.9 (We might also label it semiotic in the sense of regarding meaning-making generally and perhaps in the more pointed sense of concerning sēmata or physical signs as well.) To render the Homeric epics accurately, Parry suggests, the translator has to carry over the elemental nature of traditional expressions, the way that they were for their original audience so commonplace as to become nearly pure sound. The problem of translation here is nothing less than the rendering of the thought-world of a “traditional” culture—again, what within a few years would come more clearly into focus for Parry as an oral culture.

In light of this analysis, Parry presents several impediments to the translation of fixed epithets. First, translators often mistakenly assume that Homer is just like them (recall the bookish Lawrence’s picture of Homer as a “bookworm”). They thus treat all epithets as deliberate choices meant to enhance the scene at hand. In effect, all epithets become “particularized” epithets. The second issue is related to the first. How might a translator “make clear the crucial difference between the ornamental and the particularized epithet?” (171).10 As an illustration of this difficulty, he cites passages from both of the Homeric epics that place “ornamental” and “particularized” expressions in close proximity. He cites, for example, Odysseus’ vaunting to Polyphemus in Book 9 of the Odyssey (lines 502-505). The hero uses what Parry perceives to be an ornamental epithet, “sacker of cities” (πτολιπόρθιος), immediately before describing himself in “particular” terms as the “son of Laertes.” Parry writes, “How can we render the ornamental meaning of πτολιπόρθιον without losing at the same time the particularized meaning of the words of the following line?” (171-172). Third, even if the translator is scrupulous in avoiding the suggestion of “particularized” meanings to ornamental epithets, he or she is powerless to stop “the modern reader from following his own literary habits and looking for the specific motivation for the use of each epithet, and for some specific meaning to assign it” (171).

Earlier in the text, Parry details how a student might through repeated exposure develop an “insensitivity” to a fixed epithet, the epithet’s “meaning [losing] any value on its own,” becoming fused with the substantive it modifies (127). For these readers, the epithet isn’t nothing, as it continues to contribute “an element of nobility and grandeur, but no more than that.” In the section at hand, though, Parry worries about the average reader operating according

9 To use Walter Ong’s language in Orality and Literacy, which is in part based on Parry’s work, we might say that Parry recognized translation as brushing up against the very different “psychodynamics of orality” (1982:31).

10 The reader should note that Adam Parry’s translation here mistakenly reads: “Moreover, how could we in a translation make clear the crucial difference between the ornamental and the fixed epithet?” (171). The words that I’ve italicized flag the error. As my preceding remarks have stressed, fixed epithets are always ornamental epithets in M. Parry’s thought. There is no difference! In the French original, those words are “ornamental” and “particularized” (ornementale and particularisée).


to the standard modern modus operandi. For this audience, the fixed epithet is in essence unreadable, since reading in this case almost inevitably becomes over-reading.

In the face of these obstacles, what should a translator do? Parry doesn’t offer a clear set of instructions here or anywhere else in his writing. His project seems to be that of raising awareness of the numerous difficulties that the epithets pose. In the section’s concluding paragraph, he wonders aloud whether the “effort to find an exact equivalent for the ornamental epithet”—one that “[translates] Homer’s thought with the least addition of foreign ideas”—might be “[committing] a worse error than those who draw on their own ideas to translate the epithet” (172). An epithet that the translator hasn’t assigned a clear purpose seems likely to leave the reader “confused,” with the result that the reader will “search and find some meaning or other, and the necessary delay will break the rapid movement of Homer’s clear sentences” (172). To conclude, Parry acknowledges that the case at hand is exemplary of the “problem of translation in general,” which involves choosing between “what is obscure, but literally faithful, and what is clear, but inexact” (172). He does not tell translators how to make that choice in relation to the epithets. Perhaps unsurprisingly, one translator that we will be discussing, Stephen Mitchell, has found in Parry’s bleak assessment authorization for a strategy of widespread omission.

Once again, while the paradigm that Parry was developing here has reshaped the field, it has never been uncontroversial.11 Within a few years of the publication of Parry’s former assistant Albert Lord’s The Singer of Tales (1960), an important channel for the dissemination of Parry’s ideas, even scholars open to Parry’s theory regarding the Homeric epics’ oral backstory were calling into question his portrayal of the fixed epithets, particularly as his theory applied to the written texts of the epics in our possession. Among these respondents, once again, was Parry’s son, Adam. His “Language and Characterization in Homer” (A. Parry 1972) offers an alternative account of how the fixed epithets function in the Homeric texts that is equally illuminating for present purposes.12 

The opening section of this piece is tellingly titled “Meaning in the Fixed Epithet,” an echo of the chapter title in M. Parry’s thesis noted above. A. Parry raises the question of the meaning of Homer’s formulaic elements as follows: “Do the set pieces in which the poetry so largely consists have a meaning dependent on the individual words which are their ingredients? Or does the formulary style preclude such meaning, so that these phrases are in operation equivalent to single words?” (1972:2). In the latter sentence, A. Parry is referring back to M. Parry’s argument that the epithet so bonds to the substantive it modifies that the two become, in effect, a protracted way to say the substantive (often a character’s name).

11 For a rich reflection on the reception of Parry’s ideas, see the aforementioned John Miles Foley’s “Oral Tradition and its Implications” in A New Companion to Homer (1997). For an attempt to update Parry’s arguments in light of subsequent criticism, see Merritt Sale’s “In Defense of Milman Parry: Renewing the Oral Theory” (1996).

12 The reader should note that the question of the meaning of traditional epithets remains a live one in Homeric studies. For a very recent reflection on this matter, see David F. Elmer’s “The ‘Narrow Road’ and the Ethics of Language Use in the Iliad and the Odyssey” (2015). The piece is notable in many respects, including its arguments concerning the ethical significance of moments when epithets seem out of joint given their contexts. But for present purposes it is particularly useful for its examination of Parry’s take on the epithets and that of later scholars, including Lord.


A. Parry’s response to these questions is not a simple Yes or No. As I have highlighted above, A. Parry advocates for a less “rigid” understanding of the fixed epithet. While granting that epithets “do indeed possess metrical convenience,” he questions whether it necessarily follows that they lack meaning-in-context (4). He cites several passages to establish how what Parry classes as fixed epithets seem more than accidental deposits, including Odyssey 9.504-505, which as we noted above, his father had used as an example in the “Can the Fixed Epithet Be Translated?” section:

Φάσθαι Ὀδυσσῆα πτολιπόρθιον ἐξαλαῶσαι,
υἱὸν Λαέρτεω, Ἰθάκῃ ἔνι οἰκί᾽ ἔχοντα.

Say that Odysseus sacker-of-cities blinded you,
Laertes’ son, who makes his home in Ithaca.

M. Parry had argued that since πτολιπόρθιος was found elsewhere in the text, it must be fixed and thus ornamental in the present instance. A. Parry counters: “It is hard to think of an epithet, in fact, which serves better to reveal the nonentity in the cave suddenly as one of the greatest heroes of the epic tradition” (8). This epithet is perfectly pitched to the episode, not just the needs of the line. Indeed, A. Parry further points out, the epithet takes a variant grammatical form and sits in an unusual position in order to be included in that line. In so many ways, this epithet bespeaks a conscious choice—an effort—to place it there. The epithet defies M. Parry’s extremely rigid model: here is a stock item used in a “particularized” manner.

As A. Parry observes, one could “run Milman Parry’s argument into the ground” with such examples if one wished. But A. Parry counsels against that course. The distinction between “the fixed and the particularized adjective” remains for him a valuable one (8). There are indeed “adjectives which the evidence allows us to classify as fixed … often used in such a way that they add little to the meaning other than, as Milman Parry so well said, to remind us of the heroic nature of the world of epic poetry.” On such occasions, the epithets indeed lack individual significance. A. Parry concludes: “But the distinction is not rigid, and there is no absolute line of demarcation” (8-9).

Between the two Parrys, then, we are offered rival yet nonetheless related accounts of the epithets’ performance in the epics. What does this mean for translation? M. Parry’s system offers a straightforward answer: an epithet is fixed if it echoes across the epic or epics; if it is fixed, then it is ornamental; and if it is ornamental, then it is essentially untranslatable. As we noted above, Parry doesn’t say that the translator shouldn’t try to translate them; rather, he underscores how fraught the act of translation is in this case.

As for the consequences of adopting a view like A. Parry’s, we do not have direct word from the scholar himself. Yet the flexible view of the fixed epithet that he describes has clear implications for translation. Andrew Ford well describes the result in his introduction to Robert Fitzgerald’s Iliad: “The problem for translators … is to know whether a given word is being used for effect and when it is more generic” (2004a:xxxv). The translator’s role as a critic is thus emphasized in this understanding of the epithet: to him or her falls the assessment of an epithet’s local relevance in the Greek and, in turn, the problem of rendering that significance in English.


By turning to the translations themselves, as we will now do, we see that translators are indeed deeply concerned with the problem of signaling the significance—or insignificance—of epithets to the scenes in which they appear. In fact, several of our translators go to great lengths to demonstrate that the epithets matter, even refashioning them—linguistically and grammatically—so that their pertinence is evident. We will see epithets converted into matter and energy, the stuff of battle, the fuel and fire of action. At the head of this campaign stand Robert Fitzgerald’s efforts, which we consider next. From Fitzgerald, we will then turn to two successors in their handling of the epithets, Robert Fagles and Stanley Lombardo. Stephen Mitchell’s Iliad then allows us to see, as mentioned above, a strategy of omission rooted in M. Parry’s judgments. Finally, I examine Barry Powell’s Iliad, the most recent of the texts considered here, which attempts to outline a “middle way” (Powell’s own phrase) between earlier strategies discussed here. As noted above, I use the Iliad’s koruthaiolos, “of the flashing helm,” as my model epithet for the purposes of comparison across the translations, though the fates of others will be mentioned in passing.

Fitzgerald: The Translator as Modern Performer

For some critics, the idea of beginning our review of Parry’s influence with Fitzgerald would seem profoundly mistaken. Consider the following remarks from D. S. Carne-Ross’ introduction to the 1998 edition of Fitzgerald’s Odyssey (which, once again, debuted in 1961) (1998:lxvi):

Fitzgerald’s Odyssey was immediately recognized as a masterpiece, but it has not always pleased professional classicists, who complain that it pays no attention to the most influential contribution made to Homeric scholarship in this century, the demonstration by the American scholar Milman Parry that Homer’s poems are oral compositions.

Speaking in Fitzgerald’s defense, Carne-Ross then makes a general observation that recalls M. Parry’s remarks noted above: “the oral-formulaic style cannot be adequately reproduced in translation” (lxvii). He observes, in particular, that “the recurrent phrases” (epithets serving as his examples) are bound to seem “repetitious” to modern readers (lxvii). Fitzgerald is praised, in turn, for not trying to reproduce the oral-formulaic style: “Fitzgerald at all events does not try to pretend that he is himself composing orally and allows himself the liberties that fine verse translators have always taken from the time of Dryden and Pope” (lxvii). Among other examples of Fitzgerald’s “liberties,” Carne-Ross notes a line that often introduces Odysseus’ speeches, τὸν (or  τὴν)  δ ̓ ἀπαµειβόµενος προσέφη πολύµητις Ὀδυσσεύς. As Carne-Ross notes, Richmond Lattimore (1967), Fitzgerald’s immediate predecessor in the English Homer line, translates this line consistently as “Then resourceful Odysseus spoke in turn and answered him.” At one moment in Fitzgerald’s text, though, this line is rendered “The great tactician carefully replied” (7.258), “His mind ranging far, Odysseus answered” (13.398) at another, and later “And the great master of invention answered” (19.191), among other variations (lxix). Carne-Ross’


point is that Fitzgerald has translated the oral-formulaic style right out of the text—just as, the critic believes, a translator writing for a modern audience should.

In a 1976 interview with Fitzgerald, however, fellow poet and translator Edwin Honig argued that Fitzgerald was in fact the first translator of Homer to draw “consciously” on Parry’s ideas about “the oral tradition behind the compositions” (1985:106).13 Honig’s focus, we should recognize, was not the oral-formulaic style but what he calls “performance-invention,” referring to the improvisatory nature of ancient oral performance revealed in Parry’s writings (107). While Fitzgerald’s own lengthy postscript to his Odyssey, published for the first time in 1962, already hinted at his use of Parry in this way (along with acknowledging his debt to Parry as his “friend and teacher”), the Honig interview offers the most direct testimony to Fitzgerald’s understanding of how Parry’s ideas influenced his approach to translation.14 

Two passages stand out for present purposes. The first relates to the consequences of “performance-invention” for the translator (1985:105):

Homer, as we now know, was working in what they call an oral tradition. Now the performer— because that’s what he was—had at his disposal a great repertory of themes, narrative  and dramatic situations, and he had at his disposal a great repertory of formulae, of lines, half lines, phrases, all metrical, let it be observed, that could be modified or used in many contexts during his performance, which was always to some extent extemporary. Now, as he went along with his tale, he could and did invent new ways of handling episodes and passages that made each performance, in some way, a new thing. Do you see how this fact liberates, to a certain extent, the translator?

Over the course of the interview, Fitzgerald repeatedly frames the ancient bard’s art as an “inventive” one—as in the present passage’s claim that the bard “could and did invent new ways of handling episodes and passages.” Since there was no single, canonical version of the Iliad or Odyssey, the translator is thus not, Fitzgerald reasons, bound to the letter of the text, as she would be when dealing with the work of a writer “who had, like say Paul Valéry, labored over every line and for whom the final text in every detail had more importance than for the Homeric singer” (105). The changeable nature of the tradition thereby actively encourages “freedom in translation” (his phrase). Indeed, Fitzgerald goes so far as to suggest that this is what the “original performer” would have expected of the translator (105). In this account, the ancient kind of invention authorizes invention on the modern translator’s part. The “free” translator

13 The reader should note that although the interview was recorded in 1976, it didn’t appear in print until years later and was ultimately included in the collection of interviews conducted by Honig referenced here, The Poet’s Other Voice (1985).

14 Fitzgerald makes a number of remarks in the postscript that gesture toward his understanding of the translator as the new Homeric “performer,” but his most telling words appear in his concluding remarks about the practice of translation: “If you can grasp the situation and action rendered by the Greek poem, every line of it, and by the living performer that it demands, and if you will not betray Homer with prose or poor verse, you may hope to make an equivalent that he himself would not disavow” (italics mine; 2004b:508). I have italicized the key phrase for present purposes. It hearkens back to Fitzgerald’s earlier notes in the postscript about that “living performer,” whom Fitzgerald views as a shrewd and original deployer of his traditional resources. Yet it also characterizes any would-be translator of Homer, including Fitzgerald himself. He does not elaborate this claim further in the postscript but the phrase is richly suggestive, encouraging the reader to map the qualities of ancient oral performance— traditional yet original, improvisatory yet practiced, in Fitzgerald’s account—onto his written translation.


effectively continues the oral tradition—in spirit rather than letter. To make a “new thing” out of the received text is to be faithful to the tradition behind it.

The second passage I quoted in part in the introduction. These are Fitzgerald’s words concerning the need to understand the task of the Homeric translator as not simply the “problem” of working across languages but also across media. Here is the passage in context (108):

Homer’s whole language … was all on the tongue and in the ear. This was all formulaic, by its very nature. The phrase was the unit, you could say, rather than the word. There were no dictionaries and no sense of vocabulary such as we have. Now, the language that had grown up and formed itself on those principles is what one is dealing with, and the problem is to bring a work of art in that medium into another medium formed on different principles and heard and understood in a different way. So it’s really a larger question than merely the question of whether one is to reproduce in some standard form formulaic expressions in Greek by formulaic expressions in English. The question is how to bring a work of imagination out of one language that was just as taken-for-granted by the persons who used it as our language is by ourselves.

Hiding in the background here are translations such as Richmond Lattimore’s (first published in 1951)—the ascendant ones at the time of Fitzgerald’s efforts—that indeed “reproduce in some standard form formulaic expressions in Greek by formulaic expressions in English.” Consider his treatment of my exemplary epithet, koruthaiolos. Lattimore translates it repeatedly as “Hektor of the shining helmet,” though he makes occasional substitutions of “glancing” and “bronze” for “shining” and the abbreviated “helm” for “helmet” (2011). Fitzgerald’s “media theory” of translation implies that such a strategy fails to acknowledge the gulf between the media situations of past and present. It produces a phrase that is not only “strange” to us. It also gives the impression that Homer’s text was distant and “strange” to its original audience, too. Fitzgerald suggests that the translator’s responsibility is not to convey the medium-specific conventions of the past exactly; it is to carry over a “work of the imagination” into the language of our literate times. Here again, Fitzgerald’s awareness of the oral tradition—in this case, of its difference—licenses “freedom in translation.”

Fitzgerald’s handling of koruthaiolos offers an ideal illustration of how he put these principles into practice. If we include a handful of outright omissions among the variations, Fitzgerald translates this epithet twelve different ways over the course of its thirty-eight occurrences in the source text, including  such  variations  as  “under his shimmering  helmet” (2004:125), “his helmet flashing” (144), “in his shining helm” (146), and “in the bright helm” (282). Immediately apparent from this collection is that Fitzgerald does not follow the tradition in rendering this (or any) epithet as a static description, an honorific, as we have just seen in Lattimore’s translation. Fitzgerald thoroughly integrates the epithet into the scene at hand. Take, for example, the epithet’s first appearance, 3.83 in the Greek text. Agamemnon calls for the Argives to cease firing because “some proclamation” is going to come from “Hektor, there in the flashing helmet!” (64). The phrase “in the flashing helmet” is already noteworthy because it emphasizes that Hector is wearing his helmet this very moment. By pairing the phrase with “there” (the translator’s invention), moreover, Fitzgerald presents the helmet as a feature of the character that is conspicuous to observers within the plane of narration. On other occasions,


as my list of variations suggests, the epithet is transformed into a dependent clause that briefly turns our attention to its “shimmering” presence within the scene—as in his rendering of 5.689 in the Greek text: “Silent under his polished helmet, Hektor …” (125). Again and again, it flashes before us—and often the characters within the world of the text as well.

The fate of koruthaiolos exemplifies Fitzgerald’s “freer” approach to the epithets, and the formulaic aspect of the poem more generally, in three respects. First, Fitzgerald claims the discretion that A. Parry grants the critic to determine the relevance of the epithets on a case-by-case basis. This includes the possibility of an irrelevance that translates into omission. (This move seems designed to promote speed and drama, as in Achilles’ recollection of his directions to Patroclus at the outset of Book 18: “You must not fight with Hektor!” (430).) The second we have already noted, but now it is important to name as a stylistic marker: Fitzgerald embraces variation in the diction of his renderings. “Flashing,” “shimmering,” “shining,” “bright”: the list of adjectives is, in fact, not that much longer than Lattimore’s. Yet in Fitzgerald the sense of variation is much more apparent because there is no standard expression, no baseline from which the fluctuation occurs (as there is in Lattimore’s “of the shining helmet”). There is some repetition across the text—a whiff of the formulary, if you will—but the clear drive of the translation is to vary the epithet as much as possible.

Notably for our purposes, this aspect of Fitzgerald’s approach was flagged by A. Parry in a 1962 review of the translator’s Odyssey. A. Parry begins his analysis of the epithets by pointing out their function in Homer: “It is in fact [the epithets’] constancy that makes them resonant for us” (1962:52). He cites as an example an epithet applied to Penelope, periphrōn, which he glosses as “prudent” or “circumspect.” “It is important,” the reviewer continues, “that she remain so, because this single quality, fixed in a single word, is the background against which all her actions … take place” (52-53). A. Parry does not chastise Fitzgerald for his strategy; on the contrary, he praises the translator: “it is [Fitzgerald’s] aliveness to what goes on in each scene and the variety of his expression that make the translation sparkle as it does” (53). But A. Parry calls attention, too, to what is lost in the process—the contributions of the “constant” epithets to the development of the epic’s “principal motifs” (53). In the case of periphrōn, the epithet marks the constancy of the wife who resists the suitors’ advances over the long years of her husband’s absence.

Finally, and most strikingly of all, Fitzgerald tinkers with the epithets’ “grammatical relations,” that is, their roles in the structures of sentences, and thereby their relations to the action, their narratological functions. This titanic change is wrought with the mere swap of a preposition—from the standard “of” to “in,” “under,” or none at all (as in “his helmet flashing”). With these substitutions or subtractions, Fitzgerald draws the epithetical helmet into the scene, making it a focal point for characters and readers alike. The material that M. Parry is at pains to keep out of the action, Fitzgerald thus makes a vivid presence. For a purist, Fitzgerald might seem to commit a sacrilege with such grammatical meddling, turning an honorific into an eye-catching prop.15 One might, in short, complain that the helmet catches too much light in this translation. Yet, for our purposes, this tactic is perhaps the most illuminating of the three I have

15 As an example of such a “purist” approach to the epithets (applied in this case to Fagles’ Iliad), see John Farrell’s 2012 piece in the LA Review of Books on recent translations, “The English Iliad,” which is cited below.


named. It shows Fitzgerald not only adopting an “extemporaneous” and “inventive” approach in imitation of his bardic forebears. We also see him here intervening to reduce the strangeness of the Homeric text for modern readers for whom univocal renderings of the formulae are bound to seem, to recall Carne-Ross’ phrasing, “repetitious.” In the continuous reshaping of koruthaiolos, then, we observe how Fitzgerald works to make the epithets readable for his modern audience.

I have stressed thus far the innovation of Fitzgerald’s approach. Yet in his embrace of variation, Fitzgerald was also continuing an established practice within the tradition of the English Homer noted at the outset of this piece. He was joining a club whose membership included the likes of Chapman and Pope. Yet it is important to recognize that he was doing so on new “post-Parry” terms. His engagement with M. Parry’s thought produced what I am tempted to call the “Fitzgerald paradox.” By applying the logic of “performance-invention” to the epithets, Fitzgerald makes them more difficult to spot as epithets. The translator’s “use” of the oral tradition erases—or at least diminishes—some of the signs by which we recognize the presence of that tradition in the translated text.

Fitzgerald’s Successors: Fagles and Lombardo

Fitzgerald represents the first member of an emerging line of translators whose ruminations on the epics’ ties to the oral tradition have licensed similarly “free” approaches to the epics’ formulaic elements. Briefly, I want to consider two of the best-known of these successors: Robert Fagles (Iliad, 1990; Odyssey, 1996) and Stanley Lombardo (Iliad, 1997; Odyssey, 2000). In the “Translator’s Preface” to his Iliad, Fagles articulates with particular clarity the new terms of Homeric translation in the post-Parry age. As noted in the introduction, he calls this the “Homeric Question facing all translators: How to convey the power of [Homer’s] performance in the medium of writing?” (1990b:ix). In the corresponding “Translator’s Postscript” to his Odyssey, he notably adds the adjective “quieter” in front of “medium of writing” (1996b:489). The translator isn’t just conveying a text; he or she is also tasked with expressing the dynamism of live performance. Somehow the page must become the performer.16 

Now, Fagles does not take Parry’s insights uncritically. In that same preface, he characterizes Homer as “less the creature of an oral tradition whom Milman Parry discovered, and more and more its master, as envisioned by Parry’s son, Adam. Homer the brilliant improviser deployed its stock, inherited features with all the individual talent he could muster” (1990b:ix). Fagles’ answer to what we might call the “Translator’s Homeric Question” is, in turn, propelled by this answer to the scholar’s Homeric Question of the epics’ origins.

This Homer mixes tradition and individual talent, and Fagles translates that mixture into his approach to the epithets: “I have treated them in a flexible, discretionary way, not

16 Fagles’ translations are notable in this respect for the dynamism of their punctuation and page layouts (or what bibliographers commonly refer to as “mise-en-page”). Fagles makes rampant use of indentations, dashes, and ellipses, among other typographical fireworks, especially in his depictions of action scenes. In his Odyssey, Fagles employs at strategic moments triadic lines that recall the work of William Carlos Williams—including in his memorable description of the death of the suitor Antinous: “from his nostrils— / thick red jets— / a sudden thrust of his foot—” (1996a:440).


incompatible with Homer’s way, I think—especially when his formulas are functional as well as fixed—while also answering to the ways that we read today” (1990b:ix). Fagles’ conspicuous debt to Fitzgerald lies in his adoption of his predecessor’s grammatical-cum-narratological approach to the epithets. A case in point is koruthaiolos’ first appearance, 3.83, as noted above. Like Fitzgerald’s rendering, Fagles has Agamemnon mark Hector’s position by his helmet’s glaring: “Look, Hector with that flashing helmet of his” (1990a:131). Again, the epithet is not an honorific but a distinctive prop. In subsequent scenes, we see—à la Fitzgerald—Hector’s “eyes averted under his flashing helmet” (139), “his helmet flashing” (125), “a flash of his helmet” (208), and even “This flashing Hector” (445). As this list suggests, Fagles deviates from Fitzgerald’s practice in maintaining a consistent root across the variations, a kind of figura etymologica strung across its thirty-eight appearances, none of which—another break with Fitzgerald—are omitted in this translation.

Fagles seems to have taken A. Parry’s critique of Fitzgerald to heart, creating through this root-repetition a stronger sense of continuity across the epithet’s appearances. In the “Translator’s Preface,” he explains his approach to the epithets by way of the example of his constrained variation of koruthaiolos: “And so with Hector’s flashing helmet, in the epithet that clings to Hector’s name: I like to ally its gleaming with his actions, now nodding his head in conversation, now rushing headlong to the front lines. But a flashing helmet it is, again and again” (1990b:x). He continues (x-xi):

The more the epithet recurs, in short, the more its power can recoil. And the inevitability of its recoil for Hector is further stressed by a repeated passage in the Greek repeated verbatim in the English version. … All in all, then, I have tried for repetition with a difference when variation seems useful, repetition with a grim insistence when the scales of Zeus, the Homeric moral balance, is at issue.

On this view, the translator must maintain a degree of consistency between the variants in order to alert the reader to the epithet’s overarching thematic significance.

Fagles might be said to split the difference in his handling of the epithets—between the oral tradition and the modern reader, between the “constancy” that A. Parry champions and the variation that Fitzgerald modeled. All of this is in keeping with his notion of “Homer the performer” (as Fagles calls him) as a “brilliant improviser” of the formulaic. For Fagles as for Fitzgerald, the translator must become such an improvisatory performer.

Perhaps Lombardo has an even stronger claim to having “translated” the performer’s art, as his Iliad “began as scripts for solo performances,” as he reports in his “Translator’s Preface” (1997a:ix). As he then points out, “In this respect, the production of the translation mirrors that period in the evolution of the Iliad when writing began to shape the body of poetry that had until then existed only in the mind of the composer in live performance” (ix). Here again, the source text is positioned at the crossroads between the worlds of orality and literacy. In this case, Lombardo adds the additional wrinkle of his own mediation between the project’s beginnings as scripted oral performances of select scenes and its final form as a complete translation for readers of the printed page.


His approach to the formulaic elements arises out of the many pressures of these conditions. He tells us, for example, that for a script-toting modern oral performer speaking to an audience of literates the formulaic elements are a drag on rather than an aid to performance: “strict replication of the formulae (especially those introducing speeches) and heroic epithets would have made the performance seem less alive—stilted in style and slow in pace” (xi). Regarding his preparation of the text for print publication, meanwhile, he argues that “strict replication” of the epithets and other formulae is problematic for two reasons. The first is speed: “Greek hexameters can manage to be both rapid and direct while incorporating polysyllabic compound adjectives that would be deadly in English” (xi). The second is semantic: “no single word or phrase in one language ever completely translates even a simple word or phrase in another language” (xi). The result in the first case is some “streamlining,” and in the second variation. In the latter case, he notes that he not only attempted multiple translations; he also employed “the technique of turning an adjective into an image or an event and integrating it into the action” (xii). Lombardo observes the results on Athena’s epithet glaukopis, which has been generally translated as “gray-eyed” or “owl-eyed.” In one spot, he renders it as “eyes as grey as slate,” in another “as grey as winter moons,” and in a third “Athena’s eyes glared through the sea’s salt haze” (xii).

In all of these strategies—of streamlining, varying, and “imagizing” (if you will)—, Lombardo falls in line behind Fitzgerald. In respect to the last two tactics, though, Lombardo is no thoughtless imitator. Whereas Fagles “recoils” to a degree from Fitzgerald’s inconstancy, Lombardo outdoes Fitzgerald in diversity. Again, the example of koruthaiolos is instructive. If we include omission among the variations, Lombardo translates the epithet nineteen different ways over the course of his Iliad. This includes variations on “helmet shining”: we see it “shimmering” (1997:337), “glancing in light” (104), “collecting light” (115), “flashing light” (127), and “gathering the fading light” (135). The epithet is solidified as “his burnished helmet” (122), “Hector’s helmeted face” (154), and “Hector’s bronze mask” (231). And it is energized, too, becoming “his helmet flashed gold” and “Sunlight shimmered on great Hector’s helmet” (337, 119). In its last appearance, near the end of Book 22, it reverts to an honorific: “tall-helmed Hector” (437). If Fagles’ method might be said to comprise a figura etymologica writ large over the epic, Lombardo’s is perhaps best likened not to one figure but to the family registered under the head of amplification (the spur to Erasmus’ famous 195 variations of “Your letter pleased me mightily” in De Copia). The translator here does not just bear out different suggestions of the Greek; he revels in drawing out the aesthetic possibilities of the image.17 This helmet is repeatedly polished across the text, flashing both in the battle and on the page.

17 In her new translation of the Odyssey, Emily Wilson practices similar sorts of amplification, and, notably for present purposes, on “media theoretical” grounds. Wilson argues in her translator’s note that “In an oral or semiliterate culture, repeated epithets give a listener an anchor in a quick-moving story” (2017b:84). However, “In a highly literate society such as our own,” she argues, “repetitions are likely to feel like moments to skip” (84). In her translation, she often puts a slightly (or greatly) different spin on an epithet with each appearance. Take, for example, the rising of “rosy-fingered” Dawn. It becomes in one context, “The early Dawn was born; her fingers bloomed” (2017c:121), in another, “vernal Dawn first touched the sky with flowers” (187); and in yet another, “Dawn came, / born early, with her fingertips like petals” (301).


Mitchell: Omission is the Best Policy

In Stephen Mitchell’s Iliad (2011), we encounter a text that has been controversial from the get-go. Indeed, we might say that Mitchell made his translation controversial even before he wrote a word by basing it on M. L. West’s two-volume edition of the epic (1998, 2000).18 West’s approach is worth pausing for a moment to consider, since it, too, hinges on questions of orality and literacy. Like all of the scholars that we have been discussing thus far, West viewed the received Iliad text as an archeological site unto itself. Yet rather than go digging in pursuit of the precursor bardic tradition as Parry had done, West searched the text for signs of its corruption by a later group of oral performers, the rhapsodes. In his review of the edition, Gregory Nagy provides a superb summary of West’s account of the Iliad’s provenance (2004:40-41):

The poem was written down in the course of the poet’s own lifetime. Even during his career, the poet had the opportunity to make his own changes in his master poem. … After the master’s death, the scrolls (volumina) of his Iliad were abandoned to the whims of rhapsōidoi (ῥαψῳδοί), “rhapsodes,” who kept varying the text in their varied performances, much like the actors of a later era who kept varying the text left behind by Euripides. …

As Nagy further observes, “the opportunities for introducing more and more interpolations kept widening” in subsequent centuries, including the “Athenian accretions” that trailed the text’s formal adoption for recitation in the Panathenaia in the late sixth century (41). In contrast to the other scholars we’ve been examining, then, West’s goal was to demarcate the intrusions of an oral tradition—again, that of the rhapsodes rather than Parry’s non-literate bards.19 His goal, in turn, was to illuminate the text’s written composition, “the seventh-century Ionic text of the master poet” (41). “Wherever West has deleted or bracketed a passage,” Mitchell writes, “I have omitted it from my translation,” including the entirety of Book 10 (2011c:lvii). The goal, the translator explains, is to “[get] back to an original and a text that I could use as the basis for the most intense poetic experience in English” (lviii).

This same drive to pare the Iliad down in order to produce an “intense poetic experience” is evident in Mitchell’s approach to the epithets. In the translator’s note, he writes, “I have been quite sparing with one of the characteristic features of Homer’s oral tradition, the fixed or stock epithet: ‘flashing-helmeted Hector,’ ‘bronze-clad Achaeans,’ ‘single-hoofed  horses,’  and  so on” (2011b:lx). His explanation of the difficulty of their translation echoes M. Parry: “Throughout Homeric poetry the fixed epithet simply fills out the meter and is usually irrelevant to the context, and sometimes inappropriate to it” (lx). Mitchell further channels M. Parry’s anxiety that modern readers are prone to attribute significance to epithets where there is none.

18 In the note on the Greek text in his 2013 Odyssey translation, meanwhile, Mitchell laments that there is nothing comparable to West’s Iliad edition available and describes all of the editions on offer as inadequate to the task of ferreting out rhapsodic interpolations: “I have not been able to depend on any of them” (xxxix).

19 To use Foley’s categories in How to Read an Oral Poem, the rhapsodes would comprise “voiced texts,” which he defines as a “type of oral poetry that begins life as a written composition only to modulate to oral performance before a live audience” (Foley 2002:43). West’s edition hinges on the notion of a feedback loop between the “voicing” of the text and the text itself during the “rhapsodic” period.


Indeed, Mitchell seems to take M. Parry’s argument to the extreme in asserting that the epithets have no meaning at all—they are “simply” metrical filler. M. Parry’s suggestion that the epithets do have meaning against the background of the texts-as-a-whole isn’t mentioned in this translator’s note, a point to which we will return below. While Mitchell grants that “occasionally” an epithet can be meaningful in its immediate context (another point to which we will return), his overall judgement is clear: epithets are dangers to be avoided when dealing with readers who can’t help granting every word weight. He ultimately claims M. Parry’s authorization for his general strategy of omission: “as the Homeric scholar Milman Parry said, they are best left untranslated” (lx). Yet, as we saw above, Parry’s point was that the epithet’s functional role is untranslatable. He does not directly advise such a scorched earth approach.

Mitchell also notably cites Matthew Arnold’s counsel in On Translating Homer (1861) in defense of his strategy (lx: emphasis in first paragraph mine):

In Greek these epithets elevate the style; in English they are often merely tedious. Here again Arnold’s advice is helpful. “An improper share of the reader’s attention [should not be] diverted to [words] which Homer never intended should receive so much notice.” “Flashing-helmeted Hector,” for example, means no more than “Hector”; the poet is not calling attention to Hector’s helmet. The Trojans aren’t any less “bronze-clad” than the Achaeans. The “single-hoofed” horses are not being differentiated from any imaginary double-hoofed horses.”

Another example: at the beginning of Book 1 Apollo shoots plague-arrows at the Achaeans. The Greek says literally, “First he attacked the mules and the swift dogs.” Here Apollo is attacking all the dogs—the slow ones too, if there should be any, not just the swift ones.

We should recognize that Arnold’s immediate concern in the passage cited here is to observe that translators hatch “strange unfamiliar” phrases in English when rendering expressions that are “perfectly natural” in Homer’s Greek. His particular example is merops (µέροψ), which his rival—and partial catalyst to the On Translating Homer lectures—Frances Newman had unforgivably translated as “voice-dividing” (91). Arnold attends, in other words, to how the English rings in the ear. He would preserve what we have seen Fitzgerald characterize as the “taken-for-granted” nature of much of the Greek. As the italicized bit of the first paragraph shows, Mitchell is concerned with the way that the epithets play up certain qualities of the object in question, thereby lending them a significance that he doesn’t find in the Greek. The foregoing analysis suggests why our chosen epithet, koruthaiolos, makes the list of error-prone examples. For this epithet exemplifies how Fitzgerald and his successors make something tangible of the epithets, something eye-catching. While the Victorian Arnold is the only party explicitly named, the passage would seem to have other targets in view: Mitchell’s immediate predecessors, including the figures discussed above. By largely omitting the epithets, Mitchell quietly suggests, his translation steers the English Homer back on course.

Of Mitchell’s treatment of that epithet, we can say nothing of its active role in the text, since its lone appearance in this Iliad is in the translator’s note. In all thirty-eight of its occurrences, Mitchell omits it, leaving in its place a bare proper name, whether “Hector” or, in the case of 20.38, “Ares.” Koruthaiolos is thus exemplary of what the classicist John Farrell calls Mitchell’s “art of subtraction” (2012). Yet the translation is not completely bereft of epithets.


Mitchell’s “sparing” inclusion of them still admits, for example, the famous pair in the seventh line of Book 1: “that king of men, Agamemnon, and godlike Achilles” (2011a:1). And in the translator’s note, Mitchell himself singles out the appearance of “godlike Priam” at Achilles’ tent in Book 24 as an instance where an epithet “does have meaning” (lx). Thus, while M. Parry is the authority named in the translator’s note, critics of the more flexible A. Parry-type have left their imprint on this translation, too. In fact, the translator’s discretion might be said to be particularly on display in this translation, as the few epithets that do make the cut stand out all the more against the background of the general policy of omission. Priam appears all the more “godlike” in Book 24 as a result of the removal of the word on numerous other occasions where it appears in the Greek.20 

The translation of the oral tradition is treated here, as we saw in M. Parry’s thesis, not as  a linguistic problem. Mitchell can readily enough supply English renditions—including for epithets that appear only in the translator’s note. The issue is the reader’s likely misunderstanding of (what Mitchell takes to be) their lack of significance.21 

The value of Mitchell’s Iliad for present purposes, though, lies not only in its interpretation of Parry’s remarks on translation. It also allows us to reflect on the consequences of the strategy of omission. Here I want to begin by recalling that M. Parry held that fixed epithets do have meaning—only when measured against the epics-as-a-whole, or heroic poetry more broadly, rather than individual appearances. (Recall A. Parry’s apt phrasing that the epithets “remind us of the heroic nature of the world of epic poetry.”) This global significance included, once again, the enumeration of the specific and general qualities of the Homeric heroes. This function of the epithets is sacrificed in Mitchell’s translations, as noted in several reviews.

Consider, for example, the aforementioned John Farrell’s remarks. Farrell, let us acknowledge, voices his preference for Mitchell’s practice of cutting epithets as “dead weight” over Fagles’ of “bringing them falsely to life.” Yet he nonetheless observes that Mitchell’s strategy results in “a certain thinness to the characters’ identities” because the reader is not “ritually reminded … of their distinguishing characteristics and the names of their fathers. Those repetitions may be an artifact of oral composition, but they also reflect the concerns of Homer’s world” (2012). On this view, cutting the epithets brings more than just the loss of the air of “nobility and grandeur” that the epithets bestow even after their novelty fades (M. Parry’s argument noted above). Stripping the epithets from the text amounts to a thematic disrobing of the heroes. On this point, Foley has memorably likened an epithet’s effect to “a trademark musical theme associated with a character in a modern film or a costume that identifies a re-entering actor in a drama even before he or she speaks or is spoken to” (2007:15). Mitchell leaves out the vast majority of the fixed epithets on the grounds that they are “simply” the vestiges of ancient musical performance. Foley and Farrell would have us see that the repetition is not only mechanical; it is the drumbeat of a heroic world.

20 The phrases “godlike” or “like a god” show up a little more than twenty times in Mitchell’s text. By contrast, Fagles includes one of these expressions—which encompass two epithets in the Greek—more than fifty times. Lattimore includes more than eighty in his Iliad. And he has been recently outdone by Caroline Alexander (2015), whose translation features more than one hundred sixty such expressions.

21 The translator also observes that the mass omission of epithets speeds the text up—in keeping with Arnold’s argument that “rapidity” is one of the signatures of the Homeric style.


Again, the fate of koruthaiolos is illuminating. Mitchell does not cast Hector’s helmet out of the Iliad entirely. It still appears on several occasions in the text, including the famous meeting with his wife and son (where it frightens the latter) in Book 6. Yet the removal of thirty-seven references to Hector’s shining headgear reduces not only his grandeur but also the frequent reminders that Hector is a creature of the battle plain rather than the city (or the bedroom like his brother Paris). He is a member of Ares’ brood, as the attachment of this epithet to the War-God emphasizes. Even in the narrator’s recollection of Hector’s wedding to Andromache in 22.471—at once the last invocation of the epithet in the poem and chronologically the earliest—he is “Hector of the flashing helmet”—at least he is in the Greek original. Whether or not Fagles’ particular argument about how the epithet “recoils” over the course of the poem is correct, he is right to argue that the epithet serves as a consistent backdrop against which the audience evaluates Hector’s choices and ultimate fate. The wholesale omission of the epithet does more than dim the helmet’s luster; it alters the audience’s relation to the man under the epithet.

Mitchell quietly contends that Fitzgerald and his ilk have gone astray in making the epithets material cynosures. Better to omit, he argues, than mislead. Yet as Farrell and his fellow reviewers have argued, the epithets are far more deeply woven into the fabric of the epics than Mitchell  admits  in  characterizing them as “simply” ancient musical filler.22 The “art of subtraction” may thus avoid one pitfall but not without introducing hazards of its own.

Powell: A Middle Way?

Thus far, I have presented Fitzgerald (and his imitators) and Mitchell as rivals. Yet we must recognize that their approaches to the epithets are in fact driven by a common desire to reach modern readers. Fitzgerald and company would do this by animating the epithets, Mitchell by cancelling them out. Barry Powell’s translations of the Iliad (2014) and Odyssey (2014) present an alternative conception of the audience’s role, which underwrites, in turn, a third “post-Parry” approach to the epithets’ translation. Unlike the aforementioned parties, Powell argues that the audience must to some extent conform to the “repetitive style” of the text exactly because it springs from the oral tradition. The foreignness of the poem’s style is thus not a problem that the translator alone must solve: the onus is also in part on the reader to adapt to it. In his Iliad’s introduction (which he titles “On Translating Homer” with a nod to Arnold), Powell writes, “To enjoy our modern Homer, we must teach ourselves to accept this repetitive, formulaic

22 In his review of Mitchell’s Iliad for The New Yorker, titled “Battle Lines,” Daniel Mendelsohn astutely observes that Mitchell’s argument that the epithets were meaningless obscures their role in establishing the poem’s “authority” (2011):

For Mitchell, Homer’s famous epithets can obscure what he calls the “meaning”: “‘Flashing-helmeted Hector,’” he writes, “means no more than ‘Hector.’” But “meaning” isn’t the point. Part of the way in which the epic legitimatizes its ability to talk about so many levels of existence and so many kinds of experience is its style: an ancient authority inheres in that old-time diction, the plushly padded epithets and stately rhythms.

On this reading, the strangeness of the epithets that we have seen translators trying to remove is, in fact, key to its ancient success. The more traditional-sounding the poem, the most powerful its voice on the deepest human matters.


style, evolved in order to help the poet create his rhythmic line  on the fly in oral composition” (36).

Yet Powell is not a hard-liner. In the ensuing paragraphs, when he takes up those formulaic elements, the epithets in particular, he argues not for their consistent, literal rendering but what he calls a “middle way” that allows for both repetition and variation (37):

The translator faces the temptation to ignore these epithets entirely and translate “Achilles the fast runner” simply as “Achilles.” This would produce a translation that is not very fair to the poet- singer, obscuring the reality of the origin of these poems as oral compositions. Another strategy is to always translate the epithets in a different way, for example, “swift-footed Achilles” or “Achilles the fast runner” (for the Greek podas ôkus, “swift as to his feet”), again hiding the origin of the text as an oral poem.

I have followed a middle way: using the epithets, thus making clear that his poem is composed in an oral style, but sometimes allowing a different wording, or ignoring the epithet altogether, in accordance with modern taste. Still, we have to adjust to the repetitive style if we want to read a translation of Homer. Homer is an oral poet and he is singing in an oral style, a style utterly practical but grounded in the practicalities of oral presentation.

This passage reveals Powell to be not only a careful reader of the Greek Homer but also its varied English counterparts that we have considered. Both of the strategies that we have considered above—the arts of subtraction and diversity—are rejected here on the grounds that neither is “very fair” to the poem’s origins in the context of oral performance. In response to the subtraction camp, embodied here by Mitchell, Powell argues that the epithets must be “used” lest this vital sign of the oral style disappear. With the diversifiers, the line of Fitzgerald, though, Powell acknowledges that modern taste can tolerate only so much repetition. So “sometimes” variants are admitted. As the last words again stress, Powell wants the reader to perceive the strangeness of the poem’s style. Powell recognizes that the oral style poses a problem for translators due to its distance from modern literary practice; yet, in contrast to his predecessors, Powell sees that strangeness as an essential quality of the Homeric epics. The translator must, in turn, strive to preserve it. To do otherwise amounts to a betrayal—of the poet-singer, of course, but also of the reader from whom the epics’ origins would be concealed.

Regarding our exemplary epithet, koruthaiolos, Powell’s approach yields six variations over the course of the text, only slightly less than Fagles’ nine. But there is a pronounced and telling difference in the relationship between their variations and the sentences in which the epithet is lodged. Fagles, as we have seen, endeavors to integrate the epithet into action in diverse ways—again, at one point the epithet becomes a helmet that “flashes,” at another a “helmet flashing,” and at yet another a “flash of the helmet.” Powell, by contrast, maintains the distinction between the action taking place in the sentence and the epithet. The two translations that he most frequently uses, “of the flashing helmet” (thirteen times, once with “helm” in place of “helmet”) and “of the sparkling helmet” (nine), are applied to Hector in a general manner rather than a site-specific one. While Powell offers four other renditions of the epithet, their differences amount to the swapping of synonyms within a grammatical pattern announced in the epithet’s first appearance at 3.83: “Hector, whose helmet flashes” (2014a:94). Whether the


helmet “flashes” or “flashed,” “sparkles” or “sparkled,” these renderings are all sealed off by commas as relative clauses that do not contribute to the action immediately taking place. Thus, while Powell has clearly taken his cue from Fitzgerald and his successors in introducing a degree of variation to satisfy “modern taste,” his grammatical management of the epithets suggests that he agrees with Mitchell that translators should not convert the epithets into tangible goods. Hearkening back to Lattimore (and a long line of English translators before him), Powell wields them as character-markers. They have the air of a herald’s introduction rather than the weight of battle-scene props.

In turn, Powell is the most successful of the translators under consideration here in rendering the “ritualistic” nature of the epithets’ reappearances (to borrow Farrell’s language), their function as reminders of characters’ “distinguishing characteristics.” The risk of this tactic is that the epithets may at times (even often) seem extraneous to what is going on around them. It is the avoidance of this possibility, of course, that drives Fitzgerald to assimilate them and Mitchell to set them aside. In Powell’s case, by contrast, that scenario seems less a risk than a reward. For in seeming detached from their immediate settings, the epithets gesture toward the oral origins of the poems that Powell wishes to reveal rather than conceal. In the repetition of “Hector’s flashing helmet,” Powell would have us see at once the Homeric technique of characterization—a “capsule biography” of its wearer—and the “practicalities” of oral performance at work (2014b:37).


We have now seen that there is no simple consensus among modern translators about how the epithets ought to be handled in the wake of Parry’s theories. For some, the epithets are irreducibly strange, prompting efforts to polish them up, to make them more vivid presences and active participants in the text. That same diagnosis has licensed wiping out large numbers of them, some epithets completely. To our final translator, those tactics concede too much to contemporary taste; he argues that the work of adjusting to the epithets belongs, for the most part, to the reader rather than the translator. In their responses to the epithets, these translators employ strategies that recall those of their forebears in the English Homer tradition. Yet they debate the epithets’ merits on entirely new terms.

Those terms, as I have stressed throughout this piece, derive from Milman Parry’s linguistic excavations of the traditional roots of the Homeric poems. Through the example of the epithets, and one “flashy” specimen in particular, we have pondered the multiple issues that have trailed Parry’s discoveries. For Parry’s unraveling of the unfamiliar terms on which ancient Greek oral poetry operated—in the case immediately before us, the traditional motive for epithet- invocation—was also the revelation of the psychological gap between ancients and moderns. In Parry’s own writing, as we have seen, that gap made the epithets appear especially troublesome for translators. Beginning with Fitzgerald’s translations, we have witnessed translators adopt numerous stances in response. For all of them, though, Parry’s thought has changed the game. To translate Homer now is to do more than make judgements about diction, syntax, meter, and other familiar matters. It is to mediate between the strategies that arose within the crucible of ancient


composition-in-performance and those that hold the attention of the modern reading public. The epithets demand that the translator become a kind of medium, speaking with the dead and writing to the living.

In our review of options here, we have seen a kind of tug-of-war for the oral tradition among translators. Some view it as a problem, which must be kept from or adapted to by readers. Others have set it up as a model for their own “inventive” approaches to the text. Parry taught us to ask how an oral tradition means. In the English Homer tradition, I have tried to suggest, that question remains a live one, at once conceptual and practical. We are still working out its answer generation-by-generation, line-by-line.

Does Hector’s helmet flash? We’ll see.

Wheaton College


Alexander 2015                    Caroline Alexander, trans. The Iliad. New York: Ecco.

Arnold 1861                    Matthew Arnold. On Translating Homer. London: Longman, Green, Longman, and Roberts.

Carne-Ross 1998                    D. S. Carne-Ross. “The Poem of Odysseus.” In The Odyssey. Trans. by Robert Fitzgerald. New York: Farrar, Straus, and Giroux. pp. ix-lxx.

Elmer 2015                    David F. Elmer. “The ‘Narrow Road’ and the Ethics of Language Use in the Iliad and the Odyssey.” Ramus, 44.1-2:155-183.

Fagles 1990a                    Robert Fagles, trans. The Iliad. New York: Viking.

Fagles 1990b                    ______. “Translator’s Preface.” In The Iliad. Trans. by Robert Fagles. New York: Viking. pp. ix-xiv.

Fagles 1996a                    ______, trans. The Odyssey. New York: Viking.

Fagles 1996b                    ______. “Translator’s Postscript.” The Odyssey. Trans. by Robert Fagles. New York: Viking. pp. 489-496.

Farrell 2012                    John Farrell. “The English Iliad.” Rev. of The Iliad. Trans. by Stephen Mitchell. LA Review of Books. 10 October 2012. english-iliad/. Accessed 8 May 2019.

Fitzgerald 1985                    Robert Fitzgerald. Interview with Edwin Honig. In The Poet’s Other Voice.  Ed. by Edwin Honig. Amherst: University of Massachusetts Press, 1985. pp. 99-114.

Fitzgerald 2004a                    ______, trans. The Iliad. New York: Farrar, Straus, and Giroux.

Fitzgerald 2004b                    ______. “Postscript.” In The Odyssey. Trans. by Robert Fitzgerald. New York: Farrar, Straus, and Giroux. pp. 467-510.

Foley 1993                    John Miles Foley. Traditional Oral Epic. Berkeley: University of California Press.

Foley 1997                    ______. “Oral Tradition and its Implications.” In A New Companion to Homer. New York: Brill. pp. 146-173.

Foley 1999                    ______. Homer’s Traditional Art. University Park: Pennsylvania University Press.

Foley 2002                    ______. How to Read an Oral Poem. Chicago: University of Chicago Press.

Foley 2007                    ______. “‘Reading’ Homer through Oral Tradition.” College English, 34.2:1-28.

Ford 2004                    Andrew Ford. “Introduction.” In The Iliad. Trans. by Robert Fitzgerald. New York: Farrar, Straus, and Giroux. pp.xi-xxxviii.

Gibson 2019                    Richard Hughes Gibson. “On Women Englishing Homer.” Arion, 26.3:111-144. Gordon 2016     Kendal Gordon.“Introduction.” In George Chapman: Homer’s Odyssey. George Chapman. Cambridge: Modern Humanities Research Association. pp. 1-36.

Knox 1991                    Bernard Knox. “Introduction.” In The Odyssey. Trans. by T. E. Lawrence. Oxford: Oxford University Press. pp. ix-xxii.

Lattimore 1967                    Richmond Lattimore, trans. The Odyssey of Homer. New York: Harper and Row.

Lattimore 2011                    ______, trans. The Iliad of Homer. Chicago: University of Chicago Press.

Lawrence 1930a                    T. E. Lawrence to Bruce Rogers. 1 January 1930. Letters from T. E. Shaw to Bruce Rogers. Mount Vernon, NY: The Printing House of William Rudge. pp. 49-50.

Lawrence 1930b                    ______. 3 March 1930. Letters from T. E. Shaw to Bruce Rogers. Mount Vernon, NY: The Printing House of William Rudge. pp. 54-55.

Lawrence 1932                    T. E. Lawrence. “Translator’s Note.” In The Odyssey. Trans. by T. E. Lawrence. Oxford: Oxford University Press. pp. i-vii.

Lombardo 1997                    Stanley Lombardo. “Translator’s Preface.” In The Iliad. Indianapolis: Hackett. pp. ix-xv.

Mendelsohn 2011                    Daniel Mendelsohn. “Battle Lines.” Rev. of The Iliad. Trans. by Stephen Mitchell. The New Yorker. 7 November 2011. pp. 76-81.

Miola 2017                    Robert S. Miola. “Introduction: Looking into Chapman’s Iliad.” In George Chapman: Homer’s Iliad. Trans. by George Chapman. Cambridge: Modern Humanities Research Association. pp. 1-18.

Mitchell 2011a                    Stephen Mitchell, trans. The Iliad. New York: Free Press.

Mitchell 2011b                    ______. “About This Translation.” In The Iliad. Trans. by Stephen Mitchell. New York: Free Press. pp. lix-lxii.

Mitchell 2011c                    ______. “About the Greek Text.” In The Iliad. Trans. by Stephen Mitchell. New York: Free Press. pp. lvii-lviii.

Mitchell 2013                    ______. “About the Greek Text.” In The Odyssey. Trans. by Stephen Mitchell. New York: Atria. pp. xxxix-xl.

Nagy 1996                    Gregory Nagy. Homeric Questions. Austin: University of Texas Press.

Nagy 2004                    ______. “Editing the Homeric Text: West’s Iliad.” In Homer’s Text and Language. Chicago: University of Illinois Press. pp. 40-74.

Newman 1861                    Francis W. Newman. Homeric Translation in Theory and Practice: A Reply to Matthew Arnold, Esq. London: Williams and Norgate.

Ong 1982                    Walter Ong. Orality and Literacy. London: Routledge.

A. Parry 1962                    Adam Parry. Rev. of The Odyssey. Trans. by Robert Fitzgerald. Fat Abbot, 4:48-57.

A. Parry, 1972                    ______. “Language and Characterization in Homer.” Harvard Studies in Classical Philology, 76:1-22.

M. Parry 1928                    Milman Parry. L’épithète traditionelle dans Homère. Paris: Société D’Éditions.

M. Parry 1987                    ______. The Traditional Epithet in Homer. In The Making of Homeric Verse: The Collected Papers of Milman Parry. Trans. by Adam Parry. Oxford: Oxford University Press. pp. 1-190.

Powell 2014a                    Barry B. Powell, trans. The Iliad. Oxford: Oxford University Press.

Powell 2014b                    ______. “Introduction.” In The Iliad. Trans. by Barry B. Powell. Oxford: Oxford University Press. pp. 1-39.

Sale 1996                    Merritt Sale. “In Defense of Milman Parry: Renewing the Oral Theory.” Oral Tradition, 11.2:374-417.

West 1998                    Martin L. West. Homeri Ilias. Volume 1. Stuttgart and Leipzig: B. G. Teubner.

West 2000                    ______. Homeri Ilias. Volume 2. Munich and Leipzig: K. G. Saur.

Wilson 2017a                    Emily Wilson. “Found in translation: how women are making the classics on their own.” The Guardian. 7 July 2017. 2017/jul/07/women-classics-translation-female-scholars-translators.  Accessed 8 May 2019.

Wilson 2017b                    ______. “Translator’s Note.” In The Odyssey. Trans. by Emily Wilson. New York: Norton. pp. 81-92.

Wilson 2017c                    ______, trans. The Odyssey. New York: Norton.