Use the following persistent identifier: http://nrs.harvard.edu/urn-3:hul.ebook:CHS_Martin.The_Language_of_Heroes.1989.
Chapter 4. The Language of Achilles: Language, Formula, and Style
{146} Every hero is a performer. That is the essence of the dictum Peleus entrusts to Phoinix, who in turn reminds Achilles to be a speaker of words and a doer of deeds. [1] Between the two concepts no distinction is drawn. Both are performances. The poetry anticipates Austin and Searle in treating speech as act, part of an economy in which talk about one’s action is as important as deeds themselves, and in which no feat can survive without its afterward. Nor is recounting alone sufficient; style, in words as in fighting, marks the man: Senecan self-consciousness thrives in the world of the Iliad. [2]
I have sketched the contours of this world’s speech behavior—the poet’s traditional division of talk into epos and muthos, the former referring to utterance as such, the latter to the complete act of speaking, with its concomitants: tone, occasion, appropriateness to the audience, result—in other words, to performance. Thersites, ametroepēs, does not produce pleasing discourse, not only because the substance of his speech in Book 2 is repugnant to the elite, but also because his words are deficient as utterance. By Iliadic standards, they {146|147} are unmetrical, as we saw, lines with no clear breaks, full of slurring correptions, more so than ordinary speech in the poem. Transitions that would define the flow of speech are harsh or nonexistent.
But while the poet on occasion dwells on epea to draw attention to their unusual form or content, the main interest always appears to be the play of performances. The speech-acts of the speaker-heroes occupy the foreground. The epic itself is constructed around one vital speech-act, supplication. [3] The poem opens with pleas—both poet’s and priest’s—and rises through threat and refusal to prediction, assurance, threat, counterthreat, and grand refusal, until, with the assurance of Thetis that she will beseech Zeus for her son’s sake we reach the end of one ring-composition, on the human level, and proceed to the gods’ rendition of the same cycle. In the prelude of the poem it is not the utterances so much as the moods in which they emerge that hold us, and the poet labels these by delineating a series of muthoi—Agamemnon’s scorning of Khryses, for example, Athena’s commands to Achilles, even the words of Zeus to Hera in Book 1 are introduced with this word.
Yet these two systems for talking about speech in the Iliad are not parallel lines. At times, speaking well is the best revenge. For Achilles, whose words have exiled him from the arena of speech, creating fine epea remains his only option for enacting a muthos. It is a crucial performance.
The power of Achilles’ representation, particularly his appearance in Book 9, has persuaded readers since Plato that the words of the hero are somehow different from ordinary discourse. The ancient critics explained the difference as a matter of style, without, however, advancing beyond impressionistic descriptions of each character’s way with words: Achilles’ plain and passionate style was contrasted with the elaborate rhetoric of Odysseus, the fullness of Nestor, the laconic speech of Menelaos. [4] The only extended commentary on Achilles’ {147|148} speech, as on the Iliad as a whole, before the Renaissance comes in the scholia found on the great early manuscripts of the poem and in the copious exegesis of Eustathius, the twelfth-century bishop of Thessalonica. The latter’s remarks we might expect to be particularly valuable, as we know that Eustathius served as the imperial maistōr rhetōrōn, writing everything from eulogies to pleas for improvements in public water supplies, and thus was attuned to the spoken word. Yet in Eustathius’ view, as in the scholia, style is seen as only an extension of personality: both are haplos, or semnos, “simple” and “serious,” in the portrayal of Achilles. [5] It is not until the twentieth century, and Milman Parry’s discovery of how systematically formulaic Homeric diction can be, that the notion of a Homeric character possessing an individual way of speaking has been questioned. If the great poet himself did not lay claim to his own “style,” but rather inherited his art wholesale from generations before him, how could any figure he created come to express anything but what had been said before? Had not Parry shown that all Homeric speech was traditional? Where did this leave the hero who had made the great refusal—Achilles? Was he required to mouth the usual formulas in order to make clear his own dissenting view on the war at Troy?
The first to ask these essential questions was Milman Parry’s son, Adam. In an influential article published when he was twenty-eight, the younger Parry argued that Achilles, bound by the formulaic nature of his diction, “has no language with which to express his disillusionment” and yet succeeds in expressing it “by misusing the language he disposes of. “ [6]
{148|149} Adam Parry had posed his essay as an attempt “to explore some of the implications of the formulaic theory of Greek epic verse.” [7] On the grounds that the epic formula represents for the poet the single best expression for a given idea, Parry asserts that “the style of Homer emphasizes constantly the accepted attitude toward each thing in the world, and this makes for a great unity of experience.” [8] The argument continues with reference to Sarpedon’s famous defense of the heroic code (Il. 12.310-28), the main point of which (honor equals tangible goods) Parry finds to be agreed on by all Iliadic heroes. This universal agreement is next compared to Homer’s own formulaic style. The crucial step in the reasoning of the essay comes with Parry’s assertion that “the economy of the formulaic style confines speech to accepted patterns which all men assume to be true”— speech, thought, and reality are an undivided whole. Yet Achilles, who seems to perceive “the awful distance between appearance and reality,” who distrusts the false front of Odysseus, is unable to fit into this perfect Nominalist world. He is “the one Homeric hero who does not accept the common language, and feels that it does not correspond to reality,” writes Parry. At the same time “neither Homer in his own person as narrator, nor the characters he dramatizes, can speak any language other than the one which reflects the assumptions of heroic society,” those precepts uttered by Sarpedon. [9] Therefore, Achilles’ “misuse” of language is needed to break out of the formulaic system: it consists of his asking questions without answers and making demands that cannot be met; the former, when he questions the need for the fight with Troy, the latter, his request that Agamemnon pay back his disgrace. Adam Parry declines to discuss in detail the great speech in which Achilles allegedly misuses his language, but describes it as “passionate, confused, continually turning back on itself,” and takes this as another sign of Achilles’ inability to fit the heroic world. [10]
The description might rather have been applied to the state of scholarly discussion on Homeric style once Milman Parry’s important work began to become widely known, especially through the {149|150} writings of his collaborator Albert Lord. [11] It is helpful to recall something of this controversy in order to place Adam Parry’s challenging article in perspective; then we can examine the ensuing debates on the “language of Achilles,” and take new steps to solve the problem. The first critical responses sought to save Homer’s own “originality,” the question of idiosyncratic speech on the part of epic characters being left aside for the moment. Milman Parry’s discoveries had raised four interrelated questions: How much of Homer is formulaic? What exactly is a “formula”? Does a proof of Homer’s formulaic style mean necessarily that the epics were oral poems? And if so, how should one interpret repetitions within the poetry?
Of course, the question of whether Homer wrote had been asked and answered in various ways since antiquity. [12] Perhaps because it was such an old question, Parry’s answer received the most attention. His suggestion that Homer was an oral poet gained ground because his fieldwork in Yugoslavia far surpassed in accuracy and scope the casual observations of earlier travelers acquainted with actual oral poetry and, in addition, focused on a specific poetic technique—the deployment of formulas within Serbo-Croatian heroic songs. The analogy with Homer seemed compelling. Subsequent studies have shown that the Serbo-Croatian material, while different in many ways from Homeric verse, still has a claim to being one of the best comparanda. [13] Parry apparently solved the historical problem, through the demonstration that systematic formularity underlay Homer’s art, and the suggestion that this could only result from a long tradition of oral performance and composition. The solution for a time overshadowed the continuing and perhaps more important literary problem of the “meaning” of the Homeric poems, as it did the more technical unsolved questions about the definition of the formula and the overall “formularity” of the epics. Acceptance of Milman Parry’s conclusions seemed to carry with it a denial that one could find “meaning” at the level of the individual Homeric word, phrase, or line. Scholarship echoed contemporary politics: one was called “hard” or “soft” on “Parryism.” [14] Yet the contention of the “hard” Parryists, {150|151} that an oral poet could never express himself as subtly as a writing poet, never won a large following. Lord himself, although anxious to quash certain interpretations which he thought relied on unrealistic subtleties of formulaic repetition, had pointed out that oral poets used other means to distinguish themselves; rather than by recondite vocabulary or allusiveness, they ornamented and expanded their performances of traditional material by adding lines, motifs, and themes. [15] In the years following publication of The Singer of Tales (1960) a number of studies located Homeric innovation and creativity precisely within his tradition, in the variation, juxtaposition, and expansion of preexisting motifs and diction. [16] The poet’s capability for artful composition has been reasserted with new awareness of the levels at which “subtlety” occurs.
In short, the work of Milman Parry proved liberating. Austin has pointed out that, “far from eliminating literary criticism from Homeric studies, Parry . . . opened up possibilities that Analytic studies had made seem highly suspect. A grammar of Homeric poetics can be written, but not if we suppose that Homer is, either wholly or substantially, a victim of his metrical formulas.” [17] Although it is to this day unclear to what extent “oral,” “traditional,” and “formulaic” overlap as descriptions of Homeric poetry, at least it is recognized {151|152} that these labels do not limit the expressive power of the poet; “expressiveness” is simply posited at a different level. The ground has been cut from under Adam Parry’s contentions about Achilles’ “language.” In retrospect, it can be seen that the younger Parry made three unquestioned assumptions in applying his father’s theory. First, he used “language” to mean two very different things: as a shorthand expression for “cultural code” or “value system,” but also in the sense of “diction.” In Saussurean terms, when Adam Parry speaks of Achilles’ inability to accept the “common language,” he refers to the “signified,” that is, the heroic code. When he says Achilles has “no language” to express his disillusionment, Parry really means “no signifiers.” Of course, his concurrent work on the logos/ergon distinction tempted Parry to make this semantic slide in discussing Achilles. Thus, he was led to make a second assumption—that all Homeric language is formulaic. Milman Parry, indeed, seemed to believe that this was so (according to a remark by Antoine Meillet), but he never specified in what way such a statement might be true, other than by pointing out the existence of “formulaic systems” in Homer. [18] Nor did he indicate the essential differences between such regular syntactic patterning and the noun-epithet “formulas.” This brings us to Adam Parry’s third premise: his claim that the “economy of the formulaic style” is what predetermines how Homer’s figures speak. Even if all of Homer’s works were “formulaic” in some sense, I believe it is still the highly developed noun-epithet systems examined by Milman Parry that exhibit “economy and extension,” in a strict sense. There may be only one way to say “Odysseus” at a certain point in the hexameter line; there are a half-dozen ways of saying “Achilles was angry,” however. [19] If formulaic “thrift” is an illusion, if characters can vary their expression at will—why should Achilles “misuse” his language?
Adam Parry’s ideas on the “language” of Achilles were not contested until 1973, when M. D. Reeve briefly and persuasively pointed out that “neither an unanswerable question nor an impossible demand {152|153} is by its nature a misuse of language . . . that is, of traditional vocabulary,” as Parry had argued it was in Achilles’ speech at 9.337-38 and 9.3 87. [20] Illogical as the demand to “pay back heart-rending injury” might be, all-embracing and despairing as the question “Why must the Achaeans fight the Trojans?” may sound—these speech-acts are on a plane wholly different from that of the diction used to express them. Although they may be prompted by Achilles’ perception that there exist constraints in a highly “formulaic” system of correct behavior in war, it is a mistake to equate that system with the system of formulaic language at work in the poem.
The debate over the “language of Achilles” has continued on two levels, although most articles bearing the phrase in their titles perpetuate the original ambiguity as to whether “language” should be taken as “diction” or “thought.” Most work, like that of Reeve, has concentrated on investigating Achilles’ words to show that he makes traditional or nontraditional statements with them. David Claus’s “aidōs in the Language of Achilles” falls into this category. Claus leaves unexamined Adam Parry’s premise that a system of ideas is analogous to a system of poetic formulas. Instead, he observes the tensions and contradictions in the heroic code even within the Iliad— as when a hero must decide between two traditional modes of acting (e.g. the decision of Odysseus at 11.404 whether to retreat or be defeated). Given such explicit choices at the level of behavior, Claus notes, we might expect similar possibilities for multiple choice and meaning in the linguistic structure of the text. His commonsense argument takes Adam Parry’s reductionist theory to its logical end: if the idea of a complete unanimity of logos and ergon is pressed, notes Claus, “every statement made in the poem . . . must be one that supports entirely what are taken as the fixed ideas of the society, or it cannot be spoken.” [21] This sort of poetry would hardly lend itself to characterization. Yet Claus shows that characters are distinguished in the Iliad exactly on the basis of the way they speak about the heroic code. Sarpedon (12.310-28) expresses a vision of heroism akin to Achilles’ own view, in its assumption that the hero battles not just for his own glory and gain, but for the honor of having benefited his companions without “pay.” Achilles and Sarpedon can speak very {154} different formulas and use various rhetorical devices, and still express the same point of view: so much for the determinism of the “system,” and for Achilles’ alleged “isolation” from it.
It should be noted that Claus circumvents Parry’s argument concerning formulaic economy by moving to a more abstract level. He points out that changing contexts automatically ensure that each speaker uses the same formulas in different ways, and conversely, is able to use different formulas to mean the same thing. While this observation is no doubt true, one could make the same statement about everyday language, which, because it is constructed from the speaker’s standpoint, contains built-in flexibility in the form of linguistic “shifters” to accommodate changing contexts. [22] It is a simple fact of any temporal art that one can never step into its flow in the same way twice. The problem remains with Homer’s formulaic poetry of determining the amount of difference in meaning at each new repetition of phrase or line.
A more recent examination of the “language”—meaning “thought”— of Achilles resorts to similar abstraction. Steven Nimis also argues that formulaic language takes on different meanings according to context. He attempts to reconcile the views of Claus and Adam Parry by posing the “language of Achilles” problem in Chomskyan terms, regarding rule-governed creativity. Achilles, in this way, is a “sign-producer who wishes to change the ‘code,’ to articulate a meaning for whose communication and accurate reception no adequate conventions exist as yet.” [23] How Achilles does this remains unclear in Nimis’s exposition: the changes that the hero makes in the “code” appear to include the use of hyperbole, catachresis, and oxymoron, as well as his refusal to share a communal meal after Patroklos’ death. But surely these are no more than examples of Achilles’ ability to call upon alternatives that are equally conventional; and the “changes” themselves have nothing to do with “language” unless we persist in using that word to denote “behavior.” Hyperbole and the other rhetorical devices are operations applied to language, but in no real sense do they constitute it. [24]
{154|155} Concurrent with these probes into the world of the “signified” as it appears in Achilles’ speeches, there have been attempts, also inspired by Adam Parry’s work, to find idiosyncratic usage of “signifiers” either on the part of Achilles or by Homer in describing him: the distinction is sometimes blurred. Hogan discovers that the distribution of prin, “before/until,” is unusual in the speeches of Achilles, who doubles the conjunction. Although this quirk is neither consistently nor even frequently Achillean, Hogan nevertheless maintains that “Zeus uses the double prin when speaking of him, that Hektor attributes this emphatic form of command to him, that Homer, in the second half of the poem, at least, uses this figure almost exclusively in reference to Achilles,” and concludes that “all these elements contribute in a small but significant manner to the characterization.” [25] Scully points out that the formulas of deliberation which employ okhthēsas differ significantly when they occur in speeches by Achilles: in Achilles’ rhetorical usage, such phrases preface reflections about mortality, rather than monologues concerning the tension between personal need and social expectations (as they do in the speeches of other heroes), Thus, Achilles “uses Stereotypie patterns which outline choice in a manner that differs from other heroes.” [26] The phenomena that these scholars describe, however, could best be ascribed to conscious repetition of formulaic language on the poet’s part; it is misleading to connect this with the “language” of Achilles in Adam Parry’s sense. [27]
These reports are like suggestive sketch-marks. Taken together, they delineate a very distinctive figure, although no one characteristic stands out. Such studies, however, are vitiated by the lack of a general agreement on the background for Homeric characterization: are we to think that an audience would compile a mental dossier of such traits, noting at each turn Achilles’ deviation from a supposed “normal” {155|156} behavior? Or are these small signs only detectable by philologists’ devices, without “meaning” in the performance of the poem, perhaps accidents? What do such intriguing facts of “language” add up to? We need a real portrait, not a sketch.
James Redfield and Paul Friedrich attempted to supply a total assessment of Achilles’ language through the application of a type of linguistic stylistics developed originally for the analysis of natural speech. They found that “Achilles in the Iliad is characterized by individual speech patterns.” For several reasons, their claim demands serious attention. First, the Chicago scholars recognize (as had Claus and Reeve) that Parry’s “language of Achilles” referred to value system, not diction. They rightly label Parry’s equating formulaic language with formulaic thought “a kind of Whorfianism run wild. ” Second, they appear to acknowledge the difficulty of tracing individual speech patterns behind the scrim of a seemingly uniform poeticized speech: as they note, “all the characters in Homer ‘speak Homeric. ‘ ” But here their method falls short of the goal; for, instead of beginning with the assumption that the poet forms his characters’ speeches, Redfield and Friedrich at least seem to proceed as though the individual speakers in epic expressed themselves independently of their creator. [28]
What is more, as it turns out, the nine “distinctive features” which the two analysts find in Achilles’ speech are in fact text-linguistic aspects; they are of a different type from either the facts examined by Parry, Reeve, and Claus, or the features noted by Scully, Hogan, and Cramer. When Redfield and Friedrich include in their analysis such extrasentential structures as accumulation of detail; expanded series of statements (e.g. 9.378-86); use of hypothetical expression and similes; poetic directness in word choice, and so on, they begin to use the term “language” in yet another sense—that is, to stand for “rhetoric.” [29] All these features of Achilles’ speech, particularly of his great reply to Odysseus in Book 9, had been noticed even in antiquity, when attention to rhetoric formed the core of the study of poetic {156|157} style. Indeed, Redfield and Friedrich must concede that several other speakers in the Iliad command the same linguistic resources: Paris is “direct” in his choice of words; Aeneas uses similes; Hektor can build up a rhetorically elaborate series in his speech to Andromakhe in Book 6. In this light, Achilles stands out because he consistently uses certain devices, not because he monopolizes them. In addition, Redfield and Friedrich point out that Achilles can be contrasted with other speakers in his avoidance of certain strategies: unlike Nestor, Odysseus, or other characters, Achilles does not concede points, make distinctions, anticipate his interlocutors’ objections in argument, or offer multiple reasons for his behavior. But this is to say that Achilles is simply a different character in the poem; it so happens that literary “character” is constructed out of language; yet it would be tautologous and misleading to assert that each character thereby has a different “language.” [30]
When the analysis turns to those features that are actually “linguistic” at the level of sentence and clause, rather than at the level of discourse, it becomes less assured and more speculative. Achilles’ speeches are found to contain more asyndetic expressions; more subjunctives—perhaps he is more emotional; more elaborate and combined vocatives, more titles of address, terms of affection and abuse, emotive particles ē and dē—clear signs to the investigators of his passionate nature and dominant relation with his peers.
Redfield and Friedrich do not analyze Achilles’ speeches from the point of view of the formula, other than to note that his words do not differ from those of other Iliadic speakers in the number of formulas per line. [31] They conclude, however, with a glance at the theory that first gave rise to the “language of Achilles” debate. Apparently, they proceed from Milman Parry’s work without reference to later modifications in formula theory, when they assert that “if the choice of adjective is less meaningful than in non-traditional verse sources of meaningful variation are to be sought elsewhere—in the general shape of utterances, in the use of rhetorical devices and in the choice of particles, or particular highly-marked lexemes, or of marked syntactic {157|158} constructions.” But their search for “meaningful variation” takes them too high and too low. On the one hand, the instinct to examine discourse as a whole is admirable. It is a method I have tried to illustrate earlier in this book, one that finds its roots in the poetry itself: after all, the characters of the Iliad do not examine one another’s lexical choices, but are highly critical of one another’s “performances,” that “shape of the utterance” which we have seen designated as epos. Redfield’s and Friedrich’s attempt is actually a step toward articulating a poetics of performance. Nevertheless, it falls victim to a determination, on the other hand, to be perfectly scientific about verbal art: there remains in their study too much confidence in the lowest level of verbal behavior, a belief that counting particles and verb moods can give us insight into an overall characterization by style. Stanley Fish’s memorable critique of this method should serve to discourage stylisticians at large from leaping into the complexities of psychological portrayal from the wobbly ledge of statistics. [32]
Within a few years of its publication in Language, the brave attempt by Redfield and Friedrich to pin down Achilles’ language once and for all encountered strong criticism, in the same journal. Gordon Messing, while likewise dismissing Adam Parry’s seminal article for its method, questions several of Redfield’s and Friedrich’s procedures: the use of a rather small sample (even if it did represent all of Achilles’ words); the lack of attention to manuscript variations, even though the Homeric text is far from fixed; and most of all, the ultimate goal of finding “individualization” in the figure of Achilles, rather than “characterization” according to ēthos, the latter being the universally recognized technique of ancient writers. Messing draws on recent work in stylistics to point out the “serious error” of confusing rhetoric with speech style. Achilles’ rhetoric, of course, is a function of Homer’s own poetic craft: it may have been molded to fit the hero’s individual ēthos. But speech style—if defined as the product of such things as particle use and frequency, preference for certain syntactic constructions, choice of conjunctions, sentence length, and so forth— is unlikely to be found, since the options “might be supposed parts of over-all Homeric style rather than components in the speeches of {158|159} individual Homeric heroes.” [33] The analysis of unconscious choices among grammatical means has been quite useful when applied to dating and attributing texts, but the choices themselves cannot have been a vehicle for communicating, to a listening audience, differences in fictional characters, concludes Messing. Thus, he shows, misapprehensions about ancient poetic convention and modern stylistics damage the most concerted effort to grant Achilles his own “language. “
Because it mirrors the problems of contemporary Homeric studies, it is instructive to review the shifting debate over the “language of Achilles.” How does one interpret formulaic language? Do we need an “oral poetics”? Does linguistics, semiotics, or old-fashioned New Criticism best illuminate Greek poetry? The questions thus raised have critical ramifications for classical studies in general. But, aside from providing some insights on recent intellectual history, reviewing the work thus far shows that the essential steps have still not been taken to prepare us for any sustained study of Homer’s style, let alone such complex issues as characterization by style. Briefly, we need to know where to look for “style.” I suggest that we continue to take account of full stretches of discourse, speeches or narration, not just individual lines; Messing has shown the error of looking at the microscopic linguistic level. Furthermore, I propose that we go back to examining Homer’s formulaic art. As I have tried to show, every critic of Achilles’ language since Adam Parry has neglected to perform the close line-by-line reading that would tell us whether or not Achilles differs, in his use of formulas, from other characters. Yet it seems to me to make sense that, if a poet indeed wanted to characterize his chief hero, he would have done so by using the one feature of his art which (we now know) he could vary, expand, and shape endlessly, and which his audience would be most likely to notice and appreciate—his inherited poetic diction. The number of finely focused, formulaic analyses of Homeric poetry is absurdly small. {159|160} Since Milman Parry, only four scholars (to my knowledge) have attempted such work, resulting in examination of approximately 250 lines from a corpus one hundred times as large; no one has thought to analyze Homeric speeches at this level of specificity. [34] One can no longer avoid the task by the assertion that all of Homer is uniformly formulaic: even if it is we are obliged to examine the play of formulas. Ten years after his “Language of Achilles” paper, Adam Parry had come to see this: “The analysis of formulary diction shows us that there can be no or very little individual vocabulary and individual combination of single words. Therefore, the individuality which is so obviously there, and so much a part of the poem’s greatness, must lie in the juxtaposition of formulae. Achilles and Odysseus must use the same phrases: but they combine them into speech in separate ways.” [35] Yet neither he nor other scholars devised a way to show how this happens, or what it implies. This I propose to do now.
First, however, we need both a clearly defined notion of style, grounded in current stylistic theory, and a working definition of the formula. The lack of consensus regarding the latter has been the greatest hindrance to formulaic analysis, whether for “proving” that a text is oral or for interpreting the individual Homeric passage. My method relies on defining the formula in such a way that it can be used as an interpretive tool. In other words, I am not now interested in whether or not the Iliad is an “oral” text, by any criterion of formulaic density. [36] In fact, what I propose is a method for detecting “formulas” that does not start with a priori definitions; it is more a technique than a science, a distinction that I believe could be applied to formulaic art itself. I have linked this method to a specific view of stylistics, so it is best to mention here my working principles concerning that field.
The notion of style for me is intimately connected with performance; I have indicated earlier that “performance” is a more inclusive term for the “speeches,” so-called, made by Iliadic heroes. It has the {160|161} advantage also of being a recognizable feature of behavior in every society, so that the interpreter of poetic style as it occurs in traditional media can benefit from a growing body of work in folklore, anthropology, and the ethnography of speaking, regarding actual performances, whether poetic, social, or somewhere in between. Again, the work of Herzfeld on the “poetics” of men’s behavior in a Cretan mountain village, that of Bauman on performances in Icelandic literature and in native American cultures, and Abraham’s studies on men-of-words contain relevant and illuminating comparative material that fits with the actual picture of the role of speaking given us by the Iliad. [37] I am encouraged by one recurrent theme in their findings: the importance in each traditional culture studied of a creativity that is based entirely on the reuse and recombination of traditional themes, formulas, and “ways of speaking.”
Because the style of all traditional performances is recognizably one of new variations on old, audience-accepted performances, the stylistician working with Homeric poetry can best employ a method based on a foreground/background distinction: I have attempted in what follows to trace deviation from a norm with the understanding that “deviant” language (in this case, formula use) by itself tells us nothing about a performer or performance. It gains meaning only when compared at each turn with the apparent norm. The approach has its roots in Russian Formalist practice, and has been concisely explained by Geoffrey Leech. [38] He distinguishes three types of variation that could be investigated: first, the abnormal irregularities (such as hyperbaton; unique lexical items) or abnormal regularities (e.g. parallelism to a high degree) in a text, which stand out by virtue of being different from naturally occurring speech; second, deviations that can be found when texts are compared with the patterns of an entire canon, either that of a genre or of an author; and finally, what he calls “tertiary deviations”—those variations detectable when parts of a composition are compared against the background of the work itself. I have chosen to investigate Achilles’ language at this third level, comparing every expression in his great reply to Odysseus (9.308-429) with similar expressions elsewhere in the Iliad. [39]
{161|162} My method, then, for finding the formulas in this speech of Achilles is designed to pick out features that would make his speech uncharacteristic in terms of the composition in which it occurs. In theory, Achilles’ speech could contain no formulas, or only formulas of a certain type; or it could be filled with traditional phrases that, nevertheless, were unusual either by their meaning in context or their positioning in the hexameter, when compared with the rest of the poem. If we find any of these features, we would be entitled to speak of characterization by style—that is, there would be, in one sense, a “language of Achilles.” The only way to discover this is to undertake a thorough analysis based on a technique of indicating what is “formulaic.” Not only must we display the “deviations” from traditional phrasing and placement of diction; the analysis also involves interpreting repetition at every level by checking the content of the “formulas” against the rest of the poem.
My method of circumscribing the formula creates a unified field theory based on the two diverging trends in defining the formula since Parry first explained it as “a group of words which is regularly employed under the same metrical conditions to express a given essential idea.” [40] His definition was later pulled and tugged in two directions in the interests of quite different projects: the establishment of the “orality” of the Homeric texts, on the one hand, and, on the {162|163} other, the investigation of how an oral poet might have thought as he composed. The consequent modification of Parry’s original definition occurred particularly when critics examined the requirement of identical metrical conditions, and the grouping of words. In effect, the first trend in defining the formula came to ignore the “group of words” and the “given essential idea” as boundaries, and instead concentrated on the recurrence of metrical patterns fitted to certain grammatical and syntactical units. This method had its roots in Parry’s own observations on the “analogical” formula, which he called “formula types”—such systems as ἄλγεα πάσχων and ἄλγε’ ἔδωκε. The work of Notopoulos, Lord, and Russo stretched this idea until it became possible to call “formulaic” even a phrase that was never repeated, yet was built on an often-attested pattern. Although the fallacy of the “structural formula” was soon pointed out, and it became clear that the phenomenon is mostly due to word-localization and colon structure in the hexameter and partly due to inherited word-order rules in Greek, the consensus has been that such structural templates are an important element in the poet’s repertoire. Nagler fruitfully extended the notion in yet another direction to point out the important role of seemingly irrational phonological-metrical patterning in the poet’s employment of certain motifs. Hainsworth has included such patterns in his listing of the ten types of repetitions that have, at one time or another, been called “formulas” in Homer. The question remains open when it comes to explaining how one should interpret such structural formulas and analogical formulas in the text of the poem. Thus far, the occurrence of the type has been used mainly for arguing the high formula content of Homer. I shall deal with this problem later, when I come to speak of repetition in general. [41]
While this expansion of the meaning of “formula” was under way, a concurrent countermove to restrict the term to definable word-groups was led by Hainsworth. His valuable work proved that the metrical conditioning of the formula was not a necessary assumption: {164} instead, he suggested, we should define the formulas as a pair or group of words connected by mutual expectancy. The formula’s elements can thus occur next to one another, but they can also be separated over the space of one line or several, inflected, and expanded by the addition of other formulas. Hainsworth’s demonstration has the advantage of stressing the importance of a formula’s persistent meaning rather than its grammatical shape. In turn, identifiable semantic similarities can be exploited when one comes to interpret, say, two passages containing similar formulas within the Iliad. Hainsworth’s notion also gains support from the work of Nagy in comparative Indo-European poetics, which illustrates how formula, defined as the expression of a theme, precedes the metrical realization of the phrase historically. [42]
We can take account of these two trends in defining the formula if we simply construct a model using the structural linguistic notions of syntagmatic and paradigmatic. [43] Just as any speaker selects lexical elements from a paradigmatic axis of possible sounds and words, then combines these elements on the syntagmatic axis, so we can imagine how the poet who composes in formulas must at each point in the line face paradigmatic choices to fill out the syntagmatic structure of the hexameter line. In time, poetic traditions would develop long patterns, built on a group of words which could be repeated at certain points in the line—for instance, the epithets of gods and heroes—and which would enable the composer to make verse quickly in performance conditions. These I call syntagmatic formulas. A rough definition might be “regularly repeated phrases”; they are of the type which Parry first defined and investigated. I assume that they have all the “flexible” properties which Hainsworth attributes to them. The question of the historical evolution of such formulas appears to me moot. It could be that the contiguous co-occurrence of two elements, such as epithet and noun (the “regular occurrence” in the “same metrical conditions”), precedes their separation and mobility, or instead, the co-occurrence itself may be simply an ossification of one type of {164|165} mutual appearance (as can be paralleled at the linguistic level in the case of preverbs becoming compounded with verbs). In any event, poets could rely on such mutual expectancies to create meaning, and on audiences to gauge the extent of a performer’s talent at the play of formulas. My method for finding mutual expectancies in Achilles’ speech has been to check on all other mutual occurrences of any two words within three lines, by using a machine-readable text of the poem.
At the same time, a formulaic poet can be imagined as relying on certain patterns—the “structural” formulas, words and sounds with metrical association—at those points in the verse where he did not choose to use longer pre-made elements, or where such phrases did not yet exist. These “structural” patterns are no less traditional; the essential difference between them and the “syntagmatic formulas” is that the mere repetition of a certain phrase structure, without the repetition of exact words, could not carry intentional semantic meaning to an audience. Violating the structural patterns, however, would most likely signal something to an audience, although it may be no more than the “message” one gets when music is played at the wrong tempo, or when grammatical English is pronounced with a foreign accent. I call “paradigmatic” such things as the repeated occurrence of a word in the same metrical position, with “word” being as small an element as a particle, pronoun, or conjunction. My method for finding these paradigmatic elements in the speech of Achilles has been to check each word’s distribution in the text of the Iliad, to determine whether it usually occupies a given metrical slot. This procedure is actually another way of detecting structural formulas; it has the advantage of pinning down the attestation of the exact words used in the test passage, rather than simply showing that noun plus adjective or noun plus participle patterns exist similar to those used in Achilles’ reply.
We can assume that the evolution of the paradigmatic formulas sprang from a combination of several factors. On the one hand, the breakdown of old syntagmatic formulas would have left traces in the frequent positioning of a single word in a single metrical slot; on the other hand, discourse factors, such as the strong tendency to put anaphoric pronouns and demonstratives at the beginning of sentences, or rhetorical factors, such as the tendency to enjamb certain elements for poetic effect (e.g. patronymics, sentence adverbs, emphatic pronouns), would inevitably lead to regular use of single {165|166} words in one or two metrical slots. [44] Again, I am not so much concerned with the historical development of the “paradigmatic” formulas here; I only wish to note their usefulness for composition, if not for the interpretation of the poem. I neither claim to have discovered what is actually a formula in the tradition, nor do I believe that this step is possible in most cases (with the exception of certain epithets, perhaps). I merely wish to show what is “traditional” vis-à-vis the Iliad in Achilles’ speech and what is “innovative” so that we might get a sense of what an audience would attend to as new-sounding within the course of the performance of the poem.
Chart 1 shows all the “paradigmatic” formulas in Achilles’ speech with underlined italics. [45] The syntagmatic formulas are underlined. At a glance, my main conclusion emerges: almost all of Achilles’ great speech consists of formulas, either paradigmatic or syntagmatic. That is, the speech is traditional, in terms of the Iliad itself. [46] But, more important, the number of paradigmatic formulas far outweighs the number of syntagmatic; Achilles as a speaker (which is to say Homer when imagining how he speaks) chooses to use very small, unconnected bricks for his edifice: only occasionally can he rely on the ready-made longer phrase. Each choice of word in a given place in the line is one that can be paralleled elsewhere, but the effect of this method of composing in discrete units is one of tone: we seem to hear a man searching laboriously for the right word at every turn. I believe this is the effect that John Finley referred to when he observed: “Homer carries both heroes, Achilles chiefly, to the other side of brilliant action. The steps are hard and gradual, seemingly for two main reasons: that the poetic tradition described actions, and that the heroes, like the tradition, reached isolation painfully. It is as if neither was fully prepared for it.” In a way, I have simply quantified Finley’s telling intuition. In addition {166|167}, I suggest that the underlinings, and even more so, the frequent slash marks that separate “paradigmatic” formulas in the speech of Achilles, give us graphic proof of the primary tenet held by field-workers in oral literature, Lord in particular, that every song is both completely traditional and completely new. Having seen the formulaic poet working this way, we cannot help being reminded of an archaic metaphor: that poetry is carpentry, literally a tekhnē in which the poet consciously fits and rejoins small pieces to make a crafted whole. [47]
Chart 1. The Reply of Achilles to the Embassy (Iliad 9.307-429)
307 Τὸν δ’ ἀπαμειβόμενος προσέφη πόδας ὠκὺς Ἀχιλλεύς·
διογενὲς Λαερτιάδη πολυμήχαν’ Ὀδυσσεὐ
χρὴ / μὲν δὴ / τὸν μῦθον / ἀπηλεγέως / ἀποειπεῖν,
310 ᾗ/ περ δὴ / φρονέω / τε καὶ / ὡς τετελεσμένον ἔσται,
ὡς μή / μοι τρύζητε / παρήμενοι / ἄλλοθεν ἄλλος.
ἐχθρὸς / γάρ / μοι κεῖνος / ὁμῶς / Ἀίδαο / πύλῃσιν
ὅς χ’ / ἕτερον μὲν / κεύθῃ / ἐvὶ φρεσίν, / ἄλλο / δὲ / εἴπη.
αὐτὰρ ἐγὼν ἐρέω ὥς μοι δοκεῖ εἶναι ἄριστα·
315 οὔτ’ / ἔμεγ’ / Ἀτρείδην Ἀγαμέμνονα / πεισέμεν οἴω
οὔτ’ / ἄλλους Δαναούς, / ἐπεὶ οὐκ ἄρα τις χάρις ἦεν
μάρνασθαι δηίοισιν ἐπ’ ἀνδράσι νωλεμὲς αἰεί.
ἴση μοῖρα μένοντι / καὶ εἰ μάλα / τις / πολεμίζοι‧
ἐν δὲ ἰῇ / τιμῇ / ἠμὲν κακὸς / ἠδὲ καὶ ἐσθλός·
320 κάτθαν’ / ὁμῶς ὅ τ’ ἀεργὸς / ἀνὴρ ὅ τε πολλὰ ἐοργώς.
οὐδέ τί μοι / περί /κειται, / ἐπεὶ πάθον ἄλγεα θυμῷ
αἰεὶ / ἐμὴν ψυχὴν / παραβαλλόμενος / πολεμίζειν.
ὡς δ’ ὄρνις ἀπτῆσι / νεοσσοῖσι / προφέρῃσι
μάοτα/κ’ ἐπεί κε / λάβησι, κακῶς / δ’ ἄρα οΐ πέλει / αὐτῇ,
325 ὥς καὶ ἐγὼ / πολλὰς / μὲν αὔπνους / νύκτας ἴαυον,
ἤματα / δ’ αἱματόεντα / διέπρησσον / πολεμίζων
ἀνδράσι μαρνάμενος ὁάρων / ἕνεκα / σφετεράων.
δώδεκα / δὴ σὺν νηυσὶ / πόλεις / ἀλάπαξ’ / ἀνθρώπων,
πεζὸς / δ’ ἕνδεκά / φημι / κατὰ / Τροίην ἐρίβωλον·
330 τάων / ἐκ πασέων / κειμήλια πολλὰ καὶ ἐσθλὰ
ἐξελόμην, / καὶ πάντα / φέρων / Ἀγαμέμνονι / δόσκον
Ἀτρείδη· / ὅ δ’ ὄπισθε / μένων παρὰ / νηυσὶ / θοῇσι
δεξάμενος / διὰ παῦρα δασάσκετο, / πολλὰ δ’ / ἔχεσκεν.
{167|168} Chart 1. (continued)
ἄλλα / δ’ / ἀριστήεσσι / δίδου / γέρα / καὶ / βασιλεῦσι·
335 τοῖσι / μὲν ἔμπεδα / κεῖται, / ἐμεῦ δ’ ἀπὸ / μούνου / Ἀχαιῶν
εἵλετ’, / ἔχει δ’ ἄλοχον / θυμαρέα· τῇ παρ / ιαύων
τερπέσθω. τί δὲ δεῖ / πολεμιζέμεν / αι / Τρώεσσιν
Ἀργείους; / τί δὲ / λαὸν / ἀνήγαγεν / ἐνθάδ’ / ἀγείρας
Ἀτρείδης; / ᾖ οὐχ / Ἑλένης / ἕνεκ’ / ἠϋκόμοιο;
340 ἦ μοῦνοι / φιλέουσ’ / ἀλόχους / μερόπων ἀνθρώπων
Ἀτρείδαι; / ἐπεὶ / ὅς τις /ἀνὴρ / ἀγαθὸς / καὶ / ἐχέφρων
τὴν αὐτοῦ / φιλέει καὶ κήδεται, / ὡς καὶ ἐγὼ / τὴν
ἐκ θυμοῦ / φίλεον / δουρικτητήν / περ έοῦσαν.
νῦν δ’ ἐπεὶ / ἐκ χειρῶν / γέρας / εἵλετο / καί μ’ ἀπάτησε
345 μή μευ πειράτω / εὖ εἰδότος· / οὐδέ με πείσει.
ἀλλ’ / Ὀδυσεῦ / σὺν / σοί τε / καὶ ἄλλοισιν / βασιλεῦσι
φραζέσθω / νήεσσιν / ἀλεξέμεναι / δήϊον πῦρ.
ή μεν δη / μάλα πολλά / πονήσατο / νόσφιν ἐμεῖο,
καὶ δὴ / τεῖχος ἔδειμε, / καὶ / ἤλασε τάφρον / ἐπ’ αὐτῷ
350 εὐρεῖαν μεγάλην, ἐν δὲ σκόλοπας κατέπηξεν·
ἀλλ’ οὐδ’ ὧς / δύναται / σθένος / Ἕκτορος ἀνδροφόνοιο
ἴσχειν· / ὄφρα / δ’ ἐγὼ / μετ’ Ἀχαιοῖσιν / πολέμιζον
οὐκ ἐθέλεσκε μάχην / ἀπὸ τείχεος / ὀρνύμεν / Ἕκτωρ,
ἀλλ’ / ὅσον / ἐς / Σκαιάς τε πύλας καὶ φηγὸν ἵκανεν·
355 ἔνθα / ποτ’ / Οἶον Ἔμιμνε, / μόγις / δέ μευ / ἔκφυγεν / ὁρμήν.
νῦν δ’ / ἐπεί / οὐκ / ἐθέλω / πολεμιζέμεν Ἕκτορι δίῳ
αὔριον / ἱρὰ / Διὶ / ῥέξας / καὶ πᾶσι θεοῖσι
νηήσας εὖ / νῆας, / ἐπὴν / ἅλαδε προερύσσω,
ὄψεαι, αἴ κ’ ἐθέλῃσθα καὶ αἴ χέν τοι τὰ μεμήλῃ,
360 ἦρι μάλ’ / Ἑλλήσ/ποντον ἐπ’ ἰχθυόεντα πλεούσας
νῆας ἐμάς, / ἐν δ’ / ἄνδρας / ἐρεσσ / έμεναι μεμαῶτας·
εἰ δέ κεν / εὐπλοίην / δώῃ / κλυτὸς ἐννοσίγαιος
ἤματί / κε / τριτάτῳ / φθίην ἐρίβωλον / ἱκοίμην.
ἔστι / δέ/ μοι / μάλα πολλά, / τὰ / κάλλιπον / ἐνθάδε ἔρρων·
365 ἄλλον / δ’ / ἐνθένδε / χρυσὸν καὶ χαλκὸν / ἐρυθρὸν
ἠδὲ γυναῖκας ἐϋζώνους πολιόν τε σίδηρον
ἄξομαι, ἅσσ’ ἔλαχόν γε· / γέρας / δέ / μοι, / ὅς / περ / ἔδωκεν,
αὐτις / ἐφ/ὑβρίζων / ἕλετο κρείων Ἀγαμέμνων
Ἀτρείδης· / τῷ / πάντ’ ἀγορευέμεν ὡς ἐπιτέλλω
370 ἀμφαδόν, ὄφρα καὶ ἄλλοι ἐπισκύζωνται Ἀχαιοί
εἴ τινά που / Δαναῶν / ἔτι / ἔλπεται / ἐξαπατήσειν
αἰέν / ἀναιδείην ἐπιειμένος· / οὐδ’ ἄν / ἔμοιγε
τετλαίη κύνεός / περ ἐὼν / ἐις ώπα ἰδέσθαι·
οὐδέ τί / οἱ / βουλὰς συμφράσσομαι, / οὐδὲ μὲν / ἔργον· {168|169}
375 ἐκ γὰρ δή μ’ / ἀπάτησε / καὶ / ἤλιτεν / οὐδ’ ἄν / ἔτ’ αὖτις
ἐξαπάφοιτ’ / ἐπεεσσιν· / ἅλις δέ οἱ· / ἀλλὰ ἕκηλος
ἐρρέτω / ἐκ γάρ εὑ φρένας εἵλετο / μητίετα Ζεύς.
ἐχθρὰ / δέ μοι / τοῦ δῶρα, / τίω δέ μιν / ἐν καρὸς / αἴσῃ.
οὐδ’ εἴ μοι / δεκάκις τε καὶ εἰκοσάκις / τόσα δοίη
380 ὅσσα / τέ / οἱ νῦν / ἔστι, καὶ εἴ / ποθεν / ἄλλα / γένοιτο,
οὐδ’ ὅσ’ / ἐς / Ὀρχομενὸν ποτι/νίσ/εται, / οὐδ’ ὅσα / Θήβας
Αἰγυπτίας, / ὅθι / πλεῖστα / δόμοις / ἐν / κτήματα / κεῖται,
αἵ θ’ / ἑκατόμ/πυλοί / εἰσι, / διηκόσιοι δ’ ἀν’ ἑκάστας
ἀνέρες / ἐξοιχνεῦσι / σὺν ἵπποισιν καὶ ὄχεσφιν·
οὐδ’ εἴ μοι / τόσα δοίη / ὅσα / ψάμαθός τε / κόνις / τε,
385 οὐδέ κεν / ὧς ἔτι / θυμὸν ἐμὸν / πείσει’ / Ἀγαμέμνων
πρίν γ’ ἀπὸ / πᾶσαν / ἐμοὶ / δόμεναι / θυμαλγέα / λώβην.
κούρην / δ’ / οὐ / γαμέω / Ἀγαμέμνονος Ἀτρείδαο,
οὐδ’ εἰ / χρυσείῃ Ἀφροδίτῃ / κάλλος / ἐρίζοι,
390 ἔργα / δ’ / Ἀθηναίῃ γλαυκώπιδι / ἰσοφαρίζοι·
οὐδέ μιν / ὧς / γαμέω· /ὁ δ’ / Ἀχαιῶν / ἄλλον ἑλέσθω,
ὅς τις / οἷ τ’ / ἐπέοικε / καὶ ὅς / βασιλεύτερός ἐστιν.
ἤν / γὰρ δή /με / σαῶσι / θεοὶ / καὶ οἴκαδ’ ἵκωμαι,
Πηλεύς / θήν / μοι ἔπειτα / γυναῖκά / γε / μάσσεται / αὐτός.
395 πολλαὶ / Ἀχαιίδες / εἰσὶν ἀν’ / Ἑλλάδα / τε / Φθίην / τε
κοῦραι / ἀριστήων, / οἵ τε / πτολίεθρα / ῥύονται,
τάων ἥν κ’ ἐθέλωμι / φίλην / ποιήσομ’ ἄκοιτιν.
ἔνθα δέ / μοι / μάλα πολλὸν / ἐπέσσυτο / θυμὸς ἀγήνωρ
γήμαντα / μνηστὴν ἄλοχον / ἐïκυῖαν / ἄκοιτιν
400 κτήμασι / τέρπεσθαι / τὰ / γέρων ἐκτήσατο Πηλεύς·
οὐ γὰρ ἐμοὶ / ψυχῆς ἀντάξιον / οὐδ’ ὅσα / φασὶν
Ἴλιον / έκτήσθαι / εὖ ναιόμενον πτολίεθρον
τὸ πρὶν ἐπ’ εἰρήνης, πρὶν ἐλθεῖν υἷας Ἀχαιῶν,
οὐδ’ ὅσα λάϊνος οὐδὸς ἀφήτορος ἐντὸς ἐέργει
405 Φοίβου Ἀπόλλωνος / Πυθοῖ / ἔνι / πετρηέσσῃ.
ληϊστοὶ / μὲν γάρ τε / βόες καὶ ἴφια μῆλα,
κτητοὶ / δὲ / τρίποδές / τε καί / ἵππων ξανθὰ κάρηνα,
ἀνδρὸς / δὲ / ψυχὴ / πάλιν / ἐλθεῖν / οὔτε / λεϊστὴ
οὔθ’ / ἑλετή, / ἐπεὶ ἄρ / κεν / ἀμείψεται / ἕρκος ὀδόντων.
410 μήτηρ / γάρ / τέ / μέ φησι / θεὰ Θέτις ἀργυρόπεζα
διχθαδίας / κῆρας / φερέμεν / θανάτοιο τέλοσδε.
εἰ μέν κ‘ / αὖθι μένων / Τρώων πόλιν / ἀμφιμάχωμαι,
ὤλετο / μέν μοι / νόστος, / ἀτàρ /κλέος / ἄφθιτον / ἔσται·
εἰ δέ κεν / οἴκαδ’ ἵκωμι / φίλην ἐς πατρίδα γαῖαν,
415 ὤλετό / μοι / κλέος ἐσθλόν, /ἐπὶ / δηρὸν / δέ μοι / αἰὼν
ἔσσεται / οὐδέ κέ / μ’ / ὦκα /τέλος θανάτοιο / κιχείη.
καὶ δ’ ἂν τοῖς ἄλλοισιν / ἐγὼ παραμυθησαίμην
{169|170} chart 1. (continued)
οἴχαδ’ ἀποπλείειν, ἐπεὶ οὐκέτι δήετε τέκμωρ
Ἰλίου αἰπεινῆς· μάλα γάρ ἑθεν εὐρύοπα Ζεὺς
420 χεῖρα ἑὴν ὑπερέσχε, τεθαρσήκασι δὲ λαοί.
ἀλλ’ ὑμεῖς μὲν / ἰόντες / ἀριστήεσσιν Ἀχαιῶν
ἀγγελίην ἀπόφασθε· / τὸ γὰρ γέρας ἐστὶ γερόντων·
ὄφρ’ / ἄλλην / φράζωνται / ἐvὶ, φρεσὶ / μῆτιν ἀμείνω,
ἥ / κέ / σφιν / νῆάς τε σαῷ / καὶ λαὸν Ἀχαιῶν
425 νηυσὶν ἔπι γλαφυρῇς, / ἐπεί οὔ / σφισιν / ἥδε / γ’ / ἑτοίμη
ἥν / νῦν / ἐφράσσαντο / ἐμεῦ ἀπομηνίσαντος·
Φοῖνιξ δ’ /αὖθι / παρ’ / ἄμμι / μένων / κατακοιμητθήτω,
ὄφρα μοι / ἐν νήεσσι / φίλην ἐς πατρίδ’ / ἕπηται
αὔριον / ἤν ἐθέλῃσιν· ἀνάγκῃ δ’ / οὔ τί / μιν / αξω.
To have shown that Achilles’ composition is like that of poets in living oral traditions is a powerful confirmation of the argument put forward earlier in this book, that all speakers in the poem are “performers” in traditional genres of discourse. There still remains the possibility that Achilles’ mode of performing, his employment of repeated diction, in some way is idiosyncratic: words, after all, are not mosaic pieces; the sounds refer to something. Does Achilles use formulas in the way they are usually deployed? To answer this, we need to face yet another hard issue, that of the meaning of repetition within a formulaic art. Let me say, first, that I believe neither that every repetition in Achilles’ speech is significant, nor that the mere fact of being repeated deprives an expression in the speech of ascertainable meaning. To anticipate my conclusion slightly, I would say that Achilles’ use of syntagmatic formulas (the sole cases open to interpretation) is idiosyncratic only in that he uses expressions elsewhere used exclusively by gods in speeches, or by the narrator in diegesis. Thus there is a cohesiveness to the “deviations” in Achilles’ formula use. [48]
Achilles generally employs familiar formulas in new ways; at the same time, what seem like new and innovative uses can in fact be explained as reworkings of familiar expressions paralleled elsewhere within the poem. In the remaining pages of this chapter, I will point out as many of these reshapings as possible. But my main goals -will be to delineate another phenomenon in Achilles’ use of formulaic {169|171} art—his expansion aesthetic—and to locate the impulses that give rise to it. We shall see in this chapter and the next that it is the working of this aesthetic principle alone which creates the illusion of an independently existing “language of Achilles” in the Iliad.
Types of Repetition
My method of determining the “new” and “old” uses of formulas in Achilles’ speech depends on two assumptions: that there is a range of repetitions in Homeric poetry, and that repeated expressions do not occur in a vacuum. We will review the potential for creating meaning at each level of repetition shortly. Here I am most concerned with the level that is most amenable to stylistic analysis, and most tied to a higher tier of formulaic art. Just as formulas in the narrative are organized according to theme, as Lord showed, and themselves imply given themes whenever they recur, so formulaic expressions in speeches are organized according to “genres of discourse.” These small “genres,” which I have examined at length in Chapters 2 and 3, comprising threats, boasts, praise and blame, prayers, prophecies, and several other categories based ultimately on individual speech-acts, will be the primary tool for my analysis.
If we examine the formulas in Achilles’ speech in terms of their appropriateness to one or another genre of discourse, attested elsewhere in the poem, we can establish the larger background which is a prerequisite for making statements about the foregrounded “language of Achilles.” This technique is especially useful in those cases where Achilles mixes genres, for we can be led by a few formulas to trace the genres involved, and thus to see what Achilles leaves out in reshaping the conventional ways of speaking about certain topics. The method can help us to open up the Iliad, so as to study its construction as a monumental epic. [49] When working at this level of discourse, although I approach Achilles’ speech from the point of view of oral traditional poetics, I find myself in agreement with the insights of workers in another area of Homeric studies, the Neo-analysts, on the fundamental premise that certain portions of Homeric epic allude with intention to other specific contexts. But whereas Neo-analysts, {171|172} such as Kullmann and J. Kakridis, discuss variation of narrative motifs from poem to poem, while extrapolating backward in time from such sources as the Cyclic epics, I concentrate on “speech motifs,” if one can call them that. I believe that their repetition within the Iliad itself gives us enough material to construct a “norm” of use, against which to play Achilles’ variations. Instead of specific context as the focus of allusion, I prefer to think of specific conventional ways of talking in a given speech-genre. [50]
An example can clarify how one discovers the speech-genres that organize certain formulas. When we look for parallels to line 366 in Achilles’ speech in Book 9, the occurrence of this same whole line at 23.261 first catches the eye. It occurs in the poet’s listing of goods that Achilles brings from his tent for distribution at the Funeral Games. It seems to have no significance other than as confirmation—later in the poem we learn that Achilles does indeed have the women and iron to which he alludes in Book 9. Three of the five categories of other goods mentioned at 23.259-60 are also mentioned by Achilles in his great refusal, but separately from his reference to women and iron, and in a more emphatic manner, when he claims that “cows and stout sheep are for the taking, tripods and the tawny heads of horses can be gotten” (9.406-7). Of course, the references to these goods in Achilles’ reply are prompted by Agamemnon’s offer of tripods and gold, cauldrons, horses, and women (9.122—30). Notice that none of the three passages just cited contains the exact repetition of a list; Agamemnon, Achilles, and the poet name the goods in different order, with shifting emphasis: Agamemnon expatiates, taking three lines (9.128-30) to describe the women, five for the horses (123-27); Achilles mentions cows, without describing them, and sheep, with the epithet iphia. The poet speaks of “stout heads of cattle” (a variant of the formula Achilles uses for horses) and adds to the list mules (not referred to by Achilles). Yet below this surface multiformity and ornamentation of description there is a common ground for the very mention of lists, in the speech-act of formal declaration. And this act, in turn, can be considered conventional within a genre of discourse that is prominent in the Iliad: raiding boasts. The genre features several topics: who took what from whom; what were the precise gains; what division of spoils was made later. We can see these topics clearly {172!173} addressed in Nestor’s recollection of his initiatory raid (11.677-83, 696-705). A few of the details in that story are worth comparing with facets of Achilles’ speech. Nestor recounts the circumstances of his people’s raid on Elis, justifying the raid by citing the abuses suffered by the Pylians whose numbers had been reduced: (hēmeas hubrizontes, 11.695). Nestor’s father took part of the spoils from the raid to recompense himself (heileto, exelet’, 697, 704), and gave out the rest for equal distribution (705-6). Achilles mentions all these details, but in his view a raid has been carried out against himself. Agamemnon has taken advantage (ephubrizōn, 9.368), like the men of Elis he has selected certain things (335-36) and has distributed the rest, but not equally (333-34). In sum, Achilles uses the conventions normal for speaking about one’s relations with outsiders when he talks about his own commander. We can see this as a creative reshaping at two levels: a familiar speech-genre is redeployed for new effect; and thus, Achilles appears as a skillful manipulator of the conventional, a rhetorician. [51]
This type of repetition—formulas attached to specific “genres of discourse”—should be distinguished from another type, which is meaningful at a different level, namely, the occasional repetition at intervals within the Iliad of especially marked phrases. Whereas the first type can be detected only by looking at a number of passages in a “genre,” the second calls attention to itself. We might hesitate between calling the second type “formulaic” or merely “memorable”— as Milman Parry acknowledged, not all repetitions are formulas. I have found that such meaningful repeated phrases usually occur at prominent points in the speech and occupy a half-line. A good example comes in Book 9 at line 372: αἰὲν ἀναιδείην ἐπιειμένος certainly must remind one of the same phrase that Achilles used to describe Agamemnon early in the poem (1.149). This repetition creates meaning by providing a sense of the consistency of Achilles’ own view of the world; it characterizes the hero inasmuch as it tells us his hatred of Agamemnon remains ever fresh. The same effect can also come from {173|174} κύνεός περ ἐὼν εἰς ὦπα ἰδέσθαι (9.373), which recalls a shorter expression from the same speech in Book 1, κυνῶπα (1.159). We have stronger grounds for considering these phrases as repetitions with “contextual surplus” (rather than random formulas) since the ensemble of phrases points back to another passage in which the expressions occur together. [52] And indeed, in the next line of Achilles’ speech βουλὰς συμφράσσομαι takes us back yet again to Book 1, where the phrase is used both by the narrator and Hera to describe Thetis’ parley with Zeus (1.537, 540). Because the phrase is not found elsewhere, we can label it a highly marked repetition on Homer’s part. Moreover it is significant that the phrase is uttered only by Achilles, two goddesses, and the narrator, for this patterning of speakers accords with Achilles’ new use of other, more regularly repetitive expressions. Reminders of the quarrel scene in Book 1 occur later in the speech also: Achilles uses ἀντάξιον (9.401), a word not employed elsewhere in the context of recompense except at 1.136. There is a point to the repetition: whereas Agamemnon had demanded a gift “fitting in return,” Achilles asserts that there is nothing to exchange for his ψυχή. And Achilles’ final words, οὔ τί μιν ἄξω (9.429), surely are to be contrasted with Agamemnon’s threat to take Briseis by force (1.139); Achilles refuses to use compulsion to detain Phoinix.
A special case of the type of repetition just mentioned occurs when characters reuse phrases later that Achilles has employed in this speech: these are only “repetitions” in retrospect: they gain their full meaning only after we hear them here. One example is the phrase ἕλετο κρείων Ἀγαμέμνων (9.368), which will be repeated when Achilles recounts the same incident to Patroklos (16.58) and Thetis tells the story to Hephaistos (18.445). Achilles’ version thus becomes canonical: the repetition produces a sense of the innate truth of his view, as well as emphasizing the urgency of his case. Similarly, we can view the partial repetition of 9.342 (τὴν αὐτὸῦ φιλέει) in 9.450 (τὴν αὐτὸς φιλέεσκεν), and of 9.343 (ἐκ θυμο͂υ φίλεον) in 9-486 (ἐκ θυμο͂υ φιλέων), as a touch of realism on the poet’s part: he makes Achilles’ friend and interlocutor Phoinix pick up uncommon {174|175} phrases from Achilles’ own speech in answering it—such responsion happens in natural conversation. Yet another subclass of this sort of meaningful repetition outside of formulas per se occurs when Achilles in his speech picks up the language of those who have addressed him. Thus, line 347 in Book 9, φραζέσθω νήεσσιν ἀλεξέμεναι δήϊον πῦρ, is modeled on Odysseus’ words at 9.251, φράζευ ὅπως Δαναοῖσιν ἀλεξήσεις κακὸν ἦμαρ. Even though the lines themselves can be segmented into smaller, recurring phrases, the overriding significance comes from the recurrence of the larger structure, within a short interval. This line (347) also illustrates a common type of repetition, which may or may not have poetic meaning: the report of one character’s speech by another. Agamemnon asks the returning embassy whether Achilles intends “to ward off savage fire from the ships” (9.674), using the closest previous formulation, which happens to be Achilles’ words (presumably not heard by him). Such repetition usually arises from the poet’s needs in composition, rather than from desire to characterize any one figure. [53] But at times—for instance, when a character repeats the poet’s own previous narration—we might detect some greater meaning in this technique. Given Achilles’ other associations with the figure of poet, it becomes significant that he repeats whole lines uttered only by the narrator earlier (see 9.350 = 7.441; 9.356 is a transformation of 7.169).
This brings us to two types of repetition on levels different from either the “genre of discourse” formulas or the intentional recall of half-lines or words: namely, the repetition of whole lines and that of patterns (metrical, phonological, or syntactic). As I have already indicated, in speaking of the paradigmatic formulas, the repetition of a word or syllable in a particular metrical slot is usually without any meaning that we can trace. Although unconscious analogies of sound and meter may have had a great influence on the poet’s choice of expression as he composed a line, to find such analogies is generally unhelpful if we are looking at Homeric characterization. It is interesting, then, that the unique noun μάστακ’ (9.324) is positioned in the slot at which similar forms of the unrelated word “whip” (e.g. μάσ-τιξ’) occur, and that ἔτι ἔλπετοα (371) occupies the slot which ἐπὶ τ’ ἔλπεται fills later (24.491), but it would be absurd to argue that any {175|176} poetic meaning inhered in such repetitions: they are accidents of the system. [54]
Of more importance is the repetition of seemingly unexceptional combinations that hover between paradigmatic and syntagmatic status. One such juxtaposition of words comes at line 355: οἶον ἔμιμνε, “He awaited me when I was alone. “ [55] The phrase occurs elsewhere only at 8.80: Νέστωρ οἶος ἔμιμνε, in the narration of Nestor’s narrow escape from the Trojan onslaught. Yet in that passage, the words “he alone was remaining” denote a narrative “kernel” for the entire ensuing fifty-line episode. [56] The motif of “one man remaining” can be found at other points in the Iliad as well, in other contexts: Nestor is the last of Neleus’ twelve sons surviving to fight the men of Elis (11.693, οἶος λιπόμην) as Hektor is the last of Priam’s sons by Hecuba (24.499, οἶος ἔην). The motif has further affinities with the narrative of Odysseus’s lone survival, which, in turn, is a universal folklore theme. [57]
In sum, Achilles’ reference to a single combat with Hektor represents the embedding, in kernel form, of a narrative theme that we can recognize from its various repeated uses. What does this tell us about Achilles’ speech? It illustrates one important constraint on Achilles’ “expansion” style: Homer does not make Achilles to speak, as Nestor, ornamenting every possible statement. The narrative possibilities for a recounting of his single fight with Hektor remain unexploited, while the hint of the motif, in the two words οἶον ἔμιμνε, is still a form of expanding the speech, and effectively points an audience steeped in the traditional motifs toward another vista of experience. In this way the depiction of a chief character is shaded to give us a sense of depth.
When whole-line repetition occurs only once it is best seen as a case {176|177} of intentional recall. When a line recurs more often, and shows affinities with other formulaic lines, we are obliged to examine the context of each occurrence for variations. Near the start of Achilles’ reply we get one such line (9.314) which occurs two other times in the poem (9.103, 13.735) as the introduction to speeches of mild rebuke and advice. When Nestor and Polydamas begin in this way, there is a strong implication that the listeners (Agamemnon and Hektor, respectively) have erred. Nestor discreetly places this line after his elaborate captatio benevolentiae which explains the advantage Agamemnon will gain from hearing him out (9.96-102). Polydamas, the younger man, is less discreet as he rebukes Hektor right from the start (13.726, Ἕκτορ, ἀμήχανος ἐσσι παραρρητο͂ισι πιθέσθοα) and goes on to imply that Hektor does not have his own gift of νόος. In both speeches, the αὐτὰρ ἐγὼν line leads into brief analyses of the status quo. Nestor notes that no better plan has been found (9.104, νόος—cf. 13.732) and thus by indirection refers to Agamemnon’s faux pas; Polydamas mentions the dangerous extended position of the Trojans. Both speakers sum up their advice with exhortations to take counsel: 9.112, φραζώμεσθ’—cf. 13.741 ἐπιφρασσαίμεθα βουλὴν. Compared to this norm, Achilles’ speaking strategy is deviant: instead of stating the status quo, he leaps into the future, asserting that Agamemnon will not persuade him (9.315). The triad of denials in lines 315-16 takes us into a past-continuous tense; finally, we are shifted to a general statement of the status quo by means of another triadic structure (318-20). While these lines perform the function of the corresponding statements of grievances by Nestor and Polydamas, they take the form of general statements about types—the “good man,” the “one who stays,” and so on—the referents for which remain in doubt. The more informative οὐδέ τί μοι περίκειται (321) likewise drifts into general statement (witness the repeated αἰεί of 317 and αἰέν of 322). In fact, Achilles never states a single grievance, but floods us with a multitude. The systematic reshaping of the norm extends even to the “offer of advice” feature of the rebuke speeches. Instead of a first-person plural hortatory, Achilles uses a third-person φραζέσθω (347). This deviation fits with a larger one: unlike Nestor and Polydamas, Achilles is advising Agamemnon in absentia. The distance perhaps encourages him to heap blame on his advisee: the command to “let him take counsel” forms the summit of yet another triad, the minatory imperatives (“let him take pleasure,” 337; “let him not try me,” 345). Furthermore, I have called Agamemnon the “advisee”—{177|178} but Homer at this point has placed four men in Achilles’ presence, each of whom could be the true advisor to the hero. We might say this is another deviation, from sociolinguistic patterns, in that Achilles even presumes to use the αὐτὰρ ἐγὼν line in such company.
The full analysis of such repeated formulaic lines requires that we look also at what could have been said but was not. This is clearly a vast project once we begin to study anything more than a few lines. Yet the insight gained into Homer’s construction of character is sometimes worth the effort. Our perception of a speaker’s tone depends precisely on such cues as can be created through the poet’s selection of one variant. The line we have been investigating reminds one of a different but related formulaic line, displaying the same structure: ἄλλο δέ τοι ἐρέω, σὺ δ’ ἐνὶ φρεσὶ βάλλεο σῇσι. Although Achilles at 9.314 uses the “advising” formula, he does so with the more aggressive tone usually encountered when a speaker employs this second formulaic line. Achilles himself had used the line to threaten Agamemnon with bloodshed (1.297). With a similar hint of anger, Zeus warns Hera that she will not always get her way (4.39). The fixed value which the line seerns to have in the rhetoric of the poem appears again when Hera concedes to Zeus the right to save his own son, Sarpedon (16.443): [58]
“Do so—but all we other gods do not approve.
Yet I shall tell you something else….”
In Book 4, Hera takes advantage of the permission she is granted; Zeus, in Book 16, is forced by the threat of public blame to let his son die. The discourse strategy of relenting with a warning, which accompanies this formulaic line, occurs also in the death speech of Patroklos. He yields to fate and the god (16.849), then foretells the death of his own killer at the hands of Achilles (851-54). It is noteworthy that the death speech of Lykaon breaks this pattern: although he relents, conceding that a daimon has driven him to face Achilles (21.92-93), instead of a threat, all Lykaon can muster is an entreaty: “Do not kill me, I am not full brother of Hektor, who slew your companion” (21.95-96). Achilles, enraged rather than placated by this, is the one to utter the conventional threat; yet this too is slightly different, being a prediction of his own death, though in words that {178|179} echo Patroklos’ threat to Hektor (compare 16.852-53 with 21. no). [59] Two passages seem to contain the ἄλλο—βάλλεο formula without overt threats on the part of the speaker. Diomedes introduces instructions on horse stealing in this way (5.259), and Achilles’ instructions for Phoinix’s behavior (9.611) are similar. In both passages, the formulaic line is preceded by mild rebukes and by assertions that the speaker will not be persuaded to the course of action his interlocutor recommended. The speakers’ strategy—resist and instruct—is the reverse of the relent-and-warn tactic. Looking back to Achilles’ speech at 9.307-15 with these two patterns in mind, we can see that Homer has blended the tones and sequences of thought connected with each, in order to build up the introduction to Achilles’ finest moment. As in the first pattern, threats follow the speaker’s assertion about his own speech: Achilles uses the pattern in his threat to leave (356). Unlike the first pattern of formulaic use, Achilles’ strategy does not include an initial yielding to his audience. In its place, Homer has put the initial rebuke, which we found in the second pattern: Achilles tells the embassy not to attempt persuasion. (Compare 9.315, οὔτ’ ἐμέ . . . πεισέμεν οἴω, with 5-252: οὐδέ σε πεισέμεν οἴω.) In short, from the analysis of such repeated whole lines we can learn much about the Homeric technique of creating new expressions out of traditional usages through recombination. We can also now see the stylistic mechanism which accounts for Adam Parry’s intuition about the tone of Achilles’ speech: it sounds “passionate” and “confused” because it is the result of the fusion of less complex speech patterns, each implying a separate and different dominant—and passionate— tone.
The Reshaping of Tradition
One might think that the easiest way to detect innovations in Achilles’ speech would be to examine hapax legomena, forms or lexical items that appear unique when set against the background of Homeric Greek or Greek in general. Yet this is a more difficult operation than it first appears, and, in the end, is not of much value if we want to find characterization by style in Achilles’ language. First, there has to be a distinction drawn between lexical items that occur nowhere {179|180} else, and which therefore might be thought characteristic in a person’s speech, and uniquely occurring forms of otherwise commonly occurring items. The former are actually useful in characterizing Achillean style only if they occur in greater number in his speeches than in those of other figures in the poem. As it turns out, a count of the hapax legomena in the speeches of various characters reveals that uniquely occurring lexical items and forms are evenly distributed among the major characters in the poem. I conclude that they are not a characteristic of Achilles’ “language. “ [60] In analyzing hapax legomena, another boundary should be kept in mind—that only Achilles would have occasion to mention certain objects and places, given the plot of the poem as we have it. Once again, single unique words are valueless as a stylistic criterion.
As for the latter type, unique forms of common words, these are more valuable because they might show us stress points in the poet’s composition process, places where he needed to resort to analogizing or to the use of rarer forms in order to say something new. For example, polizemenai (9.337) occurs nowhere else in Homer, but the verb in the infinitive polizemen can be found frequently, and in the same slot as in this line. Moreover, the infinitive ending –emenai is also frequent at the point in the line where we find it in 337 (cf. line 3 56); it just happens that the ending is never attached to the stem poliz-elsewhere. Thus, the epic has produced the unique polizemenai by combining two familiar positional options. From considering this sort of explicable form, we are led to a more interesting question: why did a poet have to manufacture an expression at this point? The answer lies in the same line, 337, with the phrase ti de dei. This is a rare expression in several ways, first because questions beginning ti de are almost never voiced, and second, because it contains a word not elsewhere attested in Homer, in any form, dei. This sort of unique form is harder to explain, but we might guess that dei is used because the common khrē was ruled out either for metrical, euphonic, or discourse considerations (e.g. to avoid repeating the word with which Achilles began, in yet another emphatic position). And the conditioning behind such a choice is further related to the structure of thought and verse which the poet has devised as he constructed the rhetoric of Achilles up to this point in the speech. The postponement of the question “Why should the Argives fight the Trojans?” is the {180|181} necessary consequence of two rhetorical choices on Homer/Achilles’ part, namely the enjambement of terpesthō and Argeious. Both choices serve to produce emphasis in the total performance of the speaker. These choices, in turn, are outgrowths of the more inclusive strategy of having Achilles even ask a series of questions; and, more important, of making him ask precisely such questions, which juxtapose old ideas in new ways, as we can see from the unique collocation here of the dative Trōessi and the verb polizō in the sense “fight with” (rather than “for”) the Trojans. That is to say, the disruptions in traditional forms and placements in this line relate directly to the poet’s conception of Achilles in terms of his motivations (Achilles as one who rejects the very basis for the war against Troy) and also in terms of his character as an orator (one who constructs elaborate sets of questions). We could perform similar explications for each of the striking “deviations” represented by hapax legomena. A number of such cases will be examined in the course of my analysis of Achilles’ formulaic art. But it should be kept in mind that the importance of the expressions that are unique in this way lies in what they can tell us about the poet’s larger rhetorical aims in constructing Achilles’ speeches; they have little or no value as independent items of “style” in language.
More useful for finding what is new about Achilles’ language is the analysis of unexpected juxtapositions. These emerge only in a plotting of syntagmatic and paradigmatic formulas such as I have offered here. Unless one has an exceptional verbal memory, only this groundwork can indicate that certain phrases, which sound conventional because they exhibit the pervasive epithet-noun structure, are in fact unattested combinations when seen against the background of the poem as a whole. They are “new” in terms of the Iliad. They demand attention for this reason alone, even if they happen to be “formulas” in other poetic compositions, extant or lost. I have found the following collocations of two elements (usually noun and adjective, sometimes object and verb), which appear to be traditional in their placement, but untraditional in their juxtaposition:
aergos anēr
aupnous nuktas
ēmata haimatoenta
oarōn spheteraōn
empeda keitai
{181|182} alokhon thumarea
phileous’ alokhous
anēr agathos kai ekhephrōn
Hellēsponton ep’ ikhthuoenta
eressemenai memaōtas
esti moi
khalkon eruthron
karos aisē
psamathos te konis te
thumalgea lōbēn
gunaika massetai (or gunaika gamessetai)
kourai aristēōn
eikuian akoitin
ptoliethra montai
psukhēs antaxion
Ilion ektēsthai
lainos oudos
xantha karēna
ameipsetai herkos odontōn
kleos aphthiton
The last item on this list brings us once more to the paradox that diction which can be described as innovative, when compared on the synchronic level with the rest of the Iliad is highly traditional when considered from a diachronic perspective: a phrase can be quite “new” and yet very old. For kleos aphthiton, “unwithering fame,” represents a combination of words which dates to the Indo-European period, as Adalbert Kuhn demonstrated in 1853, comparing the phrase with the identical Vedic śráva(s) ákṣitam. The work of Gregory Nagy has now shown that the metrical shape of the phrase, as well, presupposes a common Indo-European prototype and represents within Greek “a fragment of Indo-European versification. “ [61] Despite this heritage, the phrase has been labeled by other scholars a chance innovation, because it occurs only here in Homeric poetry and employs the adjective as a predicate, with estai. In this view, it was invented because the poet sought an alternative to the more common kleos ou pot’ oleitai to avoid repetition of the verb in that formula. [62] Yet the flexibility of the {182|183} formula, often apparent in this speech (discussed later), should make us wary of arguing that any phrase using the adjective as a predicate represents an innovation by the poet. And the single attestation of the phrase, in this case, can actually be the best proof that kleos aphthiton is not an accident of composition. Instead, the phrase is used just once, at the most important moment in the most important speech of the Iliad and I believe it is used knowingly, as an heirloom from the poet’s word-hoard. [63]
We can detect similar thematic importance in the other combinations which are foregrounded in the speech of Achilles by virtue of their unique occurrence in the Iliad. First, Achilles’ speech contains an explicitly new ethical bent: it enshrines the only attestation in the Iliad of the theme of the “good man.” Line 341 gives us an opportunity to see how such new combinations spring from older, more usual phrases. Agathos frequently appears in the slot following the hephthemimeral caesura, from its use in the formulas agathos Diomedes and boën agathos. Similarly, anēr appears frequently in its slot, yet is never qualified by an adjective defining “goodness.” In this case, the new juxtaposition of two ordinary words implies an equally unparalleled way of looking at human behavior: Achilles breaks through to the abstract language of philosophical ethics. [64]
Another fresh formula in the Iliad takes on additional meaning from the surrounding narrative. The adjective in the phrase thumalgea lōbēn (9.387) is regularly used with another noun, kholos, and in a formula that has primary reference in the poem to Achilles’ own anger. Apollo urges on the Trojans by reminding them that Achilles idly “stews his soul-paining anger” (4.513—with adjective in the {184} same slot as 9.387). The other two occurrences are placed, significantly, on either side of Achilles’ speech in Book 9. Odysseus reminds the hero of Peleus’ admonition to put aside kholon thumalgea (9.260) and Phoinix includes in his paraenetic tale the detail that Meleager, too, withdrew from battle nursing this kind of wrath (9.565)—this is equivalent to calling Achilles’ own anger “soul-paining” since Meleager in the tale is so clearly modeled on Achilles. In short, when Achilles uses the adjective thumalgea with lōbēn, “disgrace,” the poet is characterizing him as rhetorician, once more. Achilles’ displacement of the adjective is an implicit answer to his elders, which asserts that his anger is not without an equally grievous motivation in the treatment accorded him. We should note, finally, that the adjective in this combination is given prominence within the speech of Achilles by being opposed to another thum– compound, also in a unique combination: thumarea (336).
Along with these touches of innovation, it is interesting that at least six of the new juxtapositions deal with the theme of women. Perhaps this reflects a lack of traditional language regarding women in the Iliad, or in heroic poetry in general. The frequency of the new phrases, however, is significant: Achilles seems obsessed by the theme. Furthermore, we have a striking bit of characterization in that these phrases attach, to the common words for women and wives, adjectives expressing tenderness and compatability: Achilles is the only hero to call a bed-mate “soul-fitting,” a woman “proper,” or to use the verb “to treat as near and dear” (phileous‘) with the object alokhous. [65]
The last-mentioned collocation actually involves more than the juxtaposition of two previously unconnected ideas, because there is a phrase alokhous te philas which occurs often enough to be called a syntagmatic formula (4.238, 5.480, etc.). In Achilles’ speech, the traditional phrase is verbalized: whereas wives commonly are called “dear,” only Achilles reactivates the meaning inherent in the adjective, by transforming the formula into a predicate. The same technique occurs at least twice more in this speech: in line 328, for example, a predicate poleis alapax’ anthrōpōn introduces action into a noun-phrase attested elsewhere, polis meropōn anthrōpōn (20.217). The formula Helenēs posis ēukomoio occurs six times in the Iliad always as an equivalent for the name Paris; Achilles has made this static epithet {184|185} work, fusing the genitive into the syntax of his sentence in line 339, rather than leaving it as a noun modifier. A third instance is at 9.400: gerōn ektēsato Pēleus substitutes a verb in place of the usual epithet hippēlata (e.g. 11.772, 9.438, 18.331). In addition, we have seen that the ancient formula kleos aphthiton is also remade by the addition of the verb estai. [66]
Thus, the close analysis of noun-phrases automatically leads us to consider how verbs are employed in the speech of Achilles, to discover whether, in general, their use is idiosyncratic (even if their placement is not). I will sketch out two ways in which Achilles’ verbal expressions in this speech differ from other such phrases in the Iliad. First, his language differs at the level of semantics and pragmatics— that is, his verbs relate to objects and events in the world of the poem in unparalleled fashion. Moreover, through his use of verbs, Achilles can often be classed with a small group of speakers, usually gods or the poet himself, who are the only other users of certain expressions. [67] Second, Achilles’ use of verbs quite often represents a deviation from patterns visible elsewhere in the poem, either of the placement of the verb within a particular speech or of the associations that the verb has with other verbal expressions. (The latter aspect is in turn related to the different rhetorical strategies that the poet chooses to give Achilles.)
The first set of deviations begins to confront us right from the start of Achilles’ reply. After a conventional greeting (9.308), he informs Odysseus that he will refuse Agamemnon’s offer completely (apēlegeōs apoeipein). The verb has two meanings in the Iliad. When not directly connected with the story of Achilles’ anger and reconciliation, it means “report” (as in 23.361, 7.416, both line-end). The other seven occurrences cluster around Books 1, 9, and 19, where there are {185|186} two recurrent patterns: first, a speaker refers to a choice of doing or refusing to do something. Thetis says to Zeus “promise or refuse” (ē apoeip’, 1.515) and Agamemnon asks “does he want to, or has he refused” (e apeeipe, 9.675), both in speaking about Achilles. Second, Achilles is said to “refuse mēnis” (19.35, 19.75), or simply, to “refuse.” The latter three occurrences, all in Book 9 at line-end (9.431, 510, 309), are surely echoic. Briefly, then, the verb apoeipein in the Iliad appears to be particularly marked whenever Achilles is the subject or speaker.
Achilles’ statement of intent continues with a rather ambiguous adding clause: “… according to the way I think and the way it will be fulfilled” (310, ᾗ περ δὴ φρονέω τε καὶ ὡς τετελεσμένον ἔσται). Aristarchus read phroneō here, but there is a variant kraneō. [68] If this reading is correct, we can say that Achilles’ usage of the verb is unique. This can be seen, first, on a comparison with the other two occurrences of the first-person singular of this verb in the Iliad. When Menelaos says he thinks that the Achaeans and Trojans are at odds on his account (3.98-100, φρονέω δὲ διακρινθήμεναι ἤδη/ Ἀργείους καὶ Τρῶας), his use resembles Achilles’ (9.310) in several respects: there is a moment of crisis in which the agreement of a principal hero is sought; Menelaos, like Achilles, says “I think” near the start of his speech. In a more significant way, however, the verb phroneō differs from 3.98 to 9.310, as can be seen when we come to 9.608: φρονέω δὲ τετιμῆσθαι Διὸς αἴσῃ. Both of Achilles’ uses of this verb are in the same setting, and both refer solely to how he thinks about himself, whereas Menelaos uses the verb with an object (Argeious . . . ). This slender evidence for Achilles’ usage of the verb in an absolute and self-directed sense stands out more clearly against the background of the word’s deployment elsewhere in the Iliad in several standard expressions. Its focus is almost always on social relations to such an extent that one can say the verb implies “thinking in relation to someone.” Thus, Agamemnon uses this verb in a speech when reaffirming a social bond with Odysseus after their mock neikos (4.361, τὰ γὰρ φρονέεις ἅ τ’ ἐγώ περ), and Zeus wishes that Hera could “think like ine” (15.50, ἶσον ἐμὸι φρονέουσα). Such wedding of minds can occur between gods or between humans, but crossovers in the Iliad are not possible. Thus, Apollo warns Diomedes: “Do not {186|187} wish to think equal thoughts” (5.440-41, μηδέ θεοῖσιν / ἶσ’ ἔθελε φρονέειν). Instead of thinking like the gods, the hero must depend on divine “thought toward” himself, as Diomedes prays at the start of his aristeia (5.116):
εἴ ποτέ μοι καὶ πατρὶ φίλα φρονέουσα παρέστης
δηίῳ ἐν πολέμῳ, νῦν αὖτ’ ἐμὲ φῖλαι, Ἀθήνη
“If ever you stood by my father in dread war,
thinking kindly, now in turn be dear to me, Athena.”
The poet describes Cheiron’s relations with Asclepius in this way (4.219, πατρὶ φίλα φρονέων), just as Iris speaks of her friendly intentions toward Priam (agatha phroneousa, 24.172.) [69] The good intentions of speakers who address the assembly is often described with a compound form of the verb (e.g. 1.73, 1.253, 2.78, 2.283).
The emphasis on the social context of thought appears frequently, as well, in an expression used to explain the motivations of various characters: ta plus phron-. The phrase can be used by either narrator or speaker in the poem. With the anticipatory pronoun ta, Homer describes Ares rousing the menοs of Menelaos, “thinking this, that he be subdued at the hands of Aeneas” (5.564), and Odysseus planning how to steal the horses of Rhesus most efficiently (10.491-92). With the same neuter pronoun used anaphorically, he explains, for example, Zeus’ intent to let Hektor fire the ships (15.603). Speakers in the poem also use the phrase for both prospective and retrospective comments: Athena agrees in this way with Apollo (7.34, τὰ γὰρ φρονέουσα καὶ αὐτὴ / ἦλθον) and Hektor tells the allies why he called them to Troy (17.225, τὰ φρονέων δώροισι κατατρύχω καὶ ἐδωδῇ / λαοὺς).
The two instances of the prospective pattern both concern Achilles. {187|188} In 23.544-46, just as Achilles is about to award the prize for chariot racing to Diomedes, Antilokhos speaks out:
μέλλεις γὰρ ἀφαιρήσεσθαι ἄεθλον
τὰ φρονέων ὅτι οἱ βλάβεν ἅρματα καὶ ταχέ’ ἵππω
αὐτός τ’ εσθλὸς ἐών
“For you are about to take away the prize,
thinking this—that his chariot and swift
horses were harmed, though he himself is excellent.”
Only here does a speaker use the phrase to refer to what is in another’s mind, as if Antilokhos has shifted into the role of narrator, explaining to the audience what Achilles is thinking. [70] Compare with this the similar formulas in which Homer foretells the dashed hopes of Agamemnon, at the beginning of the poem, and of Achilles near the end: (2.36, Dream leaves Agamemnon) τὰ φρονέοντ’ ἀνὰ θυμὸν, ἅ ῥ’ οὐ τελέεσθαι ἔμελλον and (18.4, Antilokhos cornes to Achilles to report Patroklos’ death and finds him) τὰ φρονέοντ’ ἀνὰ θυμὸν ἅ δὴ τετελεσμένα ἦεν. Agamemnon’s ignorance and Achilles’ premonitions frame the narrative but also provide vivid capsule characterizations, at two points of particular emotional intensity in the poem. The disjunction between thought (phroneont‘) and outcome (teleesthai, tetelesmena) produces pathetic irony when the concepts are juxtaposed in a single line: by contrast, Antilokhos’ reading of Achilles’ thought process (23.544—54) is dramatized with comic irony, as a piece of negotiation in which the outcome is not what has been thought. Antilokhos plays the role of an angry young man disappointed by the division of spoils—that is, the role that Achilles has just abandoned at this point in the poem. Antilokhos swears, “I will indeed be angered if you complete this utterance” (note telessēis at 23.543). He goes on to reject the alleged basis of Achilles’ decision, and calls for another gift, painting Achilles as that hero had portrayed Agamemnon: rich in possessions but unwise in their distribution (23.549-52). [71] The contrast between this scene in Book 23 and the {188|189} earlier scenes in which “thought” is opposed to outcome is all the sharper because now Achilles can be persuaded to change an outcome, himself:
” Antilokhos, if you indeed bid me to give
Eumelos something else from my own, I for my part
will bring this about.”
Thus, while thought and outcome at the narrative level are predetermined to diverge, at the level of speakers’ discourse within the poem the one can influence the other—a fine narrator’s trick for creating the illusion of fictive freedom.
The other passage in which the ta plus phroneōn phrase is used prospectively appears to have been constructed with equal attention to the characterization of Achilles. Phoinix uses the expression to explain his personal motivations for caring for the young Achilles. Thinking that a curse was preventing the gods from “bringing offspring to fulfillment” for him, Phoinix attempted to treat Achilles as his own son (9.493—95: note the disjunction between phroneōn and exeteleion, thought and outcome). The rhetorical technique here is that used by Hektor in his brief reminiscence at 17.225, cited earlier: a speaker analyzes his own state of mind at some important point in the past, in order to influence his present audience. Again, the poetic technique behind the display of this strategy aims at creating a sense of layering, by showing us characters with a permanence of memory and purpose. That the same two-word expression (ta phroneōn) is employed so consistently to introduce such rhetorical strategies gives us more reason to think each small expression in Achilles’ speech can be fruitfully compared with its congeners.
Thus far, the comparison with other occurrences of the verb phroneō has shown us that it appears in patterns that are significant for moving the narrative forward by focusing on motivations. At times {189|190} the “thought” thus described is itself ambiguous: witness the formulaic use of this verb in scenes of decision making, as when Deiphobos must decide either to join his companions or go it alone (13.458), and the poet marks the end of his hesitation with these words:
ὧδε δέ οἱ φρονέοντι δοάσσατο κέρδιον εἶναι
Thus did it seem better to him as he thought.
At other times, if the speakers are gods, the pattern built on the verb seems to function as a linguistic politeness gesture, revealing nothing of the speaker’s thought, but inviting further discourse: Hephaistos (18.426-27) and Aphrodite (14.195-96) perform this way, urging their visitors to candor by saying:
αὔδα ὅ τι φρονέεις. τελέσαι δέ με θυμὸς ἄνωγεν
εἰ δύναμαι τελέσαι γε καὶ εἰ τετελεσμένον ἐστίν
“Speak whatever you are thinking. My spirit moves me
to fulfill it, if I am able, and if it is to be fulfilled. “
As in previous patterns, the mention of “thought” prompts a comment about outcome, but unlike Achilles and Agamemnon, for whom these notions remain tragically unrelated, the gods easily complete one another’s will. Only Zeus, lured by Thetis to participate in the mortal world, cannot be divinely candid, as his spouse complains: αἰεί τοι φίλον ἐστὶν ἐμεῦ ἀπονόσφιν ἐόντα / κρυπτάδια φρονέοντα δικαζέμεν (1.541-42). His covert thought is a guarantee that the justice of Zeus is free from interference: the two notions are collocated again when Athena and Hera discuss the will of Zeus: κε͂ινος δὲ τὰ ἅ φρονέων ἐνὶ θυμῷ / Τρῶσί τε καὶ Δαναο͂ισι δικαζέτω, ὡς ἐπιεικές (8.430-31)· If we do discover the will of Zeus, it is only through the poem itself, and then, in a drawn-out fashion, for Zeus’ plan is always in process, as the proem of the Iliad makes clear, with its imperfect tense, Διός δ’ ετελείετο βουλή (1.5).
What do we learn, finally, from investigating the deployment of the fairly common verb featured in Achilles’ preliminary remarks at 9.310? I suggest that Achilles implicitly adopts the tone of Zeus himself; at least, the poet composes with the idea that the hero and the {190|191} god speak alike. For Achilles, first, uses phroneō to mean “I think,” without any hint of the meaning “I am disposed toward someone in thinking.” The depiction of Achilles even before Book 9 has prepared us for this sort of absolute isolation; I point out simply that this can be documented by formulaic analysis as well. Second, Achilles like Zeus does not reveal his thought, even when he most explicitly claims to value candor. This is an important piece of evidence for the complex characterization of Achilles as a master speaker. He appears to speak his mind, after asserting that he must deny the previous offer. But the topic he then brings up immediately is, instead, the necessity for others to speak what they think. The poet has constructed a variant of the politeness gesture used elsewhere in the type-scene of “visit”: the host invites discourse from his guest, but, in Achilles’ case, the host also maintains a tight control on the conversation, is self-assertive rather than receptive. He employs the same notions, but, instead of saying “speak what you think” and “I shall fulfill it,” Achilles reflects back on himself: “I must deny the offer … in the way that I think and the way it will be fulfilled. ” As we shall see shortly, the final phrase of 9.310 is also deviant when considered against the usual patterns and it is this that creates the arrogant tone we recognize in Achilles’ opening gambit. [72]
The principle invoked earlier—that coherent patternings of deviation from a norm make for “style”—can be successfully applied to this speech. Several other uses of first-person verbs in this speech relate Achilles to the figures of gods. Only Zeus, for example, uses the pattern of 9.397: τάων ἥν κ’ ἐθέλωμι, φίλην ποιήσομ’ ἄκοιτιν. He does so in a similar context, expressing his ability to choose whatever he wishes (in this case, whatever speech-act): ὃv δέ κ’ ἐγὼν . . . ἐθέλωμι νοῆσαι (1.549). [73] And only Zeus says “tell all as I command”: compare 2.10, πάντα μάλ’ ἀτρεκέως ἀγορευέμεν ὡς ἐπιτέλλω, with Achilles’ words (9.369): τῷ πάντ’ ἀγορευέμεν ὡς ἐπιτέλλω. In both cases, the ultimate recipient of the message is Agamemnon. [74] The god Poseidon is the only other speaker to mention {191|192} getting something by lot (15.190) as Achilles does at 9.367. Hera, finally, is the only other speaker to utter a version of 9.417. [75] She concludes her great oath, affirming that she did not induce Poseidon to aid the Achaeans, with these words: αὐτάρ τοι καὶ κείνῳ ἐγὼ παραμυθησαίμην / τῇ ἴμεν ᾗ κεν δὴ σύ, κελαινεφές, ἡγεμονεύῃς (15.45-46). Hera’s strong suggestion to Poseidon, in these words, is just like that by Achilles’ to his listeners: both advise someone to follow the will of Zeus (cf. 9.419-20).
The usual meaning or pragmatic application of verbs is changed at a number of other points in Achilles’ speech, with the result that he is characterized in certain ways, or through his word use associated with a few other important heroes. This reshaping of diction seems to occur particularly at the beginning and end of the speech; when Achilles launches into the grand rhetoric of the midsection, Homer depends more on a variation from traditional patterns to make Achilles’ speech unique.
The gnomic utterance of line 9.320: “They die alike, both the man without works and he who has done much,” is meant by Achilles to refer to his own situation: this is a continuation of the self-reflective rhetoric that we saw in the introduction to his speech. Here again, one can see that Achilles has twisted a traditional phrase so as to point it inward. For elsewhere in the poem, the full phrase καὶ δὴ κακὰ πολλὰ ἔοργε—which underlies Achilles’ words here (ὅ τε πολλὰ ἐοργώς)—always describes an enemy’s deeds: Aeneas speaks of Diomedes (5.175), Hera of Hektor (8.356), and Sarpedon of Patroklos (16.424) in this way. In the logic of battle, the man who does many things to his enemies is necessarily a boon to his friends, to be rewarded; only Achilles uses the poem’s formulas to make the argument explicit, however.
In another shift toward an interior language, only Achilles speaks of the “soul” and of his “self” in the same breath within the Iliad, both at 9.322 and 401 (ἐμὴν ψυχὴν, ἐμοὶ ψυχῆς ἀντάξιον). Once again, we can find a more formulaic precedent for conjoining the concepts, in the speechmaking of enemies: a thrice-repeated boast, uttered just before spear-casts, runs: εὖχος ἐμοὶ δοίης, ψυχὴν δ’ Ἄιδι κλυτοπώλῳ. In this line, the elliptical and ironic rhetoric of the warrior vividly contrasts gain and loss: the victim who receives the blow loses {192|193} his life, and at the same time furnishes his killer with words for the future (in which the victim ultimately gains some memorial). The man who is addressed by this line gives twice, to the living (eukhos) and the dead (Aidi). At the same time, this utterance collapses distinctions, by reducing “boast” and “soul” to counters in a game of war-exchange. The equivalence of the boast and the life which paid for it is represented iconically by the careful balance of sound and meter in the line, with eukhos initially and psukhēn—another two-syllable word with medial kh—at the other emphatic position, after the pent-hemimeral caesura. The repeated line, then, is itself a memorable piece of verbal art. [76] For a poet who had used it often, only a slight extension of the metaphor “giving life to Hades” could produce the more arresting image that Achilles uses, “gaming with my life in fighting. ” And we might detect an echo of the more traditional expression in Achilles’ reference to the gates of Hades (312).
The emotive quality of Achilles’ discourse has led others to compare him to a poet. [77] I shall explore later the way in which the hero fits the role. For now, we can note that as well as using similes more often than any other figure in the Iliad Achilles also mimics the poet’s own voice in his use of smaller phrase units and single words. [78] The expression in the second half of 9.324, for example (κακῶς δ’ ἄρα οἱ πέλει αὐτῇ) is paralleled only by phrasings of the narrator, two in particular: when Patroklos answers Achilles, to be sent on the mission that eventually ends his life, the poet comments κακοῦ δ’ ἄρα οἱ πέλεν ἀρχή (11.604). This technique of foreshadowing creates a momentary distance between audience and actors of the tragedy. In the same way, pathos results from the narrator’s interjection of a brief biography at the death of Phereklos, son of the carpenter Harmo-nides, who “had made the ships for Alexandros, the ill-beginners (arkhekakous) that were an evil for the Trojans and himself.” Achilles uses this poetic sort of expression in the same way, to increase the emotional response to his own fictional world of the simile. Within the same group of lines, another instance of Achilles’ poetic voice comes in his use of dieprēsson. The word occurs at 326 in the first-person {193|194} singular imperfect, yet in the same metrical position as the more common third-person plural imperfect, which is found in a narrative formula describing the movement of troops to battle (2.785, 3.14) and horses in a race (23.364). Through this “deviation” in referent, Achilles metaphorically depicts his lone actions as equal to those of powerful, directed entities.
Achilles’ assertion at line 9.329, introduced by phēmi, is unusual for several reasons. First, only here in the poem does this verb head an elliptical statement. Next, of the thirty-six times that a speaker utters phēmi in the poem, only six times does the statement thus introduced refer to the past: Achilles speaks this way twice (9.329 and 20.187), Priam twice (a repeated line, 24.256 = 24.494), and Nestor and Agamemnon once each (2.350, 8.238, respectively). Only Achilles and Agamemnon assert something about their own actions in the past. It is completely characteristic of them that Achilles speaks both here and at 20.187 of his personal prowess, whereas Agamemnon recalls his fulfillment of expected public social and religious functions: at 8.238, bis phēmi introduces the hypomnesis of a prayer. Occurring so closely together in the poem, in Books 8 and 9, these assertions of Agamemnon and Achilles must be intentionally contrasted, to sum up the basis of conflict between Achilles and his commander. [79]
Both Achilles and Agamemnon utter phēmi in tones that recall the normative usage of the verb within threats and boasts. Compare the tone of Sarpedon’s words to Tlepolemos (5.652—53):
σοὶ δ’ ἐγὼ ἐνθάδε φημὶ φόνον καὶ κῆρα μέλαιναν
ἐξ ἐμέθεν τεύξεσθαι
“I assert that you will have from me
murder and black death.”
Threats that begin with plēmi, as in this example, usually employ the future tense, and boasts, the present or future. [80] The unexpected {194|195} disjunctions of tense and tone, therefore, mark out Agamemnon and Achilles: only they have the authority to back up their strong assertions (labeled by phēmi) with examples of their past actions. It is not accidental that the only other speakers to make any assertions at all about the past are the elders Nestor and Priam. Achilles’ statement that he sacked eleven cities, on foot, is midway between the boast of a wairior and the exempla of the old advisors in the poem, a characteristic position for a hero who, in the retreat of Book 9, takes a contemplative stance toward his own active calling. [81]
Some other expressions show the same technique of restricting a particular mode of speaking to one or two characters other than Achilles. Only Hektor, for instance, uses ethelō as Achilles does, to speak about not wishing to do something. Hektor boasts to Ajax that his skill in war-craft prevents him from attempting an ambush (7.242-43):
ἀλλ’ οὐ γάρ σ’ ἐθέλω βαλέειν τοιοῦτον ἐόντα
λάθρῃ ὀπιπεύσας, ἀλλ’ ἀμφαδόν, αἴ κε τύχωμι
“I do not wish to strike you as you are
peeking secretly, but outright, if I might succeed.”
Achilles, too, says ouk ethelō (9.356) because he believes only in honorable fighting (unlike the brand Agamemnon prefers). Of the other eight occurrences of ethelō, four times it refers to a wish to restore something (1.116, Khryseis; 7.364, Menelaos’ goods; 9.120 = 19.138, recompense to Achilles) and three times to a desire to put things aright (8.40 = 22.184, Zeus wishes to be “mild,” ēpios, to Athena; 19.187, Agamemnon says he will swear he has not slept with Briseis, and will thus be mild—cf. 19.178, hilaos—to Achilles, as Odysseus suggested). The conciliatory use of the verb extends even to Zeus’ desire to sack cities (4.40-42) since the wish he utters remains only an abstract possibility and is voiced only for the purpose {195|196} of reaching agreement with Hera: she may harry Troy now, says Zeus; his turn will come.
Achilles, thus, shares with Hektor a pattern of speaking about his wishes. With Agamemnon (and no one else) he shares the distinction of saying “I honor” within the Iliad. At 4.257, Agamemnon claims to honor Idomeneus above all the Achaeans—the motivation for his feeling is not given, only the spheres to which it applies (war and wine in particular). It is Agamemnon, again, who provides the model for Achilles’ words at 9.378, τίω δέ μιν ἐν καρὸς αἴσῃ (line-end), when he promises in his offer of marriage to honor Achilles as a son (9.142): τείσω δέ μιν ἶσον Ὀρέστῃ (line-end). Achilles’ reformulation of Agamemnon’s words is an expressive expansion of the type we shall investigate shortly. Achilles combines the patterns, whereas other mentions of timē refer either to a generous granting of respect (4.257, 5.325-26, 6.173, 16.146, 17.576), in which case the degree of respect is explicit (e.g. malista, peri pasēs, prophroneōs), or to a withholding of respect (9.238, 13.461), in which case the fact is simply noted. Achilles denies that he feels respect toward Agamemnon, but rather than simply stating the fact, he specifies, hyperbolically, the extent of his disregard: “I honor him not a whit. “
Even the handful of examples reviewed so far has sufficed to show, first, that the poet reshapes traditional phrases to give them individualized reference, thus creating the illusion of an interiorized, Achillean language; and second, that we—the uninitiated audience—must search out such reshaping at the level of formulaic usage patterns. In other words, to read Achilles’ speech properly, we are obliged to reread every scene in the Iliad in which any phrase of that speech appears. Just as the analysis of noun formulas led us to investigate verbal expressions more closely, so this step takes us into a study of larger units of discourse. Now that we have seen Achilles’ deviation from the normative usage of certain individual verb expressions, I shall present the most important instances of his variation from these larger patterns. Whereas the study of the former showed that Achilles’ “voice” is unique, resembling as it does that of the gods and the poet, the latter set of variations will demonstrate that Achilles’ overall performance in Book 9 depends on a quite rare strategy of verbal ornamentation, which I call the “expansion aesthetic. ” Achilles “speaks” like the poet, it turns out, because, in Achilles’ words, epic poetry reveals its own precise mode of composition.
{196|197} Achilles as Formulaic Artist
Beginnings and endings carry a significance far disproportionate to that of the midportion of any temporal artistic composition. In every archaic Greek poetic genre, the start of the work is marked, either by conventional topics, grammatical devices (e.g. the vocatives at the beginning of epics), or syntactical devices (the convention of a relative clause of description, which occurs in hymns, epinikia, and epic). In the early chapters of this book I demonstrated the usefulness of considering every speech within the poem a composition in its own right, a poem within epic, subject to conventions of discourse. If this strategy is taken to its logical end, we should concentrate in particular on the beginning and end of speeches such as Achilles’ in Book 9.
With this in mind, we can turn to the other uses of khrē men, the words with which Achilles launches into his discourse (9.309). We should notice first that the Greek word is a strong expression of necessity. It demands the attention of the listener. This in itself characterizes Achilles’ explicit statement of his rhetorical strategy in the speech, to speak out plainly. [82] Furthermore, it sets off Achilles as a speaker distinct from the narrator, who never uses the word khrē in the Iliad. Achilles is concerned with the exigencies of life in a way that the omniscient Homer need not be. But besides being distinguished from the poet by his use of the word, Achilles alone among other speakers in the poem uses the combination of words, khrē men, and places the word khrē at the beginning of his discourse. Both are important facts. The particle men is key to establishing the tone of Achilles’ statement at 9.309. A rough paraphrase—”yes it is necessary, but”—shows that the combination transmits two messages at once. Something is crucial; at the same time, the speaker has other concerns: the anxious, serious tone thus created affects all the following words of the speaker. It is significant that this combination of words occurs only one other time in the Iliad, at an important juncture in the narrative. Just as Achilles meditates slaying Agamemnon, Athena intervenes: instead of stopping him, she asks, twice, for him to be persuaded (1.207, 214). [83] Her small speech moves from the {197|198} tentative αϊ κε πίθηαι to the more urgent σὺ δ’ ἴσχεο, πείθεο δ’ ἡμῖν. Achilles makes the choice of words instead of deeds; he will go on to defeat Agamemnon, symbolically, by being the best performer on the verbal level. In his reply to the goddess at 1.216-18, he affirms the theological correctness of his choice with a gnomic statement: the one who “is persuaded/obeys” the gods gets favor in turn:
χρὴ μὲν σφωιτερόν γε, θεά, ἔπος εἰρύσσασθαι
καὶ μάλα περ θυμῷ κεχολωμένον. ὧς γὰρ ἄμεινον·
ὅς κε θεοῖς ἐπιπείθηται, μάλα τ’ ἔκλυον αὐτοῦ
“It is necessary, goddess, to keep your word,
even for one angered in spirit, since it is better—
the gods pay heed to the one who obeys them.”
The particle men is never complemented explicitly by the expected de but the string kai mala per in line 217 might substitute for this. In effect, Achilles says, “I will obey (on the one hand) but I am still angry (on the other).” His immediate cessation of speech puts these words into action.
In Book 9, the topic of persuasion, of which Achilles has already shown himself to be aware, becomes the leading theme of his reply to the embassy. When, at 9.309, he begins with khrē men, just as he did at 1,216 (and as no one else does), we should notice that the two speeches form a diptych: in Book 1, Achilles grants divine speech a privileged position; in Book 9, he draws attention to his own speech, which is, paradoxically, a denial of the possibility of persuasion. As in Book 1, the context is explicitly a battle of rhetoric, Achilles’ word against that of Agamemnon. Once we have realized that khrē men is a rare turn of phrase, characteristically Achillean, we are led to examine how other speakers make mention of necessity with this word, in order to place Achilles’ use in context. It occurs twenty-six times elsewhere in the poem: half of these occurrences can be called formulaic in the strict sense of having repeated diction. I class them as follows:
- The line-end expression oude ti se khrē (7.109, 9.496, 9.613, 10.479, 16.721, 19.420, 20.133, 23.478)
- The line-end expression nun se mala khrē (13.463, 16.492, 22.268)
- Line-initial tō se khrē (7.331, 9.100)
{198|199} Achilles uses formula 1 when telling Phoinix not to curry favor with Agamemnon (9.613-14):
οὐδέ τί σε χρὴ
τὸν φιλέειν ἵνα μή μοι ἀπέχθηαι φιλέοντι.
The expression is in responsion with Phoinix’s own use of the phrase in the speech preceding:
ἀλλ’ Ἀχιλεῦ δάμασον θυμὸν μέγαν· οὐδέ τί σε χρὴ
νηλεὲς ἦτορ ἔχειν.
(9.496-97)
In a similar way he reproaches his horse Xanthos for making predictions of doom (19.420). With a slightly different formula (formula 2), which echoes Sarpedon’s instructions to Glaukos (16.492-93), he threatens Hektor before killing him (22.268-69):
νῦν σε μάλα χρὴ
αἰχμητήν τ’ ἔμεναι καὶ θαρσαλέον πολεμιστήν.
These passages show Achilles using the common formulas normally, but his speech also contains the only instance of variation for formula i, at 19.67-68:
νῦν δ’ ἤτοι μὲν ἐγὼ παύω χόλον, οὐδέ τί με χρὴ
ἀσκελέως αἰεὶ μενεαινέμεν.
This shift disturbed ancient critics: Apollonius Sophistes in his lexicon wanted to bring the passage into line by reading se instead of me. [84] Yet the line is surely referring to Achilles’ wish to give up his anger, and the variation is doubly significant because it occurs in an important speech which completes the refusal announced in line 9.307. Achilles thus differs from other users of this phrase in that he makes it refer to himself, whereas other contexts show it referring to the necessity for others to do something. Furthermore, in the other instances, the genre of discourse in which the phrase appears is either {199|200} battle advice or metalinguistic comment on another’s talk. For the first category, we can look to the speech of Apollo (16.721-25), which begins: Ἕκτορ, τίπτε μάχης ἀποπαύεαι; οὐδέ τί σε χρὴ. Similarly, Phoinix uses the phrase (9.496) in a speech designed to instruct Achilles on the need for fighting. [85] Transformations of this paraenetic function of the phrase occur at 10.479 (Odysseus urges Diomedes to act) and 7.109 (Agamemnon urges Menelaos not to fight). The same role is played by the phrase nun se mala khrē, mentioned earlier. And nine of the thirteen occurrences of khrē outside formulaic lines also appear in the context of battle advising. [86]
The second category of discourse which makes use of the notion of necessity in khrē is “talk about talk,” what linguists call a metacom-municative act. [87] We have seen this type of metacomment in the formula poion se epos phugeti herkos odontōn (4.350), which Odysseus uses in replying to Agamemnon’s testing insults in the epipōlēsis. [88] The expressions using khrē also occur in the speeches exchanged in battle, and thus the metalinguistic use overlaps the category of paraenetic expressions to some extent. Unlike the poion epos line, which usually introduces a speech of self-defense, the khrē expressions direct attention to the need for silence. Achilles speaking to Agamemnon uses the same tone as Patroklos had taken with Meriones: compare 19.149-50, οὐ γὰρ χρὴ κλοτοπεύειν ἐνθάδ’ ἐόντας, with 16.631: τῶ οὔ τι χρὴ μῦθον ὀφέλλειν, ἀλλὰ μάχεσθαι. In these lines, the necessity for stopping talk is equated with the need to start fighting. Similarly, Idomeneus reassures Meriones that he need not speak of his own battle prowess, since Idomeneus already knows how good a warrior he is (13.275) and Poseidon warns Hera not to make things difficult, after she has made a strife-rousing speech aimed at helping Achilles against Aeneas (20.133): Ἥρη, μὴ χαλέποανε παρὲκ νόον. {200|201} οὐδέ τί σε χρή. And the lesser Ajax quiets Idomeneus by talking about the other hero’s speaking ability (23.478-79):
ἀλλ’ αἰεὶ μύθοις λαβρεύεαι. οὐδέ τί σε χρή
λαβραγόρην ἔμεναι· πάρα γὰρ καὶ ἀμείνονες ἄλλοι.
Two final examples in this genre of discourse mention the need for speech: significantly, both occur in Book 9. At the end of the episode, Ajax admits to Odysseus, within the hearing of Achilles, that the muthos of the withdrawn hero must be reported (9.627). At the beginning of the book, Nestor speaks as advisor to Agamemnon. After a rounded introduction, the structure of which recalls a hymn to a god (cf. 9.97), Nestor brings up the topic of rhetoric: Agamemnon has the power to speak, to listen, and (like Zeus) to validate another’s authority to offer advice. [89] Nestor implies that Agamemnon is thus empowered to approve his own attempt to persuade Achilles. His speech is an affirmation of speech, a call for gifts and soothing words. In the foreground of his remarks, as in those of Achilles at 9.307-17, is the power of Peithō, that effective speech which makes Iliadic society cohere.
Keeping in mind the conventional deployment of khrē phrases elsewhere in the poem, we can return with fresh insight to the first words of Achilles. It now appears that his words are distinctive not only in the positioning of the “need” phrase, at the head of his statement, and in the joining of khrē with men, but, more important, in relation to larger patterning. One usually says khrē to speak about fighting, or about speaking: Achilles explicitly denies both topics in his opening statement. He will not fight, nor do his interlocutors need to speak to him about the decision. Achilles’ denial of the effectiveness of speech goes beyond the usual expression, which heroes use to silence one another for the moment. It becomes an abstract principle. If Achilles resembles any other hero in this use of an idiosyncratic pattern, it is Nestor; there is a strong hint that the young warrior deserves to be ranked with the oldest as both speaker and instructor of the niceties of speech. [90]
{201|202} The opening section of Achilles’ speech offers another example of variation from the expected pattern of phrase use in context, again one that is related to the idea of rhetoric. The phrase oude me peiseis occurs six other times in the poem, always in a reply, and, with one exception, always within the following pattern:
- A person is told not to do something
- The phrase “you will not persuade me”
- The reason for saying the phrase is given
- The intended course of action on the part of the speaker is mentioned
At 1.131-39, for example, Agamemnon tells Achilles not to deceive (1); says the phrase (2); asks whether he is to be without reward (3); and states his intention of taking a new geras (4). The same sequence occurs at 6.360-64 and 11.648-54 (two refusals of food and drink, spoken by Hektor and Patroklos, respectively), at 24.218-27 (Priam refuses to be persuaded to stay at home), and at 24.433-39 (the young man/Hermes will not be persuaded to accept gifts for guiding Priam). The only speaker in the Iliad who uses a different pattern is Achilles. At 18.126, he concludes his speech to Thetis with the phrase “nor will you persuade me.” In this speech, element 1 does occur (“do not restrain me”—124) but Achilles’ statement of intention (to reen-ter battle) and explanation (because he must die) occur in a reverse of the normal pattern, preceding elements 1 and 2 (lines 114-18). Furthermore, it is this “deviant” pattern that Achilles uses at 9.345. Elements 1 and 2 are in place, but now the “reasoning” element—why Odysseus will fail to persuade him—occupies all of the first part of Achilles’ speech (315-45); it is linked to Achilles’ position (315) that he will not be persuaded by Agamemnon. (Indeed, the change in the person of the verb to make a less common formula with the third-person singular shows that Achilles is really refusing Agamemnon, not Odysseus.) Moreover, element 4, the statement of intent, is delayed until 356 (cf. 24.223, where the statement is heralded by nun also). Meanwhile, Achilles inserts instructions to Odysseus about what action the latter should take. In sum, the usual pattern is expanded so as to be almost unrecognizable.
At times, Achilles uses phrases that are common in a narrative pattern, but rare in speeches. The word estin, “there exists,” used at the beginning of a line and sentence, most often describes the location {202|203} of topographic features: cities (6.152, 11.711), a hill (2.811), a river (11.722), a cave (13.32). Only Aeneas and Antilokhos use the word, in its existential sense, with a dative to indicate possession, and do so in expressions similar to Achilles’ at 9.364. We have seen that Antilokhos’ statement is addressed to Achilles, in such a way that we can call it an intentional recall of this scene. Aeneas speaks also of multitude, of reproaches rather than wealth, in another speech addressed to Achilles. In the entire Iliad only Achilles says “I have” in this particularly marked way, with the full force of the verb “to be.” [91]
The mention of great wealth at home (364-65), and the specific mention of gold and bronze may strike us as unexceptional. But this is because the collocation of these objects is common within the poem, according to the conventions of another genre of performance: battlefield supplications. When the sons of Antimakhos plead before Agamemnon for mercy (11.130—35), their promise of ransom begins with these:
“Many possessions lie in the house of Antimakhos,
bronze and gold and much-enduring iron.”
Adrestos, again appealing to Agamemnon, had cited the identical objects (6.47-48). Dolon, in his plea, omits mention of keimēlia but refers to the precious metals with which his father will buy his freedom (10.378—79). And Hektor leaves “iron” out of the conventional list, but adds a new feature: both his father and his mother will give gifts over and above the metals, should Achilles return his corpse to Troy (22.340-41).
The reference to a rich father, a store of goods at home, gold, bronze, and iron—all these are features of Achilles’ words at 9.364-400. But there is a change in the context with this pattern. Instead of begging for his life, Achilles is rejecting the supplication being made by the Achaeans. We have already seen that a portion of this section (line 366) seems to have been imported from the discourse associated with raiding. Now we can add that a second genre of discourse has been called on to fill out Achilles’ words. The hero uses the routine (no doubt one actually known to archaic warriors) for a new purpose, just as he had reshaped a raiding boast into the accusation of having {203|204} been raided. Of course, one could argue that Achilles is indeed begging for life since by going home he avoids death at Troy. Yet the effect of his speaking these words is different from that evoked when an Adrestus or Dolon uses the convention. Achilles can have it both ways: isolated from the fray, he yet speaks as though in the thick of battle: his private life has taken on the tone of the public conflict; the language of the field becomes his in the tent. These dualities become overt as well in the other innovation Achilles makes at this point. Not only is he using the public speech-genre for a private audience, but he also adds a reciprocal movement to the usual statement, in saying that he will take much from Troy, to add to all the keimēlia which await him in Phthia.
The simile in which Achilles compares himself to a bird (9.323) has always been singled out as a unique feature of Achilles’ “language,” an indication that he is more like a poet than any other character in the Iliad. [92] I believe the simile does characterize Achilles as a skilled verbal performer, but I wish to point out that the image itself is another example of a common motif, found in patterns elsewhere in the poem, which has been changed in the words of Achilles. The other bird similes, uttered by the poet, compare flocks of birds with troops on the move (2.459-65, 3.2, 15.690) or equate horses with swift birds (2.764). There is, however, a precedent, in discourse associated with augury, for the mention of birds by individual characters within the poem, and this more closely resembles Achilles’ creation of the bird simile. In Odysseus’ speech encouraging the Achaeans (2.308-29), the sēma made by Zeus during the layover at Aulis is described vividly. At a plane tree near the altar, a blood-red snake devoured eight sparrows while their mother, soon to be eaten herself, hovered nearby. Odysseus calls the nestlings of this bird phila tekna—an example of the pathetic fallacy consistently applied by Homeric augurs—and says the mother was grieved. [93] Kalkhas, meanwhile, had interpreted the sign on purely numerical grounds, predicting that nine years (nine birds) would elapse before Troy fell. The Trojan, Polydamas, combines sentiment with advice, as well, in his allegorical reading of another snake-and-bird sign from Zeus (12.209). In full view of the Trojans, an eagle, bitten by the snake he carries, drops it and “did not {205} succeed in carrying it to his children, to give to them” (12.222). He predicts on this basis that the Trojans will not return from a sortie against the ships. Polydamas had prefaced his prediction with the words νῦν αὖτ’ ἐξερέω ὥς μοι δοκεῖ εἶναι ἄριστα (12.215). Notice that Achilles, before comparing hinself to a bird and drawing conclusions from the equation, uses the similar phrase (314): αὐτὰρ ἐγὼν ἐρέω ὥς μοι δοκεῖ εἶναι ἄριστα. Like Odysseus, Achilles fills out his description, of the parent bird who fares ill while nourishing the young, in the pathetic manner. Like the augurs, Kalkhas and Polydamas, he draws a moral from the depiction of the bird. His individual shaping of this piece of description from the discourse of augury consists in pointing the moral toward himself, and toward the past. “So, too, I slept sleepless nights, ” says Achilles: the fall of Troy, and the outcome of battle are not his concern, as they are in the other, similar bird auguries.
Finally, the investigation of related patterns of diction can help us to answer Adam Parry’s dilemma. He adduced lines 337—41 as uniquely Achillean in content. Reeve and Glaus have replied that neither the mode of asking such rhetorical questions, nor the urge to question the reason for heroism is unique in the Iliad. In a well-known passage, Sarpedon poses a similar question to Glaukos (12.310—14). Yet there is an innovation in Achilles’ words, and it is worth noting. Only Achilles asks essentially the same question three times in succession (9.337-39). Achilles can be seen as uttering a conventional rhetorical strategy, in order to foreground his own answer to the question “Why fight?” What is not conventional, however, is the expanded form he gives to the strategy. Nor is this a minor factor. As I shall show now, it is this one innovation—an ability to ornament by expansion—that characterizes Achilles’ speech in contrast to the rhetoric of all other Iliadic heroes. It is this which explains the dozens of fine deviations from the norm that we have examined. Ultimately, it is this expansion aesthetic that shows us the motivations for Homer’s own poetry.
Footnotes
[ back ] 1. Homeric epic associates two roles not commonly conjoined: see Bauman 1978:29 on roles of performers in other societies. For a contrasting approach to mine, see Barck 1976, an examination of pervasive opposition between “word” and “deed” in Homer.
[ back ] 2. For the history of the traditional formulation “style is the man,” from Plato on, see Müller 1981, esp: 9-21.
[ back ] 3. The most detailed treatment of this theme in the poem is Thornton (1984) 113—42.
[ back ] 4. For the Platonic discussion, in which speech style equals ethical stance, see up. Mi. 3646, 3650, 3yoa. Perhaps as early as the fourth century, Odysseus, Nestor, and Menelaos were taken as models of the three rhetorical styles: see Russell 1981:137-38. Achilles, in this tradition, is not a model to be imitated. The issue of individual speaking styles in Homer is tied to the larger ancient debate over whether the art of rhetoric existed in Homeric times, on which see Kennedy 1957:23-35 and 1963:35—39. Karp 1977 argues for the existence of a kind of rhetorical art in Homer but unduly enlarges the term to include any persuasive use of language: see his comments on Achilles’ speech, 256-57. Gladstone 1874 had well compared Homeric rhetoric with the less formal art of parliamentary debate: see also Myres 1958:94-122 on this viewpoint. Instead of discussing rhetoric as a science in early Greek poetry, it is better to speak of it as one feature of universal verbal art, poetry and prose. It is this “proto-rhetoric” that needs investigation, as Horner 1983n29.
[ back ] 5. On the tradition underlying scholiastic comments about rhetoric, see Schmidt 1976:43-45. Kazhdan 1984:183-94 “as a good study of Eustathius’ rhetorical teaching; Lindberg 1977 provides the context for Eustathius’ remarks on rhetoric in the commentary. See also Kennedy 1983:316. Roemer 1914:5 points out that Eustathius held all of Book 9 to be a rhetorical contest: see Bust. 751.1 (ad 9.309) and 751.33 (ad 312) for his remarks, and also the scholia (bT) to lines 307 and 309 for similar observations.
[ back ] 6. A. Parry 1956:5-6. Similar remarks can be found in his 1957 Harvard dissertation on Thucydides, which begins with an examination of the logos/ergon distinction in early Greek. For discussion of the way in which this distinction clouded Parry’s thinking concerning formulaic language, see Claus 1975:14.
[ back ] 7. A. Parry 1956:1.
[ back ] 8. Parry 1956:3.
[ back ] 9. Parry1956: 6.
[ back ] 10. Ibid. 5-6.
[ back ] 11. See Foley (1981) 22-26 for a bibliography of Lord’s work, and Foley (1985) for a listing of some 1,000 articles inspired by Parry-Lord theory.
[ back ] 12. For a summary, see Lloyd-Jones (1981); Myres (1958); and Davison (1962).
[ back ] 13. See Krk (1962) 87-88; Vesterholt (1973); Lord (1975) 12-13.
[ back ] 14. The terms are those of Rosenmeyer (1965). Holoka (1973) traces the debates over the Parry-Lord theory in the 1950s and 1960s, a contentious series of misunderstandings and inaccuracies. For a representative selection of the work from this period, see Latacz (1979) 297-571.
[ back ] 15. See the debate with Adam Parry and Anne Amory Parry in Lord 1968 and the response of A. A. Parry 1971. Lord 1960:25-27, 68-98 discusses and illustrates the art of expansion. Milman Parry had called attention to the creativity of certain singers who knew the tradition thoroughly enough to improve it: see M. Parry 1971:335, 406-7 (hereafter abbreviated MHV).
[ back ] 16. This work took several tacks: some, like Whallon 1969 and Anne Parry 1973 found “meaning” in the so-called ornamental epithets: others, like Edwards 1980, Fenik 1968 and 1978 especially and Beye 1964 concentrated on variatio technique within type-scenes; Austin (1975) and Vivante 1970 and 1982 showed that the poet’s decision to use a formula, when the noun itself suffices, is meaningful; Nagy 1974 and 1979 starts from Parry’s insight that all Greek epic is traditional in his own explorations of the interplay among meter, diction, and theme. Hainsworth 1970:37-38 has good examples of Homeric innovation. For further developments, see Foley 1985:35—41. Recent studies in other poetic traditions show how artistic talent is revealed through the individual performer’s exploitation of traditional material: see Vesterholt 1973:75-85 and Beaton 1980:18. On the richness of Homeric style as resulting from such free variation, see Bowra 1962:32-34, Peradotto 1979:5, and Russo 1968:294. Again, Parry was not unaware of the possibilities inherent in the recombination of fixed formulas: see MHV 220, 270, 307.
[ back ] 17. Austin 1975:79-80
[ back ] 18. MHV 275-79. On the later transformations of this notion, see Hainsworth 1964:155.
[ back ] 19. The notion of “economy, ” which Milman Parry derived from earlier work on Homeric Kunstsprache, proved to be the strongest point in the demonstration of the traditional nature of the noun-epithet system. See MHV 6-16. As Adam Parry points out, the principle is open to criticism if extended to all Homeric repetitions: MHV xxxi-ii. See further Russo 1971:32-33.
[ back ] 20. Reeve 1973:194. Kirk 1976:74, while expressing reservations about A. Parry’s exaggeration of the rigidity of the system, nevertheless appears to accept his conclusion (at 207).
[ back ] 21. Claus 1975:16-17.
[ back ] 22. Jakobson 1981 contains the classic definition of such forms, which include the personal and deictic pronouns, as well as other grammatical markers.
[ back ] 23. Nimis 1986:219.
[ back ] 24. Nimis 1986:220-21. Nimis here follows Friedrich and Redfield 1978 in equating rhetoric and language: see my later discussion. Cramer 1976:301 rightly traces the differences which A. Parry picked out in Achilles’ speeches to the unique rhetorical stance of the hero, “free-wheeling but rhetorically calculating”; he cites the changes made from one speech to another in Achilles’ mode of referring to Briseis (9.340 versus 19.59) and the sea (1.157 versus 9.360).
[ back ] 25. Hogan 1976:309.
[ back ] 26. Scully 1984:24.
[ back ] 27. Scully 1984:25. Despite the title of his article, Scully seems to recognize the distinction: he explains Achilles’ idiosyncratic use as Homer’s playing against an established “pattern of expectation” at the formular level. For an analogous use of characterization by the variation of one formula, see now Olsen 1984:134-35, who observes that the andswarode formula in the Old English Andreas is restricted to Christ’s messages, whereas other characters’ speeches conclude with ageaf andsware.
[ back ] 28. Friedrich and Redfield 1978:265, 267, 266. A further methodological problem, not noticed by later critiques: Redfield’s and Friedrich’s “counter-samples” of speeches made within the hearing of Achilles, while the first effort by any scholars to provide a control on Achilles’ language, are less representative of non-Achillean speech since it has been shown recently that Iliadic replies copy, to a large extent, the diction and the structure of the speeches they answer: see Lohmann 1970:131-82.
[ back ] 29. Redfield and Friedrich 1978:271-75.
[ back ] 30. Redfield and Friedrich 1978:277-83. On the problems of analyzing “character” in modern poetics, see Rimmon-Kenan 1983:29, and on the problem of attributed direct speech in fiction, Martinez-Bonati 1981:30-32.
[ back ] 31. This was one of nearly thirty criteria, ranging from phonology to metrics to clause-structure, which they checked and found to be nonidiosyncratic in Achilles’ case. See Redfield and Friedrich 1978:283.
[ back ] 32. Friedrich 1978:284. Fish 1973. See also Pearce 1977:1-36. Plett 1985 argues for a renewed study of rhetorical stylistics to remedy the faults of statistical and structural analysis.
[ back ] 33. Messing 1981:890-94, 897. On ēthopoiia, see Kennedy 1963:90-93. An example not cited by Messing is in Eust. 752.61, where the commentator links amphiboly in Achilles’ speech with the ethos of “an angry person.” In defense of Friedrich’s and Redfield’s intuitions, it should be noted that some oral literature does in fact use grammatical and phonetic means to distinguish the talk of distinct characters: for example, the “buzzard talk” in Nomatsiguenga (Peruvian) myths, on which see Pickering 1980:21.
[ back ] 34. See Cantilena 1982:23n.9 for the list of analyses. Segal 1971 comes closest to the goal of combining literary appreciation with formula analysis. For exemplary close formulaic analysis of other oral poetic traditions, see Davidson 1983 and 1985 on the Iranian Shahnama and Barnett 1978:534-60 on Indie epic.
[ back ] 35. A. Parry 1966:12. See his further remarks on the illusion of formulaic inflexibility at A. Parry 1972:10.
[ back ] 36. On the history of attempts to prove orality by formula quantity, see Miller 1982:28-38.
[ back ] 37. Herzfeld 1985; Bauman 1978 and 1986; Abrahams 1983.
[ back ] 38. Leech 1985:45-56.
[ back ] 39. It may be objected that this comparison leaves out many formulas that can be paralleled in the Odyssey. (For lists of such repetitions and suggestions as to borrowing in either direction, see Van Thiel 1982:312-14 and Ramersdorfer 1981:108-13.) I believe that the uncertainty regarding the relation of the two monumental epics— especially in the light of such recent work as A. Edwards 1985, esp:11 — 13, regarding the competitive stance of the Odyssey-poet and the echoic nature of certain scenes and characters in the Odyssey—compels one to restrict the background, in order to avoid calling “formulaic” many lines and expressions that occur only in this speech and in a restricted number in the Odyssey. To my way of thinking, such iterata are more likely intentional reworkings of language familiar from Achilles’ speech, and composed with the assumption that an audience will recognize them as coming from a particular character and context. This is not to say that one must be an Analyst in considering the two poems: both could be oral compositions, and yet show such responsions, especially if they are indeed by the same poet, or by two poets in an agonistic performance situation. Contemporary Nigerian oral poetry affords the best example, to my knowledge, of the way in which certain themes and ways of narrating them can come to be associated with one particular poet, even though the performer in question has never set them down in writing under his name: some virtuoso oral poets among the Hausa are credited with creating a bakandamiya, “poetic masterpiece,” which they consider their favorite song and which they have reworked, expanded, and polished for years. Their reputations are based on these large-scale compositions. See Muhammed 1981.
[ back ] 40. MHV 272. This does not include echoed phrases, anaphora, or polyptoton, he notes.
[ back ] 41. Ingalls 1976 puts the structural formula controversy in context. Hainsworth 1964 had foreseen the objections of Minton 1965, who, in turn, was responding to Russo 1963. Nagler 1974:1-22 while offering some brilliant intepretations related to the formula’s associative nature ended up by seeming to deny that it was possible to invent a usable definition of the formula, as Russo 1976 pointed out. On the other definitions of “formula,” see Hainsworth 1969:19-20. The seed of the extended definition is in Parry’s work, MHV 301-9.
[ back ] 42. Hainsworth 1968, esp:39-45 outlines the various ways in which flexibility is achieved in the formula. Nagy 1974 showed that correspondence between kleos aphthiton and Vedic iravas aksitam extends to their metrical environments, and that these, in turn, suggest that the hexameter originated as an expanded lyric line. On the choice between the metrical and the semantic explanations of formula, see Cantilena 1982:45-62.
[ back ] 43. See Ducrot and Todorov 1972:139-42 for a concise discussion of these concepts.
[ back ] 44. I believe that the first factor explains the almost universal occurrence of the noun makhē after the trochaic caesura. It has this position in the most common formulas, but keeps it also when the poet does not “expand” the line by using the full formulas. See Martin 1983:67-69.
[ back ] 45. Note that this is a shift from the usual practice (e.g. Lord 1960:143) of using broken underlining for diction that resembles other formula types: my broken underlinings indicate that the word in question, or a form of the word, itself recurs in the slot in question.
[ back ] 46. In using as background the single poem, I am taking Lord’s dictum that the formula has meaning only in performance (see Lord 1960:33) to its logical end: the formula has significance as “formula” only in the space of a performance.
[ back ] 47. Finley 1979:33. On the roots of the poet-as-carpenter metaphor in Indo-European poetics see Schmitt 1967:14, 297-98.
[ back ] 48. On the need in stylistic studies for investigating whether coherence exists among deviations, see Leech 1985:50—52.
[ back ] 49. I have explained the notion of “genres of discourse” more fully in applying it to a problem in Book 8 of the Odyssey: see Martin 1984:30-32.
[ back ] 50. On Neo-analyst methods, see Kullmann 1981 and 1984. Fenik 1974:139 discusses possible non-Analyst readings of Homeric repetitions.
[ back ] 51. The formal listing of goods within a raiding context has a long history: see Watkins 1979 285-87, who views the list at 23.259-61 as a partial expression of the full Indo-European folk taxonomy of wealth as found in Hittite texts; note also the close resemblance between the expressions in Nestor’s list at 11.678-80 and the to-so lists contained in several Linear Β texts: Chadwick 1973:587, s.v. On formal declaration as the essential feature of boasts and prayers see Muellner 1976:98-99; the discourse of raiding is also connected to the particular theme of quarrel at a division of spoils: see Nagy 1979:127-30.
[ back ] 52. See, on repetitions that are not formulaic, MHV 273. Mueller 1984:150-58 borrows the notion of “contextual surplus” from Paul Ricoeur to describe the type of repetition under discussion.
[ back ] 53. On the phenomenon of such formula runs, see Janko 1981 and Hainsworth 1976.
[ back ] 54. Sometimes we get a glimmer of Homer’s reshaping of formulas from an examination of such paradigmatic occurrences, it should be noted. For example, we can guess that οὐκ’ ἐθέλεσκε μάχην (9.353), with the noun ending at the penthemimeral caesura, a slot used for it only two of twenty-seven times in the Iliad, is a transformation of, for example, οὐδ’ ἐθέλουσι μάχεσθαι (14.51, cf. οἵ μ’ἐθέλοντα μάχεσθαι at 15.722, and the same verb and infinitive at 3.241, 4.224, 6.141 in different metrical slots). This still tells us nothing about the meaning of 9.353.
[ back ] 55. For this reading see Leaf 1900-1902 at 9.355.
[ back ] 56. I borrow the term from the work of structuralist narratologists, on which see Chatman 1978 53-56 and Greimas and Courtes 1982:167, 362-64. It is no accident that recent theories of narrative stem in large part from Propp’s work in a traditional narrative genre, Russian wondertales: see Hibernian’s introduction to Propp 1984.
[ back ] 57. See Fenik 1968:232 for the use of the motif in typical battle scenes.
Compare with this 4.37-49, which begins similarly: ἕρξον, ὅπως ἐθέλεις.
[ back ] 59. Thetis (24.131-32) uses precisely Patroklos’ words to foretell Achilles’ death.
[ back ] 60. I used for this analysis Kumpf 1984.
[ back ] 61. Nagy 1974:141. See also Risch 1987. For collection and analysis of other Indo-European poetic phrases, see Schmitt 1967.
[ back ] 62. For this interpretation, see Finkelberg 1986. Nagy replies to this argument in a forthcoming work, Pindar’s Homer: The Lyric Possession of an Epic Past.
[ back ] 63. The notion of phraseology as heirloom, explicit in the Old English poetic conceit, is also inherent in Greek tradition: witness Pindar’s image of the “treasure-house” of song (Pyth. 6.8; cf. Ol. 6.65); for full explication of this image within archaic poetics, I refer the reader to the forthcoming work in this series of Leslie Kurke on Pindaric oikonomia. In this regard, it is significant that Achilles at 9.413 is actually quoting Thetis, for the phrase kleos aphthiton is thus given the authority of the speech of the immortals, an appropriate emblem for its age-old heritage. On Homer’s use of this phrase as intentional, significant archaism, see A. Edwards 1985:75—78 and also Nagy 1981.
[ back ] 64. I would argue that the phrase anēr agathos owes something to a genre of moral-didactic poetry, which eventually produces such poems as that by Simonides on the man who is “four-square and good” (Poetae Melici Graeci 542.1, 17; cf. 531.6). In other words, we do not have to consider the new phrase an entirely new creation; its novelty may lie in its being imported from an old genre for use in a new one: on this phenomenon, see Martin 1984.
[ back ] 65. On this meaning of philos and its derivatives, see Benveniste 1960,1:338-53.
[ back ] 66. Sakharnyj 1976:76 observes that a similar reshaping occurs in line 9.409: instead of the usual subject (“word”) in the phrase “passes the barrier of the teeth” we have here “spirit.” On the unusual placement and reference of lainos oudos (404), see Ramersdorfer 1981:193.
[ back ] 67. This also occurs with a few noun phrases in the speech: for instance, Achilles uses aisē (9.378) in what appears to be the older sense, “measure, estimate” (cf. Leaf {1900-1902} 1.418), as opposed to the derived meaning “fate” (cf. 22.477, 24.428—in this slot), and he is consistent in this: at 9.608 he uses it in the same way to speak of being honored “in the estimation of Zeus” (cf. 378 tiō and 608 tetimēsthai). The only other time the noun means “estimate” is in the frozen expressions kata aisan and huper aisan. An example of phraseology shared by only Achilles and the poet: thea Thetis arguropeza (410), which occurs six other times, but only in the narrator’s voice.
[ back ] 68. See Hentze 1887:158.
[ back ] 69. A similar expression is used to describe divine disfavor toward mortals: kaka phroneōn at 7.70, 12.67 and oloa phroneōn at 16.701. I suggest that the use of the phrase to describe the anger of Patroklos and Achilles, in the later portions of the poem (16.373, 16.783, 22.264, 22.320) is a significant extension of its use earlier in the composition. A transformation of the expression for friendly relations can be seen at 23.343: φίλος, φρονέων πεφυλαγμένος εἶναι. On patterns of this verb’s use in other phrases, see Lockhart 1966:99-101.
[ back ] 70. On Antilokhos as clever rhetorician, see Nagy (1983).
[ back ] 71. As often in the Iliad, we get the impression that a character has heard the previous poetic narration: Antilokhos here seems to be throwing back at Achilles the latter’s unique way of speaking about his possessions: compare 23.549, ἔστι τοι . . . with 9.364-67, ἔστι δέ μοι μάλα πολλά. Note also that Antilokhos’ impetuous argument by anticipation recalls the characterization of Achilles, who had guessed what Kalkhas had to say (1.90-91). Further touches reminiscent of Book 1 in this scene are Antilokhos’ use of ἀφαιρήσεσθοα (cf. 1.230, ἀποαιρεῖσθαι) and his echo of Agamemnon’s refusal to hand over Khryseis (cf. 1.29: τὴν δ’ ἐγὼ οὐ λύσω; 23.553: τὴν δ’ ἐγὼ οὐ δώσω.), as also the assertion that he will fight for the mare (23.554)—a contrast with Achilles’ yielding up of Briseis (1.298: χερσὶ μὲν οὔ τοι ἔγωγε μαχήσομοα). The recognition of a kindred young heroic spirit prompts Achilles’ famous smile here (23.555).
[ back ] 72. On this line, Ameis and Hentze compare 8.415: ἠπείλησε Κρόνου πάϊς, ᾗ τελέει περ in which the adverbial ᾗ modifies “he will complete” (said of Zeus’ threat). Note that, in Achilles’ version, the adverbial phrase seems to go with both verb-phrases: he makes thought and action one process, just as a god who threatens.
[ back ] 73. The more immediate model is in the previous speech, 9.288.
[ back ] 74. Furthermore, the idea of “telling all” is itself rare in the poem; it is mentioned as a task possible only for a god, not for a poet (12.176), and treated as an unusual request from his divine mother by Achilles (1.365).
[ back ] 75. Excluding the repetition of Achilles’ exact words in Odysseus’ report to Agamemnon (9.684).
[ back ] 76. On the antiquity of the formula ending this line, see Ivanov 1980:74-76.
[ back ] 77. See, for example, Friedrich and Redfield 1978:277; also King 1978:21, Maehler 1963:9-16, Whitman 1958:195. The fullest comparison is by Gerlach 1870:35-37.
[ back ] 78. Moulton 1977:100-101 lists eight Achillean similes. He compares the bird simile in this speech with other images, in the narrative, of mothers and children. See also on the image Randall 1978:74-78.
[ back ] 79. Further similarities in diction occur between 8.238-44 and 9.329-64. Agamemnon and Achilles alike view Troy as an infernal destination (see enthade errōn at 9.364 and 8.239). Both focus as well on their role in sacking cities (9.328, 8.241).
[ back ] 80. See the threats at 7.118, 10.370, 13.817, 17.27, 23.668, 23.579, and the boasts at 5.103, 10.548, 13.414, 13.785, 14.220. The noteworthy use of past tense with assertion occurs in Achilles’ boast/threat to Aeneas, 20.187. Observations of a general sort employ the present tense after the verb “I assert”: 2.248, 6.98 (two examples resembling threats and boasts); 6.488, 18.132.
[ back ] 81. Apart from Agamemnon and Achilles, only a few characters use plēmi to introduce statements about themselves. Among men, Nestor (10.548) and Meriones (13.269) use the verb to assert their lasting power in the fray; among gods, Zeus (15.165) and Hera (18.364) claim superiority using similar expression with plēmi. All these make their boasts in the present tense: Achilles thus has doubly marked deviant usage: he states his claims self-referentially, and he uses the past tense.
[ back ] 82. Cramer 1976:302 criticizes this as an insufficient theory of rhetoric, but, as I show later, the statement of candor cannot be taken at face value since Achilles proceeds to contradict it in practice.
[ back ] 83. See Chapter I for more on the primacy of persuasion within the Iliad, and Chapter 2 for a reading of this scene.
[ back ] 84. See also the remarks of Ameis and Hentze 1905 on 19.67.
[ back ] 85. On this important speech see Rosner 1976.
[ back ] 86. 2.24 = 2.61 (Dream urges Agamemnon to arm), 5.490 (Sarpedon tells Hektor to consider defense), 12.315 (Sarpedon tells Glaukos they must stand and fight), 13.235 (Poseidon warns Idomeneus to hurry to the fight), 16.631 (Patroklos tells Meriones to fight instead of making threats), 19.149 (Achilles tells Agamemnon to fight rather than chatter), 19.228 (Odysseus advises Achilles to rest so he can fight again—cf. 231-33; the same motif is at 7.331).
[ back ] 87. On the concept of metacommunication, see Jakobson 1960, Baurnan 1978:16, and Stubbs 1983:46-50.
[ back ] 88. For the conventions in this performance-genre of flyting, which Greek labels neikos, see Chapter 2.
[ back ] 89. For the notion of divine authority contained in the verb kraiaino, see Benveniste 1969,2:35-42.
[ back ] 90. On speaking as an important aspect of the Indo-European tradition of prince instruction, see Martin 1984.
[ back ] 91. On the distinction between the “existential” esti and the verb used as a copula, which is common in the poem, see Benveniste 1966:187-207.
[ back ] 92. Moulton 1977:101 points out that Achilles is not the only figure to use similes: for example, see Asius’ comparison of Achaeans to wasps and bees, 12.167—72.
[ back ] 93. On the conventions of augury talk, and the delineation of heroes by their skill at it, see Bushnell 1982.