Printer Friendly

The occasional contextual appropriateness of formulaic diction in the Homeric poems.

Through his study of epithets in the Homeric epics Milman Parry made a convincing case that we should understand and evaluate Homeric verse very differently than we do the verse of most other poets. Because the epithet-noun phrases exhibit a strict economy (there is only one way of expressing each essential idea that fills any given portion of the hexameter), Parry argued that when reading Homer we must suppress our presupposition that the poet chooses the words he does on the basis of contextual appropriateness: instead of pertaining to the particular context in which they appear, the epithets ennoble the whole of the poem. When a ship is described as fast, it is not because it is, at that particular moment of the story, moving quickly, but because all ships of the heroic age depicted in epic were fast ships.

Homeric criticism since Parry is an uneasy criticism.(1) Scholars have challenged Parry's thesis in a number of ways: arguing that, historically considered, the formulas are more contextually appropriate than they might seem,(2) that the very concept of a formula is an extremely unstable one,(3) and that many formulas are almost without a doubt chosen for specific poetic or thematic effects.(4) Nevertheless, many formulas still seem to have been selected, as Parry argued they were, solely on the basis of metrical convenience. Even the most impassioned advocate for Homer's artistry, Norman Austin, admits that some formulas are apparently chosen simply to fill a certain portion of the line (Archery 16). Critical anxiety arises in that if even some formulas are chosen on the basis of metrics alone, the possibility remains that the seemingly "artistic" formulas that critics have located--formulas appropriate to their immediate context--were also chosen simply because they fill a certain portion of the hexameter and that such formulas are therefore only, as it were, accidentally artistic. And such a possibility strikes at the core of the critical enterprise, at least insofar as that enterprise involves elucidating a poet's verbal artistry, his choice of the most fitting word or phrase from among many possibilities.

Since the presence of any formulas demonstrably chosen on the basis of metrical convenience alone raises the possibility that all formulas are so chosen, it is assumed that the ideal refutation of Parry would be to assign a nonmetrical, contextual reason for the poet's choice of each and every formula. Since we can assign a nonmetrical reason for only some of the formulas, the critical uneasiness remains--though it does not, fortunately, stop critics from arguing for the contextual appropriateness of particular epithets. I suggest that in fact we need not feel anxiety at being able to provide contextual explanation for some, but not all, of the epithets. By exposing a fundamental flaw in Parry's reasoning, I hope to justify the practice of treating some epithets as appropriate to their immediate context, even though we cannot treat all epithets in this way. My argument has larger implications for our understanding of Homeric composition. That mode of composition is, as Parry argued, significantly different from that of a nontraditional poet--but not, I maintain, in the way he argued. Because, as I will show, the bard has a degree of stylistic freedom, the audience is prepared to treat as ornamental an epithet that is inappropriate or indifferently appropriate to its context; but the audience is not therefore insensible to contextually appropriate epithets. The distinctive character of Homeric poetry lies not, as Parry claimed, in the ornamental nature of contextually indifferent formulas, but rather in the interplay between contextually indifferent and contextually appropriate formulas.

I should say at the outset a word about my method. I do not seek to refute Parry's thesis, but to complement and extend it. At two critical points, therefore, I advance my own argument by drawing out the implications of comments that Parry himself made in his seminal dissertation "The Traditional Epithet in Homer," comments whose disruptive power he seemingly failed to recognize. A note also on terminology: frequently in order to illustrate my points I refer to the lines employed to introduce direct discourse (e.g., "Swift--footed Achilles then said") with a term employed in criticism of modern novels and short stories: "tag lines."


Let me begin by stating briefly my fundamental thesis, which is a simple one. If we restrict our analysis of Homeric diction to the level of the phrase, as Parry by and large did, the diction is every bit as economical as he claimed; with very few exceptions, there is only one way of expressing any given essential idea so as to fill a particular portion of the line. But if we broaden our focus and examine Homeric diction at the level of the full line or of the passage, we find that the principle of economy no longer holds true.

For the sake of simplicity, Parry begins his demonstration of Homeric formulas by examining a special case: sentences that fill up exactly one line, a predicate in the first portion of the line and a subject in the second portion. What he showed is that no matter how much space the predicate takes up, there exists a way of expressing any subject so as to fill up the remainder of the line. So if the subject of the sentence is to be Odysseus and the predicate runs to the feminine caesura, the poet will finish the line with the phrase

[GREEK TEXT OMITTED] whereas if the predicate extends to the hephthemimeral caesura, he will finish the line with the phrase

[GREEK TEXT OMITTED] and if it extends as far as the bucolic diaeresis, the poet will finish the line with the phrase

[GREEK TEXT OMITTED] What Parry failed to consider, though, is the possibility that for the predicate portion of the line, too, there might be a set of formulas all expressing the same essential idea, but filling different portions of the line.

Although he failed to consider this possibility, he demonstrated, seemingly despite himself, that such a set of formulas does in fact exist. As the first sample sentence in his exposition, Parry chose a common tag line:


The predicate--

[GREEK TEXT OMITTED] --fills the line to the feminine caesura. Parry then demonstrated that the bard, since he had for every character in the epic a subject formula filling the space from the feminine caesura to the end of the line, was able to make the sentence "then X answered him" for any character.

The next step in Parry's argument was to show that another series of subject formulas exists for occasions in which the bard had to fill the portion of the line extending from the hephthemimeral caesura to the end of the line. The formula for Odysseus in such a case is

[GREEK TEXT OMITTED] and Parry illustrated its operation with the following line:


What he seemingly failed to realize is that the predicate hemistich of this line

[GREEK TEXT OMITTED] communicates the same essential idea as

[GREEK TEXT OMITTED] Accordingly there are two ways of using a full line to say "Odysseus replied," either



In fact there is a third, in Odyssey 15.485:


What I object to in Parry's analysis, then, is the impression that in the process of composition the epic bard simply finds himself, as if by accident, at some spot in the line and must then fill out the line with a formula of the appropriate length. Parry conveyed such an impression by statements like the following: "with very few exceptions the poet makes use of one type of formula to complete a sentence of which the predicate extends only to the feminine caesura" (Collected Works 15). True, but the poet himself controls, to some degree at least, how far the predicate will extend and therefore can deliberately choose to create a sentence which uses a desired subject formula.

If we examine the matter at the level of the whole line, then, we see that the traditional bard does possess, in some degree, the sort of stylistic freedom that Parry granted only to the nontraditional bard: he has the freedom, that is, to choose (from what is admittedly a restricted range of options, an issue I will address later) the way of expressing "then X answered him" most appropriate to the context.

Let us examine briefly how he exercises that freedom. In the argument between Achilles and Agamemnon in book 1, the poet uses two different tag lines to introduce the speeches of each hero. Agamemnon's first reply to Achilles, at line 130, is prefaced by the line


This is what we might refer to as the standard way for the poet to say "Agamemnon replied," for there are three other occasions in which this line appears (Il. 1.285, 2.369, 4.188). But in the same exchange, at line 172, the poet says "Agamemnon replied" in a different way:


This is the only time in the entire epic when a reply of Agamemnon is narrated using the formula

[GREEK TEXT OMITTED] and therefore the only time that, in replying, Agamemnon is called

[GREEK TEXT OMITTED] (though he is of course described as

[GREEK TEXT OMITTED] in connection with other actions). And the appropriateness of aranging to employ this epithet at this particular moment can be seen in two facts. First, the speech that this line introduces is the pivotal one in confirming the break between the two Achaean heroes (this is the speech in which Agamemnon threatens to take Briseis and after which Achilles contemplates killing Agamemnon and must be restrained by Athena), so the unique tag line underscores the importance of what is about to be said. Second, it is precisely his role as king that Agamemnon stresses in the speech that follows, for when he says,

[GREEK TEXT OMITTED] he knows as well as anyone else that it is only in his capacity as commander of many men, as

[GREEK TEXT OMITTED] that he is stronger than Achilles.

In describing Achilles' replies, the poet also makes use of both a standard tag line and an unusual one. Line 84 provides an example of a standard tag line:


This line is used eleven other times in the epic (counting lines addressed to female characters, which differ only by beginning with


[GREEK TEXT OMITTED] The unusual tag line appears at line 121, when Achilles first addresses Agamemnon directly (having previously been talking with Chalcas):


This line appears only one other time in the epic, at 18.181, when Iris urges Achilles to enter the war on behalf of the slain Patroclus, and he hesitates only about how he will arm himself. The two appearances of the line, then, frame Achilles' absence from battle by marking the point where he first defies Agamemnon and the point where another concern overrides his dispute with the king and he prepares to reenter the conflict. The poet is able to reserve one tag line for special thematic effects such as this only because he has another of the same metrical value that he can use in other instances. His choice of formula is therefore (to some degree) contextually motivated--as that of a nontraditional poet.

When I claim that the epic bard could choose specific epithets, I am disputing only the extent to which Parry took his argument about the virtual meaninglessness of the epithet:

It is not, then, only in cases where no relation between the idea of the epithet and that of the sentence is possible that the reader acquires this insensibility [to any particularized meaning of the epithet]. He soon comes to acquire it just as much in passages where special reasons, sometimes very good reasons, could be adduced for the choice of the epithet. (Collected Works 127)

I want to contest this point by a closer examination of the very passage that Parry adduced to illustrate it:

The novice in the study of Homer, whether he begins with the Iliad or the Odyssey, rapidly becomes familiar with certain set ways of speaking of ships, and one of the most frequent of these is the expression "swift ship,"

[GREEK TEXT OMITTED] He comes across this expression so often when the ship is at anchor, or drawn up on the beach, or wrecked, that he soon learns not to expect any particularized meaning from it. Hence when he comes to v 168, where the Phaeacians speak of the ship which Poseidon has turned to stone, and reads the expression "swift ship," it does not occur to him to look for the particular reason why this epithet was used. He will find in the phrase no statement of pity for the fate of this ship, so swift when it bore Odysseus to his home. He has invested the epithet

[GREEK TEXT OMITTED] wherever it modifies the substantive

[GREEK TEXT OMITTED] with a purely ornamental quality. He no longer reads "swift ship"; he reads "fast-sailing-ship." Having encountered so many times this single combination of words, this unity of diction, he at last attributes to it a unity of thought. The expression awakens in him a single idea, that of a hero's ship which possesses the speed characteristic of the finest ships; but in the world of epic poetry he knows only the finest ships--there are no others. So he thinks simply of ship, in the genre of epic poetry, the only kind as it seems, that there was in the heroic age. (Collected Works 127-28)

I quote at length because this is the central point in Parry's thesis that I aim to challenge. Parry here suggested that as readers grow accustomed to the fact that the poet's choice of words is governed by metrical convenience, they become indifferent to the denotative force of those words. The passage he mentions reads as follows:


One's familiarity with the epic diction, Parry argued, leads one to realize that in order to fill this particular segment of the line (from the beginning of the third foot to the hephthemimeral caesura) and express the essential idea of "ship," the poet could not have chosen any epithet except

[GREEK TEXT OMITTED] therefore it is not worth asking whether he chose the word to convey some subtle nuance of meaning, some "statement of pity for the fate of this ship, so swift when it bore Odysseus to his home." What Parry did not seem to realize, though, is that it is not by necessity that the poet had just that portion of the line to fill. He had a degree of control over how much of the line he would need to fill, as is evident when one realizes that he could have composed the line otherwise. He could, for instance, have said:


Ah me, who was it fastened our ship on the wine-dark sea? In other words, the poet has several essential ideas in this line--the ideas of "ship" and "sea" in particular--and he can expand the formula used to express one by contracting the formula used to express the other. When they appear in the positions that they do in this line, both words, ship and sea, normally appear with an epithet

[GREEK TEXT OMITTED] appears at this position twenty-one times with the epithet

[GREEK TEXT OMITTED] and twelve times with no epithet;

[GREEK TEXT OMITTED] appears in this position six times with the epithet

[GREEK TEXT OMITTED] and four times with no epithet). But as this particular line is constructed, there is only room for one of the nouns to have an epithet. The poet, I would argue, considers his options. He can say either "the swift ship was bound to the sea" or "the ship was bound to the wine--dark sea." He realizes that by choosing to say "swift ship," he can produce two striking poetic effects: an elegiac "statement of pity for the fate of this ship, so swift when it bore Odysseus to his home" (precisely what Parry argued against) and (a significance Parry did not consider, but would dismiss on the same grounds) the provocative paradox of something swift being bound. By contrast, for the poet to say "wine--dark sea" will in this context achieve no equally compelling effect. So he chooses a swift ship over a wine-dark sea, for that is the phrase most appropriate to this particular context.

In a different context, of course, to speak of a wine-dark sea might be the more effective choice. Consider, for instance, the passage in which Odysseus, eager to leave Calypso's island, states his resolve to endure the perils of the voyage:


This line could as easily have had an epithet describing the gods rather than the sea, for example,


But to speak of the dark sea is an effective way to heighten the sense of danger Odysseus faces, while to speak of some god smiting him on the sea as blessed is absurd. So the poet chooses the more contextually apt of the two phrases at his disposal.

Before continuing, I need to be clear about two points. First, by arguing that Homer was capable of arranging his lines so as to be able to employ a particular formula appropriate to the context, I am not arguing that he could--as a nontraditional poet can--make all of his language context-specific. On the contrary, in order to facilitate just a single phrase appropriate to the context in which it appears, the poet might have to sing several lines in which he uses formulas precisely as Parry argued they are always used: without regard to immediate context. The contextual appropriateness of several lines would be sacrificed, so to speak, in order to insure that one particularly apt expression might be used. Second, while I have for the sake of rapid illustration been using (and will continue to use) single lines as my examples, the principle I am describing pertains as well or better to whole passages. The general picture I am trying to paint is one of a bard who in the passage he is about to narrate has a number of (in Parry's term) essential ideas that he wants to express; and since he wants to express some limited number of these essential ideas in a particular way, he uses whatever version of the other essential ideas will leave him in a position to express his favored ideas in just the way he wants. The bard makes some formulas appropriate at the cost of others' being inappropriate or indifferently appropriate to the immediate situation.(6)

Two more examples will help substantiate (and qualify) my argument. In book 1 of the Iliad, Chryses is described praying to Apollo:


Here, in contrast with the lines I have examined up to this point, both Apollo and Leto are modified by an epithet. But I would claim that only the epithet describing Apollo is meant to convey its full denotative force. Apollo is appropriately called a king at this moment, for that denomination effectively underscores the purpose for which Chryses approaches him. Having failed to persuade king Agamemnon to release his daughter, the priest turns to Agamemnon's superior, king Apollo. There is no special reason for us to think of Leto at this time as fairhaired, so we treat this epithet as Parry would treat all ornamental epithets. In fact, a possible recast of the line,

[GREEK TEXT OMITTED] demonstrates that Leto is described by Homer as

[GREEK TEXT OMITTED] rather than

[GREEK TEXT OMITTED] solely so that Achilles can be described as a king, rather than the far less appropriate "wolf-born." This is what I mean when I claim that the poet, in order to employ a specific formula, uses whatever version of formulas nearby will allow the favored formula. We should treat those "set up" formulas, I would argue, precisely as Parry argued that we should treat all formulas: without regard to their particular denotative force.

It is possible, finally, that in some instances both portions of a line will be contextually appropriate. This is what I consider to be the case in 1.8:


Son of Atreus, ruler of men and divine Achilles.

The antithesis of


[GREEK TEXT OMITTED] sketches in bold strokes the conflict to come--a conflict in which Agamemnon pulls his rank as king of men on Achilles, who in turn relies on his goddess-mother to set into action on his behalf a divinely orchestrated chain of events. Usually, though, half of the line will have to be, as I put it, sacrificed for the sake of the other half.

Again this argument is an extension, not a refutation, of Parry's. I believe that the audience is, as Parry argued, prepared to ignore the significance of a given epithet; I diverge from Parry only when he claims that the audience always does so. I believe instead that the flexibility of the various formulas allows the bard to make some appropriate to their context--at the expense, though, of having some others be inappropriate or indifferently appropriate. But since his audience is prepared to ignore the inappropriate ones, those formulas simply form the ground or foil against which the contextually motivated formulas manifest themselves.

AVOIDING A FORMULA--BOHN ATA[theta]O[sigma] [delta]IOMH[delta]H[sigma]

The seeds of my expansion of Parry's theory, as I noted, are present in his own argument--though he did not seem to realize it. My first example was the way in which Parry himself unwittingly demonstrated that there are multiple metrically equivalent ways of saying "X replied." Exploring the second example of how Parry implicitly refuted his own argument will provide the opportunity for drawing out some of the implications of my view of formulaic diction.

Before beginning his disucssion of generic epithets, Parry summed up his argument to that point: The poet (or poets) of the Iliad and the Odyssey was so thoroughly steeped in traditional formulas that he never once ... created of his own accord an epithet revealing the personal stamp of his thought. Traces of originality remain, perhaps; but of an originality that does no more than rearrange the words and expressions of the tradition without important modifications. The poet's greatest originality in the handling of epithets would have been to use some noum-epithet formulas a little more or a little less frequently than other poets. (Collected Works 82-83)

The early portions of this quotation simply restate his fundamental thesis concerning the traditional character of epic diction. But the claim that one poet could use some formulas more or less frequently than another poet runs counter to the impression which Parry usually gives, of the bard's simply finding himself in a position in which he must use one or another epithet. Indeed, the claim that a poet could deliberately use one epithet more or less frequently supports instead the understanding of formulaic diction that I have been outlining. For how could the bard avoid (or use especially often) a particular subject formula except by deliberately controlling, through the lengthening and shortening of other formulas, the spot in the line at which he began the subject formula?

One implication that I am interested in drawing out of this argument concerns audience expectation. Parry at one point in his argument suggested that as we study Homer's diction, we should always keep in mind the apprentice bard and the process by which he masters his storytelling craft (Collected Works 56). I would suggest that we might do better to keep in mind the audience. Considering audience expectation will help show how a poet's choosing to use a particular epithet more or less often can have important contextual ramifications.

The epithet most frequently used to describe Diomedes is

[GREEK TEXT OMITTED] Yet during his raid on the Trojan camp, narrated in book 10 of the Iliad, when silence is of the essence, he is not once referred to with this epithet. One can imagine the suspense for the audience when this episode was sung. Knowing that if the poet was not careful about where he ended up in the line, he might have to describe Diomedes as

[GREEK TEXT OMITTED] and realizing that to have a great shout is a liability at this moment, they listened carefully to see if he would slip up and so describe Diomedes. The narration, in fact, parallels the event being narrated, for just as within the world of the story Diomedes must avoid making noise, so too in the real world must the poet avoid dooming him, as it were, by using that epithet.

This suspenseful poetic effect, moreover, in which the poet deliberately avoids using a particular epithet, is heightened by the frequency with which he uses that same epithet just before the episode. In the opening 283 lines of book 10, the poet describes Diomedes as

[GREEK TEXT OMITTED] no fewer than four times, or roughly once in every seventy lines of verse. Although this is, as I mentioned, the most common epithet for Diomedes, it is never used with quite this degree of frequency. Even in book 5, where his aristeia is narrated, Diomedes is only described as

[GREEK TEXT OMITTED] about once every ninety-five lines. Thus in the opening of book 10 the poet is taunting his audience, reminding them that Diomedes has a loud voice which might be a liability in the coming mission. Moreover, the poet does this right up to the moment when silence is necessary; even immediately before embarking on the raid. When Diomedes begins his prayer to Athena, he is described as

[GREEK TEXT OMITTED] Once the raid begins, though, the epithet is not used. (In this connection it is worth noting the only other time in the epics in which silence is critical: the Trojan Horse episode. Diomedes and Odysseus are the only two heroes mentioned by name in that scene, and appropriately Diomedes is again without the


The reading I have just advanced assumes that the members of the bard's audience were so familiar with the formulaic diction as to have in the back of their minds a sort of rough thesaurus or catalog of the various ways in which each essential idea could be expressed--like the tables that Parry lays out in his dissertation. Such a thesaurus would allow the audience, on some level of awareness at least, to compare the poet's actual words with a number of other ways in which he could have said the same thing. Of course, their mental thesauri would not be as complete as the poet's, so they might on occasion need to be reminded of a certain epithet--as I have shown the poet reminding them of

[GREEK TEXT OMITTED] in the opening of book 10--but once they were so reminded, they could hear (as in line 446)

[GREEK TEXT OMITTED] and recognize the appropriateness of the poet's avoiding, in this moment when silence is so critical, the standard epithet.


Though I have shown that the oral-formulaic bard had a degree of stylistic freedom in the use of his formulas, it must be admitted that that freedom is fairly limited. For example, I have shown that one can fill an entire line saying "Odysseus replied" in at least three ways:

[GREEK TEXT OMITTED] So, depending on the context, one can strss Odysseus' scheming

[GREEK TEXT OMITTED] his endurance

[GREEK TEXT OMITTED] or his divine lineage

[GREEK TEXT OMITTED] But this is still not very many choices--certainly not the range available to the nontraditional poet to whom Parry contrasts Homer, who has perhaps dozens of alternative ways of expressing any given essential idea, and whose audience, also aware of those alternate expressions, can judge how appropriate a choice he made from among them, and thus evaluate his style. Take, for example, a passage from Paradise Lost in which Milton describes Satan preparing to rouse the host of fallen angels: He [Beelzebub] scarce had ceased when the superior Fiend Was moving toward the shore; his ponderous shield Ethereal temper, massy, large and round, Behind him cast; the broad circumference Hung on his shoulders like the Moon. One might admire in this passage Milton's choice of the word ponderous to describe Satan's shield. The basis of one's admiration would be a comparison, conscious or unconscious, between the word ponderous and the other words Milton might have chosen: heavy, bulky, weighty, massive, etc. And because Milton had more choices than a Greek oral poet had, we can more easily grant that he chose le mot juste, the word that of all words best captures what we presume to be the desired nuance. Since we cannot look in the Homeric poems for one word or phrase selected for its particular nuance from a large number of similar alternatives, we might conclude that, even given my modifications of Parry's thesis, the poet is still really a traditional poet in Parry's sense of that term.

Yes and no. For if we again broaden our focus from signle words to whole lines or passages, we find that the several options which the nontraditional poet has for each word make for an almost infinite number of permutations for the line as a whole; the reader of the text has the opportunity to pause and reflect on those choices and their alternatives, and thus to evaluate the stylistic skill of the poet. The oral poet had fewer options, but on some level his audience was aware of those options. The result was that the oral poet might actually impress his audience with his stylistic skills during performance, for given the bard's restricted range of possibilities, his audience could, arguably, feel the effect of his choice of a particular formula even at the speed of the bard's performance. Thus the more limited stylistic range was eminently suited to that particular mode of presentation.


I summarize, conclude, and extend my remarks by reversing my perspective. I have so far considered the way a line or phrase can be suited to its context. Now I would like to consider what suiting a given phrase to the context does for the phrase itself. Parry's understanding of Homeric diction suggests that a given epithet, by being repeated in contexts where it fits more or less well, relinquishes its denotative force and becomes merely ornamental. But I would argue that this drift into ornamentality is by no means an irreversible, one-way process. For the more contexts in which a formula is repeated, the more opportunities it has so appropriately to fit one of those contexts as to have its meaning suddenly renewed.

Nor need such a process seem foreign to our own experience. We often say in parting, "I'll see you later," meaning simply "I'm sure I'll meet you again soon." But if one inadvertently says "I'll see you later" to a blind person, the phrase might suddenly become charged with a meaning it has long ceased to hold, and might sustain that resonance the next several times one used the phrase.

I submit that the occasional appropriateness of Homeric formulas insures their renewal as well as their fading. We accordingly ought not simply to watch, as Parry argued we should, their drift into meaninglessness (or, more precisely, we ought not to grant that almost all epithets have already drifted into meaninglessness). Instead, we ought to attend to their oscillation into ornamentality and back out again into contextually appropriate vividness. And, since (as I have shown) the bard has the ability to use a formula where he wants to and thus deliberately to jolt that formula back into significance, I suggest that we ought to attend to this oscillation as an artfully controlled process. I suggest, finally, that this particular poetic mode, in which the poet sparks back into vividness something that is drifting away into oblivion, perfectly suits the Homeric bard, whose purpose it is to grant

[GREEK TEXT OMITTED] that is, to spark back into vividness the memory of the hero, lest it fade into the oblivion of human forgetfulness.



Austin, Norman. Archery at the Dark of the Moon: Poetic Problems in Homer's Odyssey. Berkeley and Los Angeles: University of California Press, 1975.

Bakker, Egbert J., and Florence Fabbricotti. "Peripheral and Nuclear Semantics in Homeric Diction: The Case of Dative Expressions for 'Spear.'" Mnemosyne 44, nos. 1-2(1991) 63-84.

Edwards, Mark W. "On Some 'Answering' Expressions in Homer." CP 64 (1969) 81-87.

Nagler, Michael. Spontaneity and Tradition: A Study in the Oral Art of Homer. Berkeley and Los Angeles: University of California Press, 1974.

Nagy, Gregory. The Best of the Achaeans. Baltimore: The Johns Hopkins University Press, 1979.

Parry, Milman. The Collected Works of Milman Parry. Edited by Adam Parry. Oxford: Clarendon Press, 1971.

Visser, Edzard. Homerische Versifikationstechnik: Versuch einer Rekonstruktion. Frankfurt: Pepter Lant, 1987.

(1)Witness, for example, the emotion Norman Austin exhibits in the opening chapter of Archery at the Dark of the Moon regarding the implications for literary criticism of a strict understanding of formularity.

(2)See Nagy, Best of the Achaeans.

(3)See Nagler, Spontaneity and Tradition.

(4)See Austin, Archery.

(5)Austin (Archery 29) and Edwards ("'Answering' Expressions", 81) both note this apparent violation of the principle of economy, but neither draws from it the conclusions I will. Edwards, in fact, works hard to recuperate such violations into a strictly Parrian economy.

(6)Egbert Bakker and Florence Fabbricotti, following Edzard Visser, have advanced a similar argument by contending that in a given verse there is "nuclear" and "peripheral" material. They have shown that in verses that describe one character killing another, the phrase used to describe the spear is "peripheral" in that it varies depending on what portion of the line is left over when the ("nuclear") names of killer and victim have been placed. Such terminology may prove misleading, if it leads us to believe that certain elements of a verse are always nuclear, others always peripheral. (I can imagine instances in which the poet would want to insist on a sharp or shining spear and would modify the names of killer and victim to insure the proper epithet for "spear.") Accordingly, I prefer to continue using formula, a term whose limitations are already well established.
COPYRIGHT 1994 Johns Hopkins University Press
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1994 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Machacek, Gregory
Publication:American Journal of Philology
Date:Sep 22, 1994
Previous Article:Lucani Opera.
Next Article:Heracles, Deianeira, and Nessus: reverse chronology and human knowledge in Bacchylides 16.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters