Bryn Mawr Classical Review 94.09.11


Ahuvia Kahane, The Interpretation of Order. A Study in the Poetics of Homeric Repetition. Oxford Classical Monographs. Oxford: Clarendon Press, 1994. Pp. 190. $39.95. ISBN 0-19-814077-0.


Reviewed by James V. Morrison, Centre College.

The audience hears andra ("man") as the first word of the Odyssey. How does this affect the audience's reception of the epic which follows? Do repetitions of such a common word later in the epic echo the first line in any meaningful way? What if this word appears in the accusative case at the beginning of a verse -- would the added features of grammatical case and verse position more strongly echo the Odyssey's opening theme? Such questions have motivated K. to explore the significance of repetition in the Homeric epics. Although K. studies metrical shape, position, and caesura in Homer, his goal is primarily a literary interpretation examining the connections between techniques of verse composition and the issues of theme and character. As his array of evidence demonstrates, these questions ought to be asked. K.'s answers are sometimes quite persuasive, and almost always imaginative and provocative; at times the evidence is too slight to bear the weight of his interpretation.

Chapter 1, "Patterns and Verse-Making Technique," locates K.'s project within the confines of oral-formulaic theory. K. posits that the main focus of his study -- the repetition of single words -- falls outside the boundaries of what scholars call the Homeric formula (Parry's definition begins, "a group of words..."). K. then builds on work which has demonstrated that certain metrical shapes are highly localized in Homeric poetry, i.e., some words appear at certain favored positions in hexameter verse. A word with the metrical shape of u - - could appear at several points in the hexameter verse, yet in Homer Odysseus, e.g., appears at line end 95% of the time. That is, we discover a tendency of the poet to place Odysseus in the verse-terminal position. Such are the statistical facts K. wishes to interpret. In part, these phenomena are a consequence of formulaic composition, but K. believes that it is possible to discover "the semantic significance of the localization of particular words or types of words ... qua single words." (12) What do these metrical phenomena mean in literary terms? K. is not concerned with how the formulaic style was created or its capacity to facilitate the process of composition. Coming rather from the perspective of reception and response, K. wishes to learn the effects that these patterns produce.

Chapter 2, "Metrical Units and Sense-Units," explores the placement of word-endings within the hexameter line. K. shows that the most frequent breaks (or pauses) in meter coincide with the most frequent breaks in sense (K. makes use of West's work in Greek Meter on "sense-pause" -- K. maintains contra Daitz that deliberate pauses were used during performance and could be expressive). The importance of this argument is to demonstrate that metrical structure can be semantically functional. Using the data of O'Neill and others, K. calculates that the highest frequency of word-endings occur at positions 2, 3, 5, 5 1/2, 7, 8; they are least frequent at 7 1/2 (Hermann's bridge) and 11, according to the standard metrical scheme. This means that if the localization of metrical shapes overlaps the system of sense-pauses, then "there must be some significant relationship between metrical shape and location and sense." (30) This metrical regularity in turn generates expectations which the poet may manipulate.

To show the interpretive potential of this approach, K. examines two passages: Odyssey 5.1-4 and Iliad 1.1.-7. K. argues that the unconventional way of describing Eos at Od. 5.1 is context-sensitive. The opening emblem of Eos and Tithonus is thematically significant for book 5 which presents the relationship of Odysseus and Calypso and the possibility of immortality for a hero. (K.'s analysis is somewhat weakened by the fact that the description of Eos at Od. 5 is also used at Iliad 11.1, without apparent literary significance.) Noting dynamics and tension in the musical flow, K.'s fine reading of these lines (and the Iliad's opening) investigates the interplay of meter and sense-units.

Chapter 3, "Accusative Theme-Word Patterns," is the strongest section of the book. The case that metrical phenomena have a meaning is made most persuasively here. K. begins with the analogy of a musical motif assigned a meaning by repeated association with a theme. The "motifs" K. examines in the Homeric corpus are the repetition of aner, menis, and nostos ("man," "wrath," and "homecoming"). Let me recapitulate K.'s interpretation of aner in the Odyssey.

In the Odyssey, aner, the first word of the epic, occurs 32 times in the accusative case, 11 of which are verse-initial (cf. usage in the Iliad: acc. case 52x, of which 4 are verse-initial; overall occurrences of aner in the singular in all cases: 187x in Od., 222x in Il.). K. argues that every repetition of aner in the accusative case and in the initial position of the verse triggers an echo of the first line of the epic, but only when the pattern markers (accusative case, verse-initial position) are present does andra refer to the "man" who is the subject of the Odyssey. K.'s argument is built by analyzing aner in four situations: (1) pattern deixis (i.e., acc. case, verse-initial position) with reference to Odysseus. K. examines each of these 11 instances in some detail. The "man" returned by the Phaeacians to Ithaca (13.89) echoes Od. 1.1, marking the proemium to the second half of the epic. While the identity of the "man " who suffers at sea is unknown to the Phaeacians (8.139), the audience sharing "privileged information" with the narrator hears andra and recognizes Odysseus. (2) Pattern deixis with a reference to a character other than Odysseus. We find Telemach us as andra neoteron "younger man" (3.125). Here, pattern deixis echoes the similarity between the youthful Telemachus and his hero father: Telemachus has become "younger Odysseus." Later Odysseus speaks of an andra ("husband") for Nausicaa (6.181): the code spells out Odysseus himself as potential husband. When the beggar Odysseus refers to himself as andra geronta ("old man" -- 18.53), "this unambiguous reference allows the poet to put in Odysseus' mouth words that express his true identity [to the audience] ... even as he hides it [from the suitors]." (64) (As K. rightly insists -- using de Jong's terminology -- the significance of these implicit meanings is reserved for the poet/narrator and his audience/narratee.) andra then has a resonance which adds contrast or irony to particular situations. K. also examines instances of (3) aner lacking pattern markers (not accusative, not verse-initial) without reference to Odysseus, and (4) aner lacking pattern markers with reference to Odysseus.

On the face of it, K.'s scope is extremely limited. Yet using these four categories, K. cogently demonstrates that the repetition of a relatively simple word with a mundane denotation ("man") can evoke for the audience a broader epic perspective, regardless of the context. The pattern set at Odyssey 1.1 remains active throughout the work. The mere fact of a grammatical form (acc. case) functions as a marker which triggers an interpretation far beyond the local scene. K. also analyzes menis in the Iliad, which in the accusative case, verse-initial position refers to the "thematic wrath of the Iliad." One might ask to what extent does "the wrath of Apollo" (menin at 1.75) actually anticipate or evoke Achilles' anger as "the result of a chain of events started by Apollo"? (54) The third theme-word is nostos in the Odyssey (occurring in the accusative case in the Od. 38x; 19 are verse-initial; overall in the Odyssey, 59x; 7x in the Iliad). Again, noston in the acc. case, verse-initial position refers to Odysseus' "thematic" return. Of interest in category 2 (pattern deixis with reference to a character other than Odysseus): all four instances occur in Od. bk. 4 (381, 390, 424, 470) as Menelaus tells of his own attempts at a homecoming (noston) yet Telemachus (and we) want to hear about Odysseus -- and appreciate that tension by virtue of the formal pattern markers.

Chapter 4, "Patterns of the Proper-Name Vocative," limits its investigation to proper names in the vocative case (abbreviated PNV). K.'s argument is that the usage of the vocative case (in this instance, its placement in the verse) corresponds to the narrative role of the character. PNV's are found in verse-initial position approximately 50% of the time: K. calls this the "default mode." That is, while any vocative is designed to get or keep the addressee's attention, a vocative in the initial position is semantically neutral, indicating nothing of special significance to the audience. Yet the PNV's of Odysseus and Achilles -- the focus of the two Homeric epics -- almost always occur in verse-terminal position (52 of 53x combined). What is the effect of repeating these names at the end of the line? (About half of all the terminal PNV's are those of Odysseus or Achilles.) This "unparalleled (absolute) frequency of terminal positioning ... generates a reference either directly to the protagonists or to a 'protagonistic state.'" Because the "terminal PNV position coincides ... with an address to the epic protagonists," (87) K. argues that the eleven other characters addressed in verse-terminal position by virtue of the Achilles/Odysseus pattern are addressed as though they were epic protagonists. The effect of PNV diction is to generate an analogy with or contrast to Achilles or Odysseus.

Certain characters fit K.'s scheme nicely. The PNV of Hector while suited to verse-initial and verse-terminal positions is never terminal; in fact, 32 of 35 times it is verse-initial. K. interprets this as the poet "avoiding" terminal position for Hector in order to emphasize his adversarial role to Achilles: Hector is not the protagonist of the Iliad. The PNV's of the Odyssey's suitors, too, are always verse-initial: again, the effect is to distance the protagonist Odysseus from his rivals. Yet it must be said (as K. points out) that all but one of the suitors have PNV's unsuitable for verse-terminal positioning, thus weakening K.'s general point. That is, purely metrical considerations may have determined placement in the verse. Penelope's PNV is found in verse-terminal position (equivalent to "protagonistic state"), for she functions as the counterpart in cunning to Odysseus. Thus placement indicates not only the protagonist, but also someone linked to the protagonist. So far K. argues plausibly. There are, however, more doubtful cases. Telemachus's PNV appears in verse-initial position 27 of 30 times. Is the effect to emphasize that he is "a 'not-yet Odysseus' -- a character whose narrative role is defined precisely by his lack of Odyssean powers"? (100) On the face of it. Yet the metrical shape of Telemachus' PNV is unsuitable for verse-terminal position. Questionable also is the case of Eurycleia, whose role in bringing about the recognition of Odysseus is traced. Her PNV appears in verse-terminal position three out of three times, which "allows the narrator, 'using his own voice' in that code which he shares with the reader/audience, to evoke the object of these recognitions, the name of Odysseus." (95) At this point I was not confident that terminal position of a PNV still triggers an association with "protagonistic state."

The implication is certainly that usage would vary from epic to epic. Yet Odysseus has a relatively minor role in the Iliad: still 7 of 7 times his PNV is verse-terminal (in the Odyssey, 23 of 24x); the PNV of Achilles is always verse-terminal in the Odyssey (16 of 16x). Is this designed to contrast these two characters or is this formulaic composition at work? K.'s use of the term "contrast" is not unambiguous. Hector is said to contrast with Achilles structurally and narratively; thus his PNV appears in verse-initial position. Yet the PNV of Agamemnon is almost always in verse-terminal position (12 of 13x), and this, too, serves to contrast the king of Mycenae with the true protagonist, Achilles. It is true that Homer plays Achilles off against Agamemnon and Hector in different contexts, but do both metrical situations reflect that contrast? There is the danger of K. trying to have it both ways.

Chapter 5, "Patterns of the Proper-Name Nominative," (abbreviated PNN) looks at the localization of the nominative case of the proper names of heroes. While building on Parry's work of noun-epithet formulae, K. attempts to move further by asking what possible effects repetition of PNN's have at particular spots in the verse. Again K. maintains that placement in the verse corresponds to narrative role within the story. The tendency is that PNN's of major heroes appear at the end of the verse. For Achilles, Odysseus, Athene, Apollo, Agamemnon, Diomedes, and Penelope, the PNN's occur between 93-98% of the time in verse-terminal position; while the PNN's of Nestor, Aias, Hector, Hera, Ares appear in verse-terminal position only 25-61% of the time. How does K. interpret this significant disparity in frequencies? In part by pointing to formulaic composition, but K. believes this overall tendency to localize is not to be explained in simple metrical terms. Thus there is a mainstream pattern with (as K. acknowledges) a significant element of variation; overall the effect is not as pronounced as PNV usage (PNN's of Odysseus and Achilles in verse-terminal position are ~73 and 78%). Yet these repetitions come to have a meaning, namely that positioning PNN's in the verse-terminal position marks the character as heroic.

Because the material is so vast, K. begins with a close examination of each of the PNN's in Odyssey bk. 5. K. then looks beyond this book at the PNN's of Penelope, Telemachus, and Patroclus, among others. We find the same strengths and weaknesses of the previous chapter. Interesting is K.'s discussion of Eos, Artemis, and Demeter (Od. 5.121-7), whose PNN's are in verse-terminal position (marking them as heroic): "Each is the protagonist in a tale of divine-mortal liaison which is parallel to that of the central narrative of book 5. Hence each is significant as a 'heroine.'" (124) K.'s remarks regarding Calypso are intriguing (7 of her 18 PNN's are in verse-terminal position). K. reads three of the instances of verse-terminal positioning as being sensitive to the context: the poet underlines her role as a threat to Odysseus (5.263, 321, and 372 each contain the clothes which almost drown him linked with Calypso's PNN in verse-terminal position).

At times, the reasoning appears dubious. Telemachus's PNN (128x throughout the epic) is never verse-terminal. "The effect ... is to deny Telemachus the nominative markers of a 'hero.'" (135) Yet this interpretation is again undermined by the fact that Telemachus's PNN is metrically unsuitable for terminal positioning. Hermes's PNN occurs 17 of 20x in a non-terminal position. So far fine, yet the PNN of argeiphontes (while not strictly a proper name) is always verse-terminal. K.'s comments show how far he is willing to push the evidence, for he sees that "its usage may be a compromise, allowing Hermes a place on the 'fringe' of the heroic circle, and particularly in Odyssey 5, where he plays a more prominent role and is given a more rounded characterization." (127) The PNN of Eos ("Dawn") is verse-terminal 41 of 50x. How is she heroic? K.'s answer: "Eos is the personification of heroic time." (134) This appeared to me to verge on silliness. Still this special pleading does not nullify the general drift of this chapter. His reading of Patroclus's PNN (occurring 44x -- never verse-terminal) is on the mark. The fact that Patroclus's PNN does not occupy terminal position reinforces that idea that Patroclus is not heroic. In fact, Patroclus performs a contrasting role often responding to the protagonist Achilles: "his 'listening mode' particularly endears him to the audience, who are also, by definition, listeners." (140)

Given the formulaic composition of the Homeric epics, what literary questions are still available? J.M. Foley, Immanent Art. From Structure to Meaning in Traditional Oral Epic (Bloomington 1991), has recently argued for a new type of oral poetics where phrases gain meaning by reference to the entire oral tradition (their "metonymic referentiality"), thus transcending the pure denotative meaning of the word or phrase as it appears in its specific situation. In The Interpretation of Order, K. argues that the context should be the epic itself. Heroes' names and epic themes take on a broader significance within a single work. Only in the Odyssey is verse-initial, acc. case andra significant, evoking the "man" Odysseus. This may not contradict Foley so much as shift the emphasis. What is clear is that K. also insists upon a type of literary criticism which allows for context-based readings. We may interpret phrases and words with close attention to the situation in which they appear -- this operation is not exclusive to 'literary' poetry.

In sum then what does K. mean by the "Interpretation of Order"? Order in its simplest sense may mean nothing more than a word appearing in the first or last position of a verse. K. argues that this order or positioning tells us something of literary significance. The idea of order of course operates at other levels, as K. distinguishes marked positions from the default mode, the epic protagonist from peripheral characters, and epic themes from secondary issues. K. has offered the best possible argument for this sort of reading. The argument moves logically step by step. The presentation is made clear by numbered paragraphs (1.1, 1.2, 2.1.1, 2.1.2, etc.). K. is sensitive, tactful, even polite: he apologizes for the technical sections, always pointing out the reason for such statistical surveying. (Most tables are in appendices.) The work is exceptionally error free -- I found only three typos -- most important for a work of this sort. K.'s command of the Homeric scholarship -- both technical and literary -- is most impressive, as he deliberately builds upon this work. K. is at his best in chapter 3 on thematic words. As he acknowledges, chapters 4 and 5 are designed to interpret tendencies which are flexible -- overall he may be right, yet at points he is undermined by a predetermined reading. Still the questions are exceedingly important -- to raise them in such an intelligent manner is a significant achievement. K. endeavors to bridge oral formulaic theory and traditional literary criticism. Ranging over both epics, K. examines key figures and major themes, while attempting to gauge subtle effects and nuance -- all with great sensitivity to the oral basis and performative aspects of Homeric epic. Indeed, his prevailing metaphor is that we are listening to the music of Homer (my nomination for a subtitle to this work -- one of the quotations introducing chapters cites Pater: "All Art constantly aspires towards the conditions of Music."). K.'s own subtitle, "The Poetics of Homeric Repetition," indicates that by studying the effects of repetition, we come to see that the question of structure and aesthetics does not necessitate an either/or answer. K. sees beauty and meaning in the order itself.