Chapter 4. Mimesis as Performance

The reputation of the first chapter of Auerbach’s Mimesis among classicists has risen and fallen with the tides of fashion in Homeric studies. The notion of Homeric privileging of the part over the whole, argued for in the essay on Odysseus’s scar, gained wide currency at a time when scholars were concerned with parataxis and paratactic composition as the hallmark of Homer’s oral style. But the influence of Auerbach’s essay has diminished in recent years with the renewed interest in “literary” interpretations: we are now prepared to see more silence, more elliptic moments, and more “unplumbed depth” in Homeric poetry than Auerbach granted with his insistence on Homeric style as a “uniformly illuminated foreground.” [1]

Without questioning these developments, I propose to consider Auerbach’s chapter from the perspective of Homeric studies and in light of recent advances in the study of oral poetry. Fifty years after the first publication of Mimesis, Auerbach’s interpretative shortcomings are clear; but his account of Homeric style still contains a powerful core that, if reformulated, may continue to say something important about the flow and quality of Homeric poetry.

The central element in Auerbach’s discussion is his insistence on perception and illumination. Homeric poetry, according to Auerbach, is driven by a constant need for an “externalization of phenomena in terms perceptible to the senses.” [2] Moreover, Homeric narrative style “knows no background”; it fills the reader’s present entirely: “Like the separate phenomena themselves, their relationships … are brought to light in perfect fullness; so that a continuous rhythmic procession of phenomena passes by, and never is there a form left fragmentary or half-illuminated, never a lacuna, never a gap, never a glimpse of unplumbed depths. And this procession of phenomena takes place in the foreground—that is, in a local and temporal present which is absolute.” [3]

Vision and perception will also be central in my attempt to reappreciate Auerbach’s criticism. While essential to the study of the Homeric representation of reality, perception is at the same time a highly critical concept that contains all the tension inherent in the paradoxical fact that Homeric poetry is an oral tradition that has come down to us in the form of a written text. Our stance, faced with this peculiar situation, entails a crucial choice that has divided Homerists for decades: can or must we apply the standards of written literature to Homer, or does Homeric criticism require a different notional apparatus? In what follows, I will cast Auerbach’s interrelated notions of foreground and perception against the background of this dilemma. In the process, we will reconsider the very concept of mimesis and ask what it means for Homeric poetry to represent the reality on which its force depends.

Auerbach’s appreciation of Homeric poetry took shape in a context that was completely removed from the then-current trends in Homeric criticism. Nowhere in his first chapter does he mention “orality,” a concept that, by the time of Mimesis’s publication, was beginning to make its imprint on Homeric studies, having found its first, authoritative, formulation in the work of Milman Parry. [4] Studying the numerous formulaic repetitions in the Homeric poems, Parry had discovered an unmistakable regularity in the use of certain expressions, in particular the so-called noun-epithet formulas that designate the principal gods and heroes and give the language of the Iliad and the Odyssey, even in translation, its distinctive flavor (e.g. “swift-footed Achilles,” “many-minded Odysseus,” “owl-eyed Athena,” and “Zeus cloudgatherer”).

Parry was able to show that the use of these phrases obeys certain rules imposed by the structure of the verse. His first conclusion was that such a poetic “grammar” could not have been the personal style of any single person. Homeric style, Parry argued, was traditional, independent of its poet’s “free will.” In a second step, Parry came to view this more or less mechanical style as a response to the specific demands of oral poetry, which had to be composed “live” before an audience, by way of improvisation: “traditionality” had yielded to “orality.” Extensive fieldwork on the South Slavic living oral tradition provided rich comparative material for the study of Homer under this new critical paradigm.

Parry’s work would find an authoritative synthesis and extension in The Singer of Tales, written by his former student Albert Lord. [5] More importantly for my present purpose, some scholars were quick to pursue the aesthetic consequences of Parry’s findings. James Notopoulos (1949), in particular, did much to promote an “oral poetics,” arguing that Homer’s paratactic style and syntax should not be read as merely “primitive” and “archaic” (as opposed to the ideal of classical form expressed in the Parthenon at Athens and in Sophocles’ Oedipus the King); nor should it and other archaic Greek poetry be judged with respect to the “hypotactic” aesthetics of later literature (first formulated authoritatively by Plato in his Phaedrus). Instead, the privileging of the part over the whole, of fragmentation over integration, should be seen as simply a consequence of the oral nature of this kind of poetry, which had to be composed under circumstances quite different from those under which written poetry comes into being.

We may note that Notopoulos’s formulations often come strikingly close to Auerbach’s. Thus, where Auerbach calls the reality deployed in Homeric narrative “self-sufficient,” Notopoulos, discussing the propensity of Homeric episodes and scenes to become autonomous, remarks that in the Homeric similes “the poet digresses beyond the original point of comparison and finds delight in the similes per se which reveal Homer’s own world and nature.” [6] And to Auerbach’s “local and temporal present which is absolute” we might compare the following passage in Notopoulos: “The physical, technical, and psychological factors at work in the creation of oral poetry make the poet live largely in the moment and only secondarily in the larger framework of his material. When the poet composes by means of the formulaic diction, which Parry has shown in his studies, he must concentrate on the moment, on the immediate verse.” [7] A little later, Notopoulos explicitly locates the attention for detail for its own sake, not just in oral composition, but in the very medium of speech: “It is easier to digress and lose sight of the original purpose in the spoken word than in formal writing which follows an organized text with a beginning, middle, and an end.” [8]

Auerbach (1953:23) viewed the Iliad and the Odyssey (and the Old Testament) as “finished products,” a conception of their style that may seem literate and textual, especially when one considers his silence on their origins, or on their functioning in the contexts for which they have been originally designed. Yet the substance of his vision squares strikingly with the critical paradigm that sees in textlessness the essence of Homeric poetry, locating the aesthetics of the absolute present in the production of epic verse.

“Orality” in its original form necessarily implied a key role for the oral formula as the specific difference between oral and written composition, the theory’s prime axiom. Yet no two oralists could agree on the definition of the Homeric formula. On a narrow, precise definition of the concept, oral-formulaic theory could be seen by its opponents as the branch of Homeric studies mostly concerned with mechanics, structure, and production, as opposed to the literary branch, where meaning and poetic intent dominate. The Homeric poems, it was argued, were simply too good, and their overall narrative architecture too unmistakable, to be amenable to a simplistic analysis based on formulas and parataxis.

On the other hand, a broader, more flexible understanding of the formula allowed the orality hypothesis to be subsumed by an avowedly literate literary criticism: the oral background of the poems was simply taken for granted, without any real implications for criticism and interpretation. Neither definition gave much room for an independent oral poetics: In the widening split between the oralists and the “scripsists,” the concept came to be considered by the latter camp as mere terminology. In hailing parataxis as a virtue when it had previously counted as a vice, the oralists, it was thought, were paying undue attention to an unfortunate stylistic inconvenience and thus impeding the appreciation of the Homeric poems.

Though Auerbach’s ideas on Homer were conceived independently of the orality hypothesis in its original form, they shared in its decline. Yet, for all the similarity between Auerbach’s and Notopoulos’s formulations, one difference is clear: the importance of vision and perception in Auerbach’s conception. It is this aspect that comes to the fore in the revision of “orality” to which we now turn.

In the original conception of oral poetry, “performance” was the occasion when the poet puts the system of formulas into practice, a moment incompatible with textuality. [9] In recent years this emphasis on composition-in-performance has diminished. We now acknowledge that performance is not only, maybe not even primarily, an act of versification, but also an act of narration, a narrator addressing an audience. This storytelling event, moreover, is of central cultural importance because it is the occasion when the community allows a meaningful past to shape its present. Such a shift in emphasis is a kind of return to Parry’s original concept of tradition, which was being obliterated by the increasing emphasis on production and composition. [10]

From this new perspective, two aspects that have been neglected by oral-formulaic theory become increasingly important. First, the performance of Homer and other archaic poetry is not at all incompatible with writing and written texts. The acknowledgment of the existence of texts may lead, perhaps paradoxically, to a more “oral” understanding of Homeric discourse than the oral-formulaic approach has yielded. Seeing the original texts of the Iliad and the Odyssey as transcripts rather than as texts in the usual sense, the Homerist—and in general the student of any “early” work of literature—has to make the effort to thematize the notion of text, and study the poems in terms of the performed discourse of which their text is the transcription and for which it is the script. [11]

This may lead us to consider Homeric poetry not as oral poetry against a conceptual background of literate poetry, but as special speech against a background of speech, language as it functions in actual spoken use. In an earlier monograph (Bakker 1997a), I have studied Homeric poetry in this light, arguing that the conception of Homeric poetry as a reified, textual product should cede to an understanding of it as a process, a flow that must be seen in terms of speech and spoken language. We shall see below that the distinction between “product” and “process” is crucial for the discussion of perception and to Auerbach’s notion of foreground.

The second aspect of performance in its new understanding is that what makes a Homeric performance possible is not only the system of formulas described by Parry and his followers but also, from the standpoint of the performers and their audiences, the desire to recreate the heroic events of the past. In practice, this is listening to the “rumor” of the past, called kleos in Homer (Iliad II 486). Kleos encompasses the contribution of the present poet’s forebears to the tradition as well as his own, in the form of the memory of the previous performance. It is here that mimesis, the representation of reality, comes into play. Auerbach uses the term to designate the relation between a text and its referent. Yet when “performance” rather than “text” becomes the primary reality of the Greek epic tradition we may wonder what the consequences are for the concept of mimesis. In other words, what does it mean for Homeric poetry to represent reality when its own primary reality is an epic performance, which is modeled on a previous performance?

The discussion of mimesis in connection with performance, in fact, is inevitable insofar as the very word μίμησις in Ancient Greek denotes an action, a performance. The substantive μίμησις is an action noun formed from the verb μιμεῖσθαι ‘to represent’, ‘imitate’. Moreover, this verb is in the middle voice, which means that its grammatical subject is necessarily affected by the action denoted: μιμεῖσθαι is what people do, not what things are. Thus μίμησις originally does not denote a relation between a text (as finished product, see above) and its referent, but between an action (i.e. a process) and its model. [12] This “performative” sense of μίμησις is prominent in the third book of Plato’s Republic (393B–D), where the word seems to be used for the first time as a theoretical term. Μίμησις is used here for the notion of impersonation, “becoming another,” and is applied to the passages in Homer that we would characterize as “direct speech” or “character’s speech.” In these passages, according to Plato, the poet, Homer, becomes his character, assimilates himself to him or her—a scenario rejected by Plato for reasons that do not concern us here.

Μίμησις, however, does not become essential to our present purpose until we widen its scope and apply it, not only to the cases in which the performer plays the role of, say, Achilles or Odysseus, but also to the performance as such. In other words, we may use μίμησις to characterize the role of the poet, of Homer himself. [13] In this use of the term, recently propagated by Gregory Nagy, the very performance of Homeric poetry becomes an act of representation: what is represented is not only the reality of the epic tale, but also its very telling. [14] The mimesis of Homer, then, is both subjective and objective: the “divine poet” (Odyssey viii 43) reenacts the prestigious past, but is in turn reenacted himself; he becomes the model for all the performers of the future (called “rhapsodes” in the Greek context), and so is present as long as the epic tradition is alive.

We observe, then, that in its prehistory as a literary term, mimesis denotes not the properties of a textual artifact but a certain type of poetic action: the representation, re-presentation, of the previous performance, with the performer playing the role of Homer the quintessential narrator. With this new understanding of mimesis, let us now return to Auerbach’s ideas on foregrounding and realism. If Homeric poetry is primarily concerned with the dynamic business of reenacting its own previous occurrences, then what are we to do with its compulsory preoccupation with concrete, externalized detail as noted by Auerbach? First, let us examine an example of such foregrounded narrative. I have chosen a passage from the digression on Odysseus’ scar: during a hunt, young Odysseus receives the wound whose eventual scar will later be the title of Auerbach’s chapter:

This is all vivid, observable detail, in a style that ancient critics referred to as enargeia: the “graphic vividness” attributed to poetry that strives at achieving visual effects equivalent to a picture. [15] Vision, in fact, is the privileged sense in Homer in spite of its dependence on hearing (kleos); the poetics of the Homeric tradition is an optic poetics. The Homeric narrator typically adopts the stance of an eyewitness to the events he describes, [16] sometimes even staging his audience as spectators of the scene (e.g. Iliad V 85–86: “You couldn’t have told among whom Diomedes was fighting, on the side of the Trojans or among the Achaeans”). Such visual quality has been documented for other epic traditions as well. [17]

The optic quality of Homeric epic, however, is much more than a mere pretense of its performers; it is the dimension in which poetry transcends human limitations. In a famous passage, Homer contrasts the typically human medium of kleos, which is based on hearing, with the divine faculty of seeing:

The poet is privileged among mortals in having access to the divine vision of the Muses. These divinities have in later times been “reanalyzed” as symbols of poetic inspiration, but in the context of Homeric performance it seems preferable to see them in their more primordial, though no less poetic, role as agents, and warrants, of epic evidence. They were “out there,” overlooking the battlefields of the Trojan War, and witnessing the enactment of the epic events. It is to them that Homer, the proto-poet of the tradition, owes the precision and detail with which the epic events must be reenacted. And this same detail allows the rhapsodes/performers to reactivate the epic reality now, in the context of the performance. [18]

In Greek mythology, the Muses are the daughters of Zeus and Mnemosyne or Remembrance, a lineage that reveals an important aspect of Homeric poetics. In order to be represented in a new performance, the epic tale must be recalled, remembered, and it appears that the Greek epic tradition has grounded this fact in a mythical genealogy. The Muses make the poet remember, and their association with seeing, in the optic poetics of the Greek epic tradition, is particularly important in light of recent research in experimental and cognitive psychology that stresses the relation between memory and spatial and visual organization. [19]

In a recent study, the cognitive psychologist David Rubin has emphasized the importance of imagery in the stability of oral traditions. Imagery, as Rubin notes, is “one of our most powerful mnemonic aids,” [20] based on the principle that what is easiest to picture is easiest to recall. Images, or spatial relations between items, are easier to retain in memory than verbal, linear information. This fact can be experimentally proved, but is also confirmed by the actual experience of such storytellers as this Scottish bard:

The testimony of this Scottish bard is quite relevant for the Homerist. Not only does he conceive of storytelling as a picture, he also speaks of movement: The pictures come to him as he goes, as if telling the story is a kind of mental walk or hike. Homerists will point out here that the technical term for “song” in the Homeric epic is οἴμη or “song-path.” [22] It appears that this term is not just a Greek metaphor, but an experiential fact pertaining to human memory in general.

There seems to be a strong connection, then, between memory and vision, and this is probably why oral traditions all over the world are characterized by vivid, concrete, visual detail. If something is to be easy to remember, it must be easy to image and visualize. This aspect appears to be relatively neglected in the theory of oral composition and performance of Parry and Lord, with its insistence on formulaic repetitions. Some later writers on orality and literacy in Ancient Greece, however, did recognize the importance of imagery and visualization. Eric Havelock, for example, has said on the supplication scene with which the Iliad begins that “the priest does not come to promise a ransom; he carries it in his hands, and in his hand also is a golden staff with the badge of office on it. The attributes, unessential to the main story, evoke a visualization of the scene and its actors.” [23]

It is not difficult to find parallel statements in Auerbach’s essay. As with orality, Auerbach is silent on memory and performance, but that makes the convergence between his observations on the Homeric style and Havelock’s formulations all the more striking. In fact, Auerbach’s formulations may capture the essence of Homeric narrative even better than Havelock’s. In particular, we note again the notion of detail for detail’s sake:

The contention made in the last sentence of this extract is highly questionable and would be rejected by most modern Homerists. What interests me here, however, is the phrase that I have emphasized, the statement that the reality deployed in the Homeric poems is there for its own sake. Auerbach’s observation that Homeric reality is self-sufficient is powerful already, but in light of the present discussion we can go a step further. Given the function of visualization that I just reviewed, we have to say that Homeric realism is there for the sake of the recall of the story and thus for the very survival of the epic tradition: its transmission or re-performance. The “externalization of phenomena” that Auerbach so eloquently described is not a stylistic choice on the part of an author, but a phenomenon intimately connected with the functioning of the epic medium of performance.

At this point one might object that visual detail is, in and of itself, not at all confined to narrative that is to be remembered and re-presented. To stay within Greek literature and to anticipate the last chapter of this book, we may mention here famous passages in Thucydides’ History of the Peloponnesian War, such as the account of the great naval battle between the Athenian and the Syracusan fleet in the latter’s harbor:

This passage, which will concern us again and in more detail in Chapter Nine below, is intensely visual, even containing observers who are watching the spectacle from the shore, a mise en abîme that stages the reader as spectator of the historical events as they are unfolding, and draws the reader into the reality represented. This mode of narrative blurs the usual distinction between “narrative” (involving temporal progression) and “description”: What happens in time is stripped of its temporal quality, and presented as a complex tableau whose primary dimension is spatial. [25]

Yet narrative description like that in Thucydides’ passage occurs in Homer as well; we already noted the eyewitness quality of Homeric narrative and saw an example of it in the description of the boar hunt. What, then, is the difference between the two narratives? Let us note first that Thucydides’ prose is not easy to memorize nor even meant to be recalled. In antiquity its difficulty was already notorious. Homer’s poetry, by contrast, is not only formulaic but also very rhythmic. This difference is obviously important, since rhyme and rhythm are powerful cognitive constraints on memorization and recall. [26] But in the present context of imagery and visualization there is more. The formulas and rhythmic patterns that are the hallmark of the Homeric style are not only the expression of the images in the poet’s (performer’s) mind, with features that facilitate their memorization; they are also indicative of the very manner in which the image is perceived, of how the poet mentally scans the picture of which the scene consists. The flow of Homeric discourse, in other words, provides evidence of cognitive processes, and this feature is entirely absent in the visualizing prose discourse of Thucydides. Put differently, the prose narrator fictionalizes the act of perception, in order to represent the reality described, whereas presentation on the part of the epic poet is by itself an act of visualization.

To clarify this differentiation, we may turn once more to the cognitive psychologists. Guy Buswell studied the way people look at pictures and found that there are two patterns of perception: “One of these consists of a general survey in which the eye moves with a series of relatively short pauses over the main portions of the picture. A second type of pattern was observed in which a series of fixations, usually longer in duration, are concentrated over small areas of the picture, evidencing detailed examination of those sections. While many exceptions occur, it is apparent that the survey type of perception generally characterizes the early part of an examination of a picture, whereas the more detailed study, when it occurs, usually appears later.” [27] In other words, people do not take in pictures as wholes; they typically orient themselves, and then start focusing on the picture’s details one at a time. Apparently human vision is not capable of focusing on more than one detail at any single moment.

This observation has been used, among other things, by the linguist Wallace Chafe to account for the specific nature of speech. [28] Studying the physical properties of taped speech data, Chafe notes that spoken discourse typically proceeds, not in an uninterrupted flow, but in short spurts or jerks, which he calls intonation units, drawing attention to their physical, intonational, properties in the flow of speech (see also Chapter Three). Chafe argues that consciousness’s limited capacity is responsible for this more or less fragmented flow: human consciousness is incapable of having more than a very limited amount of “information” in sharp focus at any one time. A speaking consciousness will verbalize its “focus” in a short phrase that is intonationally marked, and often preceded by a brief pause during which the information is “activated” in the speaker’s consciousness. Consciousness thus imposes constraints on information flow and discourse when experience (be it actual perception or introspective vision) is turned into language. [29]

Vision and consciousness are, in fact, very similar, as noted by Chafe. [30] Both consist of a limited selection, that of which one is conscious, or the visual experience that is in sharp focus. And this focus of consciousness or of vision is constantly moving, on its way to the next focus, the next moment of experience. It appears, then, that the workings of human consciousness provide a denominator common to the acts of looking at a picture and of verbalizing a scene visualized. The way in which people describe pictures from memory (visualizing them before the mind’s eye, and turning the spatial information of the image into linear, sequential speech) strongly resembles the way in which they examine the picture when they actually perceive it.

The typical nature of vision and consciousness has important consequences for memory: These cognitive faculties will act as requirements for memorization and recall. In other words, a discourse that is to be remembered and re-presented will function better if it proceeds in accordance with the flow of consciousness. This, I propose, is the raison d’être of formulas in oral traditions. Moving beyond Parry’s analysis, we can view formulas as the speech units of the special speech of the oral tradition, the stylization of the intonation units of ordinary speech. Many poetic traditions have a verse that coincides in length with those speech units. [31]

In Homer, as I have shown elsewhere, [32] the units, often easily recognizable as syntactic and semantic atomic wholes, tend to coincide with the half lines of Homeric verse, before and after the middle caesura. Homeric discourse thus appears to proceed in little blocks of information or parcels of experience that match the working memory of human consciousness. As part of narrative description, these units of discourse represent not so much details of the image itself as the way in which visual information is processed by a perceiving or visualizing consciousness. Let us now re-present the Homeric passage quoted above, presenting each Homeric speech unit as a separate line:

We see that the construction is paratactic, with no inherent syntactic end point other than the completion of the image in question; the Greek particle de that has been translated as “and” or “but” may seem to function as a coordinating conjunction in our text but in the speech represented it serves, instead, to mark each unit as a new step in the progression of the discourse. [33] The description is preceded, as often in Homer and in accordance with the Buswell principle presented above, by a general “label” of the scene as a whole: the juxtaposition of the hunters and the boar (units 1.–4.). And in the actual description, each line represents a single detail that is often visual, a focus of vision and of consciousness, shifting with each moment and each new verbalization.

Let us now return to Auerbach’s account of the Homeric style. As we saw earlier, Auerbach considers “foreground” and lack of background to be general properties of Homeric style. We can now both appreciate the perceptiveness of this characterization and determine its limitations. Auerbach studies Homeric narrative as text and assigns to it the features of a two-dimensional picture full of detail but without perspective or background. This is mimesis in a “static” sense, a relation between a text as finished product and the reality to which it refers. Yet Auerbach’s conception of foreground as “a local and temporal present which is absolute” is enriched on the more dynamic understanding of Homeric poetry as speech. When we perceive pictures and verbalize that perception, what we focus on at any one moment is by definition “foreground”: the detail is in sharp focus at a particular moment: that of perception of speech, the present. [34] And as such it is foreground, even when it is background in terms of a static picture. In other words, in reading Homer, we witness a process, which Auerbach characterized adequately, but in terms of an inappropriate medium: as the style of the written text rather than the action of speech.

So, finally, what does it mean for Homeric poetry as performance to represent reality? To Auerbach’s static sense of mimesis, the referential quality of words and text, I opposed a more dynamic understanding of that concept, the relation between the poet’s performance and the earlier occurrences of the same discourse. The poet’s “seeing” of reality is a remembrance of earlier poets’—and ultimately the Muses’—seeing of that same reality. The vision of the Muses was detailed and specific, which in less mystic terms amounts to emphasizing the importance of imagery for the Homeric performance as a cognitive act of re-creation.