ONE

READING BEYOND AESTHETICS

REPETITION AND THE INTERPRETATION OF IMAGERY

In 1977 Robertson published an Athenian pelike acquired a few years earlier by the Boston Museum of Fine Arts¹ [Figures 1.1. and 1.2.]. He attributed it to an important painter, a very young Euthymides, in a style close to an experienced Euphronios.² These two artists were prominent members of a group of technical and stylistic innovators in late sixth-century Athens, and the appearance of a new vase from such distinguished origins is reason enough to pay close attention to it. The images and their mutual resonance make it even more compelling. On side A, two nude youths jump in midair, with fully frontal torsos, knees bent; arms are held out with elbows parallel to the shoulders, fists clenched at about waist level. Heads are turned sharply to the right. A third youth stands in profile to the left, wrapped in a himation, lifting the double pipes as he plays for his dancing companions. On side B a nearly identical youth also pipes, though here with his instrument nearly parallel to the ground. He, too, plays for two nude youths jumping in midair, with knees bent; they clench their fists. Legs are in profile. Two differences stand out: these youths show the viewer the left profiles of their faces and their backs, with arms nearly straight.

1.1. Red-figure pelike by Euthymides and Euphronios, ca. 510–500 B.C. Side A. Boston, Museum of Fine Arts 1973.88. (Youths jumping. Place of Manufacture: Greece, Attica, Athens. Height: 31.1 cm (12 1/4 in.); diameter (max.): 24.4 cm (9 5/8 in.) Robert J. Edwards Fund. Photograph © 2000. Museum of Fine Arts, Boston.)

1.2. Red-figure pelike by Euthymides and Euphronios, ca. 510–500 B.C. Side B. Boston, Museum of Fine Arts 1973.88. (Youths jumping. Place of Manufacture: Greece, Attica, Athens. Height: 31.1 cm (12 1/4 in.); diameter (max.): 24.4 cm (9 5/8 in.) Robert J. Edwards Fund. Photograph © 2000. Museum of Fine Arts, Boston.)

   In combination the two compositions appear to be intentional reverses of one another. The painter acknowledges that one viewer of an actual performance would see the back view of dancers whose fronts were visible to a colleague directly opposite. The artist extrapolates from this imagined reality to suggest that the viewer of one side of the vase will likewise see the reverse of what his companion sees. Of course the pelike is not transparent, so the pairing of reversed images is playful, a mind game.

   This short interpretation requires us to believe that the painter of the pelike expected a viewer to take the two sides of the vessel into account in reacting to it. What evidence is there, other than a hunch played out in a plausible explanation, that this was his expectation? A look at the inscriptions provides corroboration.

   On side A with the frontal jumpers are two proper names: ΑΙΝΕΑ<Σ>, “Aineas,” in front of the piper and ΚΑΛ(Λ)ΙΠ(Π)ΙΔΕΣ, “Kallippides,” above the heads of the right-hand jumper and the piper. Below the jumping figures is written ΛΕΑΓΡΟΣ ΚΑΛΟΣ, “Leagros is handsome.” On the reverse, a third name is written retrograde, or left to right, in front of the piper, (Σ)ΜΙΚΥΘΙΟ(Ν), “Smikythion.” Finally, also retrograde, is another statement that reads ΛΕΑΓΡΟΣΚΑΛΟΣΝΑΙΧΙ, “Leagros IS handsome, yes, indeed!”³ located between the piper and the right-hand jumper, curving around his knees. Like the images, the kalos inscriptions are created in response to one another: one is a statement, “Leagros is handsome,” and the other is a response, in agreement, “Leagros IS handsome, yes indeed.” The responsive relationship is made visual by the coupling of orthograde (statement, written left to right) and retrograde (response, written right to left), a verbal conjunction parallel to the combination of frontal and rear views of the jumping figures. The final affirmation, “yes, indeed!” derives its justification from its relationship to its companion statement on the other side of the vase. Although it is not absolutely clear which proper name labels what figure, there are a total of three names. If the figures on B are the same as the figures on A, the evidence from both sides provides the names of all three figures. In sum, the evidence from the inscriptions offers support for a hypothesis that the artist expected the viewer to take both sides of the pelike into account when assessing it. In fact, neglecting to take into account both sides of the vase diminishes the experience of it.

   After a recognition that slight differences in the inscriptions matter, a return to the coupled images of the jumpers provokes a round of questions: Why are the arms of the B-side jumpers straight? Why are the pipes held lower by the musician on B? Are these significant deviations, suggesting a slightly different moment in the performance, corresponding to the length of time it takes for one person to utter a comment and another to respond to it, or are they accidental?

   One way to describe the idiom adopted by the artist is “highly repetitive.” Not only are there the identical aspects of figure and inscription already noted, but the framing ornament of each scene is repeated precisely on A and B, with palmettes on the neck, a trapezoid of simple meander on top and sides, and a ground line of upright lotus buds on the belly. It is the slight deviations from exact repetition in both inscriptions and image that demand attention and create engagement with the vase. There is a tremendous payoff for observing exactly what is repeated and what is not.

   The present study starts from just this premise – noting repetition carefully yields a significant reward – so well supported by the Boston pelike. That Athenian vases are highly repetitive in shape and decorative scheme is obvious, but simply to acknowledge the existence of repetition and to note that it is fundamental to the aesthetics of Greek art is to overlook its interpretive value. Considering the conspicuousness of repetition on Athenian vases, it is very surprising that no one has undertaken a systematic study of its meaning. Virtually any discussion of the iconography of an Athenian vase must engage, either implicitly or explicitly, the issue of interpreting juxtaposed images, and when such images contain repetitions, this issue is particularly important.⁴ It is a fundamental premise of this work that understanding repetition is crucial to knowing how Athenian pottery conveyed meaning to its audience. This view is informed by what is known about the uses and purposes of repetition in all forms of communication.

   The special focus of this study is one category of ancient Greek pottery, black- and red-figure Athenian vases of the Archaic period. Most of the evidence discussed dates to about 600–480 B.C. This chronological period sees the first significant appearance of written inscriptions on vases, and analysis of repeated inscriptions is essential to understanding repeated imagery. Some evidence dates slightly earlier, some a bit later. Observable repetition in this category of vases exists on at least three levels:

1. Repetition of shapes together with repetition of schemes of decoration, including ornament and images themselves, resulting in classes or groups, called here “Types” of pottery.
2. Repetition in the imagery used to depict particular subjects, such as chariot-harnessing or Herakles fighting with the Nemean lion. Such repeated imagery, appearing on many examples, is commonly regarded as a visual “formula.”
3. Repetitions in imagery on a single visual “text,” a single vessel, either within a single field or in more than one field on a single vase.

To interpret better the meaning of the latter two types of repetition, and in particular the last, is the major goal here.

   Because repetition is an indispensable agent in virtually all forms of communication, analysis of repetition has evolved in many fields. Particularly informing this investigation are the interconnected areas of information theory, social anthropology, structural linguistics, and narratology. The remainder of this chapter is devoted to a brief account of earlier discussions of the meaning of repetition in Greek art, including on Athenian vases, and to a survey of theoretical explanations of how repetition aids transmission of information in a variety of media. These explanations serve as our charter for understanding the meaning of repetition on Athenian vases.

EARLIER STUDIES

Bronze Age Art

Morgan, in her investigation of Aegean Bronze Age art, notes that repetition functions as a linking device: “repetitions of gesture, attribute, and clothing of people establish relations between scenes, acting as a means of identifying the characters and their situations,” and she notes that the use of an unusual repeated gesture makes it likely that the same individuals are represented in two scenes.⁵ Beyond merely linking scenes, the visual repetitions create associations that “are utilized as themes which evoke an idea in the mind of the spectator.”⁶

   Morgan explains, for example, how repetition helps to express the theme of hierarchies of power. She first describes the appearance of the lion in the context of a hunt on each side of a Mycenaean dagger.⁷ She notes that the lion attacks its prey, the Mediterranean deer, on one side; on the other side, the lion is itself attacked by its adversary, the Mycenaean warrior. She believes that the reiteration is not to suggest sequence; rather, the effect of the repetition is to create a parallel between the strongest power in the animal world and man the hunter. Although she does not actually use the term “simile” or “syllogism,” she assumes that the artist directs the observer to make a comparison: lion is like man as the dominant power in his respective domain; or lion is to deer as man is to lion.

   In the Thera frescoes, Morgan sees repetition of even small details as crucial for creating coherence among different scenes. Her analysis, although stated briefly, is distinctive because its focus is so rare in scholarship on Greek art. She suggests that “repetitions of gesture, attribute, and clothing of people [the equivalent of lexical items as defined below] establish relations between scenes, acting as a means of identifying the characters and their situations.” In one case, she cites the repetition of an unusual gesture as making it likely that the same men are represented in two different scenes.⁸ Morgan’s method of interpretation essentially relies on semiotics and narratology, theoretical approaches described in more detail below.

   Marinatos analyzes art of the Aegean and Egypt to explore what she calls a “less well-explored ‘rhetoric’ ”;⁹ and she names the phenomenon observed by Morgan a “pictorial simile”: “A pictorial simile involves two dissimilar figures involved in a similar situation; the juxtaposition forces a comparison. Most common is the comparison between humans and animals. In such similes, the bravery or strength of the animal is compared to and thus transferred to the human.”

   Marinatos, like Morgan, discusses the inlaid daggers from Shaft Grave A at Mycenae. On these objects, human hunters are engraved on one side, lions on the other. At one point in her discussion, she relies on juxtaposition rather than visual repetition to make her point. Here because the scenes appear in the same position on opposite sides of the same dagger, she discusses the scenes as inviting comparison, and the human who hunts lions and the lion who hunts other prey are thereby implicitly equated with each other. At another point in her discussion, however, she does rely more on visual repetition as the basis for comparison: she points to the obvious similarities in the two compositions, as well as the repetition of vegetation motifs, to say that the similarities are reason enough to support the idea that the viewer will make a comparison, and that such a comparison was intended by the artist.¹⁰

Athenian Vases: Paradigm

Several scholars have discussed repetition in studies of post–Bronze Age art, and in particular of Athenian vases. Beazley noted the prevalence of repetition within this corpus. In lists of attributions, he often uses the phrase “the like,” or the word “each,” to indicate repetition of images on two fields of a single vase. Beazley was aware of repetition as a possible vehicle for connecting figures, especially divine or heroic figures and mortals: for example, in reference to a lekanis in Palermo, Beazley stated that “The quintet in the upper zone looks almost like a deheroized version of Paris, Hermes, and the Goddesses [in the lower zone].”¹¹ In the labeling of an amphora by the Princeton Painter, he said, “B, warriors leading women off (one would have said ‘Recovery of Helen,’ but the group is repeated).”¹² Beazley is, in these two cases, pointing out how an artist can, through repetition, vary the expected formula used to construct a scene, thereby both reminding the viewer of that formula and simultaneously eliminating it as the precise template for a particular iteration of the scene. Beazley observes a slightly different type of example on a plate fragment in Athens, “A youth leading three women to a youth...the judgment of Paris deheroized.’ ”¹³ In the last case he acknowledges that artists might create scenes with the intention of reminding viewers of a familiar formula (that does not necessarily appear on the same vase) for depicting a particular mythological event. At the same time, the artist, by omitting some features from the repetition, makes it clear that he does not represent the mythological event itself. The result is the identification of a heroic paradigm for mortal activity.

Athenian Vases: Narration

Whereas Beazley’s observation on repetition underscores the creation of paradigm, Froning discusses the role that repetition may play in early attempts to depict continuous narration.¹⁴ She discusses an amphora in Geneva by the Princeton Painter, c. 540, depicting the birth of Athena twice in two slightly, but crucially, different scenes.¹⁵ On one side, the goddess is emerging from Zeus’ head; on the other she has been born and stands on his lap. Froning maintains that although one is accustomed to speak of such pairs of images as seen on the amphora as simple “varied repetitions,” the combination of the two images on a single vase may well be for the purpose of representing two stages of a narrative. According to Froning, the artist extends the scene to two frames to stress the length of time the “actual” event takes.¹⁶ In other words, the artist extends “discourse-time” in order to suggest the length of “story-time.”¹⁷

   In another example, Froning uses methods similar to those used by Morgan to connect images painted on two sides of a vase. A black-figure cup in the Vatican depicts two very similar sets of dice-players on each side; Froning points out, focusing on small details, that different figures are depicted in each case.¹⁸ The players on the left differ with respect to details of dress and armor as well as hair length and facial hair; each right-hand player, although otherwise quite similar, has a different shield device. She concludes that the result is not a continuous narration of one story; rather two different sets of heroes are depicted on each side of the vase. However, Froning, unlike Beazley, does not propose that one purpose might be to create paradigm, that is, a situation where one sees that one set of figures is heroic or mythical and the other a mortal comparison to it.¹⁹

   In Froning’s example of the birth of Athena, the artist repeated the narrative center or “kernel” of a well-known story.²⁰ Scholars have also discussed repetitions that do not include the narrative kernel. Carpenter, for example, treats the phenomenon of repeated stock figures who observe scenes that are sometimes mythical narratives.²¹ Carpenter characterizes these figures as representing types of upper-class Greeks and suggests that in some cases they are compositional fillers only and have no real meaning; they simply fill space.²² Another view, explored here in Chapter 4, holds that the repetition of such figures is in fact meaningful, functioning in two different ways, both as signs serving to point out either similarities or differences between the scenes that they link and as indicating to whom the event is relevant and who ordains it as such.

EARLIER STUDIES IN HOMERIC EPIC

A recent current in literary analysis of ancient texts is informative because it also depends upon an understanding of repetition as producing meaning. In post–Milman Parry studies of Homer, many scholars in the twentieth century saw the frequent repetitions in epic diction as a meaningless artifact of oral composition. At least two studies present an alternative view, suggesting that repetition produces meaning as well, enhancing themes and aspects of character.²³ These examples are instructive.

   Lowenstam acknowledges that there is so much repetition in Homer it is difficult “to determine when, if ever, repeated elements become significant for the literary interpretation of the poems.”²⁴ He goes on to make the important point that “even if the use of a repeated detail can be ascribed to [composition by theme] we cannot preclude the possibility that the repetition also has literary significance.”²⁵ Here he is speaking against the position that repetition in the epic is too familiar to the audience for any particular instance to have been significant to them.²⁶ Lowenstam begins by pointing out how integral repetition is to mythical paradigms, similes, and digressions, where repetition is essential for such features to have any meaning at all in the poems.

   Admitting that noun–epithet combinations and themes are compositional building-blocks, Lowenstam explains that they are also vehicles to express the poet’s meaning even when particular instances might seem like errors, where scholars hypothesize that a phrase difficult to make sense of must have been inserted for metrical convenience.²⁷ He shows how in repetition of themes such as “the quarrel,” it is by comparison and contrast to other quarrels that the poet illuminates the main quarrel between Achilles and Agamemnon that drives the action in the poem. Lowenstam places repetition at the heart of how the audience understands what the poems are “about.” He uses the “mechanics of repetition” to argue “...that the Iliad and the Odyssey are based on repeated antitheses,” and that repetition simultaneously brings together analogous situations while highlighting their contrasting aspects.²⁸

   A second work on Homeric repetition, by Kahane, espouses a very similar view.²⁹ Like Lowenstam, Kahane maintains that there is meaning in repetition and, further, he attempts to understand how particular formal aspects of Homeric repetition express poetic themes.³⁰ His main purpose is “to expose diction which is capable of generating thematically significant ambiguities and shades of meaning.”³¹ At the heart of his argument is the idea that certain formal repetitions have semantic value. To provide a very simple and extreme example, Kahane points out that frequent repetition is an indication of a character’s importance. For example, Achilles is mentioned by name 367 times and Bienor once. Put simply, repetition of Achilles’ name reflects his greater importance in the poem.³²

   Through a very careful analysis, the author also reveals that repetition of certain words in particular metrical positions in the line has a role in conveying the poet’s meaning. One detailed analysis focuses on localization of proper names in the nominative. Heroes such as Odysseus or Agamemnon tend to have their names at the end of a line. Telemachos, when his proper name is in the nominative, never has terminal positioning. This emphasizes to the audience that he is different from Odysseus; Kahane believes that the difference includes the message that he is less important, not the main hero of the story. He concludes that repetition of a very particular sort and variation from it enhance the message the poet constructs about the characters.³³

   Kahane is saying that repetition creates a means for the poet to define character, to construct how the character has similarities to and differences from other characters, through the use of a particular pattern. In a way the process is the verbal equivalent of the visual formula, conventional in most archaic and classical art, which places the victor on the left when a struggle is depicted.³⁴ This is a visual formula that carries meaning outside of the specific figures that “fill” the formula; to place a figure at the left makes that figure, whoever he or she is, the victor. On the rare occasions where the figure who is “supposed” to have won is placed on the right – in a famous example Achilles rather than Penthesileia is on the right – the artist introduces questions of how victory is defined.³⁵

A NEW APPROACH

Many scholars, in pursuing the interpretation of either a certain image or a class of images, have suggested that repetition plays a role in constructing the meaning of a vase. Scholars likewise see verbal repetition not just as an artifact of composition, but as constructing meaning. The present study of repetition differs from these earlier studies because of its attempt to provide a theoretically grounded basis for such conclusions. Beginning with the recognition that repetition is an indispensable agent in virtually all communication, this discussion relies on the understanding of the purpose of repetition in the interconnected fields of information theory, social anthropology, structural linguistics, and narratology. What follow are brief summaries of the central ideas about repetition expressed in each of these fields.

Repetition and Information Theory

The principles of information theory can be applied to understand “any system in which a ‘message’ can be sent from one place to another.”³⁶ Such systems are diverse, including ordinary language, visual images, and even genes. Understanding a few key concepts of information theory makes clear its applicability to this investigation of repetition.

Entropy

“Entropy” is for information theorists “disorder.” A system of low entropy, with little disorder, produces a message; likewise, a system of high entropy has lots of disorder and instead of sending a message, it produces noise. A root beer float – scoops of vanilla ice cream in a glass of dark brown root beer – when first made, has relatively low entropy because it provides clear, unambiguous information and therefore “sends a message” about where the root beer and where the ice cream are. When the ice cream melts and the contents of the glass are stirred with a straw, the system has higher entropy, more noise, because the order has disappeared, providing less detail and therefore sending less information about the system as a whole. The system has become a unit of undifferentiated parts, a homogenized mixture of ice cream and root beer, and as such, it is harder to identify its components and harder to receive and understand the message of what it is.