This chapter discusses the ways in which approaches and strategies derived from anthropology inform the study of Greek tragedy. ‘‘Anthropology’’ is not, of course, an unproblematic term: there are many anthropologies, and even the definition of what discourses are part of‘‘anthropology’’ is fluid; for example, the term may or may not be taken to include sociology, cultural history, and the study of ideologies.1 The approaches and strategies derived from anthropology that are deployed in the study of tragedy range from general principles based on insights that dictate certain methodological requirements to specific strategies derived from specific anthropologies - for example, the reconstruction of polarities and binary oppositions that, according to structuralists, articulate societies’ conceptual universe, and so, some have argued, form the skeleton of tragedies (e. g., Segal 1981, 186).
To begin with principles, the most crucial of the insights derived from anthropology that underpin general methodological principles is that tragedy should be considered (not only, but also) as an anthropological phenomenon, as the product of a particular culture, performed in a specific Athenian festival. For tragedy was an organic part of the culture and contexts that generated it and gave meanings to each of its different elements in the eyes of the members of the culture, the ancient audiences, who shared with the tragedians their cultural assumptions, their conceptual map. This approach to the study of Greek tragedy (as a systematic enterprise based on the perception of tragedy as an organic part of a system, and involving also the reconstruction of the ancient audiences’ conceptual map) was pioneered by Vernant and Vidal-Naquet (Vernant and Vidal-Naquet 1972, e. g., 9-10; 1986, e. g., 10-15).
One aspect of the Vernant/Vidal-Naquet approach was the investigation of the articulations and manipulations of rituals in the creation of tragic meanings (Zeitlin 1965, 463-508; Vidal-Naquet 1972a, 133-58; Vernant 1972a, 99-131; cf. Gould 1973, 74-103, esp. 85-90). Clearly, if, for example, a sacrificial ritual helped shape the ways in which the ancient audiences - who shared knowledge of that ritual with the tragedian - made sense of a tragic segment, not to take account of this ritual is to diverge considerably from the meanings those audiences constructed. In order to
Understand how the deployment and manipulation of rituals constructed meanings for the audiences who shared the tragedians’ ritual knowledge, it is necessary first to reconstruct as much as possible of that knowledge. Anthropology helps in such reconstructions, by offering both comparative material and also some insight into the various modalities that articulate rituals in other societies. Cross-cultural comparisons can function as eye-openers to show the culturally determined nature of interpretations that seem reasonable to us, so stopping us from making sense of rituals through our commonsense assumptions. Furthermore, cross-cultural comparisons suggest a range of possibilities for making sense of whatever element is being considered, thus allowing us to construct methodologies that take such possibilities into account. But care must be taken that such possibilities are not imposed (as they sometimes were in a less anthropologically sophisticated past) as explanatory models. Nor should they be deployed as explanatory models and subsequently tested, since this strategy, which presents itself as more sophisticated, is equally flawed; for, instead of attempting to minimize cultural determination, it places at the center of the investigation assumptions that are inevitably culturally determined (about the plausibility of the comparabilities between the ancient phenomenon and the anthropological comparandum), by making them into organizing centers around which the data are arranged, and so structuring the investigation through the assumptions that underlie the model allegedly being tested - a circular, self-validating procedure (Sourvinou-Inwood 1995, 301-2, esp. 413-14). Finally, at the most basic level, anthropology teaches us to recognize, for example, rituals of transition such as initiation rituals, and so helps us consider whether certain tragedies may be structured also with the help of manipulations of such rituals or of crystallizations associated with ritual material, such as the figure of the ‘‘ephebe,’’ the adolescent in transition to adulthood (see Vidal-Naquet 1972b, 159-80; 1986b, 135-36).
As a result of the explicit deployment or implicit influence of anthropological approaches, the place of religion in tragedy is no longer underplayed. Scholars have studied the articulation of tragedy through rituals (Easterling 1988, 87-109; cf. Friedrich 1996, 269-70), stressed the character of the dramatic performances as part of a festival (Goldhill 1990, 97-129), and discussed Dionysiac aspects of tragedy (Henrichs 1994-95, 57-58, 91-92 nn. 5-7). The importance of gender as an analytical category in the study of tragedy, and in classics in general, in the last few decades is primarily due to contemporary preoccupations, above all to the discourse of feminism. But (leaving aside semantic questions as to whether or not such discourses and preoccupations are part of‘‘anthropology’’) ‘‘gender’’ is an anthropological category, and its study in the ancient world has also been based on anthropological insights, including categories such as ‘‘the Other,’’ systems of binary oppositions in which male versus female are major articulating categories, and general cross-cultural comparisons (des Bouvrie 1990 is a specifically anthropological study of gender in Greek tragedy). The influence of anthropology, then, is all-pervasive in the study of tragedy, both as a result of explicit methodological choices and through the influence of studies that deploy principles and methods based on anthropological insights.
To return to general principles that dictate methodological requirements: the insight that perception and judgment are culturally determined - an insight derived not only from anthropology but also from other disciplines - further validates and strengthens the approach that demands that tragedies be studied in their full cultural context. As noted above, awareness of the culturally determined nature of perception and judgment has helped stop us from reading Greek tragedy implicitly through commonsense assumptions, since commonsense assumptions are culturally determined (Sourvinou-Inwood 1989a, 134-48; 1989b, 141-65).
Awareness of the culturally determined nature of perception and judgment has also, however, given rise to the opposite tendency, the more general intellectual tendency of postmodernism, to which contemporary anthropology has not been immune, and which has been pulling the study of tragedy in the opposite direction. While the anthropological approaches mentioned above have been pushing in the direction of studying tragedy as an embedded part of the culture that produced it, which involves systematic attempts to reconstruct the ancient realities and cultural assumptions that shaped the perceptual filters through which tragedies were written and made sense of by audiences who shared the tragedians’ filters, postmodernism has given false methodological respectability to readings that make no attempt to reconstruct ancient realities. Postmodernism claims that we can never know anything, let alone reconstruct the ancient realities; it has thus created a presumption that attempts to reconstruct ancient realities have no validity, and therefore it is naive to embark on such attempts - the smart choice being to replace them by the commentators’ feelings and personal experience explicitly set forth as such. Such at least is the theoretical postmodernist position; in practice, postmodernist critics do not always consistently present their readings as their own modern interpretations, the distinction is often blurred between these modern readings and the ancient meanings, and claims of ‘‘truth’’ are made, explicitly or implicitly - claims which in postmodernist eyes are somehow validated through being framed by postmodernist skepticism. For one of the masks under which such strategies are given false respectability is the tenet that there is ‘‘no right reading’’ - as though this absolved us of the duty to try to reconstruct, insofar as possible, the main parameters shaping the process of meaning creation by the ancient audiences in performance - the construction of complex polysemic and multivocal meanings, probably different meanings by different segments of the audience, but nevertheless meanings shaped by certain parameters determined by cultural assumptions that are different from our own. There can be no doubt that certain aspects of ancient culture, and so at least some of the parameters shaping the ancient audiences’ creations of meanings, are definitely knowable, and can be enlisted in the attempt to reconstruct the ancient readings of tragedies.
Anyone who does not accept that such is the case should be prepared to accept that it was possible for the fifth-century Athenian audience, or a part of that audience, to understand Medea’s chariot at the end of Euripides’ Medea as a disguised helicopter, and Medea as an ordinary criminal making a daring escape, taking advantage of her distraught victim’s traumatic confusion to tell him a lot of fantastic lies; or that Euripides’ Orestes was a tragedy about a technologically advanced extraterrestrial called Apollo, who fooled the naive Greeks of the heroic age into worshipping him as a god and doing terrible things because he told them to. Of course no critic would accept this, because it involves concepts alien to the Greeks. In other words, even postmodernists would implicitly rely on the notion that we can reconstruct some ancient meanings if we take account of the ancient assumptions. It is, then, impossible to deny that we can reconstruct some of the parameters determining the process of meaning creation by the ancient audiences, who shared the tragedians’ cultural assumptions and constructed meanings through perceptual filters shaped by those assumptions.
To minimize cultural determination we must attempt to reconstruct, as far as possible, these ancient filters, or at least reconstruct the important parameters that shaped them, and so the parameters within which the process of meaning creation by the ancient audiences took place. We may or may not be able to reconstruct the meanings created by those audiences, but we can certainly reconstruct many of the factors that blocked certain meanings and facilitated, even compelled, others. And this, I believe, should be our main aim. Though it may be legitimate, as part of a modern discourse, to reread ancient tragedies through the filters of modern assumptions (a project not unlike producing, for example, Othello in modern dress), in the case of Greek tragedies, which are articulated with the help of a conceptual map alien to the modern reader, this can only result in creating a reflection of modern assumptions, while the richness, complexities, and multivocalities of the ancient tragedies are obscured and lost.
I will illustrate my own version of the ‘‘anthropological’’ approach that studies tragedy as an organic part of fifth-century Athenian culture, a version that aims at reconstructing as much as possible the ways in which the fifth-century Athenian audience made sense of the tragedies in performance, through examples from two tragedies. These examples, I hope, will illustrate how the reconstruction of two types of assumptions helps us recover some of the meanings constructed by the ancient audiences (or at least the parameters shaping such constructions). The two types of assumptions on which I will be focusing here are, first, religious assumptions, namely rituals and religious representations, knowledge of which was shared by the tragedians and their contemporary audiences; and second, a set of assumptions that is not often considered in the reading of tragedies: the audiences’ perceptions of the relationship between their world and that of the tragedies. For it is important that we do not allow ourselves to perceive, implicitly and by default, the relationship between the world of the ancient audiences and that of the tragedies on the model of such relationships in our own world. It is necessary to define this relationship, or at least reconstruct the parameters that defined it, by considering the two aspects that (crosscultural comparisons suggest) define such relationships. First, does the world portrayed on the stage have any connection to the world of the audience? Is it part of it, or part of its past, or future? Were the two similar, or were they radically different, even alien? Second, were the two worlds kept totally separate during the performance? That is, was the world of the tragedy self-contained and insulated from the world of the audience, or did it penetrate it, and if it did, what form or forms did this penetration take?
The preferred choice for the setting of Greek tragedies was the heroic age (on this and the less common settings see Sourvinou-Inwood 2003, 15-25; also 25-66). For the fifth-century audience the relationship between their world and that of the heroic age was governed by two intertwined perspectives. On the one hand the heroic age was other, distanced from the present, a time when men could have direct contact with gods and were sometimes descended from gods, and when the most prominent and important individuals became the heroes of fifth-century cult. On the other hand the heroic age, and so the world of the tragedy, was also part of the present, in that it was a crucial part of the audience’s past in which important events shaping their contemporary world had taken place; furthermore, it in some ways resembled the present. The relationship between the two worlds was not stable and inert in the course of each tragedy. I have argued - first with reference to Sophocles’ Antigone, and most recently in great detail, and more generally (Sourvinou-Inwood 1989a, 134-48; 2003, passim, esp. 15-66; cf. Pelling 1997b, 217-18, 228-29, 233-34) - that the relationship between the world of the audience and that of the tragedy was not constant throughout each tragedy; rather, it was manipulated in the course of the performance through textual devices that operated in interaction with the assumptions the audience shared with the tragedian: ‘‘distancing devices,’’ which had the effect of distancing the action from the world of the fifth-century Athenian polis, sharply differentiating the two; and ‘‘zooming devices,’’ which had the effect of bringing the world of the play nearer, pushing the audience into relating their experiences and assumptions directly to the play.
One particular category of zoomings, as will become clear in the course of this discussion, is that in which the world of the audience is penetrated by the world of the play. The passages from the two tragedies that will provide the illustration of my reading methodology involve some especially strong zoomings that have important implications for the construction of meanings by the audiences.2 They illustrate two different types of penetration. In the first, the character of the world of the tragedy as part of a ritual performance taking place in the world of the audience is activated. In the second, the superficially ‘‘other’’ heroic world of the tragedy is brought very near, indeed is presented as the same as, the world of the audience, especially with respect to its religious realities.
I will illustrate the first type of zooming (and to a lesser extent also the second) with an example from Aeschylus’ Eumenides, which takes place in the heroic age. Athens is the locale for the segment I will briefly consider here. The ‘‘other,’’ distanced from the present, nature of the world of Eumenides is manifested, for example, in the fact that in the play mortals have direct contact with gods; but that world was also part of the present, in that important events that shaped their audiences’ world took place during that period - above all the foundation of the court of the Areopagus3 and of the cult of the Eumenides.
The focus of my brief discussion is a ritual, the singing of a hymn by the chorus of the Erinyes. This singing takes place in a framework already charged with ritual elements. After his arrival in Athens Orestes supplicates Athena’s statue, and then, after the Erinyes enter in pursuit, he describes one rite, the purification he underwent at Delphi, and then enacts another rite, another prayer to Athena (282-98). After that the chorus sings a binding-song intended to put Orestes in their power (331 = 344). Normal ritual hymns present a deity’s powers, functions, and honors; here the Erinyes are doing this themselves in their own hymn; here, among other things, the Erinyes sing a hymn to themselves, about themselves. At least, they do so in the world of the tragedy, but the situation is complex. For the chorus begins their dance-song with an emphatic statement of choral self-referentiality (307-11; see Henrichs 199495, 60-65; also Wilson and Taplin 1993, 174), and then draw attention to their identity as performers in choral dance through their words and their actions - for example, the fact that they are stamping their feet on the orchestra floor while cursing Orestes (cf. Henrichs 1994-95, 62-63, 64). This self-referentiality activated the persona of the chorus as a chorus in the present, and it resulted in the ritual’s being perceived by the audience as performed not only by the Erinyes in the world of the tragedy, but also in the here and now, by the chorus of Athenian men. This double nature of the song is correlative with the oddity of the Erinyes singing a hymn to themselves, about themselves, while normally such a song would be sung by human worshippers. For the audience would not have perceived this hymn as sung only by the Erinyes in the world of the tragedy; because of the activation, through choral self-referentiality, of the persona of the chorus as a chorus in the present, they would have perceived the hymn as being sung also by the chorus of Athenian men in the present. This perception is important, because in the world of the spectators the Erinyes were indeed worshipped, and this fact was inevitably activated for them through the chorus’s singing of this hymn.
The implications are significant. After the Erinyes were defeated they threatened to blight the land (780-87 = 810-17). This marked a grave danger for Athens in the world of the tragedy. But the audience was protected from feeling this danger as symbolically too close, and threatening their present, because choral self-referentiality had activated the persona of the chorus as a chorus in the present, which entailed that the audience would have perceived the hymn about the Erinyes to be sung also by the chorus of Athenian men in the here and now, which, in turn, entailed both an act of reverence to the Erinyes in the here and now, and the activation of the knowledge that in the world of the audience the Erinyes were indeed worshipped. So the threat of the Erinyes does not affect the world of the audience; it represents a danger in the past that has been overcome. Once more Athena has protected the city; in this case she eventually, after a sustained effort of persuasion (and a veiled threat at 826-29 that she may use Zeus’ thunderbolt), convinces the Erinyes not to harm Athens and promises that they will receive worship by the Athenians.
The new relationship between the Erinyes and Athens is immediately enacted. At 902 the Erinyes ask Athena what blessings they should invoke upon Athens, and she gives a list which includes blessings that correspond to the curses uttered in 780-87 (907-9; see Sommerstein 1989, ad loc); then she herself gives Athens the blessing of military victory. Then the Erinyes deliver those blessings in song (‘‘they sing a majestic hymn of benediction upon the city’’; Lloyd-Jones 1971, 92), interspersed with comments from Athena, who stresses their power to curse and bless, and admonishes the Athenians to give them proper honor. Athena’s comments would have zoomed the world of the tragedy to that of the audience. Her description of the powers of the Erinyes at 950-55 corresponded to the audience’s own perception of the power of these goddesses, whom they worshipped in their everyday life. Above all, the statement at 993-95, that if the Athenians honor the Erinyes their city will be righteous and glorious, would have zoomed them to their religious reality, in which they did honor and worship the Erinyes - and this understanding would have brought up the expectation of the fulfillment of this promise.
The tragedy ends with a ritual, a procession establishing the cult of the Eumenides in Athens, and including torches, sacrificial victims, and the priestess and other cult-servants of Athena Polias. This procession is not a representation of the actual ritual of the cult of the Semnai Theai, but a construct deploying a variety of ritual elements, including elements evoking the Panathenaea (Bowie 1993, 27-28, 30); nevertheless, it would have zoomed the world of the tragedy to the cultic reality of the audience. For in the perceptions of the audience the distance between the cult they practiced and this heroic-age rite, in which both the deities being honored and the poliad deity took part, would have been correlative with, and so (in the process of meaning creation, implicitly would have) ‘‘accounted for,’’ the differences.
To sum up. This example from Aeschylus’ Eumenides illustrates the permeability produced by the activation of the audience’s perception of the tragedy as a performance in the present. This activation, and the zoomings to the audience’s religious realities, helped construct important meanings, not least the symbolic distance between the world of the audience and the Erinyes’ hostility to, and curses against, Athens. This symbolic distance allowed the action of the tragedy to develop as it did, and thus was critical to the process through which the audience constructed the complex and multivocal sets of meanings pertaining to the important (interconnected) themes of the tragedy and the trilogy: revenge, matricide, killing and pollution, hierarchies and balances of power among the gods, and above all the notion of justice, both divine justice and human institutions. Importantly, the activation of the audience’s awareness of the tragedy as a religious performance in the present and the zoomings to their religious realities inevitably charged the deities on stage with religious meanings and prevented the audience from perceiving them as mere theatrical devices.
Euripides’ Erechtheus offers some especially striking instances of the type of zooming through which the superficially ‘‘other’’ heroic world of the tragedy is brought very near, is presented as the same as, the world of the audience, especially its religious realities. Like Aeschylus’ Eumenides, this (fragmentary) tragedy is set in heroic-age Athens. It takes place during the reign of Erechtheus, when the Eleusinians, together with an army of Thracians led by Eumolpus, son of Poseidon, were threatening the city. Eumolpus wanted to replace Athena with his father Poseidon as the poliadic divinity of Athens. This threat is the context of fr. 351: ‘‘Raise the shrill cry, women, so that the goddess may come to help the polis wearing the golden aegis with the Gorgon.’’ My translation makes explicit the metonymic reference to Athena’s aegis. For the Athenians would have understood the expression ‘‘golden... Gorgon’’ through their own associations, which included their familiarity with Phidias’ statue in the Parthenon of Athena Parthenos wearing a golden aegis adorned with an ivory gorgoneion (e. g., Stewart 1990, 157-58). These verses, then, evoked the goddess Athena who was the poliad divinity watching over Athens in real life; the persona of Athena protecting the city was a pervasive theme in Athenian ideology and was embodied in the statue of Athena Parthenos. The metonymic reference to the aegis would have reinforced the notion of Athena’s protective power, for in myth the aegis powerfully warded off attack (cf. Iliad 21.400-401). Erechtheuswas probably produced at the end of the 420s bce or soon after (Collard, Cropp, and Lee 1995, 155). Only a few years earlier, during the Spartans’ annual invasions of Attica (Thucydides 3.2, 26, 89; 4.2, 6), the audience would themselves have invoked Athena’s help. This circumstance, and above all the fact that the function and iconography of Athena in the tragedy coincided with those of the goddess in Athenian cult, especially as represented in her most magnificent statue, would have zoomed the world of the tragedy very closely to that of the audience, whose past that tragedy was enacting.
To continue with the tragedy’s action, the oracle told Erechtheus that he would be victorious if he sacrificed his eldest daughter to Persephone. He did so, with the consent of his wife Praxithea, and two more daughters killed themselves because of an oath they had taken. The enemy was defeated and Eumolpus was killed by Erechtheus. Erechtheus in turn was killed by Poseidon. At fr. 370.55 Poseidon, enraged at the defeat and death of his son, sends an earthquake, and Athens trembles. At this point Athena appears. She first berates Poseidon and then turns to Praxithea, to whom she gives instructions to bury her daughters, adding that she, Athena, has made them into the deified Hyacinthids whose cult she is now instituting, and giving instructions about the sacrifices and other rites that should in future be performed in their honor. The expression Athena uses to describe the fate of Erechtheus’ daughters resembles closely, and thus would have recalled for the Athenian audience, the public epitaph for the men who died in the battle of Potidaea (Hansen 1983, 10.5). This would have zoomed the world of the tragedy to the world of the audience, presenting the Hyacinthids as the models for the heroization of the Athenian war dead.4
Athena proceeds to institute the cult of Poseidon Erechtheus. To be precise, she says that Erechtheus will take the name of his killer and be invoked in cult as Poseidon; this is one interpretation of the cult title ‘‘Poseidon Erechtheus.’’ She orders that a sanctuary be built for this cult in the middle of the city. Obviously, the cult itself would have zoomed the world of tragedy to the world of the audience, a zooming made stronger by the fact that at that very moment a sanctuary to Poseidon Erechtheus was under construction in the middle of the city. The Erechtheum housed, in addition to minor cults and sacred spots, the cults of Poseidon Erechtheus and of Athena Polias.5 The latter was the most important Athenian cult; it honored the very goddess who in the world of the play issues the instructions for building the Erechtheum, and who in her next words mentions her own cult; she makes Praxithea its first priestess, thus instituting the cult in the form in which it was practiced in the world of the audience. This provision zoomed the world of tragedy to the present and represented the present-day cult as symbolically anchored in the heroic world and in Athena’s will. To put it differently, this tragedy was portraying events that in the audience’s perceptions had taken place in their own past, very near the place in which they had occurred; it showed these events shaping the present as it is now, so that what happened on stage was de facto part of the audience’s contemporary context.
But, it may be objected, how can we be sure that Athena was not perceived by the ancient audiences as a simple theatrical device, an empty gesture of closure, as some modern commentators (e. g., Dunn 1996, 32-33) believe was generally the case with Euripides’ ‘‘gods on high’’? I criticize such theories in more detail elsewhere (Sourvinou-Inwood 2003, 414-22; see also 459-512), but the case against this objection, and in favor of the reading offered here, is clear even when set out briefly: by the time Athena appeared on stage the audience had already identified the goddess Athena of the tragedy as a representation of the goddess they worshipped, above all in the poliadic cult, thanks to a series of zoomings of the world of the tragedy to the world of the audience; therefore, when Athena appeared she would have been perceived as a representation of the real goddess Athena, impersonated by an actor.
Representations of deities by human actors were not unknown in Greek religion. There were three basic (and very different) types of divine appearance. The first type involved straightforward epiphany; the second, impersonation of a deity by priestly personnel; the third took place in the course of festivals of advent, in which the arrival of a deity is enacted through the arrival, or the ‘‘finding,’’ of the cult statue which had been removed from its usual place (Burkert 1988, 81-87; 1985, 134-35; 1997, 24).
‘‘Real-life’’ epiphanies (on which see Pfister 1924, 277-323; Nilsson 1974, 22527; Versnel 1990-93, 1: 190-93; Burkert 1985, 186-88; Graf 1997, 1150-51; Henrichs 1996b; Hornblower 1996, 356) took place when mortals believed that a deity had appeared to them. Though this was hardly a regular occurrence, deities were believed to appear to mortals occasionally, giving them instructions that very often resulted in the institution of a cult. These real-life epiphanies encompass two categories: dream epiphanies, in which the deity appeared when the person was asleep, and ‘‘real’’ real-life epiphanies, occurring when the person was awake. In fact, one of the tragedians was associated in the tradition with a dream epiphany that led to a cult foundation: Heracles was said to have appeared in a dream epiphany to Sophocles (Life of Sophocles 12), and informed him where to find a golden wreath that had been stolen from the Acropolis; when it was found Sophocles received a reward, which he used to found a shrine of Heracles Menutes (Heracles the Informer). Sometimes the deity or hero did not make himself known but performed a miracle (e. g., Herodotus 6.61), or some other action (e. g., Herodotus 6.69). The second type of divine appearance involved the impersonation of divinities by priestly personnel in the course of ritual, and sometimes sacred drama (Burkert 1985, 186; Mylonas 1961, 261-64; Clinton 1992, 84-95). Such ritual acts may have included an ‘‘enacted’’ epiphany (Hagg 1986, 46-47, 60-62; Kiechle 1970, 259-71; Burkert 1997, 27-28), in which the deity was impersonated by the priest or priestess.
In Erechtheus, by the time Athena appeared, dense zooming in earlier parts of the play had identified the goddess Athena of the tragedy as a representation - until then a conceptual representation - of the goddess they worshipped. Therefore the audience would have perceived the deity on the stage as a representation of the real goddess, impersonated by an actor, which would have activated the ritual schema ‘‘representation of a deity in sacred drama,’’ so that the audience made sense of the deity on stage also with the help of this ritual schema from their lived religion, especially since Athena in Erechtheus is presented as shaping the poliadic cult in its present form by appointing its first priestess, and ordering the foundation of, among others, another cult with which the poliadic cult is closely associated in the present. At the same time, Athena’s appearance would have activated for the audience the religious schema ‘‘real-life epiphany,’’ in which a deity appears, gives instructions, and a cult is instituted as a result. Thus, the audience would have seen Athena as a deity in epiphany to the tragic characters in the world of the tragedy, charged with religious significance through the activation of the ‘‘epiphany’’ schema, while at the same time this epiphany was also an impersonation of a deity by an actor, and thus would have evoked, and been perceived as partly comparable to, reenactments of cult foundations in sacred drama.
Because Erechtheus was located at the very center of the Athenian polis, it offers a very strong version of a penetration of the world of the audience by the world of the tragedy. In other tragedies the zooming may not be quite so strong, with zooming devices sometimes operating in terms of similarity, or closeness, to the world of the audience. Nevertheless, as I have argued elsewhere (Sourvinou-Inwood 2003, 301458 passim, 469-500), in those tragedies too the world of the tragedy penetrated that of the audience, so that what happened in the former was perceived as part of the latter. The deities that appeared on high at the conclusion of other Euripidean tragedies had also been zoomed to the audiences’ world and so had come to be perceived as representations of the deities in the audiences’ lived religion; they also evoked deities in ‘‘real-life’’ epiphanies, so that the audience saw them as deities in epiphany to the tragic characters, giving instructions, as a result of which a cult is instituted, which zoomed the world of the play to, and had direct relevance to, the audiences’ religious life. At the same time, correlatively with the penetration of the world of the tragedy into the audience’s own world, the ‘‘deity on high’’ would also have come to be seen as an actor wearing a divine mask acting the role of a god in a mimetic performance taking place in a sanctuary during a festival, which evoked the masked impersonations of deities by cult personnel in sacred drama and so charged the deities on high with religious significance.
Thus, when these divine appearances are considered in the full context of the performance, and of the culture that produced the tragedies, it becomes clear that the modern notion that Euripidean deities on high were simply empty gestures of closure is untenable, and that to the ancient audiences the Euripidean gods were far from simple theatrical devices.
It is clear, then, that, first, the ritual schemata and other religious knowledge which the tragedians shared with the ancient audiences, and second, the zoomings and, more generally, the shifting relationships between the world of the tragedy and that of the audience, constructed through the deployment of assumptions which the tragedians shared with the spectators, helped shape the parameters within which the fifth-century audiences constructed meanings in performance. The meanings they constructed were complex, multivocal, and possibly diverse, but such constructions were shaped and defined by the parameters created through these and the other assumptions shared by the tragedian and his contemporaries.
Among the meanings constructed during the performance of Erechtheus within the parameters shaped by the rituals and shifting relationships discussed above was, first, the exploration (at a symbolic distance) of the notion of human sacrifice, which was problematic in the religious discourse of the present, insofar as it inevitably helped color the perceptions of the gods worshipped by the fifth-century Greeks. Second, the Hyacinthids were represented as the models for the heroization of the Athenian war dead, confirming the association between the sacrifice of a virgin and the death of men in war, an association also enshrined in the cult which associates Aglauros, another virgin who sacrificed herself to save the city, with the ephebes (Kearns 1990, 330-31, 338; 1989, 139-40, 24-27; Larson 1995, 39-41). Finally, a strong symbolic anchoring is provided for the most important Athenian cult, that of Athena Polias, and of its associated cult of Poseidon Erechtheus; these cults were symbolically reinforced through the representation of their foundation in the heroic age on the orders of Athena herself.
I have argued elsewhere (Sourvinou-Inwood 2003, e. g., 513-18) that fifth-century audiences did not perceive tragedy only as a purely ‘‘theatrical’’ experience, a discrete dramatic unit, simply framed by ritual, but also as a ritual performance; and that aetiologies, deities, and other religious elements were not, for them, simply theatrical devices, but were charged with religious meanings; they were, in varying degrees and ways, part of the audiences’ religious realities. One of the arguments on which this conclusion was based pertains to the shifting relationships between the world of the audience and the world of the play in the different tragedies, relationships very different from those of modern theatrical performances, as has been illustrated here through the brief consideration of segments from two tragedies. Be that as it may, I hope that the two examples considered in this essay have illustrated the fact that, in order to make sense of the tragedies as nearly as possible to the ways in which the ancient audience did, it is necessary to reconstruct - among other assumptions - the ritual and other religious knowledge which the tragedians shared with their audiences, through the deployment of which meanings were constructed in performance, and also the shifting relationship between the world of the tragedy and that of the audience as perceived by the latter.
Therefore, these illustrations have also demonstrated the general methodological point that any non-anthropological approach, in the sense of any approach that does not study tragedy as an embedded part of fifth-century Athenian culture, and as a performance in the Theater of Dionysus during the Dionysia, inevitably produces meanings that are very different from those created by the ancient audiences. Our task, I have argued, is to try to make sense of the tragedies as nearly as possible to the ways the ancient audiences did.