Navigation – Plan du site
Études

Networks and Narratives: A Model for Ancient Greek Religion*

Esther Eidinow
p. 9-38

Résumés

La polis religion est devenue le modèle dominant pour décrire l’activité rituelle des communautés de la Grèce antique. En fait, les chercheurs ont invoqué cette notion pour tenter de résoudre la question très débattue de la définition de la « magie » en regard de la « religion », en assumant que les pratiques magiques et leurs acteurs n’appartenaient pas à la religion collective de la polis. Néanmois, la relation entre polis religion et pratique « magique » comme l’écriture de defixiones, est certainement plus complexe, tout autant que d’autres activités cultuelles, comme celles qui relèvent du culte de Dionysos, et ce de diverses manières. Un examen approfondi montre qu’il devient difficile de définir ce que cela suppose, pour une activité rituelle, d’être intégrée au schéma de la polis religion quand on analyse la variété des organisations cultuelles et les différents niveaux et types d’implication de la polis. Cet article montre que la social network theory permet de dépasser ces difficultés conceptuelles. Une telle approche offre une construction alternative, plus fluide, de la religion grecque antique, en permettant de prendre en compte des réseaux d’activités rituelles qui coexistent et se recouvrent parfois.

Haut de page

Texte intégral

Introduction: Thrasyllos, an Example

1The subject of Isocrates’ speech, the Aeginetikos, is the rightful inheritance of a large sum of money. The son of a man called Thrasyllos, a Siphnian from the Cyclades, is defending his inheritance from the claims of a half-sister. In the course of this speech, we learn about the rags-to-riches life story of Thrasyllos himself:

  • *  An initial version of this paper was given at a conference in honour of Christiane Sourvinou-Inwoo (...)
  • 1  Isokrates, 19 (Aiginetikos), 5-7 (trans. Norlin [1980]).

Thrasyllos, the father of the testator, had inherited nothing from his parents; but having become the guest-friend of Polemaenetos, the soothsayer, he became so intimate with him that Polemaenetos at his death left to him his books on divination and gave him a portion of the property which is now in question. [6] Thrasyllos, with these books as his capital, practiced the art of divination. He became an itinerant soothsayer, lived in many cities, and was intimate with several women, some of whom had children whom he never even recognized as legitimate, and, in particular, during this period he lived with the mother of the complainant. [7] When he had acquired a large fortune and yearned for his fatherland, he left this woman and the others as well, and disembarking at Siphnos married a sister of my father.1

  • 2  See Flower (2008) for an excellent overview of the ancient Greek seer.
  • 3  For example, Meyer – Mirecki (1995), esp. the summary on p. 2. On how scholars have used the term, (...)

2Thrasyllos made his money on the road. He inherited some soothsaying books and practiced this art across many cities, before returning to the island of Siphnos, where he finally settled down. The evidence suggests that Thrasyllos was not remarkable in his lifestyle, at least not in the combination of his profession and his wanderlust. From a variety of ancient sources we glimpse a world of travelling ritual specialists, who base their expertise on various different claims—of training, family inheritance, geographical location, divine favour.2 Perhaps it is because of their number and variety that scholars still find it difficult to decide how we should categorise, or even how we should think about these ritual specialists. Attempts have largely focused on the anomalous nature of the practices rather than the people: how do these activities fit into, or alongside, our concept of ancient religion? And from there we are quickly sucked into the labyrinthine problem of defining ‘magic’ itself, with its many twists, turns, even dead ends—including the question of whether the term should be used at all.3

3In this paper I take a different approach: I use evidence for these independent practitioners and their practices, but instead of problematising ancient Greek magic and its definition, my focus is on our current understanding of ancient Greek religion—specifically the concept of polis religion, as outlined in two articles ‘What is Polis Religion’ and ‘Further Aspects of Polis Religion,’ by Christiane Sourvinou-Inwood.

  • 4  A brief and very clear description of a network can be found in the introduction to Constantakopou (...)

4My argument is that while polis religion offers a useful schema for understanding some aspects of ancient Greek religious activity, it cannot provide a comprehensive account of ritual practice across and within ancient Greek communities. If we assume that it does, then we are left with a range of ancient actors and activities that appear anomalous, ‘loose ends’ for which other categories (such as ‘magic’ or ‘marginal’) have to be found. Instead, I propose that rather than conceptualising ancient Greek religion in terms of a single entity, the polis, we might better picture it in terms of a social network, that is, at its simplest as sets of nodes (representing individuals or groups), linked by ties, usually multiple ties (representing relationships of various different sorts).4 Among those relationships are likely to have been many that occurred within or because of polis-centred ritual activity, but, importantly, not all of them did so. These ties or relationships might co-exist, develop or change over time; moreover, these ties may also have helped those involved to create meaning. To illustrate these ideas, this paper offers two case studies which provide different perspectives on ancient Greek religion: the first, the creation and use of binding spells, is usually described as a practice of individuals; while the second focuses on a group activity—cults relating to the worship of Dionysos. This paper explores how, in each case, these two ritual practices and their practitioners challenge the framework of polis religion, and how an approach that uses social network theory might offer a more nuanced perspective on their relationship to the polis, and vice versa.

  • 5  Horden – Purcell (2000); Malkin (2003); for 2006 conference see above, n. 4.
  • 6  Collar (2009).
  • 7  Collar (2009), p. 151. She poses the question in terms of ‘“vulnerability” to a new faith’ and exa (...)

5The idea that ancient Greek religion might better be described in terms of a dynamic network seems timely for a number of different reasons. First, although the idea of Mediterranean interconnectivity is not new to ancient history and network theory has been a recognised methodology in Mediterranean archaeology for at least 20 years or so, in ancient history, the particular use of network theory has only really come into its own relatively recently. Peregrine Horden and Nicholas Purcell have explored networks of connectivity in the Mediterranean; Irad Malkin has used network theory to think about colonisation and identity; and in 2006, Christy Constantakopoulou and Katerina Panagopoulou held a conference encouraging scholars to use the theory in discussions of different aspects of ancient history.5 In that context, Anna Collar presented an inspiring paper, later published in the conference proceedings, which used network theory to think about the spread across the Roman Empire of the cult of the monotheistic Theos Hypsistos.6 Her paper focused on describing the kinds of nodes and ties across which information may move, showing how network theory might help to map the development of a cult. As she observed, there are questions still to be answered about the ideology and utility of religion—the ‘attractiveness’ of a belief system and its role in its spread.7 This paper will try to offer some answers to some of these questions.

  • 8  For example, Guettel Cole (1995), with response by Burkert (1995).
  • 9  A selection of examples of scholars reconsidering aspects of polis religion: in Kernos, Bremmer (2 (...)

6A second reason for a new model lies in changes taking place in the study of ancient Greek religion itself. While the concept of polis religion provides a valuable starting point for considering the polis as a religious organisation, nevertheless, it has been debated and nuanced since its inception.8 Recent work suggests that some scholars are also questioning the focus it has given (or been used to give) to classical, usually Athenian, religious practice: some have noted that this risks limiting our view of what counts as ‘religious’, and may also mean that later ritual practice is viewed in terms of disorganisation and decline from an apparent ideal religious form.9 These criticisms suggest that it may be time for an explicit reconsideration of our model of ancient Greek religion—not in order to abandon the polis religion schema altogether, but so as to build on it, and what we have learned from its use, especially in the context of broader work on ancient Greek culture and society.

  • 10  I will draw particularly on White (2008), see further below. In its focus on the interactions in a (...)

7In this paper, in putting the case for social network theory as a basis for a model of Greek religion, I will draw on the work of the sociologist Harrison White, suggesting that, in addition to describing interaction between individuals, such an approach may also help us to examine the mentality behind it. In this model, networks are not only social but ‘cognitive’, and participation in a network helps to create meaning, offering ways of thinking about, describing and explaining life experiences.10 Through this process of creating meaning, networks offer a way of constructing individual and group identities—and vice versa. Sometimes these cognitive aspects—the creation of different ties, meanings and identities—act in concert, but they may also conflict and they are likely to change over time. In putting forward this proposal for one possible new approach, this paper is intended to prompt discussion about the assumptions that underpin current approaches to ancient Greek religion, and to provoke others to offer further alternative models.

1. Problems Presented by Polis Religion

  • 11  See above, n. 9.

8We begin with some of the initial problems presented by polis religion. The difficulties and weaknesses presented by this model have been considered elsewhere, including recent articles in Kernos itself, so this paper will confine itself to a brief overview.11

9The schema of polis religion brings a powerful structuralist analysis to bear on the plethora of religious institutions, roles and experiences that co-existed and interacted across the ancient Greek world. At the centre of this analysis is ‘the polis’ described in Christiane Sourvinou-Inwood’s first formulation of this approach (in the paper, ‘What is Polis Religion?’) as holding a central and centralising role. It is:

  • 12  Sourvinou-Inwood (2000a), p. 19; she does add that the comparison with Christianity is misleading.

… the institutional authority that structured the universe and the divine world in a religious system, articulated a pantheon with certain particular configurations of divine personalities and established a system of cults, particular rituals and sanctuaries… it was the ordered community, the polis, which assumed the role played in Christianity by the Church.12

10In Sourvinou-Inwood’s second key paper on the topic, ‘Further Aspects of Polis Religion’, this role has been extended: all religious behaviour is construed around and embedded within, the central entity of the polis:

  • 13  Sourvinou-Inwood (2000b), p. 51

In the Classical period polis religion encompassed all religious activity within the polis. All such activity was perceived as symbolically legitimated through the religious system of the polis, which shaped the perception for the gods and articulated the relationships between men and the divine… and the polis was the authority which sanctioned all cult activity within its boundaries and mediated it beyond them… A point that needs to be stressed is that all cult acts, including those which some modern commentators are inclined to think of as ‘private’ are (religiously) dependent on the polis.13

  • 14  Sourvinou-Inwood (2000a), p. 18: ‘Each person was a member of this Panhellenic group in virtue of (...)
  • 15  Sourvinou-Inwood (2000a), p. 17: ‘Greek religion, then, consists of a network of religious systems (...)

11The schema focuses implicitly on—and perhaps works best for—one particular polis, namely Athens; or rather, for a particular image of classical Athens in which many religious activities are configured around the structure, and sub-structures, of the polis (for example, demes, phratries, gene, etc.). If we stop at this point of our analysis, then the schema of polis religion offers a coherent, essentially continuous, internally consistent structure, aligned around, and defined by a single attribute—that of belonging, or not belonging, to the polis. Membership of a polis allowed access to being a Greek, an identity ‘cultically expressed in, and reinforced through, ritual activities in which the participating group was ‘all the Greeks’.’14 Although Sourvinou-Inwood does in fact use the term network to describe this, the image is more net than network, a static ‘net’ of shared Greek religiousness.15 This binary categorisation enables the construction of a coherent conceptual system with which we can make swift sense ofthe relationship between the ancient Greek cultural imagination and the operation of its institutions.

  • 16  For ethne, see, for example, Morgan (1997). Sourvinou-Inwood (2000a), n. 1: ‘I cannot consider eth (...)

12But how far can this overriding schema be retained as the details of all ancient Greek ritual activity are explored? Quite apart from the fact that this approach appears to pay no attention to ritual life within other political structures such as ethne, it also risks depicting ancient Greek religion within the Greek city as a single, more or less static, monolithic system.16 It draws attention away from the dynamic complexity of ritual activity among and between individuals and groups, both within and outside their polis communities, and shows scant awareness of developments over time.

2. A Brief Overview of a ‘Social Network Theory’

  • 17  Quotation from Knoke – Yang (2008), p. 4.
  • 18  Breiger (2004), p. 505.
  • 19  Breiger (2004), p. 519.
  • 20  White (2008): I hope that this very brief overview of some of White’s key ideas does not do them t (...)

13Social network theory ‘explicitly assumes that actors participate in social systems connecting them to other actors, whose relations form important influences on one another’s behaviour’ where actors ‘may be individuals, small groups, organizations, nation states’ or, we might say in this case, poleis.17 It explores the ‘patterning of relations among social actors’—that is, the ties between them.18 Over the last decade, various branches of this discipline have attempted not just to model or analyse the resulting network structures, but also to describe them, conceptualizing more clearly the role of ideas, beliefs and values in shaping network structures.19 This paper draws in particular on the work in this area of Harrison White, who has tried to understand how networks operate in lived experience.20

  • 21  White et al. (2007), p. 553.
  • 22  Personhood as a distinct technical meaning within White’s theory; fundamentally, ‘persons are form (...)
  • 23  White (2008), p. 17.
  • 24  White (2008), p. 5, 9-10: for example (p. 5), ‘when the clothes that classmates insist upon as the (...)
  • 25  White (2008), p. 17-18.

14Where the theory of polis religion stresses a cohesive and stable context, organised and legitimated by a central authority, White’s network theory describes how social formations emerge from ‘unending struggles for control to attain footing.’21 Rather than discussing integral persons, the building blocks of White’s theory are ‘identities’.22 He identifies four meanings of ‘identity’: first, the identity that one works to establish on first entering a new situation (for example, starting a new school, going to a conference); an identity that arises from settling into a particular context, which we might describe as a role, such as ‘mother’ or ‘doctor’, or ‘class clown’; the third meaning of identity is ‘the pathway a person, entity or place takes through social time’23; it arises as we carry an identity from one setting into another setting (where it is often inappropriate, and may result in a ‘mismatch’24). Finally, there is the meaning of identity that is in colloquial use, one that is constructed from the process of reviewing for oneself one’s pathway through social time. White offers an example appropriate for this paper: ‘if the third sense [of identity] is, for example, the detailed account of the misfortunes of Oedipus’s life story, the fourth sense is the realisation that he failed.’25

  • 26  Godart – White (2010), p. 570-571.
  • 27  White (2008), p. 6.
  • 28  White (2008), p. 7-9.
  • 29  White (2008), p. 7-10.
  • 30  Examples from Steiny (2007), p. 612-614, and White (2008), p. 63-109.
  • 31  The overlap of identities is what creates ‘persons’ in the ordinary sense, who manifest what White (...)

15As this suggests, identities emerge in response to contingency, and from interactions with other identities, all of which create a sense of uncertainty.26 To mitigate that uncertainty we attempt to exert some control over our context. In this context, ‘control’ does not necessarily mean dominance or power; rather it concerns the effort of finding a social footing, with regard to other identities.27 White argues that individuals do this by creating relationships or ties with others in each social space we enter. He calls these social spaces ‘network domains’ (or ‘netdoms’).28 Within network domains, in turn, we find areas of specific joint activity, or ‘disciplines’, which identities shape and are shaped by during the struggle for control.29 Examples of disciplines might include playing hopscotch, a staff meeting, or Google: disciplines supply order, rules, and valuations for comparing identities.30 As we move between and among disciplines, between and among relationships or ties, between and among network domains, the struggle for footing continues, forging identities through the creation of relations or ties with others. So, identities are situational, emerging from our experiences within and among relationships. Importantly, as this brief description suggests, this process is dynamic: because the world moves around us—and we move around the world from context to context—the need to establish who we are is constant.31

  • 32  Mohr – White (2008), p. 493. We may make up stories alone (e.g., in diary entries), or in conversa (...)
  • 33  White (2008), p. 31.
  • 34  White (2008), p. 37.
  • 35  ‘Story-telling sets human interaction apart’, as Podolny (2001), quoted by White (1992), p. 20: re (...)
  • 36  Godart – White (2010), p. 572.
  • 37  White (1992), p. 40.

16As part of this process of creating an identity, and thus the formation of social networks, White emphasises a further, crucial dimension—the emergence of stories. Stories create and transmit meaning; they ‘are told to oneself and shared with others about the immediacy of events, actions and agents that are known, witnessed and participated in.’32 Stories make networks explicit: as White notes ‘anything about which you tell a story can get reflected in a relation.’33 And of course many stories can be told about any particular relation, so stories occur in sets or repertoires, with many different stories about particular ties and interconnections of ties, which offer different options for fulfilling different story-lines that can accommodate events.34 The reach of such stories is formidable: they may appear to describe individual ties (‘dyads’), but, in fact, since they evoke experience, they should be understood as interwoven across the ‘mesh of interrelations’ in a network, capturing the complexity of relationships, and giving social ties context and order. They may endow them with a beginning, middle and end (although not necessarily), and assign them significance.35 Such stories may also involve a temporal dimension: stories can evoke the traces of past ties and imply the possibilities of future ties—as well as adapting to change over time. But they may also be atemporal, for example rules of thumb or business analytical frameworks, both of which enable the struggle for control.36 Although stories may appear to be simply descriptive, they play a significant role in the struggle for control, with different story-lines offering, and so constraining, different possible outcomes. As White puts it ‘Stories cite behavior. Behavior guides stories… Story goes beyond behavior to weave interpretation and meaning into and around relationships.’37

  • 38  Godart – White (2010), p. 572. Mische – White (1998), p. 702: These so-called ‘network domains’ ma (...)

17Introducing this discursive aspect helps to take the idea of a network structure beyond a simple description of physical location or even just individual or group behaviour. The introduction of stories, story-lines and story-sets begins to help us to understand how network structures and culture interrelate. This joint activity of story creation, embedded in a network between its members, is a key part of how we make sense of, and shape our social environment.38

  • 39  The difficulty of defining institutions is discussed by Mohr – White (2008), p. 488.
  • 40  It comprises a series of dualities: at the individual level, between mind and body; at the group l (...)
  • 41  Mohr – White (2008), p. 508-509.

18This process of story-telling or giving an account also plays a key role in the creation of larger social formations, and two of these are particularly relevant to this paper. These are ‘institutions’ and ‘regimes’. ‘Institutions’ are enduring social formations made up of interlocking networks, including social and cultural networks, whose members share in the creation and expression of that institution’s ‘style’.39 As identities are maintained by stories, so institutions are sustained by ‘rhetorics’. Stories offer descriptions of the relationships created by ties between identities, while rhetorics describe the shared understandings of ‘realms’—a subsystem comprising many different disciplines and styles. Unlike stories, rhetorics are general, abstract accounts of how the world works, and they are collective: they enable the arrangement of fixed sets of stories, according to the sets of values that they communicate. Mohr and White have argued that it is the interpenetration of cultural values and social networks that renders institutions robust—they give, as one example, the institution of caste in a field of villages in central India.40 Nevertheless, as they point out, it is important to remember that institutions rarely exist in isolation—they are more likely to co-exist and overlap, offering individuals contrasting sets of values.41 Developing from this idea of institutions are ‘control regimes’, which as their names suggest, are concerned to ‘generate control over the controls being attempted by identities.’ As institutions are sustained by rhetorics, control regimes coalesce around a narrative—an overriding blueprint of sets of values, perceptions, organisation and action into which rhetorics or stories from across the network population are fitted. But, importantly for this paper, control regimes are neither static nor constant: they still comprise the network relationships described above, each of which entails new struggles for control.

3. Networks and Ancient Greek Religion: Two Case Studies

19The theory we have examined emphasises a process and product of social formation that does not consist of inherent qualities or attributes, but instead involves relations and patterns of relations. It does not evoke an ongoing static structure but draws our attention to the ways in which the formation of society comprises continual turbulence and change. Moreover, it is the ongoing construction of meaning, in terms of stories, rhetorics, narratives that sustain these social structures, and vice versa. Whereas the schema of polis religion implies that changes in religion occur only at particular moments of disruption, for example, the definitive introduction of a cult or being conquered by another state, the implication of social network theory, as described above, is that variety and change were likely to be continuing within polis communities more generally. To illustrate this idea, this paper will offer two brief case studies to suggest some ways in which we might view ancient evidence for ritual activity through the prism of a social network theory. In some ways both of these return us to the example of Thrasyllos the Siphnian, our travelling specialist in supernatural services, with which this paper began. We turn first to one of his likely ritual activities: the creation of binding spells.

3.1. Case Study 1: Binding Spells and Polis Religion

  • 42  Plato, Republic, 364c-e. Carastro (2006), ch. 6 argues that Plato is intent on associating the pra (...)
  • 43  These salesmen apparently also provided sacrifices and incantations that can expiate and cure misd (...)
  • 44  Plato, Laws, 933a.
  • 45  Plato, Meno, 80b. Meno tells Socrates that he fears he has been bewitched by him, and that Socrate (...)
  • 46  Carastro (2006) and Flower (2008).

20Binding spells were inscribed on strips of lead (curse tablets); they were created and used across the Greco-Roman world, from the sixth century BC to roughly the eighth century AD. These texts tell us quite a lot of information about the targets of these spells, but, in general, very little about their writers. The one often quoted passage that provides some insight into at least one of their modes of production and/or dissemination appears in Plato’s Republic.42 In this passage, binding curses, or as they are called here katadesmoi, are one of the supernatural services offered by travelling celebrants of orphic rites or orpheotelestai.43 In contrast to most of the other services listed, katadesmoi are described as methods for harming an enemy. This is confirmed by another passage in the Laws in which Plato provides a list of ways to harm someone by supernatural methods in which katadeseis, another word for binding spells, appears alongside sorceries and incantations.44 This may have been why, as a passage from Plato’s Meno suggests, there was a chance that these supernatural salesmen would not have been warmly welcomed by all cities or citizens.45 It appears that some cities may have dealt quite harshly with those who sold supernatural services. Yet the material evidence suggests that the practice of writing binding spells was widespread—and both literary and material evidence indicates that Thrasyllos was far from an anomaly in ancient Greek society.46

  • 47  Sourvinou-Inwood (2000b), p. 42.

21How and where might we locate such individuals, and their activities, in our understanding of ancient Greek religion? We start first with the framework of polis religion, and a couple of statements made by Sourvinou-Inwood herself—one about curses and one about the people who provided them. Turning to the latter first, with regard to wandering manteis, there is evidence that the schema of polis religion was intended at least to include such itinerant characters. In ‘Further Aspects of Polis Religion’, we find ‘charismatic individuals who offered solicited and unsolicited advice’, placed alongside the exegetai of the polis, prophets like the Pythia who were ‘institutionally appointed in an established cult and sanctuary’, and those who interpreted omens on military expeditions.47 But unlike the individuals occupying other positions—who do indeed seem to conform to her description of being religious personnel with specialised duties—her description here is too brief to clarify how she understood the role played by these charismatic individuals within the polis. Nor does it make clear to which institutions she would have argued that they belonged, in order to be considered as polis ‘personnel’ with some religious authority. Nor whether she expected this to be uniform across all communities (the passage from the Meno suggests otherwise).

  • 48  Sourvinou-Inwood (2000b), p. 44.
  • 49  Ozolian Lokris: IG IX, 1²3:609 (Meiggs and Lewis, no. 13); treaty with curse for anyone who destro (...)
  • 50  See discussion and examples in Eidinow (2007), p. 165-191.
  • 51  See Faraone (1992), p. 74-85, and App. 4; and Graf (1985), p. 81-98. Olympiodoros of Thebes: FHG I (...)

22The second statement, from the same essay, is about the supernatural practice of binding itself, and its relationship to polis practice. Sourvinou-Inwood explains that the ‘individual was without doubt the primary, the basic, cultic unit in polis religion’ a system in which ‘the individual’s act of worship is not different in nature from that of the group’s’. In this context, she states that ‘Personal curses have a counterpart in polis curses’.48 However, although we can certainly see the relationship between individual and group activity in the case of conditional curses, for example, in the role of arai in civic legal codes, or in public curses, immediate polis parallels for binding spells are more difficult to find.49 There is no doubt that individuals wrote binding spells, while some curses allow us to deduce the presence of groups, for example the mention of several targets on one legal team might presuppose the presence of several cursers from the opposing legal team.50 However, can we identify the use of binding spells by, or on behalf of, a polis? As Christopher Faraone and Fritz Graf have shown, there is some, albeit later, evidence that some ancient states and communities resorted to a kind of practice of binding to protect themselves from potential risks—both mortal and immortal.51 What this suggests is that individual binding curses may have some counterpart in certain kinds of binding rituals in communities—at least to the extent of binding an object or person that presents a risk—but it is hard to say more than that.

  • 52  Graf (1997), p. 27; Flower (2008), p. 68 also follows this definition.
  • 53  Parker (2007), p. 122 questions the evidence, but does seem to imply that it is an object of socia (...)
  • 54  Parker is drawing on modern analyses of the term ‘magic’ to evoke ancient attitudes: (p. 122) ‘it (...)
  • 55  In particular, such approaches bring to mind the distinction of Frazer (1911), p. 11 between magic (...)
  • 56  Consider two of the earliest ethnographic works on magic: among Malinowski’s Trobriand islanders, (...)
  • 57  Compare for example, the binding spell by Phila from Pella in Macedonia (Voutiras 1988), an oracle (...)
  • 58  On magic-workers see the discussion in Frankfurter (2002). Graf (1997), p. 229, although arguing f (...)
  • 59  See n. 30.

23For further help in locating these itinerants and their activities, we can turn from Sourvinou-Inwood’s own work to that of other scholars. In general, we find a consensus that the practice of binding, and its practitioners, are not considered to have had a place in polis religion: particular emphasis is placed on binding as a practice of individuals, which distinguishes it from the collective nature of polis religion. For example, for Fritz Graf, binding spells are one of a number of practices peddled by characters like Thrasyllos that are, as he argues, ‘alike in not belonging to the collective religion of the polis.’52 Robert Parker’s analysis of the practice goes a little further: rather than thinking in terms of it being un-collective, his commentary implies that it is anti-collective. Emphasising the vicious nature of binding (calling it ‘black magic’), and describing how it is the ‘furthest extreme from civic norms’, he draws attention to the distinction between the religion of the city and practices like binding that are introduced by what he calls ‘unlicensed religious professionals.’53 His language elides modern categories and the attitudes of the ancients—cursing becomes ‘black magic’ with all the social meaning that implies.54 But care needs to be taken with these approaches, which are underpinned by assumptions that date back some time: many of these concerns about the nature of magic vs. religion, of lone, marginalised individual vs. collective, of disruptive vs. cohesive activity have their origin in early anthropological definitions of ‘magic’.55 These descriptors have since been decisively challenged by ethnographies that have revealed a wide variety of magic-workers operating in different ways within diverse cultures, while attempts to distinguish ‘magic’ from ‘religion’ are famously debated.56 When we turn to the ancient world, and to our binding-spell sellers, these assumptions remain hard to sustain. Even a brief examination of a range of ancient communications with the gods, some classified as religious (prayers); some as magic (for example, a curse and an oracle), can demonstrate how difficult it is to distinguish them using these descriptions.57 Some of the ritual specialists that sold these services could be described in scholarly discourse as occupying a liminal space, moving between poleis, on the fringes of cities, waiting, literally, to be invited in to serve a particular purpose.58 Nevertheless, as Thrasyllos’ somewhat busy personal life suggests, we need to be careful that we do not confuse a theoretical liminality with actual physical and social isolation. Here was a man who seems to have lived and practiced successfully from within a community, indeed, within several different communities, before he finally returned to Siphnos. Moreover, Robert Parker’s term ‘unlicensed’ is also surely misleading, since we lack evidence either for any licensing process, or, as he himself admits, that these ritual activities were subject to strong uniform disapproval.59 Finally, we might challenge the idea that seems to shape the modern approaches described above, that the writing of binding spells should be viewed as simply a disruptive and divisive activity in contrast to the stabilising force of polis religion. For example, if groups within the city were either cursing or feared that they would themselves be cursed, that may have helped to promote a sense of group cohesion. Moreover, the surreptitious expression of such emotions among individuals may have constrained more open attacks of aggression, and so encouraged a certain level of harmony. In turn, just as curses might be seen as methods for promoting cohesion rather than destroying it, as is usually assumed, it is also possible that religion in the polis was not itself a cohesive force—we will return to this possibility later in the paper.

  • 60  See Faraone (1991), p. 5 and Eidinow (2007), p. 145-147. Gordon (1999) has argued that the present (...)
  • 61  DTA 24 and 103, DT 60, SGD 9, 14 and 48; see Eidinow (2007a), p. 168-173.

24In the end we are left struggling with ambiguity: if we take ‘polis religion’ to indicate that the polis was the only ritual authority of ancient culture, and, in turn, that only collective ritual activity qualified as polis religion, then we are left with a series of questions about how to set the practice of curse-writing and its practitioners in the schema of ‘polis religion’. The practice of writing binding spells occurs within the polis community, but its role as a collective force is unclear. It appears not to be mediated or legitimised by the polis as a collective, but is spread by itinerant ritual specialists, and by members of the polis themselves, seemingly without concern for official legitimacy. Nevertheless, although this practice was not a focus of polis regulation, curse tablets were integrally linked with life within a polis. Reasons to think this come from the texts themselves. The formulae of binding spells seem to have been shaped by community conventions: for example, the use of particular phrases and verbs in the texts.60 More obviously, perhaps, many of the texts revolve around situationsthat could be described as polis-related, for example, some curses target individuals in their capacity as the holder of a formal or informal post within a particular polis institution—as litigants or dikastai in the law courts. There are also curses that seem to have a political motivation, which is apparent from the identity of their targets.61 But while encompassing polis concerns, these are hardly ‘collective’ in approach. In general, they reveal the attempts by some individuals to build allegiances, or at least to establish shared enmities, with supernatural entities, in order to maintain personal, social and political standing while doing down that of others. The aspects of polis life they express are darker and seemingly more malicious than those usually discussed under the heading of polis religion. The schema of polis religion suggests an ideal that was stable, cohesive, and collective. The evidence of the binding spells suggests a more ambiguous, dynamic, turbulent reality, which requires a model fluid enough to be able to take account of such social complexity.

  • 62  Eidinow (2007a).
  • 63  Godart – White (2010), p. 570-71.
  • 64  Bremmer (2010), p. 21-22.
  • 65  NGCT 14 (later fourth century; sanctuary of Pankrates, Athens) is a text found in the sanctuary of (...)

25The social network theory described above may be able to offer resolution for a number of the problems described above. Although much of our evidence for Greek ritual practice may draw our attention to collective activities, binding spells offer evidence for other kinds of relationships within the polis: first of all, we might say that they reveal to us a network domain where threatening circumstances have prompted the need to reassess and realign particular relations. Those who were involved in creating curse tablets were part of the network population of the polis—the polis was the network domain or context in which the relational situations occurred that produced the struggles that, in turn, produced these texts. The spells reveal one aspect of how, within various network disciplines—be it civic performance, the law courts, politics or other situations of risk62—identities created, modified and broke relational ties in their struggle for control, in a context of uncertainty.63 Beyond the relational ties between mortals, we also see ties between humans and supernatural entities invoked, reminding us that the network population included the invisible realm. Moreover, importantly, as Jan Bremmer has observed, the gods named in these spells ‘constitute a motley of divinities’, but do not really warrant condemnation as ‘aberrant’. Rather, he suggests, it ‘shows the Athenians could use formulaic enumerations of deities as provided by a professional selling them the tablet, or they could personalize their divinities.’64 I would add to this the suggestion that the choice of gods on some tablets may show regional variations—reinforcing the idea of a network of local ties, comprising both mortal and immortal.65

  • 66  Godart – White (2010), p. 575.
  • 67  Jordan (2000), no. 24; very early fourth century BC.

26More generally, these texts reveal to us the overall storylines that helped individuals to make sense of their experiences, in particular experiences of misfortune. The formulae of binding spells allow us to glimpse ‘explanations encapsulated in time-frames and sequences’: these texts show us the ‘story-lines’ which not only provided some sense of stability and coherence for their users, by making sense of their experiences, but also offered individuals a means for exerting control over their lives.66 I offer a very brief example of what I mean from a binding spell dated to the fourth century BCE: an individual believes he has been cursed, and is cursing, in turn, his unknown opponent, whoever they may be.67 In network terms, the text is evidence for that individual’s sense of a change in his social footing. Some event prompted him to perceive a shift in his relations (with mortals) within the polis, and in his metaphysical stance, with regard to supernatural danger. In response, he attempted to change the nature of both his mortal and immortal relationships. By means of a binding spell, he aimed to position himself differently in his networks, making new ties with the gods, and giving himself control over his (presumed) enemy. (Through the process of writing or commissioning the spell, he may also have forged social ties between himself and a professional spell-writer, which provided him with the knowledge and skills he needed.)

27Through this process, the author of the spell created new meanings, rewriting the narrative of his experiences by trying to reformulate the (presumed) relationships in this situation: the spell tells a story in which he moves from victim to winner, trumping the hostile actions of his anonymous rival. As well as this possible individual story, this text also provides an example of a culturally shared storyline—one that provided an explanation for experiences of misfortune, and a method for overcoming them. The text suggests that the author of the tablet perceived himself to be a victim of jealousy on the part of a rival—suggesting that he had suffered some kind of trouble, which he put down to an occult attack; this prompted him, in turn, to attack his unknown assailant.

  • 68  Olive – Myers (2003), p. 183-184. This would be in contrast to a broadcast network where a large n (...)
  • 69  As Olive – Myers (2003), p. 191 notes, ‘people’s attitudes are shaped by those of the people to wh (...)

28As for those who may have sold curse tablets or other supernatural services, as described above, our evidence is sparse. But in a networked view of ancient Greek religion, whether we regard them as ‘professionals’ or not, such characters need no longer hold an ambiguous position in relation to the polis and its religion. Instead, they can be described in terms of their position within and across different network populations. In network terms it seems likely that the practice of binding spells, for example, spread through a model of information diffusion, via a node-to-node network, that is, where Actor A communicates with Actor B, etc.68 In terms of their ritual activities they would have moved between network domains of users, across poleis, as well as being embedded in networks of practi­tioners. If indeed most individuals that they came across shared the same kinds of cultural storylines about the origins of misfortune, then this would have increased the likelihood that they would have bought a curse tablet.69 Nor was it necessary for this information always to have been spread by a professional curse-writer: although there is evidence for their role, a number of spells suggest that binding spells were being composed by non-professionals, as well, which would have increased the potential for information flow. It is worth bearing in mind that attitudes to such individuals would have been fluid: relations within and between networks were dynamic, continually changing through interactions among their constituents. Any individual, during the course of his/her life, may hold multiple identities across multiple networks. So, an itinerant professional diviner, such as Thrasyllos, may have held an identity of expertise in some network domains, and been perceived as a charlatan in others. Mention of Thrasyllos also reminds us that these individuals were, of course, likely to have identities within and across different social networks—which, in turn, may have increased the diffusion of information through other social groups.

29However, although we may be able to explain binding spells in terms of network theory, it is unclear what implications this might hold for the polis religion model: perhaps binding spells were the anomaly—some kind of polar opposite to what was otherwise a collective and coherent religious structure—the exception that, in effect, proved the rule? The second case study offers further material for reflection on the networked nature of religion in the ancient Greek city, and starts with some further reflection on individual ritual experts.

3.2. Case Study 2: The ‘“Orphic” Type’

  • 70  Sourvinou-Inwood (2000b), p. 55.
  • 71  A structuralist approach like that noted by Burkert (1983), p. 2.
  • 72  Burkert (1983).
  • 73  Col. 20, l. 4 in Betegh (2004), p. 43; for the translation see Parker (2007), p. 116. The word tra (...)
  • 74  Hippocrates, On The Sacred Disease, 2, 12-13 and 4, 36 ff. Versnel (1991a), p. 197, n. 40. This pa (...)
  • 75  Graf – Johnston (2007), p. 178-181 also suggests that the ritual activities of the Orphics should (...)

30Binding curses are not the only ritual activity that raises questions about polis religion. Indeed, Sourvinou-Inwood herself mentions Orphic religion (‘non-institutionalized sectarian discourse of the Orphic type’), which in ‘at least some manifestations … may have been perceived as lying outside the authority of the polis discourse—some important aspects of which they did in fact challenge’ (although she does not elaborate on what such a challenge might imply for the schema itself).70 Her formulation suggests again the binary categorisation with which we started—polis or not-polis.71 But closer examination of the evidence indicates a more complex environment than this description suggests. For example, research and new discoveries suggest that ‘Orphic religion’ covered a range of ritual practices that varied across time and place, overlapping, in various aspects, with Bacchic mystery cult, and comprising a range of offerings from a variety of travelling ritual specialists.72 Support for this view is found in the Derveni papyrus, which seems to contain a theogonical poem of Orpheus, along with a commentary, and which describes an environment in which there is little or no control over who was selling supernatural services. Indeed, the author himself professes to be concerned that some of those who ‘make a profession out of rites’ simply do not know what they are doing.73 The note of competition that we hear in this text reminds us of the well-known, competitive Hippocratic text, On the Sacred Disease.74 Like the writer of the Derveni Papyrus, the writer of this text also does not seem to have formulated his own practices in opposition to polis-centred activities, or to regard them as having greater authority than his own; indeed, they do not seem to have been regarded as a point of reference at all. Instead, these two writers seem to have looked to the activities of those like themselves.75

31The allusions in these two texts suggest that this process of comparison and competition concerning ritual practices was occurring across society, among different groups, in different contexts—a rich and complex arena of ritual practice that the schema of polis religion overlooks. In network terms, however, this material provides us with a glimpse of a complex of overlapping networks of ritual activity, which, in turn, interacted with social networks, and other social groups. At first sight, it appears that the ritual experts of the Derveni Papyrus and the healers of the Hippocratic text were almost certainly travellers, moving from community to community, selling their services—and this suggests, again, a marginal role, with little impact. Alongside the other evidence for these supernatural salesmen, Plato’s orpheotelestai in particular, and the scattered nature of much of the epigraphic evidence for ‘orphic’ ritual practice, we get a picture of individual, even isolated initiators and initiates, operating at the margins of community activity. But, as Thrasyllos’s own story reminds us, this description may give only a particular account—and provide a misleading impression of the social position of these individuals, and of their role in the formation of local ritual practice.

  • 76  Euripides, Hippolytos, 952-954. However, Burkert (1983), p. 11, does draw attention to the presume (...)
  • 77  Theophrastus, 16, 11; Plutarch, Saying of the Spartans (Mor., 224e-f): for example, in a retort gi (...)

32In terms of the orpheotelestai themselves, some ancient evidence may suggest the possibility that these individuals were living within communities, rather than wandering between them: when Euripides’ Theseus mocks Hippolytos, he lists a number of characteristics of an individual who ‘celebrates Bacchic rites’ and follows Orpheus, but he does not include wandering amongst them.76 Certainly, Theophrastus’ ‘superstitious man’ joins in their rites every month, and takes along his wife and children, suggesting some stability of time and place. Other anecdotes may suggest that such priests were well known within their communities, and were considered useful in difficult circumstances.77

  • 78  Ansell (2003), p. 126-7.
  • 79  Possibility of communities of initiated: Graf – Johnston (2007), p. 163-164. The term ‘orphikoi’ r (...)
  • 80  Public and private cults somehow linked, see Henrichs (1969) and Seaford (1981), discussed Graf – (...)

33If we try to picture some of these characters as living and practising within a polis community it is difficult to fit them within the schema of polis religion, but within a networked model, we can see how these individuals may have created sub-networks of ritual practice that were embedded within their communities, and within other social networks.78 There may indeed be indications for more stable groups of followers, perhaps as autonomous communities, but if not, then perhaps as thiasoi within poleis.79 Indeed, some scholars have argued that members of private thiasoi for a particular god may simultaneously have been members of a public cult of the same divinity—and that such cult organisations may have been linked.80 In this context, we will turn to the evidence, albeit fragmentary, for the public and private worship of Dionysos.

  • 81  Sourvinou-Inwood (2000b), p. 55.
  • 82  If it means the foundation, then there is, indeed, evidence that this was sometimes the case, at l (...)
  • 83  See Versnel (1990), p. 149.
  • 84  Sourvinou-Inwood (2000b), p. 54 states that both Dionysiac rites and Eleusinian Mysteries were enc (...)
  • 85  Pausanias, X, 4, 3 (Athenian) and X, 32, 7 (in Parnassos); Pausanias, VI, 26, 1 (a feast outside E (...)
  • 86  In contrast, there does seem to have been a polis-instituted cult of Dionysos: goat sacrifices to (...)

34Sourvinou-Inwood asserts that Dionysiac rites, in classical Athens at least, were ‘wholly integrated into the socio-religious structures of the polis.’81 But it remains unclear just what integration means here—control, oversight or foundation, or all of the above.82 Turning to Athens, evidence suggests that some Dionysiac cults may have been self-organised: on Lysistrata’s wish that her fellow women had been invited to a Bacchic feast (she also names other religious festivities), the Scholiast remarks how women ‘used to celebrate many feasts outside the state ceremonies, and they convened privately.’83 If we are to take this as describing rites that were ‘integrated into the socio-religious structures of the polis’, then this must prompt questions about the meaning and significance of ‘polis’ in polis religion: is it enough that these women celebrated their private rites within the polis for them to count as polis religion?84 And what about when those rites took their celebrants out of the polis?85 When we are told by Pausanias that every other year women of Athens went to Delphi to worship Dionysos with his female followers there, the Thyiades, can we still place such celebrations squarely within the schema of polis religion?86

  • 87  A peace-making role has been evoked for the Sixteen Women at Elis, who seem to have had a reputati (...)

35A network model provides some resolution for these questions, allowing us to describe these events in terms of ongoing sub-networks of religious affiliation which connected individuals across different poleis, individuals who also had network identities within those poleis. We can also draw on White’s terminology to suggest that the Dionysiac rites could be described as institutions, offering a particular set of values and styles that characterised membership. Moreover, we have some evidence that raises questions about how those institutions may have been regarded by those who did not belong to them. Thus, Plutarch relates a story about a band of Thyiads from Phokis: the women became lost and ended up wandering in the dark. Finally, they reached the enemy city of Amphissa where they fell asleep. When they awoke, rather than attacking them, a group of local Thyiads gave them protection and food, and accompanied them home. Such a story may not provide us with historical data, but it may present evidence for some attitudes to such cults. Although we might see in these women’s activities a peace-making role, the idea that the Amphissan Thyiads chose to help out their lost sisters, rather than support their own city, may suggest that to those who did not belong to these institutions, membership may have been perceived as holding the potential to generate compromised allegiances.87

  • 88  See Aristophanes, Birds, 874-875; Wasps, 9; Lysistrata, 387 f., (where the rites of Adonis are lik (...)
  • 89  Demosthenes, 18 (De Corona), 129 and 259-260, and 19 (On the False Embassy), 249 and 281.

36Similar indications of a concern that certain cult memberships might mean a lack of trustworthiness appear in relation to the cult of the god Sabazios, a Thraco-Phrygian god whose worshippers entered an intoxicated, ecstatic state. Sabazios appears to have held a somewhat ambiguous status within the polis during the classical period, insofar as we can tell from the allusions made by some of Aristophanes’ characters.88 Moreover, further evidence suggests that in the courts, before your fellow citizens, connections to such a network might be introduced by one speaker as a way of discrediting another—as Demosthenes does with regard to Aeschines.89 These accusations occur in an agonistic con­text, suggesting very different views on what counted as a respectable religious affiliation, whatever the apparent legitimacy granted by the polis. In terms of networks, we can say that this explicit conflict, in which an opponent’s membership of a particular institution is used by one speaker to attack the other, reveals an implicit struggle for control, in which identities, vying for superiority, introduce the rhetorics of their institutions as part of that struggle.

  • 90  Inscription of Miletus: LSAM 48 (see n. 82 above and discussion in Henrichs [1969]); Versnel (1990 (...)

37This type of individual interaction may also be glimpsed at a larger social level. An inscription of Miletus cited by Sourvinou-Inwood as indicating the total integration of Dionysiac rites into a polis can also be read in other ways. The inscription proclaims: ‘Nobody may convene a thiasos before the thiasos of the city is assembled.’ It goes on to give instructions for the money to be given to the priestess by any woman who ‘wishes to perform initiations in order to honour Dionysos Bakcheios.’90 This suggests that it was possible that thiasoi were or might be being conducted privately, and that the city was attempting to make some kind of connection with, and exert control over these groups and their rituals, if only temporal and financial. Rather than the straightforward integration of ritual activity within the polis schema, this suggests, at least, a more dynamic, more nuanced interaction between polis and thiasoi concerning ritual activities. To put it in network terms, it indicates a struggle between two institutions for control over the relationships in their networks. It is just this idea of institutional struggle that, I would like to suggest, may help to clarify the apparent prominence of polis religion in both primary and secondary sources. The final section of this paper turns to a possible description of polis religion in a network model.

4. The Narrative of Polis Religion

  • 91  Sourvinou-Inwood (2000b), p. 48.

38The itinerant characters, like Thrasyllos, with whom we started, who moved among and between communities, are obviously hard to pin down within the schema of polis religion. But, as suggested by this paper’s brief overview of cult activity relating (in various ways) to the worship of Dionysos, ambiguity continues to characterise ritual practice as we move into the polis itself, and examine the different cults/groups who worshipped within it. This seems to be the case even in what we might expect to be the most distinct of category divisions, citizens/non-citizens. The schema of polis religion draws particular attention to the distinction between citizens and strangers—politai and xenoi—living within a polis. Sourvinou-Inwood describes this difference as regulating the religious identity of individuals at every level from polis to panhellenic: ‘In every polis non-citizens could only worship as xenoi.’ ‘But’, the argument continues, ‘not all xenoi had the same status: some were admitted into the system more than others.’91

  • 92  Deme-level activities discussed in Humphreys (2004), who suggests, for example, that the state bec (...)
  • 93  Aleshire (1994) provides a thorough and convincing analysis of the difficulties of describing ‘sta (...)
  • 94  The difficulties of describing variations in the details of ritual activity within the schema of p (...)

39On examination, however, it is hard to identify a single system, with different levels of admission: rather, the evidence evokes a more fluid set of various and varying relationships between individuals and groups. Within the category of ‘citizens’, it appears there are further sub-categories to be made with regard to cult activity, some bearing more autonomy or exclusivity than others. A polis might attempt to exert control over some particular aspect of ritual activity, or it might negotiate with those involved, but it might also simply ratify, tolerate, or even just ignore them.92 Indeed, defining what it means for ritual activity to be integrated within the schema of polis religion becomes increasingly difficult as we examine the variety of cult organisations and the different levels and types of involvement the polis might have. Moreover, this in turn raises the question of what is meant here by ‘polis’—for example, are decisions by the demos to be regarded as the same as decisions by a genos or a thiasos?93 By trying to assimilate everything within one supreme category, polis religion risks eliding significant distinctions and variations in ritual activity.94

  • 95  This idea of multiple connections is also conveyed by Sourvinou-Inwood (2000b), p. 46, who describ (...)
  • 96  A point also made by Aleshire (1994), p. 13.

40In light of these observations, what actually seems to be at issue is not a single binary relationship between individual and polis (and therefore panhellenic) identity, but many different relationships across and among individuals and a wide variety of cultic groups, resulting in individuals with ties to some polis members but not others; groups within the polis with ties to similar groups outside the polis; itinerant groups and individuals connecting across territories. Individuals may exhibit diverse, and sometimes conflicting identities within and among these different networks.95 Similarly, across these networks, there was a range of ways in which the polis may or may not have been involved in the ritual activities of groups and individuals.96 The resulting configuration suggests the need for an alternative model for ancient Greek religion, one which does not seek to deny the significance of the role of the polis, but which tries to assimilate the elements of ancient Greek ritual experience overlooked or elided by the polis religion schema, and some of its interpretations.

  • 97  Using the network theory language of Barabási – Albert (‘Statistical mechanics of Complex Networks (...)
  • 98  Mohr – White (2008).
  • 99  The process by which the polis consolidated its power through absorption of social and political s (...)

41Nevertheless, it is still the case that even if we view ancient Greek religion as a networked religion, many of the most powerful ties were established within and around certain (polis) communities. For modern scholars, one of the most prominent of these is the religious network that constituted the polis of Athens, which, in turn, has provided the basis for the concept of polis religion.97 Within a network theory using White’s terminology, as noted above, the polis religion of Athens seems to have behaved like an ‘institution’. It linked different social systems across different levels of organisation (both local and extra-local), and integrated values and social relations, offering powerful ‘rhetorics’ (in place of, say, a dogma or creed) that provided members of the institution with a shared set of meanings.98 But I think we can go further than this and argue that polis religion provides us with a particular example of an institutional system, that is, a ‘control regime’, whose members ensured that, in the struggles for control, they maintained dominance.99 One of the expressions and tools of this struggle was the ultimate ‘narrative’ of polis religion, which, although it may have com­prised multiple rhetorics from different sources, subsumed them (or attempted to do so) in a single account of shared values and style: now, as then, this narrative is highly persuasive.

  • 100  Josephus (Against Apion II, 267) offers a list of men ‘put to death by the Athenians because they (...)
  • 101  In the trial of Socrates, for example, we learn about his connections to politically dangerous ind (...)
  • 102  Graf (1997), p. 59: Cicero, Laws II, 21; Paulus Sententiae V, 23, 15; FIRA II, 410; and Codex Theo (...)

42As mentioned above, control regimes tend to contrast their values with those in another realm, and the examples given above (the suspicion surrounding membership of certain cults; the possible punishment meted out by some cities to travelling seers) have already suggested that this was an ongoing process within Athens, if not within other cities. Athens also provides some more specific examples: the impiety trials of the fourth century, most famously perhaps, that of Socrates, but including a number of other famous men, and less well-known women reveal some aspects of the ongoing maintenance and construction of the institutional system (and its rhetoric) of polis religion.100 Attention to the inciden­tal details of the charges against these individuals suggests that it was not simply the introduction of new gods or rituals that was problematic for the Athenians: there were other factors at play. The details of these trials indicate that decisions about acceptable ritual activity (rhetorics about value) were entwined with defini­tions of acceptable social identity (social status).101 Acceptance within or rejection from the network of Athenian religion rested on a complex duality of social and cultural forms that involved the social and political ties and identities of those involved. On the one hand, the outcomes of these trials support an institution that we now recognise as polis religion: they show the members of an institutional system rejecting the anomalous behaviour of a few; we see a single cohesive system exercising control and overriding authority. And yet, on the other hand, these trials also provide evidence for other activities, relationships, identities and systems of meaning. As with those involved in the practice of binding and, for the most part, the groups of Dionysiac worshippers, these other groups may have co-existed peacefully within the more dominant institution for much of the time. However, it appears from the evidence described above that there were times when all of these activities could be taken as opposing or threatening to undermine it—and then action was taken.102

  • 103  Sourvinou-Inwood (2000a), p. 37.

43This network approach to ancient Greek religion builds on existing elements of the polis religion schema. Most specifically, I think it can help us to understand some further ways in which, as Sourvinou-Inwood put it, ‘the role of the polis in the articulation of Greek religion was matched by the role of religion in the articulation of the polis.’103 The notion of networks can help us to see just how that ‘articulation’ may have occurred, suggesting the individual and institutional struggle for identity and, above all, the creation of meaning (through stories and rhetorics) that this involved. It implies that what we call ‘polis religion’ was a budding institution that emerged from a meshing of social and cultural networks. It also reminds us that, alongside the polis, there were other networks and interactions, which, at the very least helped to define it.

  • 104  There is not room to consider this in depth here, but some specific examples for which network the (...)

44There is, of course, more work to be done—not only the refinement of the ideas within this paper, but also by looking beyond the polis to encompass, for example, ethne contexts where we find religion providing identity in an absence of political structures, and beyond the mortal realm to encompass relationships between gods and mortals, gods and gods, and even the development of various identities of particular divinities.104 But this has been an initial attempt at describing how polis religion both comprised, and was embedded within, a dynamic, networked environment of ritual practice.

Haut de page

Bibliographie

DTA: R. Wünsch, Defixionum Tabellae Atticae Inscr. Gr. III.3, Berlin, 1887.

DT: A. Audollent, Defixionum Tabellae, Paris, 1904.

LSAM: F. Sokolowski, Lois sacrées de l’Asie Mineure, Paris, 1955.

NGCT: D. Jordan, ‘New Greek Curse Tablets (1985–2000),’ GRBS 41 (2000), p. 5-46.

SGD: D. Jordan, ‘A Survey of Greek Defixiones Not Included in the Special
Corpora,’ GRBS 26 (1985), p. 151-197.

SGDI: O. Hoffman, ‘Die Orakelinschriften aus Dodona,’ in H. Collitz, F. Bechtel (eds.), Sammlung der Griechischen Dialekt-Inschriften, Göttingen, 1899.

S. Aleshire, ‘Towards a definition of ‘State Cult’ for Ancient Athens,’ in R. Hägg (ed.), Ancient Greek Cult Practice from the Epigraphical Evidence, Stockholm, 1994, p. 9-16.

C. Ansell, ‘Community Embeddedness and Collaborative Governance in The San Francisco Bay Area Environmental Movement,’ in M. Diani, D. McAdam (eds.), Social Movements and Networks: Relational Approaches to Collective Action, Oxford, 2003.

I. Arnaoutoglou, Thusias heneka kai sunousias: Private Religious Associations in Hellenistic Athens, Volume 37, Issue 4 of Academy of Athens: Yearbook of the research centre for the history of Greek law, Athens, 2003.

A-L. Barabási, R. Albert, ‘Statistical Mechanics of Complex Networks,’ Reviews of Modern Physics 74 (2002), p. 47-97.

G. Betegh, The Derveni Papyrus: Cosmology, Theology and Interpretation, Cambridge, 2004.

D. Boedeker, ‘Family Matters: Domestic Religion In Classical Greece,’ in J. Bodel, S.M. Olyan (eds.), Household and Family Religion in Antiquity, Oxford, 2008, p. 229-247.

H. Bowden, ‘Oracles for Sale,’ in P. Derow, R. Parker (eds.), Herodotus and his World, Oxford, 2003, p. 256-274.

R.L. Breiger, ‘The Analysis of Social Networks,’ in M. Hardy, A. Bryman (eds.), Handbook of Data Analysis, London, 2004, p. 505-526.

J. N. Bremmer, ‘Manteis, Magic, Mysteries and Mythography: Messy Margins of Polis Religion?,’ Kernos 23 (2010), p. 13-35.

W. Burkert, ‘Craft vs. Sect: Problems of Orphics and Pythagoreans,’ in E.P. Sanders, B.F. Meyer (eds.), Jewish and Christian Self-Definition: Self-definition in the Greco-Roman World, Philadelphia, 1983, p. 137-160.

—, ‘Greek Poleis and Civic Cults: Some Further Thoughts,’ in M.H. Hansen, K. Raaflaub (eds.), Studies in the Ancient Greek Polis. Papers from the Copenhagen Polis Centre 2, Stuttgart, 1995 (Historia Einzelschriften, 95), p. 201-210.

M. Carastro, La cité des mages. Penser la magie en Grèce ancienne, Grenoble, 2006.

A. Collar, ‘Network Theory and Religious Innovation,’ in Constantakopoulou – Malkin – Panagopoulou (2009), p. 144-157.

C. Constantakopoulou, I. Malkin and K. Panagopoulou (eds.), Greek and Roman Networks in the Mediterranean, London, 2009.

M. Detienne, ‘Les chemins de la déviance. Orphisme, dionysisme et pythagorisme,’ in Orfismo in Magna Grecia, Atti del Quatordicesimo Convegno di Studi sulla Magna Grecia, Naples, 1975, p. 49-79.

M. Dickie, Magic and Magicians in the Greco-Roman World, London, 2001.

J. Dillery, ‘Chresmologues and Manteis,’ in S.I. Johnston, P. Struck (eds.), Mantikê: Studies in Ancient Divination, Leiden, 2005, p. 167-231.

K. Dover, ‘The Freedom of the Intellectual in Greek Society,’ Talanta 7 (1976), p. 26-54.

E. Eidinow, Oracles, Curses, and Risk Among the Ancient Greeks, Oxford, 2007a.

—, ‘Why the Athenians Began to Curse’, in R. Osborne (ed.), Debating the Cultural Revolution: Art, Literature, Philosophy, and Politics 430-380 BC, Cambridge, 2007b, p. 44-71.

—, ‘Patterns of Persecution: “Witchcraft” Trials in Classical Athens,’ Past & Present 208 (2010), p. 9-35.

A. Erskine, ‘Epilogue,’ in J. Bremmer, A. Erskine (eds.), The Gods of Ancient Greece: Identities and Transformations, Edinburgh, 2010, p. 505-510.

E.E. Evans-Pritchard, Witchcraft, Oracles and Magic among the Azande, Oxford, 1937.

C. Faraone, Talismans and Trojan Horses: Guardian Statues in Ancient Greek Myth and Ritual, Oxford, 1992.

—, ‘The Agonistic Context of Early Greek Binding Spells,’ in C. Faraone, D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion, Oxford, 1991, p. 3-32

N. Fisher, ‘Greek Associations, Symposia and Clubs,’ in M. Grant, R. Kitzinger (eds.), Civilization of the Ancient Mediterranean: Greece and Rome II, New York, 1988, p. 1167-1199.

M. Flower, The Seer in Ancient Greece, Berkeley, 2008.

D. Frankfurter, ‘Dynamics of Ritual Expertise in Antiquity and Beyond: Towards a New Taxonomy of ‘Magicians’,’ in P. Mirecki, M. Meyer (eds.), Magic and Ritual in the Ancient World, Leiden, 2002, p. 159–178.

J.G. Frazer, The Golden Bough: a Study in Magic and Religion. Part 1: The magic art and the evolution of kings, London, 1911.

V. Gabrielsen, ‘Brotherhoods of Faith and Provident Planning: The Non-Public Associations of the Greek World,’ Mediterranean Historical Review 22:2 (2007), p. 183-210.

F.C. Godart, H.C. White, ‘Switchings under Uncertainty: The Coming and Becoming of Meanings,’ Poetics 38:6 (2010), p. 567-586.

B. Goff, Citizen Bacchae: Women’s Ritual Practice in Ancient Greece, Berkeley, 2004.

R. Gordon, ‘Imagining Greek and Roman Magic,’ in B. Ankarloo, S. Clark (eds.), Witchcraft and Magic in Europe: Ancient Greece and Rome, London, 1999, p. 159-275.

F. Graf, La Magie dans l’antiquité gréco-romaine. Idéologie et pratique, Paris, 1994.

—, Magic in the Ancient World, Cambridge MA, 1997.

—, Nordionische Kulte, Rome, 1985 (Bibliotheca Helvetica Romana,21).

—, S.I. Johnston, Ritual Texts for the Afterlife, London, 2007.

S. Guettel Cole, ‘Civic Cult and Civic Identity,’ in M.H. Hansen (ed.), Sources for The Ancient Greek City-State. Symposium August 24-27, 1994, Acts of the Copenhagen Polis Centre, 2, Copenhagen, 1995 (Historisk-filosofiske meddelelser, 72), p. 292-326.

M.H. Hansen, The Trial of Sokrates—From The Athenian Point of View, Copenhagen, 1995 (Historisk-filosofiske meddelelser, 71).

—, ‘Introduction: The Polis as a Religious Organisation,’ in M.H. Hansen, T.H. Nielsen (eds.), An Inventory of Archaic and Classical Poleis, Oxford, 2004.

A. Henrichs, ‘Die Mänaden von Milet,’ ZPE 4 (1969), p. 223-341

—, ‘Changing Dionysiac Identities,’ in E.P. Sanders, B.F. Meyer (eds.), Jewish and Christian Self-Definition: Self-definition in the Greco-Roman World, Philadelphia, 1983, p. 137-160.

—, ‘Between Country and City: Cultic Dimensions of Dionysus in Athens and Attica,’ in M. Griffith, D.J. Mastronarde (eds), Cabinet of the Muses, Atlanta, 1990, p. 257-277.

P. Herrman, ‘Teos und Abdera im 5. Jahrhundert v. Chr.,’ Chiron 11 (1981), p. 1-30.

J. Holzhausen, ‘Poetry and Mysteries: Euripides’ Bacchae and the Dionysiac Rites,’ Electronic Antiquity 12.1 (November, 2008). see http://scholar.lib.vt.edu/ejournals/ElAnt/V12N1/holzhausen.pdf

P. Horden, N. Purcell, The Corrupting Sea: A Study of Mediterranean History, Oxford, 2000.

S. Humphreys, ‘A Sense of Agency: Religion in the Attic Demes,’ in The Strangeness of Gods: Historical Perspectives on the Interpretation of Athenian Religion, Oxford, 2004, p. 130-196.

P. Ismard, ‘Les associations en Attique de Solon à Clisthène,’ in J.-Chr. Couvenhes, S. Milanezi (eds.), Individus, groupes et politique à Athènes de Solon à Mithridate, Bordeaux, 2007 (Perspectives Historiques, 15), p. 17-34.

D. Jordan, ‘New Greek Curse Tablets (1985–2000),’ GRBS 41 (2000), p. 5-46 (NGCT).

—, ‘A Survey of Greek Defixiones Not Included in the Special Corpora,’ GRBS 26 (1985), p. 151-197 (SGD).

E. Kearns, ‘Between God and Man: Status and Functions of Heroes and their Sanctuaries,’ in Le Sanctuaire grec, Geneva, 1992 (Entretiens sur l’Antiquité classique, 37), p. 65-99 and discussion, p. 100-108.

—, ‘Religious Practice and Belief,’ in K. Kinzl (ed.), A Companion to the Classical Greek World, Oxford, 2010, p. 311-326.

J. Kindt, ‘Polis Religion—A Critical Appreciation,’ Kernos 22 (2009), p. 9-34.

D. Knoke, S. Yang, Social Network Analysis, London, 2008.

B. Latour, ‘The Trouble with Actor Network Theory,’ Soziale Welt 47 (1997), p. 369-381.

—, ‘On Recalling ANT,’ in J. Law, J. Hassard (eds.), Actor Network Theory and After, Oxford, 1999, p. 15-24.

B. Malinowski, Coral Gardens and Their Magic: a Study of the Methods of Tilling the Soil and of Agricultural Rites in the Trobriand Islands, vol. 1: The Description of Gardening, London, 1935.

I. Malkin, ‘Networks and the Emergence of Greek Identity,’ in Mediterranean Paradigms Classical Antiquity, Special Issue of the Mediterranean Historical Review 18 (2003), p. 56-74.

M. Mauss, H. Hubert, A General Theory of Magic, trans. R. Brain, London, 1972 [1902].

M. Meyer, P. Mirecki, Ancient Magic and Ritual Power, Leiden, 1995.

J.W. Mohr, H.C. White, “How to Model an Institution,” Theory and Society 37 (2008), p. 485-512.

C. Morgan, ‘The Archaeology of Sanctuaries in Early Iron Age and Archaic Ethne,’ in L. Mitchell, P. Rhodes (eds.), The Development of the Polis in Archaic Greece, London, 1997, p. 92-109.

A. Mische, ‘Cross-Talk In Movements: Reconceiving The Culture Network Link,’ in M. Diani, D. McAdam (eds.), Social Movements and Networks: Relational Approaches to Collective Action, Oxford, 2003, p. 258-280.

—, H. White, ‘Between Conversation and Situation: Public Switching Dynamics Across Network Domains,’ Social Research 65 (1998), p. 295-324.

G. Norlin, Isocrates with an English Translation in three volumes, Cambridge, MA, 1980.

P.E. Oliver, D.J. Myers, ‘Networks, Diffusion and Cycles of Collective Action,’ in M. Diani, D. McAdam, Social Movements and Networks: Relational Approaches to Collective Action, Oxford, 2003, p. 173-203.

H. W. Parke, The Oracles of Apollo in Asia Minor, Oxford, 1985.

R. Parker, Athenian Religion: A History, Oxford, 1997.

—, Polytheism and Society at Athens, Oxford, 2007.

J. Podolny, ‘Networks as the Pipes and Prisms of the Market,’ American Journal of Sociology 107 (2001), p. 33-60.

J.Z. Smith, ‘Trading Places,’ in Meyer – Mirecki (1995), p. 13-27.

L. Rubinstein, ‘ “ARAI” in Greek Laws in the Classical and Hellenistic Periods: Deterrence or Concession to Tradition?,’ in L. Gagliardi, E. Cantarella (eds.), Symposion 2005. Akten der Gesellschaft für griechische und hellenistische Rechtsgeschichte 19, Vienna, 2007, p. 1-20.

M. Scott, Delphi and Olympia: The Spatial Politics of Panhellenism in the Archaic and Classical Periods, Cambridge, 2010.

R. Seaford, ‘Dionysiac Drama and the Dionysiac Mysteries,’ CQ 31 (1981), p. 252-275.

H.A. Shapiro ‘ “Hêrôs Theos”: The Death and Apotheosis of Herakles,’ CW 77 (1983), p. 7-18.

C. Sourvinou-Inwood, ‘What is Polis Religion?,’ in R. Buxton (ed.), Oxford Readings in Greek Religion, Oxford, 2000a, p. 13-37.

—, ‘Further Aspects of Polis Religion,’ in R. Buxton (ed.), Oxford Readings in Greek Religion, Oxford, 2000b, p. 38-55.

D. Steiny, ‘H. White, Identity and Control (2nd ed.), Cambridge University Press, Cambridge (2008),’ Social Networks 29:4 (2007), p. 609-616.

R. Styers, Making Magic: Religion, Magic, and Science in the Modern World, Oxford, 2004.

H. Versnel, Inconsistencies in Greek and Roman Religion vol. 1: Ter Unus. Isis, Dionysos, Hermes, Leiden, 1990.

H. Versnel, ‘Some Reflections on the Relationship Between Magic–Religion,’ Numen 38 (1991a), p. 177-197.

—, ‘Beyond Cursing: The Appeal to Justice in Judicial Prayers,’ in C. Faraone, D. Obbink (eds.), Magika Hiera: Ancient Greek Magic and Religion, Oxford, 1991b, p. 60–106.

E. Voutiras, διονυσοφωντοσ γαμοι: Marital Life and Magic in Fourth Century Pella, Amsterdam, 1998.

H.C. White, Identity and Control: A Structural Theory of Social Action, Princeton, 2008.

—, J. Fuhse, M. Thiemann, L. Buchholz, ‘Networks and Meaning: Styles and Switching,’ Soziale Systeme 13 (2007), p. 543-555.

Haut de page

Notes

*  An initial version of this paper was given at a conference in honour of Christiane Sourvinou-Inwood held at Reading in 2008; a second version at a conference on communities and networks in the ancient Greek world at University College Dublin a year later. The paper has been greatly improved by comments from attendees at both conferences, and in particular by the detailed reading of Claire Taylor, Kostas Vlassopoulos, Robin Osborne, Simon Hornblower, and the referees of Kernos.

1  Isokrates, 19 (Aiginetikos), 5-7 (trans. Norlin [1980]).

2  See Flower (2008) for an excellent overview of the ancient Greek seer.

3  For example, Meyer – Mirecki (1995), esp. the summary on p. 2. On how scholars have used the term, see Smith (1995), p. 16 and 20, while Styers (2004) argues that definitions of magic espoused by academics were, and continue to be, inherent in the formation of the idea of modernity. The heuristic value of the term is supported by Versnel (1991a); its usefulness is dismissed by Gager (1992), p. 24-25.

4  A brief and very clear description of a network can be found in the introduction to Constantakopoulou, Malkin and Panagopoulou (2009). The idea of a network may be implicit in some of Sourvinou-Inwood’s own descriptions of polis religion: for example, her account of how individual poleis, and individuals or groups from within poleis, participated in rites in other cities or at a panhellenic level, see 2001a, 15-17; see further below.

5  Horden – Purcell (2000); Malkin (2003); for 2006 conference see above, n. 4.

6  Collar (2009).

7  Collar (2009), p. 151. She poses the question in terms of ‘“vulnerability” to a new faith’ and examines evolutionary and network approaches to answering this question, suggesting that the two should not be considered as mutually exclusive (p. 152).

8  For example, Guettel Cole (1995), with response by Burkert (1995).

9  A selection of examples of scholars reconsidering aspects of polis religion: in Kernos, Bremmer (2010) and Kindt (2009), both of whose papers originated, like this one, in a conference at the University of Reading exploring the idea of polis religion (see n. 1); see also Kearns (2010) for a systematic appreciation of what polis religion can and cannot account for; Scott (2010) for reflections on Delphi as Sourvinou-Inwood’s ‘litmus test case for the primacy of the polis in the organisation of the religious experience’ (p. 271, n. 94); Boedeker (2008) for consideration of domestic religion in the context of polis religion (describing polis cult and family cult as ‘an interlocking set of practices, asymmetrical though often complementary’ p. 244); Hansen (2004), p. 130-134, for a thorough discussion of the role of the polis as a religious organisation, and the development of that role; Morgan (1997) for exploration of religious development in ethne as opposed to poleis. Erskine (2010) persuasively explores the implications of the polis focus.

10  I will draw particularly on White (2008), see further below. In its focus on the interactions in a network, this approach may recall the Actor Network Theory (ANT) of Bruno Latour. However, ANT ascribes agency to both humans and nonhumans (Latour 1997 and 1999), which is not the case with White’s approach.

11  See above, n. 9.

12  Sourvinou-Inwood (2000a), p. 19; she does add that the comparison with Christianity is misleading.

13  Sourvinou-Inwood (2000b), p. 51

14  Sourvinou-Inwood (2000a), p. 18: ‘Each person was a member of this Panhellenic group in virtue of being a member of a polis. It is not simply that being a citizen of a particular polis guarantees one’s Greekness; as we saw, the polis mediated participation in Panhellenic cult.’

15  Sourvinou-Inwood (2000a), p. 17: ‘Greek religion, then, consists of a network of religious systems interacting with each other and with the Panhellenic religious dimension.’ She also describes this participation in terms of ‘the openings and the closures of the polis towards outsiders’ where she allows some dynamism over time, noting that they vary according to city and time period (2000b, p. 47); a little later in the same description (p. 48), the ‘opening’ between a colony and mother city is described as ‘a link.’

16  For ethne, see, for example, Morgan (1997). Sourvinou-Inwood (2000a), n. 1: ‘I cannot consider ethnos religion here. The differences between ethnos and polis religion do not impinge on our investigation.’ Nevertheless, she does mention ethne in the rest of the paper, often in the formula polis/ethnos, as if the two were equivalent. (See n. 8, a discussion of permission needed to make dedications by the polis/ethnos which owned the sanctuary; and p. 16, in which she discusses ‘Amphictionies or Leagues, associations of poleis/ethne.’ The subsequent discussion however focuses on the ways in which a polis would mediate the activities of its members within these associations.)

17  Quotation from Knoke – Yang (2008), p. 4.

18  Breiger (2004), p. 505.

19  Breiger (2004), p. 519.

20  White (2008): I hope that this very brief overview of some of White’s key ideas does not do them too much harm, and might, at least, provide an illustration of how they might be deployed to help describe the physical, social and emotional ties that made up ancient Greek religion.

21  White et al. (2007), p. 553.

22  Personhood as a distinct technical meaning within White’s theory; fundamentally, ‘persons are formed from compounds of identities,’ White (2008), p. 129 (see further n. 25 below).

23  White (2008), p. 17.

24  White (2008), p. 5, 9-10: for example (p. 5), ‘when the clothes that classmates insist upon as their badge of belonging, are disdained by a parent at home who resists purchasing them.’

25  White (2008), p. 17-18.

26  Godart – White (2010), p. 570-571.

27  White (2008), p. 6.

28  White (2008), p. 7-9.

29  White (2008), p. 7-10.

30  Examples from Steiny (2007), p. 612-614, and White (2008), p. 63-109.

31  The overlap of identities is what creates ‘persons’ in the ordinary sense, who manifest what White (2008), p. 112-170 calls ‘style’ across and between their identities. Style presents as ‘sensibility’ in experience: White observes that it is ‘rather like Bourdieu’s habitus, but where Bourdieu proposes habitus as the signature of a person… [White argues] for person as style’ (p. 114). ‘Styles can be thought of as a fundamental specification of how individual agents live their lives through an ongoing process of combining understandings of situations with sets of practices arrayed across lives embedded within social networks (netdoms)’ (Mohr – White [2008], p. 491). As well as descriptors of patterns of enactment, White (2008), p. 115 also argues that styles can be understood to provide ‘envelopes’ that limit and funnel attempts at control by identities.

32  Mohr – White (2008), p. 493. We may make up stories alone (e.g., in diary entries), or in conversation with others (e.g., in gossip), or we may learn them from other sources (e.g., onlookers, the media). For these and other examples see Steiny (2007), p. 611-612, and White (2008), p. 30.

33  White (2008), p. 31.

34  White (2008), p. 37.

35  ‘Story-telling sets human interaction apart’, as Podolny (2001), quoted by White (1992), p. 20: relationship ties between people are ‘prisms for meaning as much as they are pipes for connectivity.’ The role of discourse in linking culture and networks is discussed in Mische (2003) and Mische – White (1998). The need to consider the ‘mesh of interrelations’ is discussed in White et al. (2007), p. 546, as part of an analysis of Niklas Luhman’s work and its relation to White’s social network theory. Godart – White (2010) describes the ways in which stories and meaning are related.

36  Godart – White (2010), p. 572.

37  White (1992), p. 40.

38  Godart – White (2010), p. 572. Mische – White (1998), p. 702: These so-called ‘network domains’ may also include other forms of discourse, including symbols, idioms, registers, grammatical patternings.

39  The difficulty of defining institutions is discussed by Mohr – White (2008), p. 488.

40  It comprises a series of dualities: at the individual level, between mind and body; at the group level, between story and network; at the institutional level, between rhetoric and systems of social organisation: see Mohr – White (2008), p. 496, and the Indian caste system, p. 496-500.

41  Mohr – White (2008), p. 508-509.

42  Plato, Republic, 364c-e. Carastro (2006), ch. 6 argues that Plato is intent on associating the practice of writing binding spells with magoi—itinerant specialists from the East who arrived in Greece during the fifth century—and therefore developing a foreign notion of mageia. This, in turn (he argues), has given rise to modern theories that this practice was Near Eastern in origin and introduced by the magoi (he gives Gager [1992], preface and p. 10, and Graf [1994], p. 194-198 as references), whereas, in fact, curse tablets were ‘des produits de la culture grecque qui préexistent à l’arrivée de ces personnages venus de Perse’ (p. 187). Carastro’s stimulating argument and conclusions raise some questions: first, although he is surely right to observe that Plato evokes hostility towards travelling ritual specialists (apparent in other ancient Greek sources), the argument that by the fifth century the term magoi would have indicated a specialist of Near Eastern origin is disputable (see Dickie [2001]). Second, Plato notes only that curse tablets were sold by these individuals (along with other products which had long existed, such as healing incantations and oracles from oracle books), he does not say they introduced them. Neither Graf nor Gager argue for this idea: Graf (1994), p. 194 argues against the idea that these spells were simply spread from the Near East by travelling magoi; while Gager (1992), p. 4 suggests that professionals may not have been much involved in writing the majority of the early Greek curse tablets. Both (see Gager [1992], p. 26-27 and Graf [1994], p. 194-198) give Near Eastern parallels or influences for the practice that date to much earlier than the fifth century. Carastro agrees with the scholarly consensus that binding spells seem to have spread from west to east, and (p. 183) suggests that fear may account for this (for an alternative view, see Eidinow [2007b]). He does not offer an explanation for the similarities between Greek and early Near Eastern spells.

43  These salesmen apparently also provided sacrifices and incantations that can expiate and cure misdeeds by an ancestor; oracles from oracle collections of Musaios and Orpheus; and initiations for delivering individuals and cities, from ills in the next world. There is some debate about the overlap between travelling sellers of oracles (chresmologoi), and those who sold other supernatural services (manteis): Dillery (2005), p. 170, argues that they were thought of as being separate terms, although he admits that they could overlap and ‘even be used to describe the same man.’ He argues that by the fourth century, the chresmologue has almost totally disappeared, leaving the mantis to absorb any popular hostility directed at these independent diviners (p. 219-221). For a similar view on the separate sense of these terms see Flower (2008), esp. p. 58-65, and, in contrast, Bowden (2003); while Parker (2007) offers a carefully balanced view.

44  Plato, Laws, 933a.

45  Plato, Meno, 80b. Meno tells Socrates that he fears he has been bewitched by him, and that Socrates would be well advised not to travel out of Athens, since, if he behaved like this in another city where he was a stranger, he would probably be taken away to prison as a goes (‘magician’).

46  Carastro (2006) and Flower (2008).

47  Sourvinou-Inwood (2000b), p. 42.

48  Sourvinou-Inwood (2000b), p. 44.

49  Ozolian Lokris: IG IX, 1²3:609 (Meiggs and Lewis, no. 13); treaty with curse for anyone who destroyed the text: Argos, IG IV 506; public curse of the city of Teos (c. 470 BC), see Meiggs and Lewis, no. 30 (also repeated at the Teian colony of Abdera, see Herrmann [1981]); foundation decree of Kyrene, see Meiggs and Lewis, no. 5. Rubinstein (2007) argues that the threat of divine punishment continued to be taken seriously into the Hellenistic period.

50  See discussion and examples in Eidinow (2007), p. 165-191.

51  See Faraone (1992), p. 74-85, and App. 4; and Graf (1985), p. 81-98. Olympiodoros of Thebes: FHG IV, 63, 27; oracle to Syedra at Claros: Parke (1985), p. 157-158; Inscription from Thrace: Palatine Anthology IX, 805. Bound gods: Artemis Eurynome at Phigalia (Pausanias, VIII, 41, 6); Artemis Ortheia (‘Lygodesma’) at Sparta (Pausanias, III, 16, 7-11); Aphrodite Morpho at Sparta (Pausanias, III, 15, 10-11). Shut in a box rather than bound: Dionysos Aisymnetes (Pausanias, VII, 19, 6-9).

52  Graf (1997), p. 27; Flower (2008), p. 68 also follows this definition.

53  Parker (2007), p. 122 questions the evidence, but does seem to imply that it is an object of social disapproval.

54  Parker is drawing on modern analyses of the term ‘magic’ to evoke ancient attitudes: (p. 122) ‘it is generally agreed today … magic differs from religion … merely by negative social evaluation.’ See also Versnel (1991b), p. 62 who argues that the use of curse tablets was not socially condoned.

55  In particular, such approaches bring to mind the distinction of Frazer (1911), p. 11 between magic and religion, which states that these practices can be distinguished not only by the number of practitioners, but by the nature of their intentions, and the approach used to attain their ends. Thus, magic is described broadly as a private ritual act (often coercive in approach) designed to foster individual ends, as opposed to the public and sacred practice of religion, which aims to persuade the gods in order to render assistance. Similarly, the idea that ‘magic’ is a private practice similarly draws on early anthropological reflections on the magician as ‘a being set apart’, as described by Mauss – Hubert (1972 [1902]), p. 29.

56  Consider two of the earliest ethnographic works on magic: among Malinowski’s Trobriand islanders, garden magicians practice their rites for the benefit of, and along with, the community (Malinowski [1935], p. 62-68) and (p. 153) ‘There seems also to be a general feeling … that it is not the right thing to carry out private magic’—an aspect of Trobriand life that Malinowski attributes to the importance of gardening, and the potential for competition to cause dangerous envies and jealousies; among the Azande, as described by Evans-Pritchard, although magic was importantly usually performed in private by individuals, its use was widespread ([1937], p. 388-544 passim; for example, p. 432, ‘every Zande, except small children, whether old or young … is to some extent a magician’), and it was practiced by associations. Of course, good magic in these cases is different from witchcraft (which uses psycho-social powers, rather than medicines) and sorcery (bad magic), but good magic can have destructive functions (p. 389) and Evans-Pritchard observes that opinions vary when categorising magic as good or bad (p. 406); all this complexity is part of the point.

57  Compare for example, the binding spell by Phila from Pella in Macedonia (Voutiras 1988), an oracle tablet from Dodona (SGDI 1597) and the prayer of Chryses (Homer, Iliad I, 43-49), all of which beseech the gods for a personal advantage, while both spell and prayer also ask the god to act against particular individuals. As Graf (1997), p. 222-229, has demonstrated, we need to bear in mind the ways and places in which coercion appears in ritual texts, and consider its development in the context of wider social and political changes.

58  On magic-workers see the discussion in Frankfurter (2002). Graf (1997), p. 229, although arguing for the magician’s isolation as a deliberate reversal of traditional (religious) ritual, nevertheless provides a balanced interrogation of this image of the magician, by exploring his role as a healer (p. 229-32).

59  See n. 30.

60  See Faraone (1991), p. 5 and Eidinow (2007), p. 145-147. Gordon (1999) has argued that the presentation of victims in Athenian curses as lists of names is an allusion to that city state’s lists of atimoi, the public debtors, deserters, orthose condemned for homicide, which were set up on tablets for public view (Lysias, 26 [Euandros], 10).

61  DTA 24 and 103, DT 60, SGD 9, 14 and 48; see Eidinow (2007a), p. 168-173.

62  Eidinow (2007a).

63  Godart – White (2010), p. 570-71.

64  Bremmer (2010), p. 21-22.

65  NGCT 14 (later fourth century; sanctuary of Pankrates, Athens) is a text found in the sanctuary of Pankrates, Attica, includes a plea made to Palaimon (who had a presence in the sanctuary); the group of binding spell texts found in the temple of (and dedicated to) Demeter in Knidos (DT 1-14). These particular texts seem to have engaged with a specific cult within the polis.

66  Godart – White (2010), p. 575.

67  Jordan (2000), no. 24; very early fourth century BC.

68  Olive – Myers (2003), p. 183-184. This would be in contrast to a broadcast network where a large number of people get information from a single source, a model that might better fit some aspects of group religious practice. As Olive – Myers (2003), p. 185 points out, the two models may also work simultaneously (for example, a node-to-node network may multiply and deepen broadcast information flows, and information received via node-to-node connections tends to much more influential) and may affect each other.

69  As Olive – Myers (2003), p. 191 notes, ‘people’s attitudes are shaped by those of the people to whom they have network ties, and in particular that the degree of influence will be affected by the homogeneity/heterogeneity of the opinions in the networks to which one is tied.’

70  Sourvinou-Inwood (2000b), p. 55.

71  A structuralist approach like that noted by Burkert (1983), p. 2.

72  Burkert (1983).

73  Col. 20, l. 4 in Betegh (2004), p. 43; for the translation see Parker (2007), p. 116. The word translated as ‘profession’ is techne—a term that, as Parker observes, carries the sense of a craft or an expertise, which is (as the author of the papyrus observes) used as a profession or trade.

74  Hippocrates, On The Sacred Disease, 2, 12-13 and 4, 36 ff. Versnel (1991a), p. 197, n. 40. This passage is seen by some as providing an example of an ancient view of the opposition between ‘magical’ spells or charms vs. pious supplicatory prayer.

75  Graf – Johnston (2007), p. 178-181 also suggests that the ritual activities of the Orphics should be viewed as ‘supplemental’ rather than alternative.

76  Euripides, Hippolytos, 952-954. However, Burkert (1983), p. 11, does draw attention to the presumed isolation of Hippolytos.

77  Theophrastus, 16, 11; Plutarch, Saying of the Spartans (Mor., 224e-f): for example, in a retort given by Leotychidas to ‘Philip, priest of the Orphic mysteries’, who had, apparently, something of a reputation not only for his poverty, but also for his habit of complaining about it. In the Phaedrus, 244d-e, Plato’s discussion of curing madness through purifications and sacred rites at least suggests that the suppliers of these services may not always have been so resented and reviled as they appear in the description given above from the Republic, discussed by Burkert (1983), p. 5. The Phaedrus passage mentions Dionysos rather than Orpheus (implied in the description of the Republic). The evidence often creates an association between followers of Orpheus and followers of Dionysos, and it is not apparent that a clear distinction can be made; see discussion in Graf – Johnston (2007), p. 142-143.

78  Ansell (2003), p. 126-7.

79  Possibility of communities of initiated: Graf – Johnston (2007), p. 163-164. The term ‘orphikoi’ read on one of the bone plates from Olbia may also suggest this, see discussion Burkert (1983), p. 4. Private cults: see Henrichs (1983), p. 147, with p. 225, n. 99, who accepts Herodotus, IV, 78-80 as evidence for private and ecstatic cults of Dionysos at Olbia in the fifth century. For women, see Aristophanes, Lysistrata, 1, with the scholium; further discussion in Versnel (1990), p. 149. For the co-existence of a private male cult with a female ecstatic cult, see Euripides, Bacchae, 465-490, with Henrichs (1983), p. 147, although it is difficult to know to what extent this can be treated as evidence for actual practice.

80  Public and private cults somehow linked, see Henrichs (1969) and Seaford (1981), discussed Graf – Johnston (2007), p. 141. Plato, Phaedo, 69c contains a verse of Orpheus which differentiates between those who are worshippers of Dionysos and those who are Bacchoi—perhaps suggesting a similar smaller private cult within a larger public one, a view supported by Graf – Johnston (2007), p. 143.

81  Sourvinou-Inwood (2000b), p. 55.

82  If it means the foundation, then there is, indeed, evidence that this was sometimes the case, at least in the Hellenistic period, as suggested by an inscription from Magnesia, I Magnesia 215 (Parke – Wormell II [1956], no. 338), an oracle from Magnesia that describes the establishment of a sanctuary of Dionysos, and involves the fetching of maenads from Thebes: part of the ‘official religious apparatus of the city,’ see Versnel (1990), p. 142 for quotation and further discussion. Sourvinou-Inwood (2000b), p. 55, n. 60, gives as evidence LSAM 48, a decree from Miletos, discussed below—it does seem to suggest that not all Dionysiac ritual was polis organised.

83  See Versnel (1990), p. 149.

84  Sourvinou-Inwood (2000b), p. 54 states that both Dionysiac rites and Eleusinian Mysteries were encompassed within polis religion.

85  Pausanias, X, 4, 3 (Athenian) and X, 32, 7 (in Parnassos); Pausanias, VI, 26, 1 (a feast outside Elis); see also Plutarch, Mulierum Virtutes, 13 (Mor., 249e), discussed further below and see Versnel (1990), p. 138. Maenadic activity seems to be referred to at Sophocles, Antigone, 1149-1151; Euripides, Ion, 550-552; Aristophanes, Clouds, 603-606; Aeschylus, Seven Against Thebes, 498 and 836. Supernatural revels that evoke the cult of Dionysos referred to at Euripides, Bacchae, 306-309 and Sophocles, Antigone, 1126-1130; see for discussion and references Holzhausen (2008).

86  In contrast, there does seem to have been a polis-instituted cult of Dionysos: goat sacrifices to Semele and Dionysus are recorded in a fourth-century sacrificial calendar from the deme of Erchia, see Sokolowski LSCG no. 18, A 44-51, 33-40; see discussion in Henrichs (1983), p. 144.

87  A peace-making role has been evoked for the Sixteen Women at Elis, who seem to have had a reputation as peacemakers, according to the stories recorded by Plutarch and Pausanias. On the origins of the sixteen, chosen from sixteen cities in Elis, in order to help promote harmony between them see Pausanias, V, 16, 4-6. Plutarch (Mulierum Virtutes, 15 [Mor.,251e]) relates a story, dating to the third century BC, which does identify the women with the welfare of the people (specifically the women) of Elis, but also raises questions about allegiance to political authority. See discussion by Goff (2004), p. 191-193. Henrichs (1983), p. 220, n. 57 suggests that these congregations and the Delphic and Attic Thyiads ‘reflect ancient clan structures’. See Henrichs (1983), p. 150 for a discussion of the ways in which cults of Dionysos seem to have encouraged a sense of intra-group identity among worshippers.

88  See Aristophanes, Birds, 874-875; Wasps, 9; Lysistrata, 387 f., (where the rites of Adonis are likened to those of Sabazios); Kratinos, Horai, fr. 566 (ed. Kock), with scholia.

89  Demosthenes, 18 (De Corona), 129 and 259-260, and 19 (On the False Embassy), 249 and 281.

90  Inscription of Miletus: LSAM 48 (see n. 82 above and discussion in Henrichs [1969]); Versnel (1990), p. 145 suggests this document demonstrates ‘a drastic routinization of maenadism in the context of the official cult practice.’

91  Sourvinou-Inwood (2000b), p. 48.

92  Deme-level activities discussed in Humphreys (2004), who suggests, for example, that the state became increasingly involved in managing cults because of its role in sanctuary loans to the state (p. 139), and discusses the apparent autonomy of the deme festival calendar. Aleshire (1994) argues that state control of cult was primarily focused on financial affairs. See also Faraone (2008) and Boedeker (2008) for discussion of the interaction between polis and smaller units of society, including gene and oikoi.

93  Aleshire (1994) provides a thorough and convincing analysis of the difficulties of describing ‘state cult’, and/or identifying the ways in which a polis exerted control over religious life.

94  The difficulties of describing variations in the details of ritual activity within the schema of polis religion are indicated in some of the language used by Sourvinou-Inwood herself: for example, she describes metic religious as being ‘“encompassed”, in the diluted sense of being sanctioned, by the polis’ (2000b, p. 49).

95  This idea of multiple connections is also conveyed by Sourvinou-Inwood (2000b), p. 46, who describes the ‘manifold identity elements, not all of which were dependent on the others’ of each individual. The ‘network or web’ pattern of cult connections for an individual within Athens is also described by Aleshire (1994), p. 10: ‘the cult associations of a given individual were multiple … cult associations could and usually did cut directly across the political hierarchy.’

96  A point also made by Aleshire (1994), p. 13.

97  Using the network theory language of Barabási – Albert (‘Statistical mechanics of Complex Networks,’ Reviews of Modern Physics 74 [2002], p. 47-97, cited by Collar [2004], p. 147), we might call Athens a ‘hub’, that is, a node that is massively well connected.

98  Mohr – White (2008).

99  The process by which the polis consolidated its power through absorption of social and political social formations is vividly described by Gabrielsen (2007) and Ismard (2007).

100  Josephus (Against Apion II, 267) offers a list of men ‘put to death by the Athenians because they uttered a word against their gods.’ How many of these men were actually prosecuted is famously debated, see, in particular Dover (1976). The women are Phryne, Theoris of Lemnos, and Ninon: see Eidinow (2010).

101  In the trial of Socrates, for example, we learn about his connections to politically dangerous individuals and his association with young men: see Hansen (1995). On Ninon, Theoris and Phryne, see Eidinow (2010).

102  Graf (1997), p. 59: Cicero, Laws II, 21; Paulus Sententiae V, 23, 15; FIRA II, 410; and Codex Theodosianus IX, 16, 7. The Peiraeus decree IG II2 1177 = LSCG 36 (mid-fourth century) constrains possible religious/ritual behaviour during the Thesmophoria, including forbidding anyone to assemble thiasoi; an Athenian decree (IG I3 78a, 55-59 = LSCG 5, c. 422 BC / 418 BC) dictates that no one should place an altar in the sacred territory of the Pelargikon without official consent. See also a lex sacra from the cult of Asklepios at Erythrai (380-60 BC), LSAM 24, I. Erythrai 205.

103  Sourvinou-Inwood (2000a), p. 37.

104  There is not room to consider this in depth here, but some specific examples for which network theory might provide a particularly useful approach, range from the conceptual questions involved in the figure of Herakles as both god and hero (Shapiro 1983); or to the specifics of cult practice (for example, such as Pausanias describes [V, 13, 3] in which a cult complex contains a hero and a god, where some types of participation in the cult of the hero renders the worshipper unfit to come before the god, see Kearns [1992] p. 86 and 87). Also, to explore the variety of ways in which communities represented their gods as being involved in the life of the polis (this varied, as Guettel Cole [1995] has demonstrated).

Haut de page

Pour citer cet article

Référence électronique

Esther Eidinow, « Networks and Narratives: A Model for Ancient Greek Religion », Kernos [En ligne], 24 | 2011, mis en ligne le 11 janvier 2014, consulté le 28 août 2014. URL : http://kernos.revues.org/1925 ; DOI : 10.4000/kernos.1925

Haut de page

Droits d'auteur

Tous droits réservés

Haut de page