Holy Places, Vital Spirits

Before shrines, of course, it was the ekkl ē sia, the religious assembly itself, that served as such a theater for the articulation of spirits. Already in the first century, such gatherings seem typically to have led to possession by “angelic” spirits (1 Cor 13; Rev 1–3). But it is a letter of Firmilian of Caesarea (256 c.e.) that describes in especially dramatic terms the dynamics that we will see in late antique shrines. Firmilian notes how, in the wake of a series of disasters in the region, a woman in his group began to enter an ecstatic state and prophesy “as if stirred by the Holy Spirit.” The group seems to have attended closely to this spirit's words about imminent signs and prodigies. The possessing spirit sanctified the Eucharist, performed baptisms, walked barefoot in the snow, and announced it was returning to its home in Judea. The congregation, on its side, was rapt. When an outside exorcist showed up, however, he denounced the spirit as, rather, a “demon,” demonstrating (according to Firmilian) “that that which was before thought holy was indeed a most wicked spirit [nequissimum spiritum].”Footnote ⁸ Firmilian tries to balance the spirit's appearance of holiness (i.e., its behavior within the roles expected for holy spirits) with his subsequent recognition of its demonic identity. These oscillating perspectives were made all the more acute given that the woman was performing sacraments.Footnote ⁹ But of most historical significance in this letter is Firmilian's depiction of the real ambiguity of possessing spirits and the imprecision of people's attempts to classify spirits as either “holy” or “demonic/wicked.” The possessed woman functioned with considerable authority for some time as the vehicle of a prophetic spirit, and even after the exorcist's pronouncement there was probably some conflict over the definition of the spirit, reflected in Firmilian's own ambivalent tone. We note also the diachronic process of performance, negotiation, and interpretation through which the spirit gained one identity, then another, fulfilled esssential roles like prophecy and then seemed to switch roles to embody the demonic itself—that is, to oppose the central spirit, Christ. This process, I would argue, is a perennial one, due to the inherent ambiguity of spirit possession (especially in new apocalyptic movements) and the urgency that polarized classifications lend to the definition of spirits.

The appearance, authority, and interactions of this spirit would have owed much to the historical and social circumstances of ekklēsiai in third-century Caesarea: apocalyptic literature that implied the resurgence of biblical prophecy, ecstatic leadership in the New Prophecy movement, rumors of persecution and martyrdom, and beliefs in the ekklēsia as site for human/angel interaction.Footnote ¹⁰ All the same, the social dynamics of possession that are discernible in Firmilian's letter do prepare us to comprehend phenomena of the following centuries that revolve around not congregations, but particular shrines. Thus we shift to mid fourth-century Egypt, where Athanasius of Alexandria was lamenting (370 c.e.) a new practice among Egyptian Christians in the countryside of congregating in tombs that had acquired reputations as sites of martyrs’ relics, for a curious ritual process. “Many people,” notes Athanasius, “who had unclean spirits gain healing in the martyria, claiming that the martyrs’ spirits come upon the demons [] … and the demons cry out.” However, he complains, people also “seek to see the demons [],” apparently through possession at these tomb shrines. Then, with the demons manifest in devotees’ bodies, “these people give glory to them and ask them about what will happen [] … they dare to question the unclean spirits!” They assume “the demons are the prophets of the martyrs. But the martyrs did not confess that they would speak through the demons! [].”Footnote ¹¹

Although he reports people going to tomb-shrines to experience demons, Athanasius is clearly not describing simply a cult of exorcism, of the expelling of spirits, nor does this seem to be the positive type of spirit possession that we know from the New Prophecy movement of Asia Minor and North Africa. There is clear ambiguity in these Egyptian Christian possession forms. As Athanasius depicts the situation it is demonic spirits that emerge at the shrines, due to the awesome spirits of the martyrs. Once they become manifest, however, the “demons” become oracular voices—even “the prophets of the martyrs”—to which devotees could address their questions. The spirit possession Athanasius observes evidently involved some fluidity in function—that is, in social utility—between spirits deemed “demonic” and spirits deemed “martyrial.”Footnote ¹²

Less than a century later (mid fifth century c.e.) the abbot Shenoute of Atripe echoes some of these complaints about the popular relic cult, but he aims his sermon at those illegitimately establishing topoi (places) for relics, in or by churches, where people claim to have seen lights appear. After detailing how people “sleep in the tombs to gain visions and … question the dead about the living”—familiar and longstanding incubation practices—Shenoute refers to “other sorts of places—of divination [] or of mediumship [].” It would seem that the “places” are, again, shrines, while the word eggastrimythos denotes ecstatic speech, such as the medium of Endor provided for Saul in the Bible, but now provided, Shenoute says, by “Christians and clerics (who serve) in the house of God.”Footnote ¹³ While in the biblical account the medium of Endor was supposed to have channelled the spirit of a dead Samuel, Christian writers of the second and third centuries debated whether such a voice would have to be intrinsically demonic, and Shenoute would certainly be commenting on the ambiguity of possession by spirits in his time: Are they dead martyrs or demons? Thus spirit possession of some sort appears to have arisen again (“in this time []”), Shenoute says, and in connection with relic veneration and new martyr shrines.Footnote ¹⁴

As we saw possession initially occuring in the social context of the sectarian congregation, so too in late antique Egypt the locus of possession and of the interpretation of spirits could shift between spontaneous activity at actual shrines, such as Athanasius and Shenoute report, and social bodies: in this case, the socially-embedded clairvoyant powers of charismatic monks. One story from the fourth or fifth century involves an Egyptian monk near the village of Boushēm in whom “a spirit of divination [], that is [] the demon of falsehood, came to dwell, so that he would tell people about numerous events that were going to happen to them: he would say ‘they will happen,’ and they happened, and he spoke about the waters of the Nile and about many other worldly events that he learned about from the spirits.”Footnote ¹⁵ Once again this Coptic author communicates the ambiguity of a spirit, revered by local villagers yet reviled by certain church leaders.

This phenomenon was hardly restricted to late antique Egypt, however. The same ambiguity of spirits’ functions—potentially both harmful and clairvoyant—appears in the west as well, in legends associated with the cult of St. Martin in Tours. The late fourth-century legend of Martin's life describes a time of great public anxiety about imminent barbarian invasion, when

Here the possessing spirit provides a helpful service for a town in panic by clarifying the nature of a real crisis: the possibility of imminent invasion. Other places associated with Martin likewise continued to inspire these manifestations even in the sixth century. There is a “man accustomed to utter many words, frequently to speak the language of unknown people, to prophesy future events, and to confess crimes” whose demon is expelled shortly after arriving at Martin's cell; while another demoniac announces in the middle of the Martin church that the Roman general (and later Frankish “king”) Aegidius had triumphed over enemies due to Martin's assistance.Footnote ¹⁷ Again, the oracles, albeit demonic in origin, answer questions, reframe misfortunes, and resolve social tensions.

The final two examples come again from the eastern empire: Constantinople and its environs. The fifth-century church historian Sozomen reports (393 c.e.) that a “certain person afflicted by demons arose in the temple of God [that is, the Theodosian shrine to John the Baptist's head in the outskirts of Constantinople] and insulted John the Baptist” and then proceeded to give an oracle about the current war: “You conquer me and lay snares for my army,” as if in the voice of the enemy's genius. “The persons who happened to be on the spot and who were waiting impatiently to learn some news of the war, were amazed.”Footnote ¹⁸ Such performances were not unusual in early Byzantine Constantinople, it seems, for the seventh-century Miracles of St. Artemios depicts a time when “there were a very large number of possessed [δαιμоνιῶντας] in many churches” in the city to whom one might go for oracles. In one story a man robbed of his clothes is advised to head to the St. Panteleemon shrine where “someone [τινα] was there dispensing information [ἐπίστασιν διδόντα]” who might reveal the robber. But when he gets near and hears “the cry of the possessed one,” he balks, realizing that “now [he was] forsaking God and approaching demons” to resolve his robbery. Thus demon-possession—here of the afflicted, residing hopefully in a healing shrine—had gained a perennial civic function, with some possessed becoming local manteis for the desperate.Footnote ¹⁹

In these cases from across the Mediterranean world the possessed individual is not simply cured, the demon expelled, and the holy man or shrine victorious. Rather, the possessed person, in full embodiment of some otherworldly spirit, becomes a valuable mouthpiece for oracles and an instrument of divination. And he notably performs these services—enacts the spirits—in a shrine of some recent vintage. What does this phenomenon signify?

Literary and Phenomenological Considerations

There is, of course, a literary character to these gnomic pronouncements by unclean spirits, reminiscent of the Gerasene demoniac's utterance in the gospel of Mark, “My name is Legion, for we are many” (5:9). This literary character certainly informs the dramatic exorcism scenes in the Martin legends. Might such biblical typologies have so determined the many stories of oracular demons that no historicity can be assumed at all—that they were created out of whole cloth? This conclusion would be too extreme, for biblical typology does not in itself negate historicity; and indeed the persistence of the very motif of the prophetic demon in the literature of Christian thaumaturgy would have sanctioned the behavior itself in areas where such texts were read. But even more, the range and consistency of the witnesses under discussion suggest that we are dealing with a type of historical occurrence, not just a literary trope. Athanasius and Shenoute offer fairly direct witnesses to real oracular phenomena at martyria in late fourth- and early fifth-century Egypt, while even the more legendary sources like the Life of Martin describe “demons” offering oracles that are immediately embedded in historical events and socially relevant.Footnote ²⁰ The late antique texts thus place demonic utterances in historical context, as if the possessed person were responding to actual situations of political anxiety. They reflect, if not particular historical occurrences, a type of possession performance that authors and their audiences knew to exist.Footnote ²¹

Assuming a degree of general historicity, then, let us consider the nature of the possessions themselves, and more specifically their labelling as “demonic.” It would seem that for the authors and for their cultures the term daimōn conveyed both censure, referring possession and mantic phenomena to an official demonology, and a fundamental imprecision in the identification of possessing spirits in late antique Christianity.Footnote ²² We certainly see this imprecision in the elision of martyrs’ spirits and “demons” in Athanasius's complaint about possession at tomb-shrines. In a similar vein the fourth-century Cappadocian writer Gregory of Nazianzus celebrates the martyrs “by whom demons are expelled and diseases cured, by whom occur apparitions [ἐπιφάνειαι] and predictions [πρоρρήσεις],” a remarkable series of prodigies notable both for their diversity and for their dependence on the bodies of shrine devotees for their performative expression. A single ritual sphere thus serves as theater for both possession and useful prophecy, martyrs’ spirits and “demons.”Footnote ²³ The possession, therefore, is in every case ambiguous: socially beneficial, but brought by spirits at best peripheral to the Christian pantheon.

Furthermore, as we have seen, the historical phenomenon we are describing involves both semi-spontaneous forms of possession, such as the texts describe among laity at saint-shrines, and a kind of beneficial spirit mediumship, an ongoing social role, like that of the prophetess in third-century Caesarea and the monk from the Egyptian village of Boushem. Further cases of this kind of mediumship appear in late antique Christianity: a Christian priest in Gaballōn, for example, whose great “signs and wonders” won him great regional renown—until his powers were reassigned to a demonic archontikon.Footnote ²⁴ Such individuals gained renown as regional oracles or prophets by virtue of their possession by, or familiarity with, certain spirits. At some point in the period when the hagiographer recounts them, the oracles and their spirits fell afoul of church authorities—unlike, for example, the late fourth-century Egyptian monk John of Lycopolis, whose “gift of prophecy” [πρоφητείας χάρισμα] allowed him to maintain a credible reputation with church and civil authorities for “reveal[ing] things hidden in the future … and predict[ing] the rise and fall of the Nile.” He too “used to foretell when some divine threat was going to [occur] and exposed those who were to blame for it.” But despite largely identical claims and services to those deemed demonic, John's spirit (or basis of clairvoyant authority) remained legitimate.Footnote ²⁵ Thus charismatic authority, especially when achieved through the claim or reputation of familiar spirits, tended to be evaluated in conflicting and probably even oscillating ways.

These latter “mediumship” cases in particular show us that the term daimōn itself did not indicate a distinct form of possession performance but rather served as one possible evaluation of possessing spirits.Footnote ²⁶ These were spirits whose performance in possession evidently could fit quite neatly into local needs and styles. And yet, like modern Middle Eastern zar and jinn demons, they were ambiguous enough to be potentially dangerous. They required either mastery or exorcism.Footnote ²⁷

Historical Considerations

How do these phenomena fit into the larger religious situation of the late antique Mediterranean world, whose landscapes were assimilating a newly institutional Christianity and its shrines? First of all, in accounts from places like Tours and Constantinople, spirit possession is explicitly linked to historical uncertainty or, in the cases of Athanasius and Shenoute, to larger conflicts about the locus of the holy in a Christianizing landscape. It is no surprise that such social anxieties would precipitate a need to see and hear the spirits, to witness their clairvoyance and authority in settling matters of great social and political import. Yet it seems that the actual identity or class of the spirits that appear emerges only through a process of negotiation between possessed performer, audience members, institutional emissaries, and so on. What might begin as a martyr redivivus in one setting might end up as a dangerous archontikon; what might begin as an ancestral or earth spirit might end up as a spirit of prophecy with full ecclesiastical sanction.Footnote ²⁸ And each such process of spirit-definition would likewise culminate in a ritual performance: exorcism for the one, spirit mastery and oracular performance for the second.Footnote ²⁹ Thus the resolution of uncertainty would be sealed with drama.

Secondly, the clustering of these witnesses in the late fourth and fifth centuries reflects quite immediately the religious changes occurring in local Christianities of the period. In Egypt, Gaul, and elsewhere a thriving local religious culture had persisted into the Roman period, with diverse shrines, demarcated points in the landscape, varieties of beneficial, ambivalent, and hostile spirits, and the ritual traditions to deal with these supernatural beings.Footnote ³⁰ But, with the decline of traditional cult centers, the development of an autochthonous Christian local piety, and the struggle of an institutional church for hegemony, much attention came to be directed to the locations and powers of holy places as sites that embraced local experience and that situated the new pantheon of Christian “spirits” in the landscape.Footnote ³¹ Now, in the late fourth and fifth centuries, new local religious cultures had to comprehend and mediate the new religious system in locally authentic ways, all in a system presided over by bishops.

This brings us to the third historical factor in these cases of spirit possession after the third century: its predominant occurrence in shrines. We see spirits appearing in new, ad hoc tomb shrines in upper Egypt;Footnote ³² in churches and shrines associated with Martin of Tours;Footnote ³³ and in the new “temple” [νεώς] of John the Baptist's head outside Constantinople (and the church of St. Panteleemon within). These were buildings dedicated to invoking the presence of the saints whose bones lay beneath. The eyes and postures in wall-paintings would signify saintly presence no less than the false-doors inside Egyptian tombs of another era, through which the soul of the deceased was supposed to enter and exit. Visitors were familiar with an iconography that mediated supernatural beings. Sermons and stories, too, not only brought martyrs’ sufferings to life but envisioned their heavenly bliss and promised their eternal contact. Through clouds of incense and the dim light of oil lamps the martyr's spirit would become tangible to those crowded inside.Footnote ³⁴ And then, already brimming with the power of these new (if often home-grown) martyrs’ spirits, the martyria came to attract, as Athanasius describes it, the manifestation of other spirits.

The Performative Construction of Demons

But who were these “other spirits”? There exist a few cases in which the spirits are said to identify themselves—as, say, Jupiter and Mercury in one description of demoniac possession associated with the Martin cult.Footnote ³⁵ But more generally these “other spirits”—these “demons” peripheral to the now-central Christian pantheon—were not recognized as the “old gods,” specific ancestral spirits come to invade and depart via dramatic exorcisms. The “demons” that emerged at these fifth- and sixth-century shrines reflected, I would argue, the incoherent possession forms of a culture conscious of its abandoned gods but utterly invested in the power of the saint. They were fundamentally Christian performances, meant to articulate the power of the saint; but in constructing and fleshing out the spirits of the periphery they also involved quite local acts of bricolage. That is, they drew from memories and anxieties of the local sphere, and from a folklore combining exotic biblical legends with familiar landscape spirits. Again, we might compare them to modern Middle-Eastern zar and jinn spirits, whose identities with peripheral lands and peoples are likewise constructed in performance and in counterpoint to the central power of sheikh or marabout. Possession allowed people to participate in Christianization, sometimes critiquing the displaced pantheon through mimesis of its spirits, sometimes inventing an antithesis of the new pantheon with some type of chaotic spirit. One might even acclaim the new pantheon, using the peripheral spirits—those rendered marginal and ambiguous through Christianization—as the voice of ancestral tradition itself. For example, one demoniac in Rome proclaimed “the God of St. Callixtus,” while in late fourth-century Nola (Italy) the demoniacs would declare that St. Felix was afflicting them. Indeed, it seemed here that the demons themselves were announcing “that Christ shines out in the person of his saint—and they prove[d] it by trembling with shaking limbs and wagging heads.”Footnote ³⁶ By such spectacles local audiences could witness Christ's (or a saint's) reception, as it were, by the familiar spirits of neighborhood and village. There is inevitably a ludic, creative element in articulating through possession a shift in pantheons.

Of course, there is also a distinct locative aspect to these phenomena, where the place of the martyr or saint becomes the place also of the demonic. For the early Christian laity the demonic was envisioned more in geographical than spiritual terms—as a threat revealed and distributed in landscape features, especially the urban landscape, where a casual visit to the baths or the theatre (obvious demon habitats) might precipitate demonic possession.Footnote ³⁷ In this worldview the holy topos of the martyr served not only as counterpoint to demonic spaces but also as the very resolution of a polarized landscape: here the demons would emerge and be vanquished. But on another level the new cult centers that Christianity brought to regional cultures had to accommodate the full range of local peoples’ supernatural experiences and potentialities. Just as at Delphi, where (according to Plutarch) an unprepared Pythia might get overtaken by a “mighty and baleful spirit [ἀλάλоυ καὶ κακоῦ πνεύματоς]” instead of the oracular god,Footnote ³⁸ so too at the Christian shrines the declaration of Christ and his saints’ immanence through liturgy, paintings, and incense invited the complementary embodiment of other spirits: peripheral spirits, “demons.” The space of the saint became the space of all the spirits, in conflict and collusion.Footnote ³⁹

There was obviously an element of spectacle in the construction of demons, for people went out to the shrines in order to experience the new pantheon and its powers through dramatic possessions; and naturally they expected some consequences of the demons’ appearance, whether exorcisms, oracles, or simply drama. Jerome describes his friend Paula's experience in a late fourth-century saint's shrine in Palestine, among “men howling like wolves, baying like dogs, roaring like lions, hissing like serpents, and bellowing like bulls”—a veritable theater of demoniac possession.Footnote ⁴⁰ Paulinus of Nola marvels at a demoniac who hangs upside-down from the balustrade next to St. Felix's tomb, his robe miraculously staying up to cover his genitals.Footnote ⁴¹ A man visiting St. Artemios's shrine in Constantinople for a testicular condition turns out also to have an “evil spirit,” leading him to hang suspended before the icon of Artemios “as though his hands were tied by chains, hovering one cubit above the floor, and yelling loudly, so that all [those supplicants likewise incubating in the shrine] were astonished by the sight and were cowed by fear.”Footnote ⁴² Such performances served as the authenticating spectacles of shrines, initiated through the centralizing experience of the shrine itself.

Finally, behind these performative and architectural contexts for the definition of these peripheral spirits, we must recognize the very agency of the local milieu in appropriating the shrine—often an official structure—as the place of the spirits and declaring it to be the site for possession or the resolution of possession, with or without the systematic encouragement of bishops. The spirits at the shrine of St. Felix in Nola do not simply release their victims but amazingly seem to prolong their possessions. While Paulinus attributes this phenomenon to the saint's effort to expiate more fully the victims’ sins, we may perceive some popular agency in maintaining the drama, not simply acquiescing to the power of Felix.Footnote ⁴³ Whether martyrs’ bones or John the Baptist's head, St. Martin's body or patriarchs’ remains, it was the possession of the laity and its experts, even more than the reading of martyrological passions, that served to articulate the praesentia and potentia of the relics inside.

The focus in this paper on local audiences and pilgrims as principal agents in the representation of demons (whether oracular, heathen, or animal-like) obviously sets aside that central figure in early Christian literature, the exorcist. It is the exorcist who is typically credited with the authority to govern lay possession. Literature of the Roman period, from Luke-Acts to Philostratus, tends to focus on this figure as the essential dramatic counterpoint to the demonic. Through his technical ability to identify and subdue chaotic spirits, or simply through his representation of Christian power, the exorcist becomes both ritual expert and hero, and the literature depicts the possessed as acting largely in aggressive, then passive, counterpoint.Footnote ⁴⁴ Christian ritual manuals from the third, fourth, and later centuries outline such exorcisms in practical terms, yet the late antique materials discussed here ignore or diminish the role of exorcists in actually triggering demonic displays in others.Footnote ⁴⁵ Instead, they depict the forms of the possession state as initiated through the mere presence of holy men or saintly relics. It is possible that scholars have put too much historical emphasis on the roles of exorcists as instigators of possession states—that instead we should regard them as shapers of demonic possession, authorities in the definition of spirits that laypeople initially construct themselves.

Towards a New Framework for the Christianization of the Mediterranean World

The late antique spirit-possession phenomena described in this paper would not have required prior traditions of beneficial spirit possession, like shamanism, ecstatic oracles, or possession “cults.” Nor did they necessarily involve the resurgence of prior religious forms, as many have viewed (for example) New Prophecy in the context of Phrygian cults.Footnote ⁴⁶ To be sure, literature of the Roman period is rife with cases of individuals possessed by “spirits,” “holy spirits,” “pythonic spirits,” and sometimes “gods” who gained prestige as prophetic oracles.Footnote ⁴⁷ There were certainly areas of the ancient Mediterranean world that viewed the embodiment and mantic voice of an ancestral spirit as an essential role in society. There may well have been traditions of ecstatic mediumship in the regions of these saints’ shrines; but the approach in this article to the upsurge of demons in shrines does not depend on it.Footnote ⁴⁸

Rather, spirit possession in late antiquity was an historical epiphenomenon of the process of Christianization itself, involving three critical historical developments: first, the increasing hegemony of a new, totalist pantheon that, second, polarized ancestral spirits as liminal and chaotic, and that, third, was “realized” in the landscape through the fixity of particular places as the loci of spirits and the manifestation of spirits. In this final section, these stages of Christianization will be explained in more detail, in light of both the late antique materials introduced above and current anthropological work on “conversion” and spirit possession.

Let us begin with the supernatural world in advance of Christianization. What used to be called “paganism,” but which is better described as (Egyptian, Gallic, or other) local religion, involved perpetual negotiation with a range of ancestral and landscape spirits, as well as more well-known gods. A gold lamella from Romania adjures the “daimōn menacing here” to go and instead(!) afflict Julia Cyrilla; while a Sicilian Jew needs an elaborate amulet to protect him from the local Artemis.Footnote ⁴⁹ In Roman Egypt alone we find one letter extolling “our ancestral gods [who] continually assist us, granting us health and safety” and another lamenting how “one of the gods sent [my daughter's illness] in malice, and likewise let him remove it!”Footnote ⁵⁰ Here too, apotropaic gods like Bes, Toutou, and Horus stood against a range of chaotic and capricious beings, some of which we find listed in magical and mortuary texts.Footnote ⁵¹ For a woman of the first century it was a local Osiris to whom one should address a complaint about a husband's sexual reticence.Footnote ⁵² Negotiating among these gods and spirits involved family and local traditions, individual self-determination, and the appeal to temples and ritual experts for aid. By what god are my children stricken? How can I assure my ancestor's continued favor on my family? To whom should I place my vows for a successful trip to Alexandria? It is this kind of culture of religious agency and the negotiation of spirits that we must imagine encountering Christianity.Footnote ⁵³

Now, any new religious ideology brought into this kind of local culture—by religious experts or holy men, for example—will propose and often bring about a reorganization of pantheon and cosmology as well as the classification and recognition of supernatural forces. Which spirits are beneficial, which harmful? By what ritual means are people truly protected, and which rituals amount to sorcery? This is a pattern repeatedly observable in ethnographic studies of religious conversion, but it especially covers the radically totalist Christianity of the fourth century, which made a profession of discerning and opposing the demonic.Footnote ⁵⁴ This reorganization often simply marginalized spirits associated with the landscape, ancestors, or divination, while the new transcendent spirits were centralized.Footnote ⁵⁵ Occasionally in history this marginalization has become more extreme, turning into a process of polarization in which missionaries, prophets, and religious reformers render ambiguous spirits (daimones, jinn, even orishas and loas in Afro-Caribbean cultures) as utterly and actively evil, to be shunned and exorcised for the sake of the central God. While exorcisms might eventually allow more innovative, even helpful performances by these evil spirits, their initial experience among groups unprepared for the “domestication” of such spirits may be terror.Footnote ⁵⁶

The possession materials in this article demonstrate above all that, in receiving this reorganized or polarized Christian pantheon, people in local communities were not passively acquiescent. Instead, they actively engaged in the mimesis and articulation of the new configuration of spirits through ritual performance and possession. This is the crucial element of agency in the religious transformation of the local sphere. Rather than depending on missionaries and prophets, it is quite often laity themselves who demonstrate the new spirits in action, the old spirits in submission or resurgence, or combinations of both—an overall dramaturgy of religious transformation, Christ's victory, and often the continued relevance and potency of ancestral spirits. Another aspect of this local dramaturgy is the opening up of new social roles: for possession specialists, like some Egyptian holy men, and for nonspecialists, whose possession often took the form of affliction.

But the late antique materials show that possession is also a process. As the letter of Firmilian of Caesarea shows most vividly, initial possession states, especially among nonspecialists, have to be shaped and defined so that people recognize the possessing spirit as a “demon” rather than a “god,” or a “holy spirit” rather than an angel. But how do people distinguish the demonic from the holy in possession? While efforts to establish hard-and-fast rules for discerning spirits have considerable antiquity, the recognition of demons and holy spirits in fact has always involved a process of negotiation—sometimes over years—between the possessed person, the community, and (often) institutional experts.Footnote ⁵⁷ Out of that process emerges an established role for the beneficial spirit or else the dramatic exorcism of a demon—or in many historical cases the first followed by the second.Footnote ⁵⁸ Numerous accounts of spirit possession in early Christian literature show the debates involved in evaluating and labelling the spirits themselves over time—especially when those spirits (and their bearers) had gained some regional cachet before others pronounced them false or demonic.Footnote ⁵⁹ In the stories of the spirit-bearing monks mentioned above, possession led to credible service as a master of spirits, and it was only later ecclesiastical opponents who pronounced their spirits “demons.”

This process of negotiation does not end if a spirit is pronounced—and then popularly regarded as—a “demon.” Rather, this recognition simply ushers in a whole new set of questions for audience and experts alike. Are demons simply to be expelled, or should they first “perform” their evil, their allegiance to Satan, their affiliation with heathen cult, or their erotic or violent natures? Perhaps the performance of demonic characters comes to serve as a regular shrine spectacle, as we saw at Nola and Tours. In one notable modern case young Sinhalese women journey to the Catholic shrine in Kudagama and there manifest demons to the fascination of the crowds: demons named after Hindu gods, who put the girls through vividly erotic performances.Footnote ⁶⁰ The persistent manifestation of demons at shrines thus can maintain the vitality of the enchanted landscape. It may not be vital in the old way, where tree- and house-spirits had their names, habitats, and modes of supplication, but even as evil demons, devils, and opponents of Christ, their acknowledgment in ritual and possession demonstrates to the community that a spirit world remains. We can learn from the many ethnographic studies of the Christian demonization of local spirits that the articulation of the demonic in ritual offers a real reassurance in the tangibility of spirits, even if they are only to be dominated and bound by the authority of a St. Martin, St. Antony, or St. Sissinios.Footnote ⁶¹ If contemporary jinn and zar rituals can again shed light on these phenomena, the ranks of demonized spirits can still play a role in local culture, even lending themselves to healing and divination rites unattributable to the transcendent spirits of the central pantheon—to Christ or Allah. Indeed, their penetration of the lives of villagers often involves a moral dimension: popular values are affirmed through interaction with demonic spirits.Footnote ⁶²