Civilization
The Kadagi
The oracle of the eastern Georgian highland shrines — the man or woman seized by a khati and made its first-person mouthpiece, the one shrine office nobody applies for, rendered at the public-record distance.
In the shrine-year of a Pshav or Khevsur village there comes a moment when a person’s grammar changes. A man or a woman who a breath earlier stood among the others, saying what the shrine wanted said, leans forward and begins to speak in the first person of a god. It is no longer the khati asks but I took you; the demand, the reproach, the warning of a coming trouble all issue from a human throat as the words of the being seated in the stone. The Georgian highlands call this person the kadagi — the mouthpiece of the khati, the one through whom a sanctuary of the eastern mountains says aloud what it wants. And the office begins not in a decision but in an affliction. No one becomes a kadagi by applying for it. The god does the appointing.
The mouthpiece of the khati
The word carries the office. Kadagi (ქადაგი) belongs to the ordinary Georgian verb for proclaiming or preaching, and a kadagi is, at the root of the term, the one who announces — where what is announced is the will of a divinity. The being that announces is a khati (also hati), one of the khvtishvili, the children-of-god who hold the shrines of the highlands and mediate between the remote high god Ghmerti and the human communities beneath. In the ethnographers’ description the kadagi is a person, of either gender, permanently possessed by one of this class of local powers — a khati, a jvari (“cross”), or a saghmto (“divinity”) — and in the oracular state the indwelling khati foretells the future in a secret, sacred idiom of its own. The oracle does not consult the god so much as house it. Between festivals the kadagi lives an ordinary life in the village; when the god comes down, the person recedes and the khati speaks.
The office is attested in three kinds of record, and they are best held apart. It comes down first through the living tradition of Pshav-Khevsureti, where the shrine world is still worked and its own people will say, half in jest, that the whole of highland culture comes to two things — tsatsloba, the old courtship custom, and kadagoba, the oracle-speaking. It comes down second through the early-twentieth-century record: the ethnographies of Sergi Makalatia, whose Pshavi appeared in 1934 and whose Khevsureti followed in 1935, and the folklore collections, which describe the kadagi as an institution then still active, its oracles still going into trance at the summer shrine-feasts. And it comes down third, and most fully, through the scholarship — Georges Charachidzé’s structural analysis of 1968, Vera Bardavelidze’s and Zurab Kiknadze’s placement of the oracle within the shrine’s human community, Nunu Mindadze’s dedicated study of the institution, and Kevin Tuite’s Caucasus fieldwork. The three registers do not always tell the same story, and where they diverge the divergence is itself part of what is known.
What the kadagi speaks in trance is not, in the tradition’s account, ordinary Georgian. The oracles and the shrine’s clergy share a secret register the ethnographers recorded as the jvart-ena, the “language of the cross” — described as the sacred speech of the divinities of northeastern Georgia and, in practice, of their oracles and priests, some of whose words carry no sense in the modern tongue. A possessed oracle is called a meene, which the same record glosses as a kind of “linguist,” one who has the god’s language; and among the terms preserved is kandara, in everyday Georgian a perch for birds — a fitting figure for the oracle who is the divinity’s alighting-place. The kadagi is thus not only a voice but a translator, rendering into the community’s hearing a speech the community cannot itself command.
When the god says “I”
The organizing fact of the office is that a mute high god’s order becomes audible speech through it. Morige Ghmerti orders the world from the ninth sky and is scarcely addressed; the khvtishvili hold the shrines and administer the valleys; but a precinct of stone and banner and sacred oak cannot itself answer a question put to it, cannot name the taboo a household has broken, cannot call for the beast it wants sacrificed or foretell the sickness or the raid still over the horizon. For any of that, the khati needs a throat. The kadagi is that throat, and the strangeness of the office is concentrated in what happens to the pronoun when the god arrives. The person stops reporting and starts being spoken through. In the record the oracular speech runs in the first person of the divinity — I took you, the khati says of the one it has seized, I forbid you — the god issuing its claims and prohibitions directly, in its own voice, out of a human mouth. The demands and the warnings are the god’s; the mouth is the oracle’s; and to those who have gathered to hear, the distinction has collapsed.
The kadagi goes into that state on two kinds of occasion. The first is the shrine’s own festival, the kadagoba, when the community stands before its sanctuary at the height of summer and the seated god addresses its people through the leaning oracle. The second is the crisis — a sickness in a house, a death, a failed harvest or threatened one, a matter that troubles the whole commune — when the khati is asked, through its oracle, to say the cause and the remedy. A great shrine such as Gudanis Jvari, holding authority over a wide stretch of Khevsureti, could speak this way to a whole province through its kadagi. The office, in other words, is not a spectacle staged once a year; it is the standing channel by which a silent sanctuary remains answerable to the people who keep it.
What the god says through the oracle covers the whole width of a community’s need. In the record the kadagi tells the people why the divinity is displeased and what will turn its anger; foretells the things a mountain year turns on — a sickness, a death, the yield of the harvest, the coming weather; recalls to memory a rite that has been let slip and must be kept again; and settles the frictions of daily and social life that no elder’s ruling had quieted. It is at once an oracle, a diagnosis, and a court of last resort, all issuing from the one leaning figure. The kadagi does not deliver a doctrine or expound a scripture — the highland religion has none to expound — but answers the particular case, the specific household, the exact broken thing, in the god’s own naming of it. That is why the office could not simply be replaced by the elder’s learning or the banner’s authority: those govern the general order of the shrine, and the oracle speaks to the single instance no rule anticipated.
That channel is not the same as the shrine’s government, and the tradition is careful to hold the two apart. The kadagi is not the khevisberi — also called the khucesi — the elder-priest whom the community elects and who alone approaches the altar, offers the sacrifice, and brews and blesses the sacred beer of the feast. The elder governs the shrine and performs its rite; the oracle utters its will. Around them stand the further offices of the sanctuary — the dasturi, the stewards who keep the precinct and its stores, and the medroshe, who bears the cross-topped banner in procession — the whole apparatus that belongs to the jvari-khati order and is cataloged there. Within that order the kadagi occupies a place unlike any of the rest, because the rest are functions a person is chosen or elected into by the community, and the kadagi is not. The elder is raised up by his fellows for his knowledge and his character. The oracle is seized by a god.
Here the office parts company with every priesthood one applies to, trains for, or is ordained into, and the parting is the heart of the matter. The kadagi’s vocation arrives as an affliction. The chosen person falls ill — at any age, though the tradition watched the young for the signs, marking out those who might be destined for the calling before they were grown — and, in the account the shrine-world gives of itself, will not recover until the calling is accepted. The illness is the god’s hand laid on its chosen; recovery is the yielding to it. And the yielding is rarely eager. The one the khati wants most often resists, for the office is understood as a life-consuming election rather than an honor sought: a kadagi could be spoken of, in the frame’s own idiom, as a slave of the khati, bound to the god’s demands and set apart from the ordinary course of a life. What passes between the god and the one it takes — the induction into the office, the rite by which the calling is fixed — belongs to the closed life of the shrine and is set down here only as public record, at the ethnographer’s distance, not as any instruction. The vocation, the resisted call, the sickness that lifts on acceptance: these are named as the tradition names them, and no further.
One thing must be said in the tradition’s own register, because the outside record so often refused it. When observers reduced the kadagi’s seizure to “hysteria” or “epilepsy,” to a fit or a nervous disorder, they were describing something they had already decided not to credit — fitting the affliction to a diagnosis rather than hearing what the shrine-world says it is. Within that world the sickness is not a symptom to be explained away but the mark of a god’s choosing, and the person who submits to it becomes the one indispensable thing a heaven too high to be addressed otherwise lacks: a voice on the ground. Georges Charachidzé, whose Le système religieux de la Géorgie païenne (Paris: F. Maspéro, 1968, with a preface by Georges Dumézil) remains the fullest analysis of the highland religion, gave sustained attention precisely to this knot — the oracle’s vocation, the making of a kadagi, and the tie between the calling and illness — reading the whole as a coherent system rather than a pathology. The affliction, in that reading, is not the failure of the office but its door.
The office under pressure
The trajectory of the kadagi has to be told honestly, because it runs downward across the twentieth century even as the shrine world around it kept working. The cults Charachidzé analyzed were those practiced by the mountaineers of eastern Georgia in the early twentieth century — already, when the great records were being made, a tradition under strain. Soviet hostility to religion pressed on the shrine practice through the middle of the century, driving much of it into the quiet; and in the early 1950s the resettlement of much of Khevsureti to the lowlands emptied the very villages whose sanctuaries a kadagi served, so that the district that had counted thousands in the nineteenth century held only a few hundred by the modern census. The sober summary is that the living oracle grew rare across the century. The festivals continued, and continue; the elder still presides and the banner still comes out at the sanctuaries that are still kept; but the standing, indwelt voice — the person permanently possessed, ready to be spoken through — is the part of the system that thinned most. No invented vitality will do here. The record shows an office attenuated alongside a shrine world still worked.
The scholarship, meanwhile, keeps returning to it. Charachidzé read the kadagi structurally, in the manner of Dumézil and Lévi-Strauss, as one term in an ordered whole. Kiknadze set the oracle within the saqmo, the shrine together with the human community bound to it, so that the kadagi is understood not as a solitary ecstatic but as an organ of a congregation. Kevin Tuite’s fieldwork on the Pshav-Khevsur summer festival cycle, carried out across the late 1990s, and his study “Lightning, Sacrifice, and Possession in the Traditional Religions of the Caucasus” (Anthropos 99, 2004) placed the highland oracle within a wider Caucasian pattern in which people and animals are appropriated by divine beings, and possession, sacrifice, and the strike of lightning belong to one order of thought. Nunu Mindadze devoted a monograph to the institution directly, and more recent scholarship, including a study by L. Solovyeva in the Studia Religiosa Rossica, has taken up the oracle again as a subject in its own right. An office that has grown rare on the ground has not grown less legible on the page.
One strand of that scholarship is genuinely contested, and neither reading is endorsed here. Tuite has argued that in Pshav-Khevsureti the oracular role underwent, over time, a professionalization and masculinization: the sanctioned kadagi, the authorized spokesman of the god, came to skew male, while spontaneous possession occurring in women was increasingly ascribed to demons rather than to the khati — the same current, in his reading, by which a female divine figure was subordinated to the male St George in the shrine’s own myth. Against this, Mindadze records the kadagi as an office held by persons of either sex, and the shrine-world’s own tradition keeps a female oracle strand emphatically alive: Samdzimari and her sister-goddesses, seated at Khakhmati, are the demon-born oracle-priestesses whose particular work is to restore to the shrine-priest his lost contact with the gods. Whether the historical oracle was one gender or both, and whether the record of women’s possession reframed as demonic marks a real shift or the angle of the sources, is debated and not resolved. Behind it stands the deeper and equally disputed matter of whether this whole system preserves an archaic pre-Christian layer or is largely a medieval-and-later development — the archaism-or-innovation argument that runs through the study of Ghmerti and is not re-fought here.
What is not in doubt is the shape of the thing the highlands built. The Pshavs and the Khevsurs are living Georgian Orthodox communities whose shrine practice runs alongside the church, and the fuller histories of that life belong to the Pshavs and the Khevsurs. What the kadagi added to it was a solution to a specific difficulty — how a people can be answered by gods who do not, as a rule, speak. Their answer was to let a god take a person. The one so taken did not ask for the burden and, by the tradition’s own account, would as soon have refused it; but once the khati had seized its chosen and the sickness had lifted into speech, the mountains had what a silent sky cannot supply on its own — a mouth through which the order of heaven could be heard as a demand, a warning, and a name spoken in the first person, into the middle of an ordinary human day.
→ Related: Kviria · Iakhsari · Adgilis Deda · Pirkusha · Shatili · Mutso · Anatori · Gudanis Jvari · The Khevsurs · The Pshavs · Jvari Khati
Sources
- 'Oracular speech — gvrini and kadagoba' (batsav.com) — the kadagi as oracle of the khati; the calling through illness; the future oracle watched for young; the trance at festivals and crises; the god speaking in the first person; the 'slave of the khat'i'
- 'The Language of the Cross' — ჯვართ ენა, jvart-ena (batsav.com) — the secret sacred idiom of the shrine's oracles and priests; the terms meene ('linguist/seer') and kandara
- L. Solovyeva, 'The institution of the oracles (kadagi) among the highlanders of Eastern Georgia,' Studia Religiosa Rossica: Russian Journal of Religion 3 (2019): 92–111 (with citation of Nunu Mindadze's study of the institution)
- Georges Charachidzé, Le système religieux de la Géorgie païenne: analyse structurale d'une civilisation (Paris: F. Maspéro, 1968; préface de Georges Dumézil) — publisher page
- Kevin Tuite, 'St George in the Caucasus: politics, gender, mobility' (posted paper, 2017) — the professionalization and masculinization of the oracle role; women's possession reframed as demonic
- Kevin Tuite, 'Lightning, Sacrifice, and Possession in the Traditional Religions of the Caucasus,' Anthropos 99 (2004): 143–159 (continued 481–497) — JSTOR record
- 'Khevisberi' — Wikipedia (the elected shrine-elder/priest who alone approaches the altar and blesses the sacred beer, distinct from the oracle)
- 'Khevsureti' — Wikipedia (the jvari/khati/salotsavi sanctuaries; the shrine offices; the twentieth-century depopulation)