1. Introduction: Two Commonplaces and One Identification

“The oracle of Empress Jingū was shamanism.” “Himiko’s kidō was shamanism.” Both claims are true, and both are nearly empty. “Shamanism” has long been a loose label: in everyday speech it blurs into “sorcery,” “magic,” and “the occult,” and even in scholarship it has served as a catch-all for any dealings with spirits. This paper replaces the label with a coordinate system. Once that system is in place, the two claims acquire content — and they turn out to mean opposite things. Under the same name of “shamanism,” the two shaman-queens of early Japan occupy entirely different points on the plane.

The two claims are also linked historically. The Nihon shoki itself identifies Jingū with Himiko: its Jingū chapter cites the Wei zhi in interlinear notes at the thirty-ninth, fortieth, and forty-third years of the reign, aligning her chronology with the embassies to Wei[1]. But the identification is chronological only. Typologically, as this paper will show, the compilers shaped Jingū as Himiko’s exact opposite. Why date her as Himiko and yet portray her as Himiko’s inverse — by failure, or by design? This is where the coordinate system earns its keep: it turns a vague incongruity into questions that can be decided axis by axis.

The argument proceeds as follows. Section 2 states, in summary, the coordinate system whose full derivation is given in the companion paper[2]. Section 3 sketches the religious substratum of early Japan: Mount Miwa, the solar cult of Amaterasu, the ritual system of the early Kofun period. Sections 4 and 5 place the two shaman-queens on the plane as a dual pair of case studies — Jingū’s oracle as peripheral possession, Himiko’s kidō as central possession, the latter further refined by three axes of authority, ritual, and social function. Section 6 draws the verdict on the compilers’ identification. Section 7 shows that the same coordinates describe, with equal consistency, the later Japanese transformations of shamanism, from the institutional miko, the Sumiyoshi cult and kagura to Shugendō and Noh. Section 8 concludes, and adds one determination from the side of content, owed to Henri Ey: what both queens lack.

2. The Instrument in Brief

The coordinate system is a plane spanned by two axes, both properties of the religious office. The first is the axis of technique, derived from Eliade’s strict definition of shamanism as the “technique of ecstasy”[3], read as a Weberian ideal type: at one pole the ecstatic form, in which the soul goes out of the body and the human being remains the active master of spirits; at the other pole possession, in which both variables are inverted — the spirit comes in, and the human being is its vessel. The second is the axis of the seat of authority, after I. M. Lewis[4]: central possession, incorporated into an established seat and legitimating the community’s order, against peripheral possession, seatless, oblique, borne case by case by the socially marginal — taken here as a description of position only, without Lewis’s functionalist explanation of motive. Society-level variables — mode of subsistence, lineage of tradition — are recorded as annotations of a located point, not as axes; Campbell’s familiar dichotomy of hunting and planting religions, which bundles the levels together, is decomposed rather than adopted, and it is precisely Himiko who wrecks it, as Section 5 will show. Finally, the coordinates deliberately do not measure the individual’s state: with Karl Jaspers, the quality of what an individual undergoes and the institutional seat that a society designs are held apart as different dimensions throughout[5]. The derivation, defence and Greek testing of this instrument occupy the companion paper[2]; here it is put to work.

3. The Religious Substratum of Early Japan

Before the two case studies, the ground must be laid: what religious order already existed in the archipelago when Jingū’s oracle breaks into the record? Its most instructive element is the cult of Mount Miwa (Nara prefecture). Miwa is the classic instance of a mountain worshipped as the very body of the deity: the Ōmiwa Shrine at its foot has no main sanctuary building, for the mountain itself is venerated directly — an archaic form of Japanese nature worship. The god of the mountain, Ōmononushi, appears as a serpent; in the chronicle of Emperor Sujin, the princess Yamato-totohi-momoso-hime discovers that the husband who visits her nightly is a serpent[6][1]. The serpent is a thoroughly agrarian symbol — a creature that “rebirths” itself by sloughing its skin, a dweller of the waterside and hence a deity of water and rain — and the same chronicle shows the god sending pestilence and being appeased by correct ritual: the land-deity as guarantor of order and, at once, punisher of ritual negligence[1]. Crucially, the Miwa cult is a cult bound to place: a particular holy mountain, a particular precinct, prescribed procedure. It corresponds exactly to the priestly, planting-type religion of Campbell’s scheme[7]. Momoso-hime herself, transmitting the god’s will as his consort, was an oracular figure of the miko kind — but hers was an institutional oracle, rooted in the Miwa precinct (her genealogy, and the marriage-myth itself, the author has examined in a companion study in Japanese[8]).

The second core is the solar cult of Amaterasu. The sun governs the agrarian calendar, and the myth of the Heavenly Rock-Cave — the sun-goddess’s concealment and return — has long been read as a calendrical myth of the sun’s “death” and “rebirth” at the winter solstice[9]. Institutionalized at Ise, with its rule-bound liturgy and the ritual rebuilding of the shrine every twenty years, the cult is among the most refined priestly systems on record; and the saiō, the consecrated princess who serves it, is no shaman — her office rests on institutional purity and the observance of taboo, not on personal trance. Yet one trace of an older stratum remains. In the Rock-Cave myth, Ame-no-uzume dances before the assembled gods, disrobing, in a state approaching divine possession: a hint that the planting-type system did not expel the ecstatic substratum but absorbed and transformed it[7][9].

By the early Kofun period (later third to fourth century), the Yamato paramountcy had consolidated this order into a fully institutional system: standardized keyhole-shaped tumuli; the organized distribution of triangular-rimmed bronze mirrors; the regalia of mirror, curved jewel and sword established as sacred objects whose transmission ritually legitimated succession[10]. Religious authority was concentrated in the king and in hereditary ritual lineages such as the Nakatomi and the Imbe; the room left for personal, charismatic authority of the shamanic kind was institutionally narrow.

This is the ground against which the two case studies must be read — and what makes Jingū’s oracle the anomaly it is. When a deity seizes a human mouth in the middle of this ordered, priestly world, the chroniclers mark the event as extraordinary; the coordinate system will now say precisely in what way it was.

4. Case Study I — The Oracle of Empress Jingū: Possession-Type, Peripheral

4.1 The scenes in the sources

In the Kojiki the oracle is a séance of three roles[6]. At the palace of Kashihi in Tsukushi (northern Kyūshū), Emperor Chūai plays the koto, the zither whose music induces the trance; the minister Takeuchi no Sukune sits in the saniwa, the “sacred court” where the divine will is discerned; and the empress, entered by the deity, speaks with the deity’s voice, promising a land to the west laden with treasure. The emperor does not believe the words; he pushes the koto away. Urged by the minister, he draws it back and plays grudgingly — and presently the sound ceases. When the lights are raised, the emperor is found dead.

The Nihon shoki tells the tale twice. In Book VIII the god speaks through the empress: “There is a better land than this, a land of treasure … the land of Silla of the coverlets of paper-mulberry.” The emperor climbs a high place, sees only ocean, and objects: “We have looked all around, and there is sea, and no land. Can there be a country in the Great Void?” The god’s reply is a sentence of deposition: “But as thou, O King! hast spoken thus, and hast utterly refused to believe me, thou shalt not possess this land. The child with which the Empress has just become pregnant, he shall obtain it.” The emperor persists in his campaign against the Kumaso, fails, and dies suddenly[1]. In Book IX — the annals of the regency — the empress then conducts the counter-séance in her own person: “The Empress, having selected a lucky day, entered the Palace of worship, and discharged in person the office of priest. She commanded Takechi no Sukune to play on the lute, and the Nakatomi, Igatsu no Omi, was designated as Saniha.” After seven days the possessing gods consent to be named — among them the three sea-deities Uha-tsutsu no wo, Naka-tsutsu no wo and Soko-tsutsu no wo, the gods later enshrined at Sumiyoshi[1]. (Quotations are Aston’s translation.)

The structure is constant across the tellings: an instrument-player who induces the trance, a medium through whose mouth the god speaks, an interpreter who discerns and tests the words. This three-role séance is a typical pattern of North Asian shamanism — and it stands in the record like a foreign body within the priestly order described in Section 3.

4.2 First location, and a source-critical caveat

On the axis of technique the location is immediate. Nothing here is soul-flight: no journey, no ascent, no land of the dead. The movement runs inward — a god arrives from outside and occupies a human vessel — and agency lies with the god, not the medium. In the exact sense fixed in Section 2, the oracle of Jingū is possession-type: both variables of the ecstatic type inverted. On Campbell’s first-pass scheme, likewise, the scene belongs to the shamanic religion of the hunting type — direct commerce with the supernatural through personal charismatic experience — and the record itself dramatizes the collision with the indigenous priestly order: the emperor who refuses the oracle dies by it.

One caveat must be entered. The historicity of Jingū has long been doubted, and at least part of her portrait may be a construction of the eighth-century compilers — above all, her chronological identification with Himiko, which the Nihon shoki itself performs in the interlinear notes cited in the Introduction. If the fictional reading is adopted, the séance scenes record not fourth-century practice but the compilers’ understanding of how an oracle is conducted — the seventh- and eighth-century court’s image of ideal possession. Either way, the typological analysis stands: what this chapter locates on the plane is the structure of the scene as narrated. Under the fictional reading the conclusion simply transposes into an independent piece of testimony — that the court of the compilation era imagined the ideal oracle as possession-type and peripheral. But the identification leaves a strangeness in place: dated as Himiko, Jingū is shaped as Himiko’s typological opposite. That question is adjudicated in Section 6.

4.3 The Korean connection

The Kojiki itself traces the empress’s maternal line to Ame-no-hiboko, a prince of Silla[6]. The comparison that matters here is Korean musok: the mudang enters trance to the beat of drums and gongs and speaks the words of the god — a ritual structure whose kinship with northern shamanism was argued systematically by Akamatsu and Akiba in their pioneering study of 1937[11]. The pairing of instrument-player and entranced medium matches the Kashihi scene with striking exactness. Silla’s elite culture, moreover, shows marked affinities with the Scytho-Siberian steppe — the gold tree-form crowns and antler ornaments of its royal tombs[12] — so that a Silla-descended empress embodying a northern, shamanic mode of religious authority is historically coherent. (Whether musok may itself be called “shamanism” depends on the definition adopted; under the strictest ecstatic definitions it falls outside the term, a difficulty which the coordinate usage of this paper avoids.)

4.4 Peripheral possession: the structural basis of mobility

On Lewis’s axis the location is equally sharp. Jingū’s oracle occupies no established seat. It is not attached to a sanctuary, a lineage of mediums, or a calendar; it erupts at a military crisis, on the march, far from the center, to legitimate one decision, and it moves with its bearer — a mobile, occasional authority. This is peripheral possession in the precise sense of Section 2: the description of a position, with no claim about motive.

The basis of that mobility deserves to be stated exactly, because it is structural, not ethnic. The Silla lineage of §4.3 is one historical bearer of this peripherality, an explanatory variable — not its ground. What makes the authority mobile is not the descent of its bearer but the fact that the authority itself is settled in no central seat. Blood explains, at most, how such an authority came to hand; the axis explains why it could travel. Where a figure such as Jingū may be a construct of the compilers, her recorded Silla ancestry measures not blood but design — what the compilers chose to give her: the coordinates locate modes of narration, not persons. On the same axis, one pole away, stands a queen whose possession was anything but mobile — and to her we now turn.

5. Case Study II — Himiko’s Kidō: Possession-Type, Central

5.1 Kidō as central possession

The Wei chronicle’s account of the Japanese archipelago in the third century records that, after years of disturbance, “the people agreed upon a woman for their ruler. Her name was Himiko. She occupied herself with magic and sorcery, bewitching the people. Though mature in age, she remained unmarried…. She had one thousand women as attendants, but only one man. He served her food and drink and acted as a medium of communication” (tr. Tsunoda)[13]. The chronicle calls her art kidō — the “way of the spirits.” The analysis of that art as an apparatus of rule — the joint enthronement after the great disturbance of Wa, the diarchy of sacral queen and administering brother — the author has given in a companion study (in Japanese)[14]. The present chapter takes up the other level: the religious practice itself, and where it falls on the plane.

On the axis of technique, the location is quickly secured. The chronicle records no soul-flight, no celestial journey, no descent to the dead[13]; Japanese scholarship has accordingly understood kidō as possession — a deity descending upon the queen — and Carmen Blacker’s ethnography turns this from a silence into a positive finding: the implements of the ancient miko — mirror, bells, curved jewels, bow — are devices for summoning the spirit to the medium, and the bird-and-flight symbolism that saturates the Siberian shaman’s costume is wholly absent from her equipment; the costume draws the spirit to her, it does not send her to the spirit[15]. (On the semantics of the word kidō itself — an oracular authority embodied in a person, distinguished within the chronicle from the technical bone-divination that Chen Shou compares to the “method of the burnt tortoise-shell” — see the companion study[16].) The question, then, is not whether kidō is possession, but what kind of possession it is.

On Lewis’s axis the answer is the mirror image of Jingū. Himiko’s possession is central: it is the moral and cosmological order of the community itself that her office guarantees. The circumstances of her accession — jointly established at the symbolic center of a confederation, after a crisis of order — confirm the centrality from the institutional side[13]. Two naive confusions dissolve at once: kidō must not be conflated with the ecstatic shaman of Eliade’s strict definition, and it must not be conflated with the peripheral possession of marginal female practitioners (the lineage that leads to the later itako and yuta). Himiko’s possession is not contestation from the margins; it is the guarantee of order at the center. And here Campbell’s dichotomy, taken as it stands, breaks: her kidō arose inside an agrarian society, within the indigenous ritual sphere confirmed in Section 3, yet it is possession-type in technique and occupies an established, inheritable, institutional seat. On Campbell’s scheme this combination should not exist — agrarian religion is supposed to be priestly, formalized, incompatible with the charismatic and personal mode of possession. Himiko sits on the “impossible” coordinate; it is her case, above all, that forces the decomposition of the scheme into levels of description[2].

5.2 Three axes: authority, ritual, function

The located point has an internal structure, which three further axes measure. The first is the mode of authority: charismatic (grounded in personal spiritual experience) or institutional (a transmitted office), in Max Weber’s sense[17]. Himiko shows both faces — she “bewitches the people,” yet she is jointly established — and the succession decides the question. When she died, “a king was placed on the throne, but the people would not obey him. Assassination and murder followed…. A relative of Himiko named Iyo, a girl of thirteen, was then made queen and order was restored” (tr. Tsunoda)[13]. This is Weber’s routinization of charisma enacted in a single sentence: a charisma on which a community has come to depend cannot die with its bearer, and is converted into an inheritable seat. A thirteen-year-old successor can occupy the seat because the authority now belongs to the seat, not the person — the same structure as the mhondoro mediumship of Shona society, in which the royal ancestor spirit takes a new medium when the old one dies and the office continues[4].

The second is the mode of ritual: ecstatic (centered on trance) or liturgical (centered on prescribed rites). Himiko’s own rites are unrecorded, but the chronicle’s notice of bone divination among the Wa suggests a mixture of the two. The third is the social function, in the sense of Hubert and Mauss[18]: divinatory — legitimating particular decisions, activated on private and occasional occasions — or pacificatory, the public and obligatory maintenance of cosmological order. Hubert and Mauss’s criterion is deliberately social rather than psychological: what distinguishes magic from religion is not the attitude of the practitioner (Frazer’s coercion versus supplication) but the position a rite occupies in the organized cult of the community. On this axis Himiko’s kidō is a composite: born divinatory — the joint enthronement legitimated a settlement — it became, as the established seat came to carry the order of the whole confederation, public and obligatory. The three axes together describe the point’s interior — institutional in authority, mixed in ritual, dual in function — without replacing the plane: the work of distinguishing the two queens was done by the two axes of technique and seat.

5.3 State and office

A methodological distinction now becomes indispensable, and it is Jaspers’s[5]: the state — the quality of experience that occurs in an individual — and the office — the institutional seat that a society designs — are different dimensions, however similar their surface expression. Japanese material supplies the two extreme cases that prove their independence: the itako of northern Honshū, who retain the office without the state — a full institution of apprenticeship, initiation, implements and fees, but no initiatory crisis, and a trance that careful observers judge imitation — and the founders of the modern new religions, Nakayama Miki, Deguchi Nao, Kitamura Sayō, who present the state without the office: the full shamanic biography, but no prepared seat, so that the seat has to be built afterwards[15]. State and office are thus two independently varying quantities — and in the succession from Himiko to Iyo, only the office was at stake. It is this independence that entitles the present chapter to define kidō without any access to Himiko’s inner experience.

5.4 The gradient of institutionalization, and a definition

An ethnographic sketch gives the typology flesh. In the village league of the northern Yamato basin from which the author’s family comes, a dozen settlements sharing a reservoir and its channels formed a single community of fate in matters of water, and the timing of the cutting of the sluice, as the rainy season advanced, was the league’s weightiest decision. In an old story the author heard from his grandmother, it was not any established functionary but a woman of comparatively low standing who, entering trance, climbed a tree naked and cried: cut the sluice. Both axes are folded into the one scene. The bearer is peripheral in status; the climbing of the tree carries an echo of ascent — of the ecstatic motif — while the trance-utterance is possession, the two poles mixed in a single case, exactly as ideal types lead one to expect; and a person peripheral in standing, borrowing possession, inserts one binding word into the central decision of the community. Lewis’s oblique strategy of the powerless here is the central function.

The two poles of Lewis’s axis may accordingly be reread as the two ends of a gradient of institutionalization. At the base, a peripheral village medium carries the central function for one moment, by spontaneous and transient possession; as scale and institutionalization increase, the function crystallizes into a standing seat; at the summit, standing and office fuse — Himiko. That the gradient can actually be climbed is on record: in 1838 Nakayama Miki, standing in one night for an absent professional medium at a healing rite, was seized in violent possession, announced “I am the General of Heaven,” and did not return to herself until her husband had yielded her body to the god as its “shrine” — a disposable stand-in at the bottom of the gradient becoming, in one night, the seat of a new center[15]. (The sketch is family oral tradition and proves no continuity with the third century; it serves as ethnographic analogy for the persistence of a structure.)

Kidō may now be defined. It is neither an ecstatic technique nor a peripheral possession, but the institutionalization of central possession: the ritual core of a sacral kingship, held as an inheritable office irreducible to any individual’s state — a definition that joins, without remainder, the analysis of Himiko’s government as an anti-state diarchy given in the companion study[14].

6. The Compilers’ Design

The two queens can now be set side by side. They agree on one axis only: both are possession-type — a woman through whose mouth a god speaks. At that coarse resolution, and only at it, the compilers’ identification of Jingū with Himiko was possible at all. On every other axis they fall on opposite sides: seat of authority (central / peripheral), mode of authority (institutional / charismatic), mode of ritual (mixed / ecstatic), social function (a composite reaching the pacificatory pole / purely divinatory). Had the compilers tried to copy the Himiko of the Wei chronicle and failed, the misses would scatter axis by axis, like noise. A systematic inversion across every axis does not look like noise. It looks like design.

The instrument that reads the design is the function axis, for only it connects a figure’s type to its narrative task. What task does the oracle of the Jingū chapters perform? The legitimation of a single expedition; the switch, through Chūai’s death, to the line of Ōjin; in short, the opening of a dynasty. That is the work of the divinatory pole — a rite activated on a private, particular occasion to legitimate one decision — and it is work that the guardian of a standing order cannot perform. One who maintains an order does not open a dynasty. The Kashihi séance, with its koto, its saniwa, its inquiry into one military decision and its doubting king struck dead, satisfies Hubert and Mauss’s description of the private and occasional rite almost to the letter[18]; the Himiko whom the Wei envoys observed was a steady-state apparatus carrying the order of a whole confederation. The divergence on the function axis is then no failed copy: it is the sign that no copy was needed. The identification was made at the level of chronology; at the level of type the compilers were executing another task — and executing it consistently across every axis.

One methodological remark. In Section 2 this paper took from Lewis the description of position and renounced the functionalist explanation of motive. That renunciation concerns real persons, whose psychology cannot be guessed from outside the sources. What is measured here is not a real person’s blood or motives but the compilers’ design — and a designer’s purpose is not guessed but read off the structure of the document. The functionalist reading forbidden for the living miko is legitimately unsealed for the designed narrative. What is obtained is a verdict of elimination — the hypothesis of unwitting failure is excluded. Whether a further intention lay in the reversal — whether rewriting the memory of a centrally seated queen as a peripheral, one-time oracle was itself a way of taming the oracle, placing it outside the royal line — is a possibility this paper leaves open, for its adjudication requires reading the oracle notices of the chronicles systematically as sources for the religion of the compilation era.

7. The Japanese Transformation of Shamanism

What became of the institution? Neither queen’s mode of authority survived intact. The shamanic elements of the early record were absorbed into the developing religious system of the archipelago, transformed, and redistributed — and the redistribution can be described on the same coordinates as everything above, as movement in two directions at once: integration into central seats, and subsidence into peripheral possession.

The clearest instance is the bifurcation of the miko. Japanese folklore studies have long distinguished two lineages: the shrine miko, attached to a sanctuary, dancing kagura and assisting at rites; and the kuchiyose medium — the itako and yuta — who calls the spirits of gods and of the dead into her own body on request[19]. Lewis’s axis gives this descriptive dichotomy its theoretical footing[4]: the shrine miko continues the central lineage — incorporated into an established seat and legitimating the standing order — while the kuchiyose medium continues the peripheral one, seatless, engaged case by case, borne by the socially marginal. In the vocabulary of §5.4, the two lineages are the same oracular function crystallized at the upper and the lower ends of the gradient. Historically the movement ran downward and inward together: as the ritsuryō state formalized its cult, the trance element was pushed to the margins; the medieval kuchiyose medium was expelled into folk religion, while the shrine miko was absorbed into the institution as a dancer — Eliade’s general pattern of institutionalization and marginalization[3], stated more exactly as the two directions of Lewis’s axis.

The cult of Sumiyoshi shows the central direction in its most direct form, for its gods are the gods of Jingū’s own oracle — the three Tsutsu sea-deities named in the séance of §4.1, enshrined at Sumiyoshi together with the empress herself[20]. A deity once summoned by mediumistic rite at the moment of sailing became the fixed patron of navigation; the shrine acquired a port, domains, and a calendar of regular rites; the trance vanished from the official cult, and only the mythic bond of the god with Jingū and Ōjin preserves the memory of its shamanic origin. The same centralizing logic produced the institution of the saniwa itself: a standing expert who discerns whether possessed speech is truly divine is a mechanism for placing possession — always potentially peripheral — under the control of a central seat[21]. Through such channels (the Hachiman oracle tradition among them), possession was not abolished but administered.

Buddhism added a further transformation, and with it the most interesting case: Shugendō, the mountain asceticism that fused esoteric Buddhism with indigenous mountain religion[22]. The yamabushi acquires his powers through austerities — waterfall, fast, fire-walking — whose structure is that of the shamanic initiation, death and rebirth through ordeal, down to Eliade’s “mastery of fire.” Blacker divides the Japanese shamanic personnel accordingly into two poles: the miko, into whom the spirit comes from without, and the ascetic, who travels to the other world for power — possession-type and ecstatic-type, each with its own bearer[15]. And she argues that the two were originally one: the primitive Japanese shaman combined the medium’s voice with the master’s power to subdue. It was the esoteric, Shugendō form of Buddhism, on her account, that appropriated the active, subjugating side of the complex for the male ascetic, downgrading the medium’s role to passive, almost automatic utterance — taking from the miko her ancient dignity, and the music and dance with which she had once summoned the gods. For the coordinate system this is a remarkable by-product: the split runs exactly along the second variable of the technique axis, the locus of agency. A distinction constructed as an ideal-typical axis crystallized, in one religious history, into an actual social division of labor.

One channel of preservation remains that has no living bearer at all: the Noh theatre. In Aoi no Ue, the possessing spirit is summoned by the miko Teruhi with her catalpa bow — the invoking, vessel pole — and subdued by the holy man of Yokawa — the active, exorcising pole; the division of labor of the preceding paragraph is written into the plot itself, for the miko can call the spirit but cannot expel it, and the second figure must be sent for. Between the two worlds runs the hashigakari, the bridge-corridor of the stage. Blacker opened her study of Japanese shamanism with that bridge, and closed it with the stone bridge of the play Shakkyō — the narrow, moss-slick arch that none may dream of crossing without the help of the gods[15].

Thus each pole that the coordinate system distinguished has landed, in the long history of Japanese religion, on a bearer of its own: the yamabushi who travels (the ecstatic pole); the saiō and the shrine miko in their established seats (the central-possession lineage); the itako and yuta, seatless and engaged case by case (the peripheral lineage). The coordinates describe not merely a third-century episode but a longue durée.

8. Conclusion: The Dual Pair

The religious transformation of early Japan can now be stated as what it was: the contact, collision and fusion of two fundamentally different religious types. The Yamato paramountcy of the early Kofun period had established a priestly, institutional order grounded in an agrarian society — Mount Miwa, the solar cult, the regalia (Section 3). Into this order, from the fourth century onward, northern shamanic elements entered by way of the Korean peninsula, figured in the record by Empress Jingū; and the record itself dramatizes the collision, in the emperor who refuses the oracle and dies (Section 4). What followed was not expulsion but redistribution: integration into central seats and subsidence into the periphery, along the routes traced in Section 7 — the bifurcated miko, Sumiyoshi, the saniwa, kagura, Shugendō, and the Noh stage.

The core of the paper is the dual pair. Jingū and Himiko are both possession-type, and they stand at opposite poles of the axis of the seat: the pure type of central possession — a seat jointly established, carrying the order of a confederation, inherited by a successor — against peripheral possession — seatless, mobile, activated crisis by crisis. The two poles, moreover, are ends of one gradient of institutionalization, along which the same oracular function appears at every scale of agrarian society, from the village medium in the tree to the queen of Wa. On the question raised in the Introduction, the verdict of Section 6 stands: the compilers’ identification of Jingū with Himiko was an operation at the level of chronology, and the systematic inversion at the level of type is the shape of design, not of noise — the task assigned to the Jingū narrative, the opening of a dynasty, demands the divinatory pole, and the guardian of a standing order cannot perform it.

Two determinations frame the pair from outside. First, the grid is no local artifact of the Japanese material: in the Greek field assembled by Dodds, the Pythia at Delphi fills Himiko’s cell, the wandering god-men the ecstatic-peripheral, the maenads the possession-peripheral — and the fourth cell, ecstatic and central, is vacant in Greece as in Japan[23][24]; the comparative argument, including a hypothesis on why the unverifiable resists central institutionalization, is given in the companion paper[2]. Second, a determination from the side of content, owed to Henri Ey’s posthumous essay on the shaman: Ey placed healing at the centre of the shaman’s office — “his principal function is to heal”[25]. The Japanese material has the healing function in abundance, in the healer-ascetics of Blacker’s ethnography and the yorigitō at which Nakayama Miki was seized[15]. But the two queens lack it. No healing is recorded of the Himiko whom the Wei envoys observed, nor of the empress at Kashihi; of Ey’s pair they hold only divination — the oracle. The lack demands no fourth axis, for the axes measure mode and position, not content; it works instead as an index that cuts the shaman-queen out of the broader concept of the shaman: an oracular specialization without healing.

No religious system anywhere was formed by purely indigenous development, any more than any civilization was built without borrowing. The contact and fusion of religious elements of different origins is the motor of religious history, and the transformation of early Japan — a priestly agrarian order receiving, resisting, absorbing and preserving a northern shamanic intrusion — is an East Asian realization of that universal structure. This paper has not sought final conclusions. It has sought placement: to set the two shaman-queens, at last, upon a single plane.

Notes and References

Citations are to original-language editions; English translations are noted where they exist. Works available only in Japanese are marked.