Report I — The Meiji Rupture and the Closure of the Japanese Social Imaginary
Prefatory Chapter: The Ground Before 1868
How to use this chapter
This chapter has one job. It establishes enough of the pre-1868 religious landscape to make the events of 1868 legible, and to disarm two widespread stories about that landscape that would otherwise distort everything that follows.
It is not a history of Japanese religion. It moves at treetop level and skips whole centuries. It also does not attempt to answer whether something recognizable as "Shinto" existed before Buddhism arrived. That question is genuinely open and is the subject of Report II. What this chapter establishes is weaker but sufficient: that the categories "Shinto" and "Buddhism," as bounded and mutually exclusive things, poorly describe the pre-modern situation. That can be shown without settling origins, and it is all that is needed to proceed.
Terms are bolded on first appearance and defined. Japanese words are given in romanization with an English gloss. Citations appear in parentheses at the point of claim; freely accessible sources are hyperlinked. A glossary and an annotated bibliography follow, and a final section flags claims that are contested or that I could not verify.
0.0 A working chronology
Period
Dates
Relevance here
Jomon
Yayoi
Kofun ("old mounds")
Asuka
Nara
Heian
Kamakura
Muromachi
Edo / Tokugawa
Meiji
c. 14,000–300 BCE c. 300 BCE–250 CE c. 250–538 CE 538–710
710–794 794–1185 1185–1333 1336–1573 1603–1868 1868–1912
Pre-agricultural; treated in Report II
Wet rice, metallurgy, continental contact
Monumental tombs; emergence of the Yamato polity
Buddhism transmitted; Chinese-style state building
Law codes; state Buddhism; Kojiki and Nihon Shoki
Aristocratic court; Tendai and Shingon; Engishiki
Warrior government; new Buddhist movements
Yoshida Shinto; sectarian consolidation
Temple registration; Confucian and nativist scholarship
The rupture; Chapters 1 and 2
Landmark dates: Buddhism officially transmitted from the Korean kingdom of Baekje in either 538 or 552 (the sources disagree); the Taiho Code promulgated 701; the Kojiki completed 712 and the Nihon Shoki 720; Saicho died 822; the Engishiki completed 927; Shinran 1173–1263; Yoshida Kanetomo 1435–1511; temple registration consolidated in the 1630s; Motoori Norinaga 1730–1801; Hirata Atsutane 1776–1843.
0.1 The problem: two symmetrical romanticisms
Two stories about pre-modern Japanese religion circulate widely, including in serious general writing. They appear to be opposites. They are not.
The harmony story. Japan enjoyed roughly a millennium of peaceful syncretism in which Buddhism and Shinto coexisted and enriched one another, producing a distinctively tolerant religious culture. The Meiji state then wrecked it. This story is congenial to Buddhist modernists, and it contains real truth about the violence of 1868.
The pollution story. Japan possessed an indigenous religion of the kami (deities, numinous powers) that was overlaid and corrupted by a foreign import. Buddhism was a contaminant. Meiji did not destroy anything; it recovered something. This descends from eighteenth-century nativist scholarship and remains active in contemporary Japanese cultural nationalism.
The two are mirror images of a single assumption. Both require that there were, at some point, two discrete, bounded, self- identical traditions — one Japanese and one foreign — which then either fused or contaminated one another. They disagree only about whether the mixing was good or bad. Neither questions that there were two things to mix.
That shared assumption is what this chapter dismantles. Once it goes, both stories lose their subject matter at once, and there is no need to decide which was right.
This matters for Report I specifically. The legislation of 1868 did not merely favor one existing tradition over another. It commanded the separation of things that had never been separate, on the authority of a historical claim — that a pure kami tradition existed prior to and independent of Buddhism, and could therefore be restored. Whether that claim was true is Report II's question. What matters here is that the administrators of 1868 had to act as though it were, and that acting on it required enormous physical and institutional destruction.
0.2 The vocabulary problem
0.2.1 The word "Shinto" before the modern period
Shinto is written with characters meaning "way of the spirits" or "way of the gods," from the Chinese shendao, a term with a long prior life in Chinese texts where it means something closer to "the numinous order of things" than to any particular religion.
The word appears in Japanese sources very early and very rarely, and its meaning is disputed. The two most-cited instances are both in the Nihon Shoki (720). Emperor Yomei is said to have believed in the Law of the Buddha and revered the Way of the Kami. Emperor Kotoku is said to have honored the Law of the Buddha and despised the Way of the Kami, the evidence being that he cut down the trees at Ikukunitama Shrine (Aston, Nihongi, 1896). Note that in these passages the two are named as things one can hold different attitudes toward, but nothing establishes what the second one was.
The historian Kuroda Toshio (1926–1993) argued in 1981 that in these passages the word most plausibly means something like local or popular cult practice, not a distinct organized religion. He went further: the idea of Shinto as an ancient, continuous, indigenous tradition is, in his phrase, a ghost image produced by a word linking together unrelated phenomena. On his account the shrines, rituals, priesthoods, and concepts of kami were, through most of Japanese history, Buddhist in their organizing framework, and Shinto as an independent religion is a modern construction serving nineteenth-century nationalism.
Kuroda's article is the most influential piece of scholarship in this field, is freely available, and should be read first (Kuroda 1981). Its conclusions are widely contested in detail and widely accepted in outline; it is the departure point for essentially all subsequent Western-language work (Teeuwen and Scheid, Tracing Shinto in the History of Kami Worship, 2002).
The consequence: "Shinto" cannot be used to name a pre-modern religion without begging the question at issue. For the pre- modern period this text says kami cult, kami worship, or the jingi system instead.
0.2.2 The words for mixture
Three terms, and the differences carry weight.
Shinbutsu shugo ("kami-buddha amalgamation") is the ordinary scholarly term today, used descriptively and neutrally. It is largely a modern analytic coinage.
Shinbutsu konko ("kami-buddha commingling") is older and pejorative. It suggests muddle and adulteration. This was the vocabulary of critics, first Confucian and then nativist, and it is the vocabulary the Meiji edicts inherited. A document from the 1860s using konko is not describing a combination; it is naming a defect.
Shinbutsu hanzenrei ("orders for the clear distinction of kami and buddhas") is the standard Japanese name for the 1868 edicts; the process is shinbutsu bunri ("separation of kami and buddhas"). The word hanzen — clear, distinct, unambiguous — is worth noticing. The legislation was framed not as reorganization but as clarification, as though the distinction were already real and had merely been obscured. That framing is itself a historical claim.
0.3 Institutional entanglement rather than theological fusion
The harmony story imagines a spiritual blending. What the record shows is more concrete and more consequential: kami cult and Buddhist institutions were fused at the level of buildings, personnel, land tenure, revenue, and law. That is a stronger form of entanglement than theological borrowing, and far harder to undo.
0.3.1 Shrine-temples and shrine monks
From the early eighth century, Buddhist temples were built within or beside shrine precincts. These were jinguji ("shrine- temples"). Their stated purpose was frequently the salvation of the kami itself: the kami was understood as powerful but unliberated, still bound to the cycle of rebirth, for whom sutra recitation could be efficacious. This is a striking inversion for readers coming from Indian Buddhism, where local deities are subordinated but not usually treated as objects of soteriological rescue.
The personnel followed. Shaso ("shrine monks") were Buddhist clergy who served at shrines and recited sutras before the kami. Betto ("superintendent") was the title of the administrator of a shrine-temple complex, typically a Buddhist monk. In a great many cases, the person holding legal and financial authority over a major shrine was a Buddhist cleric.
The court institutionalized this. Seven monks were ordained to recite sutras for the three kami of Aso as early as 794, and in 850 seventy monks were ordained to recite for a corresponding number of major kami. Donating sutra scrolls, buddha images, and Buddhist relics to prominent shrines became customary, including on the enthronement of a new emperor (Teeuwen and Rambelli 2003, introduction).
0.3.2 Honji suijaku: what it is, and what it is not
The doctrine organizing this arrangement is honji suijaku ("original ground and manifest trace").
The honji ("original ground") is the true identity of a divine being. The suijaku ("trace") is the localized form in which that being appears in a particular place, adapted to local capacities. Applied to Japan, a given kami was identified as the local trace of a specific buddha or bodhisattva. Hachiman became Hachiman Daibosatsu, "Hachiman the Great Bodhisattva." The paired entity was called a gongen.
Four things about this doctrine are routinely misunderstood, and each matters.
First: it is not the two truths. This is the most important clarification for a reader with Buddhist philosophical background, because the vocabulary invites the error. The two truths — ultimate and conventional — concern the status of assertions and phenomena, and in a Madhyamaka frame the ultimate is empty and is emphatically not a ground in any substantial sense. Honji suijaku concerns the relation between a manifesting being and its manifestation, and its "ground" is a specific named buddha, a positive locus. These are different axes. A third axis, institutional priority, is discussed below. Conflating them produces the heavy metaphysical reading that the historical evidence does not support.
Second: the machinery is not hierarchical in origin. The origin/trace distinction comes from Zhiyi (538–597), the Chinese Tiantai systematizer, who used it to organize the Lotus Sutra — the eternal Buddha of the later chapters against the historical Shakyamuni of the earlier ones. In that system, origin and trace are complementary and interpenetrating opposites, not superior and inferior terms (see the free Columbia reference dictionary JAANUS, entry honjibutsu). The standard account holds that honji and suijaku form an indivisible whole which in theory should have equal standing, though in practice it often did not. This is precisely why the relation could later be run backwards without doctrinal scandal.
Third: it was plural, not a single system. At least four frameworks competed:
Ryobu Shinto ("dual-aspect Shinto"), built on Shingon esoteric cosmology, mapping the Ise Inner Shrine and Amaterasu to the
Womb Realm mandala and the Outer Shrine and Toyouke to the Diamond Realm, with Dainichi Nyorai (the buddha
Mahavairocana) as ultimate ground.
Sanno Shinto, the Tendai counterpart, centered on Hiyoshi Taisha at the foot of Mount Hiei.
Watarai Shinto at Ise, discussed below. Yoshida Shinto, from the fifteenth century.
Note what this plurality does not show. It is sometimes offered as a refutation of Kuroda, on the grounds that Tendai did not monopolize the framework. But Kuroda never said Tendai. His term is kenmitsu taisei, "the exoteric-esoteric system," chosen precisely because he was describing a shared esoteric framework spanning Tendai and Shingon both. Ryobu and Sanno as parallel productions of the two esoteric schools illustrate his thesis rather than undermining it.
Fourth: the direction was reversible, and could also be refused. In the fourteenth century, Tendai-trained figures including Jihen (active c. 1332–40) used the flattened ontology of original enlightenment to argue the reverse: if local phenomenal reality is already identical with ultimate reality, then the kami require no foreign justification. This position is called shinpon butsujaku ("kami as ground, buddha as trace"), or han-honji suijaku, "inverted honji suijaku."
A more radical move was also available. The Watarai priests of the Outer Shrine at Ise appear to have argued not that the relation should be inverted but that Amaterasu should be placed outside honji suijaku conventions altogether — bypassing the frame rather than flipping it (Andreeva, introduction to Assembling Shinto, Harvard, 2017).
But observe what even the most kami-primary position of the medieval period is made of. Inversion was constructed out of hongaku logic by Tendai-trained monks using Buddhist conceptual equipment. That is Kuroda's strongest single card, and it should be held in view when Report II evaluates him.
And then the institutional point. Recent scholarship reads honji suijaku less as metaphysics than as an instrument of property and jurisdiction. Establishing which buddha a kami "really" was established which temple held ritual authority over that shrine, and therefore which institution controlled the land and revenues attached to it. Pairings were unstable, frequently rewritten, and fought over.
Resist the reduction, though. The argument is not that the metaphysics was a sham covering material interest. It is that in a society where legal title to land was religiously encoded, ontology and jurisdiction were the same operation. Deciding what a kami really was was deciding who owned the estate (Rambelli, "Honji Suijaku at Work," in Teeuwen and Rambelli 2003).
This is the single most important thing to carry into Chapter 1. When the Meiji state ordered kami and buddhas separated, it was not primarily legislating about belief. It was dissolving titles.
0.3.3 The mutual dependence of the two laws
The accompanying political theory was obo buppo soi, "the mutual dependence of the Royal Law and the Buddha Law." Obo is
the law of the sovereign; buppo is the Dharma. Each supported and legitimated the other.
Kuroda proposed two models of the medieval order, frequently confused in secondary literature, and worth separating:
Kenmitsu taisei ("the exoteric-esoteric system") describes the religious order: Tendai and Shingon esoteric Buddhism
supplied the shared conceptual and ritual framework within which all significant religious activity operated, including kami
cult.
Kenmon taisei ("the system of ruling elites," literally "gates of power") describes the political order: medieval Japan governed
not by a single sovereign authority but by interlocking blocs — imperial court, warrior houses, and great temple complexes —
each with lands, armed retainers, and jurisdiction.
Both are Kuroda's; both are load-bearing; they answer different questions.
0.3.4 Ise as the alleged exception
Ise is routinely offered as the case proving a boundary existed. The evidence is real.
The Engishiki prescribes imikotoba ("taboo words") for the Saio, the imperial princess serving at Ise. Buddhist vocabulary was forbidden in the precinct and replaced by circumlocutions: a buddha was nakago ("middle child"), a sutra somegami ("dyed paper"), a temple kawarabuki ("tile-roofed"), a monk kaminaga ("long hair"), a nun mekaminaga. This is deliberate boundary maintenance, in law, at the most prestigious shrine in the realm.
Two qualifications. The taboo targets pollution within a specific precinct and office, not the doctrinal falsity of Buddhism. And Ise was not sealed: Ryobu Shinto mapped both Ise shrines onto Shingon mandalas, and a Kamakura-period text recast Amaterasu in a thoroughly Buddhist frame as a judge of the dead alongside a lord of the underworld of Indian derivation (Teeuwen, in Teeuwen and Rambelli 2003).
The honest summary: Ise maintained a real ritual boundary and did not maintain a doctrinal or institutional one. That is more interesting than either romanticism can accommodate.
0.4 Strands that fit neither category
0.4.1 The jingi apparatus and its separate law
Jingi ("the heavenly and earthly deities") was the official term for the deities of the court cult, drawn from Chinese ritual vocabulary. The Jingikan ("Department of Divinities") was the organ of state responsible for them, established under the Chinese-modeled law codes. In protocol it ranked above the Daijokan, the Great Council of State that ran the actual government, though its head held a modest court rank — a discrepancy that tells you something about ceremonial precedence versus real authority.
The legal treatment was asymmetric in a way that bears on whether the state distinguished these domains. Kami affairs were governed by the Jingiryo ("Code on Deities"), administered by the Jingikan. Buddhist clergy were governed by a separate instrument, the Soniryo ("Code on Monks and Nuns"), administered not by any religious body but by a civil bureau under the Ministry of Civil Affairs.
This cuts against the strongest form of Kuroda's thesis. From the eighth century the state maintained two distinct legal and administrative regimes. That is not two religions, and the jingi system was thoroughly Chinese in bureaucratic form. But it is real institutional differentiation, predating the medieval combinatory order and outlasting it — and it is one of the materials the Meiji state would later work with.
0.4.2 Purity law
The most durable non-Buddhist element is the pollution system.
Kegare ("defilement," "pollution") denotes ritual impurity, and must be distinguished sharply from sin and from anything in Indian Buddhist vocabulary. Kegare is not a moral category and does not track intention. It is contracted by contact — with death, birth, blood, disease — regardless of whether the contact was voluntary or blameworthy. It is removed not by repentance but by harae (purification) and misogi (ablution). It is distinct from tsumi, transgression, which does involve fault. It is also contagious: the Engishiki specifies that defilement passes from person to person through up to three degrees of contact.
The Engishiki ("Procedures of the Engi Era") is the key document. Ordered in 905, completed 927. Of its fifty books the first ten concern the Jingikan. Book 8 contains twenty-seven norito (liturgies addressed to the kami). Books 9 and 10 comprise the Jinmyocho ("Register of Deity Names"), listing 2,861 shrines and 3,132 kami receiving court offerings. Book 3 sets out a graded hierarchy of pollution: thirty days' abstinence after the death of a person, seven after a birth, six after the death of a domestic animal, three after the birth of one (Scheid, "Death and Pollution as a Common Matrix of Japanese Buddhism and Shinto"; English translation of the text: Bock, Engi-shiki, 1970–72).
Note what the hierarchy encodes: human death defiles more than human birth, which defiles more than animal death or birth. This is a systematic ritual logic, and it is not derived from Buddhist ethics. Buddhist doctrine has a great deal to say about death and nothing at all to say about death rendering one ineligible to approach a shrine for thirty days.
0.4.3 The outcast orders
Because contact with death and animal carcasses conveyed kegare, the tasks involving such contact — corpse disposal, leatherwork, butchery, executions — fell to hereditary groups placed outside ordinary society, designated by terms including hinin ("non-persons") and eta (characters that read "abundance of defilement"). The inujinin attached to the Gion shrine complex in Kyoto, under the great Tendai monastery Enryaku-ji, performed corpse disposal and shrine cleaning while holding formal shrine affiliation.
Amino Yoshihiko (1928–2004) argued that in the medieval period the status of such groups was genuinely ambivalent rather than simply degraded — marginality carried a sacred charge as well as stigma — and that the hardening into systematic early- modern discrimination was a later development (Amino, Rethinking Japanese History, 2012).
The structural point: this system is generated by the intersection of two distinct logics, a purity code with no Buddhist derivation and Buddhist prohibitions on killing, administered by institutions that were simultaneously shrines and temples. It belongs to neither tradition, and appears in neither romanticism.
0.4.4 Onmyodo
Onmyodo ("the Way of Yin and Yang") was a system of divination, calendrical calculation, directional taboo, and apotropaic ritual derived from Chinese yin-yang and five-phases cosmology. It was not folk practice at the margins but an organ of state: the Onmyoryo ("Bureau of Yin and Yang") sat within the Ministry of Central Affairs, and its diviners governed the timing and direction of court activity.
For a Heian aristocrat, whether to travel on a given day was an Onmyodo question, whether a death in the household required seclusion was a kegare question, and whether to commission sutra recitation was a Buddhist question. These were not felt as three religions but as three departments of one practical order. (See the Onmyodo special issue, Japanese Journal of Religious Studies 40/1, 2013, free at the Nanzan archive.)
0.4.5 Mountain asceticism
Shugendo ("the way of ascetic practice") combined kami cult, esoteric Buddhism, Daoist elements, and mountain pilgrimage into an organized tradition with its own lineages, ordinations, and territories. Its practitioners were yamabushi ("those who lie down in the mountains").
Shugendo matters here for one reason above all: it is the clearest case of a tradition with no possible home in either post-1868 category. It was not a shrine tradition and not a temple tradition; it was constitutively both. It was targeted directly by the separation orders and abolished outright in 1872. Gaynor Sekimori's work on the Haguro complex traces what that meant on the ground.
0.4.6 Confucian and Neo-Confucian statecraft
Under the Tokugawa, Neo-Confucianism — the systematized Confucian metaphysics and ethics associated with Zhu Xi (1130– 1200) — supplied much of the vocabulary of governance and moral education. It is customary to treat this as secular political philosophy rather than religion, but the figures involved did not observe that boundary.
Two facts reset a common assumption.
Anti-Buddhist polemic in Japan was Confucian before it was nativist. Hayashi Razan (1583–1657), the founding Confucian scholar of the Tokugawa establishment, and Yamazaki Ansai (1619–1682), the most influential Zhu Xi scholar of the century, both began as Buddhist monks and both turned against Buddhism after encountering Zhu Xi's critiques. The intellectual attack that fed into 1868 has a Confucian tributary long predating Kokugaku.
The resulting kami theologies were themselves hybrids. Ansai developed Suika Shinto, identifying Neo-Confucian metaphysical principles with the deities of the Japanese chronicles, and received his Shinto religious name from a disciple of the Yoshida lineage. So a Neo-Confucian anti-Buddhist reading of the kami was transmitted through a Shinto school built out of Buddhist materials.
0.5 For readers coming from Indian Buddhism
Several key terms look familiar here and mean something different. Four matter for Report I.
0.5.1 Kami is not deva
The pairing of kami with buddhas superficially resembles the Indian assimilation of local deities, where devas are real, powerful, still within samsara, and subordinate to the Buddha. Two differences. Kami are not a defined class within a settled cosmological hierarchy; the category is loose enough to include ancestors, natural features, abstractions, and the emperor. And the direction of authority was reversible in Japan in a way it never was in India.
