Japanese

The kami are counted at eight million, which is a way of saying they are not counted. That is the first structural fact: the pantheon is open at the edges, and every mountain, river, shrine, exceptional tree and notable ancestor can hold one.

The second is that a genealogy was laid over this openness. The Kojiki begins with heavenly deities, moves to Izanagi and Izanami and their generation of the islands, and arrives at the descent of the imperial line from Amaterasu. That line is the spine of the written mythology and the justification for a dynasty, and it does not describe how kami were actually venerated across the provinces.

Whether Amaterasu heads the kami generally is genuinely arguable. She heads the imperial account, and that primacy was pressed hard during the period of State Shinto. Local cult has always had its own centres of gravity, and Inari and Hachiman have far more shrines than any reading of the Kojiki would predict.

Succession is by delegation rather than conflict. Izanagi, returning polluted from the land of the dead, purifies himself, produces the three noble children, assigns them their domains and passes from the narrative. There is no war between generations of kami.

The door from human to kami is wide and has been used repeatedly. Sugawara no Michizane, exiled and dead, was enshrined as Tenjin after disasters were attributed to him. Tokugawa Ieyasu was enshrined at Nikko. Enshrinement of the dead is a live institution rather than a closed mythological episode.

The two chronicles that carry the kami mythology were commissioned by the court, within eight years of each other, and they do not agree. That disagreement is the most instructive thing about them.

The Kojiki, presented in 712, and the Nihon Shoki, completed in 720, tell the same broad sequence differently, and the Nihon Shoki habitually records variant versions side by side with the formula that another book says otherwise. A tradition that preserves its alternatives in the same volume is telling the reader something about how settled it was.

Both were produced to serve a political need: to establish the descent and legitimacy of the ruling line at a moment when continental models of statehood were being adopted wholesale. They are therefore excellent evidence for eighth-century court theology and much weaker evidence for what anyone in the provinces had been doing.

Two other bodies of material partly correct that. The fudoki, provincial gazetteers ordered at the same period, record local deities and place legends that the chronicles ignore. The norito, the liturgical addresses preserved in the tenth-century Engishiki, are the actual words of ritual rather than narrative about it.

Finally, the long entanglement with Buddhism has to be held in view. For most of Japanese history kami and buddhas were worshipped in the same institutions under theories that identified them, and the separation of the two was imposed by government in the nineteenth century. Much of what is now presented as ancient and purely Shinto dates from that separation.