<論説>神社建築の形成過程 : 平安時代前期・中期を中心に

Abstract

天武朝の官社制創出に伴って、律令神祇祭祀制度が確立されるとともに、神社の施設が整備されるようになったとされている。しかし神社建築が七世紀後半以後直ちに全国に一律に流布してゆくわけではない。官社制の下での神社建築は、いつ、どのような過程で全国に流布していったのか。神社数の変化、神社修造・清掃政策、国史に見える神社施設の実態、発掘遺構などから知られる建築の形式などの検討を通じて、平安後期までに定型的な神社建築が徐々に建設されるようになってゆく過程と、神社建築の多様性の淘汰の過程を検討した。具体的には以下の点を明らかにした。平安前・中期に、国家は神社建築の造営・修理・維持管理や清浄に保つことを全国の諸社に命じ続けた。この対象となる神社は、限定的なものだったと考えられるし、その命令は迅速には貫徹しなかった。そのような中で、徐々に神社本殿等の施設が全国の神社に浸透していった。しかし、平安時代中期になっても神社の建築に関しては、本殿を持つ神社も、持たない神社も混在し、さらに発掘遺構などから見て本殿があったとしてもいまだ画一的な形式とはならず、七世紀以前の古風な建築形式を模倣したかと思われる本殿も含まれていた。神社本殿を建設することが定着し、限定的な建築形式に収蔵するのは少なくとも十二世紀に降る。官社制の成立が直ちに一律に神社本殿の建設をもたらしたのではなかった、その具体的様相を解明した。

Accompanying the creation of the system of government shrines during the reign of Tenmu, the Ritsuryo state's system of Kami worship was also established and shrine structures began to be maintained. However, this does not signify that shrine architecture spread uniformly throughout the country following the latter half of the seventh century. When and by what process did shrine architecture spread throughout the country under the government-shrine system? Through an examination of the actual state of shrine structures as seen in national histories and the forms of architecture as seen in archaeological sites, I examined and the gradual process of the construction of a style of shrine architecture that became fixed in the late Heian period and the process of the winnowing this type out of the multiplicity of shrine architecture. I specifically clarified the following points. First, in tracing the change in the number of shrines from the eighth to ninth century, I demonstrated that the numbers increased over time. Conversely, I point out that only a small number of shrines among those of the five capital provinces and seven circuits are seen in the national histories of the eighth century. Furthermore, I investigated the policy of repairing and maintaining the cleanliness of the shrines from the eighth through the tenth century. As indicated in previous studies, the state continuously ordered shrines throughout the country to build, maintain and repair shrine architecture and maintain their purity. Judging from the aforementioned estimate of the number of shrines, it can be surmised that the number of shrines that were the object of such orders was limited to a few, and it appears, on the basis of their repeated issuance, that the contents of these orders was not immediately adopted nor implemented by the shrines. This means that architectural structures had not completely spread to regional shrines and that governors and local religious organizations that had responsibility for building and maintaining structures were not functioning fully. In addition, one also sees cases of previously built shrine structures not being sufficiently maintained and ceasing to exist. Next, I sought examples in historical sources that would confirm the actual state of shrine architecture. At shrines that were not influential to the state, specifically named shrine architecture first appeared in the middle of the ninth century, and in contrast, it is common to find cases where there was no shrine architecture, and when there was an effort to build a structure, it was frequently a Buddhist temple associated with a shrine. I also found the case of the natural scenery substituting for architecture because were no architectural structures, as can be seen in record of the case of deity of Kozushima in lzu from Jowa 7 (840). On the one hand, this is an example of a shrine without shrine architecture, but it also suggests the circumstances of the spread of architectural structures to a certain extend in the area of the shrine. In order to grasp the reality of architectural structures, about which knowledge from written sources is limited, I chose archaeological remains that had a high probability of being those of shrine architecture from archaeological surveys, and examined them to recreate the architectural form. In addition, I took examples from written records such as the eleventh-century Kozuke no kuni kotai jitsuroku cho from which one can learn the actual form of the architecture, and I pointed out the fact that there was no fixed form for the main honden, and there were honden among them that appeared to be modeled on the archaic style of architectural forms from prior to the seventh century. This differs from the situation of extant shrine architecture from the medieval period onward that is concentrated in either the nagarezukuri or Kasuga-zukuri styles, and thus the prior to the eleventh century the form of shrine architecture had not been formalized. The fact that theNenju gyoji emaki, which is thought to be from the mid-twelfth century, depicts only the nagarezukuri or kasuga-zukuri styles, indicates that the style of the shrine honden had by then become uniform, and I hypothesize that the formalization of the honden style was linked to the formation of the system of shrine officials that occurred at the same time. The establishment of government shrine system did not immediately bring about the construction of a fixed style of shrine honden that spread throughout the country in a uniform manner. This article has elucidated concrete aspects of this fact.