When Prince Sadahito 貞仁 ascended the throne in 1072 and became Emperor Shirakawa 白河 (1053-1129), no one could have predicted that this young prince would dominate the Japanese state for the next four decades. His father, Emperor Go-Sanjō 後三条 (1034-1073), was the first emperor in 170 years who did not have a Fujiwara mother, and the early years of Shirakawa's reign did not suggest the grandeur that was to come. Many scholars have addressed the so-called retired emperors' (Insei 院政) period; some have written historical accounts of a particular retired emperor (in 院) and his policies. In this article, I will reconsider the historical process whereby "rule by regents" (sekkan seiji 摂関政治) developed into "rule by retired emperors" (insei seiji 院政政治), question the validity of these two categories, and approach the latter from the point of view of monastic lineage to formulate a new characterization of the retired emperors from Shirakawa to Go-Shirakawa 後白河 (1127-1192).
Whether or not to consider the retired emperors between 1086 and 1185 a new system of rule is a question requiring great methodological subtlety. Reflecting upon this question within the context of shared sovereignty, I will avoid the fixed categories "Michinaga's rule" or "rule by retired emperors." Rather than looking for an absolute ruler, I place the institution of the retired emperor at the center of a network that as a whole constituted the state. Indeed, throughout Japanese premodern history, no single entity, whether person or monastery, could ever claim to hold absolute sovereignty. What we do find, however, are particular individuals (for example, Fujiwara no Michinaga 藤原道長 [966-1027] or Retired Emperor Shirakawa) or monastic institutions (Kōfukuji or Enryakuji) whose power increased and decreased within the [End Page 239] general framewor a state characterized by shared sovereignty. Thus, in the same way that historian Takayanagi Mitsutoshi 高柳光寿 referred to the imperial system as the main constituent of public matters (kuji 公事), I regard the retired emperor as being at the center of a state network. 1 The question of the relationship between the retired emperor and the state is thus not dissimilar from the current debate on the relationships between kasei 家政 and kokusei 国政 and between the kenmon 権門 (see below) and the state. Ihara Kesao 井原今朝男 defines kokusei as the polity on a macro level that originally developed out of the ritsuryō 律令 state, and kasei as the polity on a micro level that regulates a particular kenmon. In applying these modern categories to a premodern context, he discusses two characteristics of the premodern concepts of public and private that are significant in relation to the question raised above. First, Ihara argues that these spheres overlapped in premodern Japan, lacking their modern antithetical separation. Second, he defines kasei as a body that internally regulates a kenmon, but stresses that it is impossible to separate this kasei out of the larger kokusei framework. 2 In other words, what constitutes the state is a network of different actors.
Contrary to the work of Ishimoda Shō 石母田正 and the negative view of the retired emperors as despots prior to the rise of the warrior class, this article does not present the institution of the retired emperors as an outcome of the actions of Fujiwara no Michinaga. 3 Harkening back to the early scholarship of Hayashiya Tetsusaburō 林屋辰三郎 and inspired by Ihara's description of the relation between the various units within the larger framework of kokusei, I call attention to several parallel developments between different kasei of which certain regents, Michinaga, the retired emperors, and the large southern and northern monastic complexes were an integral part. 4
Through an analysis of specific temples, lineages, and rituals, I trace the evolution of the retired emperors back to Emperor Uda 宇多 (867-931), rejecting the common assumption that the retired emperors' power grew out of a new form of sovereignty (ōken 王権) established by Michinaga. 5 I believe it makes more sense to explain particular institutional and doctrinal developments under the retired emperors from the point of view of the newly established imperial vow temples (chokuganji 勅願寺) surrounding Ninnaji 仁和寺 and their connections with Nara/Nantō Buddhism from Emperor Uda on. The creation of...