1993 年 102 巻 11 号 p. 1947-1972,2070-
While we have seen significant development recently in the study of castles in medieval Japan, the research on these spaces during the early medieval period leaves a lot of areas to be explored. Also, such research themes as the castle in relation to graves and holy ground have brought to the forefront the important mental or psychological aspects of the castle. In the present paper, the author examines the conditions of existence for and mentality surrounding the castle in early medieval society. First, the author uses such source materials as the Azumakagami and Heike Monogatari to investigate the actual conditions of castles during that time, and finds that what is called jo 城 and jokaku 城郭 (with the exception of these terms being used for Kyoto) does not appear in the realm of everyday life, but rather at extraordinary times of war and rebellion. The castles that appear here were usually built on open plains for the purpose of combat between individuals on horseback. In many cases their structural character included a relatively simply one story line of defence encompassing a particular space. The phase jokaku wo kamau 構城郭 (lit. build a castle) actually meant to take a stand within this defensive position, or space. The author then investigates the various ways of thinking at the time about where to set up defensive positions, in order to discover the territorial character of the places where warriors "took their stands". In many cases areas deeply related to religious belief like temples and shrines were chosen. For example, Kinugasa-jo built by the Miura family during the Genpei War was set up on a mountain considered to be holy ground and a sutra mound (kyozuka 経塚) was even constructed within the emplacement. Miura Yoshiaki's reason for taking a stand in this particular place (according to the Engyo 延慶 version of the Heike Monogatari) was to die in a castle so well known that both his ancestors and the enemy would not fail to hear of it. Mt. Kinugasa was chosen as a defensive position because it had important religious significance (including ancestor beliefs) for the Miura family. There are many cases in which this very psychology worked in the selection of holy or spiritual places as castles in early medieval Japan. Taking this particular mental aspect into consideration, we can also see the spacial significance of Akadani in Echigo Province, were the Jo family built a castle in which they prayed to the bodhisattva, Sudristi (妙見菩薩; Buddhist deification of the North Star),to bring misfortune down on the Genji forces. The author also points out that both Mt. Kinugasa and the area around Akadani had strong connections to mountain beliefs and ascetic practices (Shugen 修験), and surmises that warriors who built their defenses in mountain areas in many cases called upon mountain ascetics (yamabushi 山伏) to guide them into holy, sanctified ground. This is how, for example, the Satake family of Hitachi Province took their stand at Kanasa Shrine and then fled after their defeat to the hallowed ground of Mt. Hanazono, the Jo family built their castle at Mt. Totsusaka, the holy ground of Haguro asceticism, and why during the rebellion of Kiyohara Iehira (Go-sannen-no-Eki; 1083-1087), Fort Kanezawa was built on the sacred site of Gionji temple, the starting point for mountain pilgrimages.