Monograph No. 68
KIBIIKE ABANDONED TEMPLE EXCAVATION REPORT INVESTIGATION OF THE SITE OF KUDARA NO ODERA
CONTENTS
Chapter I Introduction 1
1. Circumstances regarding the excavation 1
2. Organization of the excavation 3
3. Publication of this report 4
Chapter II The investigation 6
1. Area of investigation 6
A. Location of the site 6
B. Previous investigations 9
C. Surveying 11
D. Division into sectors 12
2. Outline of each excavation 14
A. Excavation no.81-14 (main hall) 14
B. Excavation no.89 (pagoda, cloister southern sector) 15
C. Excavation no.95 (cloister southern and western sectors) 16
D. Excavation no.105 (priests’quarters, cloister eastern sector, main hall) 17
E. Excavation no.111 (middle gate, priests’quarters) 18
F. City of Sakurai, Excavation no.9 (priests’quarters) 19
G. City of Sakurai, Excavation no.11 (northern periphery) 20
H. City of Sakurai, Excavation no.12 (southern periphery) 20
3. Excerpts from excavation logs 21
A. Excavation no.81-14 21
B. Excavation no.89 23
C. Excavation no.95 25
D. Excavation no.105 26
E. Excavation no.111 29
F. City of Sakurai, Excavation no.9 31
G. City of Sakurai, Excavation no.11 32
H. City of Sakurai, Excavation no.12 33
Chapter III Archaeological features 34
1. Location and topography of the site 34
2. Descriptions of separate features 38
A. Main hall SB100 and features in its environs 38
B. Pagoda SB150 and features in its environs 46
C. Middle gate SB320 and features in its environs 51
D. Cloister southern sector SC160 and features in its environs 53
E. Cloister western sector SC200 and features in its environs 57
F. Cloister eastern sector SC300 and features in its environs 61
G. Priests’quarters and features in their environs 62
H. Features in the northern periphery 70
I. Features in the southern periphery 72
Chapter IV Artifacts 76
1. Tiles and bricks 76
A. Eave tiles 76
B. Miscellaneous roof tiles 90
C. Round tiles 93
D. Flat tiles 101
2. Other ceramics 122
A. Pottery 122
B. Haniwa 138
3. Stone implements, clay objects, metal implements, etc. 143
A. Stone implements 143
B. Clay objects, etc. 145
C. Metal implements 146
D. Wooden implements 147
E. Plant and animal remains 148
Chapter V Interpretations 149
1. Inference of the temple name and its historical change 149
A. Kibiike abandoned temple and Kudara no odera 149
B. Historical changes of Kudara no odera and Takechi no odera 150
C. Location of Kudara no odera 153
D. Location of Takechi no odera 155
SUMMARY
The Kibiike abandoned temple is located near the southeastern edge of the Nara basin, at Kibi in the city of Sakurai, Nara prefecture. As seen from the Asuka region, where palaces of the seventh century emperors concentrate, it lies to the northwest, on the opposite side of the hilly mass of Mt. Kaguyama.
While it had been previously known that roof tiles from the seventh century were recovered in this area, the view taking it to be the remains of a kiln was formerly dominant. But when an archaeological excavation was conducted in 1997, on the occasion of repair to the dike of the pond, it became clear that these were the remains of a large temple of the seventh century. The attention this received as an important discovery from the perspective of ancient history is still a fresh memory. Since then academic investigations, including the initial discovery, have been conducted over five years up to 2001, yielding many significant results. The current volume is the formal report on those activities.
1. Archaeological features
Main hall. The main hall had an enormous tamped earth podium, about 37 m east-west by 25 m north-south, and with an original height of 2 m or more, built into a depression prepared for it to a depth of approximately 1 m. One portion of the southern side of the podium extends out toward to the south. In addition, the podium’s circumference was ringed with a gutter having stones on its bottom, which apparently drained to the west together with water from the interior of the depression made for the podium. Although no pillar base stones survive, nor was it possible to detect traces left by them, from the podium’s size and the ratio of its long and short dimensions, a building of seven bays on its main axis and four bays transverse to it may be presumed. A skirting pent roof was possibly added to the outer side of this structure. Although no trace remains of the finished surface, from the condition of the archaeological features a finished surface made of wood is inferred. This is assumed to apply as well for the pagoda, and middle gate and cloister, which are described next. As pebble paving was detected on the podium’s eastern side, it is thought that originally this type of pavement was laid down around the podium.
Pagoda. The pagoda located to the west of the main hall has an enormous podium for which no depression was dug in preparation, tamped earth layers being piled directly on the leveled ground to an original height of 2.8 m, and measuring approximately 32 m on a side. In the central portion of the podium is a large hole left from where the central pillar base stone was removed, and it is thought that this stone was of the above-ground type, with its upper surface protruding up above the podium. Also, on the western side of the podium, a slope for hauling the central pillar base stone up onto the podium was detected within the tamped earth layers. While neither pillar base stones nor traces related to these were left, other than for the central pillar base stone, from the size estimated for the latter it is highly possible that the four pillars forming the inner corners of the pagoda, surrounding the central pillar, were also stood upon the central pillar base stone. The podium exceeds all other temples of the Asuka period (592-694) in scale, and even the seven-storied pagodas of Todaiji (in the city of Nara) in the Nara period, and equals the nine-storied pagodas of Daikandaiji (at Koyama, in the village of Asuka, Takaichi county) of Emperor Monmu’s reign, and Hwangyongsa temple of Silla. It is reasonable to think that the pagoda of the Kibiike abandoned temple was also nine stories. The pagoda’s first story layout is inferred to have been seven bays square, with equally spaced pillars 3.3 m apart.
Middle gate and cloister. The podium of the middle gate had been leveled, but from the encircling rain gutter it is estimated to have been approximately 12.0 m east-west by 9.8 m north-south. It was probably of gabled-roof construction, three bays along the ridge by two bays transverse to it. But the gate’s location is off to the west from the main hall’s center, from which the possibility is inferred of there being two middle gates, one each corresponding to the main hall and pagoda. Also, regarding the cloister, the southern, eastern, and western sectors have been ascertained, and the podium width can be reconstructed as 5.4 m, with a beam width of 3.3 m. The distance separating the eastern and western sectors of the cloister measures 156.2 m between the outside rows of pillars, and may be regarded as planned at 440 taishaku (1 taishaku approximately equaling 0.355 m). While the space enclosed by the cloister is thus very wide, the architecture of the middle gate and cloister is on a surprisingly small scale in comparison with the main hall and pagoda. This may be related to the arrangement of two middle gates, as previously surmised.
Priests’ quarters. To the north of the main hall, and to the west from that point, three embedded pillar buildings used as living quarters for the priests were ascertained. All were buildings with ridges running east-west and having transverse widths of two bays, and represent the oldest remains of priests’ quarters in the country that have been confirmed through excavation thus far. Of two which line up north and south, building SB 340 has a ridge eleven bays long, and assuming their western ends were aligned, SB 400 is also inferred to have had a long and narrow plan of nine bays in length. They were probably used by laying a low floor and partitioning the interiors. The distance from the northern line of pillars of SB 400 to the outer line on the southern face of the cloister is 440 taishaku, the same as the cloister’s east-west span. From this planned arrangement, and with regard to the central axes of the main hall and the temple compound as a whole, it may be supposed that a number of such priests’ quarters were laid out. While there are a number of points of morphological difference in comparison with eighth century priests’ quarters, these finds can be regarded as extremely valuable for ascertaining the development of this type of structure.
Temple precinct and building arrangement. The extent of archaeological features relating to the Kibiike abandoned temple reaches more than 260 m north-south by 180 m east-west. The center of the cloister assumes the so-called Horyuji style arrangement, with a southern-facing main hall and pagoda lined up east-west, and represents the oldest example of this style. The main buildings were placed along a plan based on multiples of ten taishaku, based in principle on the outer line of pillars of the cloister. The centers of the main hall and pagoda are 100 taishaku (35.7-36.5 m) from the cloister’s eastern and western sectors, respectively, and the main hall’s center is located 150 taishaku (53.5 m) from the cloister’s southern sector. The central area enclosed by the cloister is long in the east-west direction, and its east-west scale greatly exceeds all other domestic temples of the same period, equaling that of Hwangyongsa temple of Silla. Also, the distance of 240 taishaku (84.1 m) between the centers of the main hall and pagoda is distinctively large, though this possibly derives from the height of the pagoda. While the lecture hall and the northern sector of the cloister were not detected, from the large scale of the leveled ground and from recoveries of roof tiles, the existence of these structures can be seen as certain. The degree to which the full temple complex was completed may also be regarded as high.
Features postdating abandonment. After the temple at Kibiike was abandoned, the surrounding area was subsumed within the Fujiwara capital (694-710), and a street plan based on a rectangular grid was laid out. In addition to the roadside gutters, perimeter walls and buildings of house lots, partitioning ditches, and so forth were found in the excavations. Subsequently, after the move to the Nara capital (710-784), it is thought that division into agricultural plots under the jori system was undertaken, and the area converted to paddy. From recovered roof tiles it appears that a small tile-roofed Buddhist building was erected around the first half of the eleventh century.
2. Artifacts
Eave tiles. Four types of round eave tiles (IA, IB, II, III), and three types of flat eave tiles (IA, IB, III) were recovered, but excluding the Heian period type III items (four round and two flat eave tiles), the preponderance are limited to round eave tiles having eight-petal lotus designs with simple petals (43 items of type IA, 23 of type IB, 25 indeterminate items), and flat eave tiles with stamped palmette designs (12 items of type IA, 1 of type IB), all dating from the time of the temple’s construction. Tiles used for reproofing are entirely lacking, so it is almost certain that the temple was moved to another location. The tiles recovered from the Kibiike abandoned temple site are considered to be items discarded on that occasion as incapable of being reused. Round eave tiles of types IA and IB, in terms of style, have elements that slightly precede the tiles used in the construction of the main hall of Yamadadera temple (at Yamada, in the city of Sakurai), begun in 643. In contrast, flat eave tiles of types IA and IB were made with molds of flat eave tile type 213B from Ikarugadera temple (also referred to as Wakakusagaran of Horyuji, in the town of Ikaruga, Ikoma county), and from the condition of damage on the molds, they were clearly made after those of Ikarugadera. Accordingly the ages of the eave tiles used at the temple’s founding may be judged as falling between the 630s and the beginning of the 640s. Additionally, the molds used for flat eave tiles of types IA and IB were afterwards taken to the Kuzuha Hiranoyama tile kiln, and used to produce tiles for Shitennoji temple (in the city of Osaka), and the IB mold was further used to make tiles for the founding of Kaieji temple (in the city of Sennan, Osaka prefecture). It is highly likely that these types of tiles from Shitennoji were related to maintenance work on the complex conducted in conjunction with the construction of Naniwa capital during the reign of Emperor Kotoku (645-654), which in addition to strengthening the chronological argument given above, is suggestive of the positions held by Shitennoji and the Kibiike abandoned temple during this period.
Round and flat tiles. A total of 2,649.4 kg (576.1 kg of round tiles, 2,073.3 kg of flat tiles) were recovered. More than 99 percent of the round tiles are type 1 round tiles from the temple’s founding, being made on a single wooden mold (which served as the core around which a slab of clay was folded) all the way to the lip on their butt ends. The thick-walled type 1 flat tiles from the temple’s founding, with which these items were paired, represent approximately 90 percent of all flat tiles. The remaining flat tiles are nearly all the thin-walled type 2 flat tiles, for which there is no corresponding type of round tile, and it is surmised that tiles from the time of the founding of Abedera temple (at Abe, in the city of Sakurai) were brought from there. Assuming the ratio of type 1 to type 2 flat tiles is 9:1, and dividing the total weights of round and flat tiles by the respective weights taken for a single tile of that type, thereby converting into numbers of complete items, the ratio of round to flat tiles is calculated in the range from 1:2.62 to 1:3.34. These are entirely appropriate figures for the tiles used on a building roofed entirely with tile. Also, the amount of tile recovered at the Kibiike abandoned temple is 68.3 kg/100 ㎡ on the whole, with areas reaching several times that level, as seen in the environs of the main hall (282.3 kg). By contrast, for ancient temples in the Asuka and Fujiwara districts, recovered tiles may exceed a metric ton per 100 ㎡, whereas conditions close to those of the Kibiike abandoned temple are presented in the central portion of the Fujiwara palace. There the amounts per 100 m2 are on the 200 kg and 300 kg levels, and it is certain that this shows the results of the greater portion of roof tiles being sent to another location. It is therefore deduced that tile-roofed structures existed at the Kibiike abandoned temple, and were in like manner removed to another location.
Pottery. Although the amount of recovered pottery filled eighty-two wooden boxes used for sorting, small sherds were numerous, and good materials few. Dirt used in leveling the site at the time of the temple’s founding included pottery from the first half of the seventh century, and of the pottery recovered from the depression dug in preparation for the main hall podium and from the gutter surrounding the podium, items for which the date can be ascertained fall within the first half of the seventh century. These materials indicate the terminus ad quem for these contexts, and it may be thought that the establishment of the Kibiike abandoned temple comes very close to this in time. Pottery included in the postholes of the priests’quarters indicates the same age. By contrast, the hole left by the removal of the central pillar base stone of the pagoda yields pottery from the second half of the seventh century, allowing an inference of the date of the abandonment. In addition, pottery dating from the latter half up to the end of the seventh century is included in the rain gutters of the cloister, and in the holes left from removing the pillars of the priests’ quarters.
Artifacts predating the temple’s founding. As artifacts preceding the Kibiike abandoned temple in time, in addition to knife-shaped stone tools of sanukite (Paleolithic period) and projectile points (Jomon), large bifacially beveled stone axes (Yayoi), and other items were recovered. While none of these accompany archaeological features, they indicate that the area was the site of activities of livelihood from early times. In addition, haniwa (funerary ceramics, Kofun period) over nearly the entire area of investigation. It is presumed that mounded tombs are present atop the small hills spreading from the temple complex’s north to the southeast. Apart from this, crucibles, spruecups, iron slag, scorched earth, and sharpening stones were recovered in concentrated fashion from the area around the priests’ quarters, indicating that workshops related to metal crafts existed at the time of the temple’s founding.
3. Kibiike abandoned temple and Kudara no odera, Takechi no odera
The Kibiike abandoned temple’s name. As seen above, the Kibiike abandoned temple is thought to have been founded between 630 and the early 640s, and later moved to a separate location. The scale of the main hall and pagoda, clarified through archaeological excavation, exceeded all other domestic temples of the same period, and came close to that of Hwangyongsa temple of Silla, or Daikandaiji of Emperor Monmu’s reign. In the same manner that these were founded on a grand scale as state temples, the Kibiike abandoned temple may also be seen as resulting from the wish of a Japanese monarch ( tenno or okimi). In that case, only Kudara no odera would match with the date of founding. Both the Nihon shoki (720) and the Daianji garan engi narabi ni ruki shizaicho (747) record the establishment of this temple by Emperor Jomei in 639, the first to be founded on the wish of a monarch. Jomei died shortly after the start of construction, but his consort and successor Empress Kogyoku continued its construction, and it appears to have been completed to a certain degree by the reign of Emperor Kotoku (645-654). Subsequently Emperor Tenji, son of Jomei and Kogyoku, donated the joroku Buddha, made in the dry lacquer sculpture technique, which later became the principal image in the main hall of Daianji temple, and the completion of the temple appears to have been proceeding in relatively satisfactory fashion. Further, Kudara no odera was moved in 673 to Takechi, becoming known as Takechi no odera. In this regard as well, the condition of the artifacts and archaeological features of the Kibiike abandoned temple agree well with the historical record. In addition, as it is known from the Nihon sandai jitsuroku (901) that Kudara no odera stood on the bank of “the Kudara river of Tochi county,” this is thought certain to correspond with the river currently known as Yonegawa, which flows to the south of the Kibiike abandoned temple. There is almost no doubt that the Kibiike abandoned temple is Kudara no odera.
The significance of Kudara no odera. Kudara no odera, the first temple to be erected by the imperial family, was the product of a sense of rivalry with Asukadera, the clan temple of the Soga line. The nine-storied pagoda which far surpasses that of Asukadera is the very symbol of this. It is also highly possible that the existence of priests, and of others returning after study abroad, from Tang China, and Paekche and Silla on the Korean peninsula, is related as a background factor to its construction. In particular, information about the nine-storied pagoda which saw completion in 645 at Hwangyongsa temple in Silla, was surely of great import for Japan at the time. It is not difficult to imagine that the founding of Kudara no odera was inseparable from relations to these conditions in the East Asian world.
The location of Takechi no odera. The move of Kudara no odera from “the site of Kudara” to “the site of Takechi” appears to be related to the building of Emperor Tenmu’s palace, Asuka Kiyomihara no miya; subsequently Takechi no odera was renamed Daikandaiji in 677 (the Daikandaiji of Tenmu’s reign), while in the reign of Emperor Monmu (697-707) a separate large temple complex was built (the Daikandaiji of Monmu’s reign). The precise location of Takechi no odera is still undetermined, but a rice field in Yabe village, Takechi county appears in the Nihon sandai jitsuroku as the former temple site of “Takechi Daikanji,” and it can be seen to have been in close proximity to the Daikandaiji of Monmu’s reign. Meanwhile, a record appears in the Ruijusandaikyaku (eleventh century) indicating that between the Daikandaiji of Monmu’s reign, and the Asuka river to its west, there lay an additional temple site. That location corresponds to the place named “Takechi no sato,” and roof tiles older than the Daikandaiji of Monmu’s reign, such as flat eave tiles with concentric arc patterns which are seen in common at Daianji in the Nara capital, are distributed over that area. The possibility that Takechi no odera was located in this “Takechi no sato” is great.
2003年3月28日 発行
吉備池廃寺発掘調査報告
－百済大寺跡の調査－
奈良文化財研究所創立50周年記念学報第68冊
