ISHIGAMI SITE Ⅰ THE FINDSPOT OF THE ANCIENT STONEWORK

CONTENTS

Chapter I Introduction　1

1　Background of the Excavation　1

2　Excavation Teams / Organizations　3

3　Preparation of the Excavation Report　3

Chapter II Historical Environment　6

1　Location and Topography of the Site　6

2　Archaeological Sites in the Surroundings　10

3　Creation of the Site Image and the Previous Excavations　14

Chapter III Excavation　21

1　Surveying and Grid System　21

2　Outline of the Excavations　25

3　Excavation Logs (Abstract）　29

Chapter IV Archaeological Features　47

1　General Stratigraphy　47

2　Fills and Confirmation Surface of Features　50

3　Later Modification of the Topography　52

4　Descriptions of Archaeological Features　54

Ａ　Phase I (First Half of the Seventh Century AD）　54

Ｂ　Phase II (Mid to Late Seventh Century AD）　68

Ｃ　Phase III (Late to End of the Seventh Century AD）　77

Ｄ　Phase IV (After the Nara Period）　82

Ｅ　Kofun Period (Before the Sixth Century AD）　84

Ｆ　Features related to the Stonework　85

Chapter V Artifacts　87

1　Pottery and Clay Objects　87

Ａ　Pottery from Phase I Archaeological Features　88

Ｂ　Pottery from Phase II Features and Fill B　98

Ｃ　Pottery from Phase III Archaeological Features　110

Ｄ　Pottery of the Nara and Heian Periods　129

Ｅ　Blackened Haji Ware imported from Northeastern Japan　131

Ｆ　Inscribed Pottery with Spatula　132

Ｇ　Ceramic Inkstones　132

Ｈ　Silla Pottery　135

Ｉ　Clay Objects　136

Ｊ　Pottery of the Kofun Period　137

Ｋ　Pottery of the Medieval Period　139

2　Roof Tiles and Bricks　143

Ａ　Round Eave-end Round Tiles　143

Ｂ　Flat eave-end Tiles　150

Ｃ　Rafter-end Tile　152

Ｄ　Round Tiles　153

Ｅ　Flat Tiles　156

Ｆ　Tiles with Inscriptions　166

Ｇ　Discs cut from Rooftile　167

3　Wooden Objects　171

Ａ　Wooden objects from SE800 　171

Ｂ　Wooden objects from the Other Archaeological Features　182

4　Metal Objects　183

Ａ　Iron Objects　183

Ｂ　Copper Objects　215

5　Artifacts related to Metallurgy　217

Ａ　Earthen Tuyeres　217

Ｂ　Molds and Crucibles　219

Ｃ　Copper Material, Copper Droplet, Copper Slags, Iron Material　221

Ｄ　Iron Slags　223

6　Mud Plaster　224

7　Stone Objects　224

Ａ　Whetstone　224

Ｂ　Small rounded Pebbles　233

Ｃ　Other Stone Objects and Beads　234

Ｄ　Muro Volcanic Slate　235

Chapter VI Scientific Analyses　241

1　Residual Lipid Analysis of Haji Jars from SE800 241

2　Tree Species of wooden objects from SE800 250

3　Depositional Environment and Flow Direction of the Ditch SD335 255

Chapter VII Discussions　257

1　Pottery　257

Ａ　Dating of Pottery from the PhaseA-3 Features　257

Ｂ　Haji Jars from the Well Frame of SE800 263

2　Roof Tiles　267

Ａ　Dating of Ishigami Types A, B, D, and E　267

Ｂ　Dating of Ishigami Type F　272

Ｃ　Background of Ishigami Types A, B, D, E, and F　272

Ｄ　Background of the Roof Tiles of Asuka-dera Temple　273

3　Metal Objects　277

Ａ　Distribution of Iron Objects and the location of "Oharida Hyogo"　277

Ｂ　Relationship between the Iron Objects and the Soldiers' Equipment　279

4　Reconstruction of the Chronological Transition of the Features　284

Ａ　Previous Excavations and the Periodization　284

Ｂ　Reconsideration of the Periodization and the Rationale　287

Ｃ　Relative Sequence of Archaeological Features　290

5　Archaeological Features　292

Ａ　Typology of Deposits in Pillar Holes　292

Ｂ　Construction Method and Function of Masonry Ditches　294

Ｃ　Changes of the features in the First to fourth Excavation Areas　295

Ｄ　Changes of the Water Channel Network and Function　300

Ｅ　Features destroyed by Natural Flow NR310　　302

Chapter VIII Conclusions　303

Appendix: Ancient Stonework from Koaza Ishigami　307

Appendix Tables　319

English Summary　329

PLAN

PLATE

PHOTOGRAPHS

Summary

　This first volume of the final official report publishes the results of the first to the fourth excavations at the Ishigami site (石神遺跡） and the Asuka Fujiwara 214th excavation, as a supplementary investigation of the first excavation covering the southeastern part of the area where the Nara National Research Institute for Cultural Properties (hereafter, Nabunken）has conducted excavations continuously. Among them, the rice field at Asuka 287 (飛鳥287番地） is well-known for the provenance of the Sumeru (須弥山石）, a sacred mountain of ancient India, and the stone statue (石人像） in Meiji 35 and 36 (1902-1903）. At the same location, Ishida Mosaku (石田茂作） carried out an excavation between May and June of Showa 11 (1936）. Therefore, this is regarded as not only the first volume of excavation reports published by Nabunken but also a culmination of the 120-year research history on the Ishigami site, starting from the excavation of the stonework. In this volume, the authors propose a new periodization based on the current research and revise the chronological transition of the confirmed features shown in the annual bulletin 2009 on several important points, although this new interpretation should be evaluated after the publication of the other volumes of the excavation report.

1.　Background of the Excavations

　The Ishigami site, a seventh-century archaeological site situated in Asuka, Asuka Village, Takaichi district, Nara prefecture, Japan, is located to the north of the Asukadera Temple (飛鳥寺） and the east of Amakashi-no Oka (甘樫丘）. In the rice field of Koaza Ishigami, the stonework so-called Sumeru stones and the stone statue were unearthed 120 years ago, between Meiji 35 and 36 (1902-1903）.

　In the research history, two main hypotheses have been highlighted. The first hypothesis interprets the site as the banquet hall (or guest house） of Empress Saimei (reign: 655-661）, based on the characteristics of the Sumeru. Another hypothesis explains the Asuka-Kiyomihara Palace (飛鳥浄御原宮） of Emperor Tenmu (reign: 673-686） was situated at the Ishigami site and the surroundings. These hypotheses have been argued respectively without any mixture. The ongoing excavations by the Nabunken during approximately three decades are closely affected by these hypotheses.
Until around 1980, the current Ishigami site was called the "estimated site of Asuka Kiyomihara Palace," because of a hypothesis that the palace was once located in Asuka Elementary School and the eastern neighboring land by Kida Sadakichi (喜田貞吉）. Notwithstanding the very narrow area, the excavation in Showa 4 (1929） revealed a stone pavement in the territory of the elementary school, which was regarded as the remain of Asuka Kiyomihara Palace by the excavators. In the rice fields of Koaza Ishigami, the east of the elementary school, a hypothesis that identified pieces of the unearthed stonework with Sumeru in the Saimeiki (斉明紀） was proposed in Meiji 36 (1903）, immediately after the discovery.

　The excavation in Showa 11 (1936） produced a masonry ditch and a stone pavement near the findspot of the stonework, which were regarded as remains of the banquet hall of the Saimei era in the later period. In sum, Asuka Elementary School and the surroundings have long been considered to be important sites occupied from the Saimei to the Tenmu eras through examination of toponyms and partial excavations. Nevertheless, the entire site remained unclear for a long time due to the limited excavations.

　After WWII, the continuous excavations at the Asuka-dera Temple caused academic interest in the estimated site of Asuka Kiyomihara Palace to the north of the temple. Specifically, an identification of the northern boundary wall of Asuka-dera Temple (1977） revealed that the rice fields of Koaza Ishigami were included in the northwest corner of the temple precinct. Since it is assumed that the rice fields also contained the intersection of the Nakatsu-Michi road(中ツ道）, which likely passed in front of the west gate of Asukadera Temple, and the "Asuka-dera North Road"(「飛鳥寺北路」）mentioned in the article of the Jinshin Rebellion (壬申の乱, 672） in the Nihon-Shoki (『日本書紀』）, comprehensive research was required from a new perspective. In addition, a remain of the water clock during the Saimei era was found at the Mizuochi site(水落遺跡）, in the vicinity of the estimated site of the palace. Based on these results, Nabunken started excavations at the estimated site of the palace in Showa 56 (1981）. At that time, the site was first called "Ishigami Site".

　As mentioned above, the continuous excavations by Nabunken started at the rice fields of Koaza Ishigami (Asuka 287）. A series of excavations continued until the 21st excavation (2008） in one rice field borrowed per fiscal year. When the organization of excavation data for the final report started in 2009, the authors faced a significant contradiction that the east-west fences (SA600 and SA560） and the east-west ditch (SD347） confirmed in the third excavation have not been identified in the east-neighboring area of the first excavation. Among these, pillar holes of SA600, likely bending northward along the masonry ditches (SD334 and 335）, had to be confirmed in the area of the first excavation .

　In Reiwa 5 (2023）, the re-excavation at the area was carried out after 42 years of the first excavation, with the consent of the rice field's owner (Asuka Fujiwara 214th excavation）. As the excavation revealed that SA600 turned north before SD335, we partially revised the results of the first excavation. Primarily based on the original records of the first to fourth excavations, the authors revised the previous assessment of the Ishigami site through the reinterpretation of the records and the current re-excavation.

2.　Major Archaeological Features

New Periodization:　The previous excavations at the Ishigami site have adopted the five-phase periodization: A-1, A-2, A-3, B, and C. This periodization has been constructed through revisions in each excavation after the fourth excavation. According to the Fujiwara annual summary reports (『藤原概報』）, Phase A was mainly dated to the Saimei era (mid-seventh century）, Phase B was assigned to the Tenmu era (late seventh century）, and Phase C corresponded to the period of Fujiwara Palace(藤原宮）. However, the excavation reports show changes in the periodization before the seventh excavation (1987）, in which the general periodization was almost established. Since this volume reports the results of excavations mainly before 1984, the periodization cannot be used without any revisions . In addition, the re-excavation of the first excavation area (Asuka Fujiwara 214th excavation） caused modification of the chronological assessment of the main features . Given the situation, the authors decided to adopt a new periodization in this volume rather than the conventional one. This new periodization consists of five phases, I-1, I-2, II-1, II-2, and III, to each of which the archaeological features of the seventh century are allocated . The following describes these phases and their main features.
Phase I: This is the period comprising the features earlier than the first fill (Fill A）. Based on the overlapping features, the phase is subdivided into two: Phases I-1 and I-2. Although both subphases correspond to the conventional Phase A-1 and 2, the formers date to the period earlier than the Kogyoku (皇極朝） and Saimei (斉明朝） eras. The southern fence in Phase I, which was as large as the great fence(Ogaki 大垣）of Fujiwara Palace, turned north as SA4650 at the eastern end of SA600. The southeastern corner, located in the rice field at 287 Asuka (1st excavation area）, bordered a large area in the first half of the seventh century. SA4650 turned east at the northern end and was likely to connect to the east-west fence SA1460 (in Ishigami Site 1990-1）. A masonry ditch consisting of SD334, SD335, and SD435 along the large fence (SA600 and SA4650） turned east in the north toward the Cho-dori(丁通り）area. The main water channel, extending to the east of the Ishigami site, was likely arranged to border the exterior side of the large fence. Although the entire area inside the great fence remains unclear, some buildings, such as SB745, SB750, SB810 and SB811, SB450, SB530, and SB735, were confirmed. Based on the overlapping features and sediments of pillar holes, SB745 and SB810 were assigned to Phase I-1, and SB450, SB530, and SB735 were assigned to Phase I-2. In the present volume, the masonry ditch consisting of SD334, SD335, and SD435 and the well SE800, which were once regarded as features functioning until the conventional Phase A-3, dated back to Phase I based on the revised chronological assessment of the pottery assemblages from them. AsSA600 was assigned to Phase I as well, it dated to the earlier period than the water clock (Rokoku漏刻） at the Mizuochi site. Furthermore, the period of SB735, originally assigned to the conventional Phase B , has been revised to Phase I based on the elevation of the stone pavement.
Phase II: The features built between the first fill (Fill A） and the second fill (Fill B） were allocated to the phase.. Based on the overlapping features, the phase is subdivided into two: Phases II-1 and II-2. Corresponding to the conventional Phases A-3 and B (Tenmu era）, the present phase dates to the Saimei era, although some features of the conventional Phase A-3 were dated back to Phase I (see above）. After the large southern fence SA600 in Phase I was totally abandoned , SA560 was built at almost the same location. A masonry ditch from the south, consisting of SD330, SD331, and SD332, turned west before the north-south wall SA560 and turned north again to run into the interior space . Whereas it is assumed that this masonry ditch, contemporaneous to SD335, was used during the entire Phase A in the previous interpretation , the authors reinterpreted that the same masonry ditch functioned in Phase II-1 in the present volume because the frequent removal of the lateral stones allowed us to estimate the higher elevation of the construction than SD335 . As with SD332, SB620 was assigned to Phase II-1. After scrutinizing the chronological sequence between SD730 and SD332, the authors estimated that SD332 was built older than SD730 . In other words, it is assumed that the straight ditch of SD730 was built to replace the bent ditch consisting of SD330, 331, and 332 to penetrate between the pillars of the fence SA560.
　The buildings of Phase II include SB620 (II-1）, SB510, SB520, SB580, SB736, and SB770 (II-2）. In the conventional interpretation, SB400associated with the stone pavement in the surroundings , was dated to the Tenmu era.. In the present volume, SB400 was regarded as the building of Phase II-1, the Saimei era.
Phase III: This period dates to the second half to the end of the seventh century, which is later than the second fill (Fill B） . The phase roughly corresponds to the conventional Phase C (Fujiwara Palace period） . Given that the Asuka IV pottery assemblage, which has been found at the site in large quantities, may be dated as late as the Fujiwara Palace period, the present phase covers from the Asuka Kiyomihara Palace period to the Fujiwara Palace period. During the phase, the east-west wall SA311 and the east-west ditch SD347 running from the east turned north and continued northward as SA751 and SD640, respectively. Only SB742 was situated inside the boundary, which was drastically changed from before Phase II. In the northwestern part of the fourth excavation area, SB830, as a building with internal pillars, was built in the area bordered by SA780 and SA781. Although the pillar holes of the Phase III buildings were buried with the brown soil containing charcoal, which may have been derived from Fill B by digging from the upper surface, these remains were confirmed in the lower part of the brown soil, equivalent to Fill B, according to the logs of the third and fourth excavations.
Phase IV (after the Nara period): After the abandonment of the buildings and walls of Phase III, the natural flow NR310 ran down to the west in the eighth century, whichdestroyed the features of the seventh century. The south stone pavement SX327 attached to SB400 and the masonry ditch SD330 were eroded by the natural flow. Small post-inground structures SB440 and SB460 were built after NR310 was buried with sediment in the ninth century.

3.　Artifacts

Huge amount of excavated pottery:　Notwithstanding a paucity of mentions in the interim reports of the Ishigami site, the large amount of excavated artifacts characterized the site impressively. The pottery sherds excavated in the area, which was reported in the present volume, totaled 780 wooden boxes for the organization. This number occupied 16.3% of4,788 boxes which contain all the pottery sherds and clay objects excavated until the 21st excavation. Most of the pottery sherds found in Fill B and the features of Phase III were classified as Asuka IV, while a small amount of pottery sherds were found in the features of the Saimei era, the conventional Phase A-3. .
The same situation can be observed in the features of Phase II after the renewal of the periodization. Therefore, the large amount of the pottery sherds may lead us to conclude that the site was most flourishing during Phase III, the period from the Asuka Kiyomihara Palace to the Fujiwara Palace. In addition, the current analyses revealed that the Silla pottery, frequently related to the venue of the royal feasting in the Saimei era, and the smoked Haji ware imported from northeastern Japan (Tohoku region） were contained in the vast amount of the Asuka IV pottery sherds rather than the features of the Saimei era. Furthermore, the large amount of iron implements found in the focused area was likely contemporaneous with the Asuka IV pottery assemblage.
　From the well SE800 in Phase I, a large number of Haji jars and Sue jars and pots were found in the inside of the frame. Although it was assumed that SE800 was continuously used from Asuka I to IV based on the pottery chronology, the present volume does not adopt this interpretation. The typological features of the Haji tableware imply that the well was buried from Asuka I to II. As the stable carbon isotope ratio analysis of the adhering carbides showed (see Section VI-1）, many sooted and charred Haji jars were used for boiling C3 plants, Given that these jars were dumped into the well after the abandonment,, SE800 likely functioned before the Saimei era, when the well was filled by waste such as the pottery assemblage . The previous papers have described SE800 as a representative feature of the conventional Phase A-3, which was connected with the masonry ditch SD900 extending north from the fifth excavation area. The current revision of the chronological assessment and the burying sequence will affect the interpretation of the chronological transition of the features in the forthcoming volumes .
The dating of rooftiles:　While a huge amount of pottery has been excavated, the amount of roof tiles is small: 155 round eave-end tiles, 9 flat eave-end tiles, 1 rafter tip tile, 381 round tiles (49.63 kg）, and 1,715 flat tiles (116.19 kg）. This quantity is outstanding in the Ishigami site as roof tiles, but less than other artifacts. As discussed in Chapter II, Ishigami type A, B, D, and E were newer than Asuka-dera type VII and were most likely made at the same time as tiles from Ikaruga area using a common technique. The upper limit of the production date of these types is from the end of the 6th to the beginning of the 7th century, the period during which Asuka-dera temple was built, and the lower limit is around 639, when the construction of Kibiike-haiji temple is considered to have begun. Production period of these types ended during Phase I of Ishigami site.
Ishigami type F is thought to have been produced at the same time as Ishigami A, B, D and E. On the other hand, most of the tiles of Asuka-dera temple were excavated from flow channel NR310 and the later cultural layer, suggesting that these tiles had once been roofed on the Ogaki of Asuka-dera temple but were fragmented into smaller pieces and flowed into the Ishigami site.
Wooden Products from SE800:　Most of the wooden products described in the volume were found in the well frame of SE800. Stratum1 (gravel sediment） contained wedges, skewers, and processing rods . Stratum2 (sandy gravel sediment） yielded wedges, stoppers, cutting objects, and processing rods. Strata 3 (brown organic soil sediment） and 4 (clay sediment） as the lowest layers included sword handles, wedges, and stoppers.
Metal products, metallurgical artifacts:　Iron products consist of mainly arrowheads, agricultural tools such as sickles, axes, chisels, halberds, knives, and chisel-shaped implements, and others such as spindle wheels, needles, nails, and sickles. Most iron products found in "pits" and "large pits" are likely derived from Fill B, although the northsouth ditch SD640 contained these artifacts. The variety and large number of iron weapons characterize the assemblage of iron products from the Ishigami site.
　Artifacts related to metallurgy include tuyeres, molds, crucibles, iron slags, and ingots of copper and iron. Given that tuyeres and crucibles were also found from the dark brown soil overlaying the stone pavement of the neighboring Mizuochi site, both sites were likely leveled by the same fills, which were Fill B prior to the renewal in Phase III. A large amount of pottery sherds, iron products primarily comprising arrowheads, and metallurgical artifacts date approximately to the same period. The assemblage may include whetstones mainly from the area covered by Fill B. Focusing on the relation between iron weapons such as arrowheads and whetstones, the authors pointed out the possibility that an armory was situated in the vicinity of the Ishigami site (see Section VII-3）. Although this hypothesis almost agrees with the "Oharida-Hyogo (armory situated in Oharida area）" hypothesis mainly by Aihara Yoshiyuki (相原嘉之）, SB735 and SB736, regarded as remains of the armory, were reassigned to Phases I and II, respectively, in the present volume. Consequently, the presence of the Oharida-Hyogo at the Ishigami site is not necessarily asserted. Even so, it should be kept in mind that the evidence shown in the present volume does not deny the possibility that the armory was located near the Ishigami site. The provenance of the iron products found in the fill should be identified in conjunction with an enormous amount of pottery sherds and artifacts related to metallurgy.

4.　Conclusion

Background to the discovery of stonework:　Given that this volume reported the results of the excavations, including the findspot of the stonework, the authors summarized the background that pieces of the stonework were unearthed and donated to the Tokyo Imperial Museum in 1904 (Meiji 37）. The related documents described that, Mori Kunimatsu (森 国松）, the mayor of Asuka Village, bought the stonework from Tsujimoto Ukichi(辻本宇吉）, the owner of the rice field, and unearthed them from the field. Another report by Takahashi Kenji (高橋健自, in 1903） and a copy of the notification documents, probably left by Mori, revealed that the Sumeru stones were dug up to obtain the construction materials for the expansion of Asuka Elementary School. The documents also mentioned that Tsujimoto Sadashiro (辻本定四郎）, the son of Tsujimoto Ukichi, proposed Ishida to excavate the findspot of the stonework. In short, the stonework was preserved and the research of the Ishigami site was promoted through a cooperative relationship between local residents and researchers from the museum. The Sumeru stones and the stone statue were designated as Important Cultural Properties of Japan in 1980 (Showa 55）. These pieces of stonework are now exhibited inside the Asuka Historical Museum 50 years after its opening.
Evaluation of the hypothesis of Asuka Kiyomihara Palace:　Excavations at the Ishigami site by Nabunken started as the research of "the estimated site of the Asuka Kiyomihara palace." After being named "Ishigami " in the second excavation, the site has never been called the name. In both conventional and present assessments, the sparse distribution of the features during the Asuka Kiyomihara Palace period never allows us to advocate the presence of the palace at the site. Given that the features of Phase IIIb in the Asuka Palace site can be regarded as the current persuasive candidate for the palace, the hypothesis of the Asuka Kiyomihara Palace by Kida should be rejected. Therefore, there is no evidence that the palace was located in the Ishigami site.
Evaluation of the Ishida's hypothesis:　The revision of the periodization and the pottery chronology resulted in reclassifying some features to the period older than one in the conventional periodization.The east-west fence SA600, considered to be used through the conventional Phase A, was newly regarded as the southern great fence of Phase I-1, which was not the boundary of the Saimei era. Based on the pottery chronology and the simultaneity with SA600, the masonry ditch comprising SD334, 335, and 435 was reevaluated as the water channel of Phase I-1, by which Ishida and his colleagues imagined the water supply to the "fountain tower". The hypothesis of the "banquet hall of the Saimei era" by Ishida and others cannot be validated. Furthermore, the chronological assessment of pottery enabled us to arrange SE800 to Phase I-1, which was buried before the Saimei era.
　Since the features mentioned above were newly dated before the Kogyoku era, these are never regarded as constituents of the "banquet facility" in the conventional Phase A-3a. The boundary fence SA560 (Phase II-1）, simultaneous with the water clock at the Mizuochi site, was once arranged to the conventional Phase B. Consequently, the rearrangement of the features implies the presence of an unidentified important facility in Phase I, surrounded by the boundary fences as large as the great fence of Fujiwara Palace.
Toward the forthcoming volumes:　The forthcoming volumes will adopt the new periodization of the Ishigami site proposed in the present volume. The new periodization leads to some changes in the chronological transition of the features expected in the conventional periodization. Eventually, these changes will affect the evaluation of the entire Ishigami site. Because of the report on the findspot of the stonework, the present volume included a summary of the background that the Ishigami site came to be regarded as "the banquet facility of the Saimei era." Since the background is deeply related to the evaluation of the main buildings confirmed in the areas of the fifth to eighth excavations, which will be dealt with in the second volume, the authors recognized the necessity to clarify the topic in advance.
　The interpretation of the Ishigami site has been shaped by a mixture of the hypothesis of Sumeru first advocated by Takahashi and the hypothesis of the "banquet hall of the Saimei era" primarily proposed by Ishida, and the assessment of the entire site by Nabunken based on the mixed hypothesis. This interpretation has been supported as an established theory during the past 120 years. The new interpretation leading to a better understanding of the entire Ishigami site remains unfinished in the present volume. As the second volume will focus on the areas of the fifth to ninth excavations, where the central facility of the conventional Phase A-3, identified as the "banquet facility" and "guest house," was located, we hope to finalize the interpretation of the site in the next volume titled "ISHIGAMI SITE II."