Monograph No. 26
NARA (HEIJO) IMPERIAL PALACE SITE EXCAVATION REPORT VII
SURVEYS IN THE “DAIRI NORTH EXTERIOR PRECINCT” CARRIED OUT DURING 1962-1975
ENGLISH SUMMARY
NARA (HEIJO) IMPERIAL PALACE SITE EXCAVATION REPORT VII
SURVEYS IN THE “DAIRI NORTH EXTERIOR PRECINCT “‐ 1976 ‐
CONTENTS
Chapter I Introduction 1
A. Development of the Excavation surveys 2
B. Preservation Work and Physical Layout of the Site 4
C. Committees Handling Research and Facilities 5
D. Compilation of the Present Report 7
Chapter II General Outline of Research Work 8
1. Areas Surveyed 8
2. General Development of Research Work 9
A. Survey No.10 9
B. Survey No.11 10
C. Survey No.13 11
D. Survey No.20 12
E. Surveys No.82-7 and 95-6 13
3. Calendar of Excavation Work and Major Findings 14
A. Survey No.10 14
B. Survey No.11 15
C. Survey No.13 16
D. Survey No.20 18
E. Surveys No.82-7 20
F. Surveys No.95-6 20
Chapter III Sites and Remains 18
1. General Remarks 21
A. Types of Remains 21
B. Numerical Evaluations of Building Size 22
C. Stratification 23
2. Specific Remains Classified by Geographic Location 24
A. Remains Dividing Western, Central and Eastern Sectors 24
B. Remains in Western Sector 29
C. Remains in Western Half of Central Sector 32
D. Remains in Eastern Half of Central Sector 38
E. Remains in Eastern Sector 46
3. Relative Chronology of Remains 50
4. Ichiniwa Kofun (Burial Mound) 53
Chapter IV Artifacts 55
1. Wooden Tablets (Mokkan) 55
A. Mokkan from SK820 55
B. Mokkan from SK870 58
C. Mokkan from SK2101 59
D. Mokkan from SK2102 59
E. Mokkan from SK2107 59
2. Tiles and Bricks 60
A. Roof-edge Tiles 60
B. Demon Tiles, Round and Curved Tiles, Bricks 71
C. Green-Glazed Tiles, Bricks and Tiles with Chinese Characters 74
3. Pottery 75
A. Pottery from SK820 77
B. Pottery from SK2101 87
C. Pottery from Cultural Layer in Area Q 90
D. Pottery from SK2113 90
E. Pottery from SK.370 94
F. Pottery from SK2102 96
G. Pottery from SE715 96
H. Pottery from Other Remains and Cultural Layers 98
I. Lead-Glazed Ceramics 99
J. Inkstones 101
K. Pottery with Ink Writings or Drawings 103
L. Pottery with Incised Writings or Drawings 108
M. Pottery Decorated with Thin Engraved Lines after Fired 108
N. Earthenware Lamps 109
O. Haniwa 110
4. Wooden Objects 111
A. Manufacturing; Processes 111
B. Objects from SK820 112
C. Objects from SK2102 and Elsewhere 128
5. Fibers and Charcoal Remains 132
Chapter V Problems and Analysis 135
1. Roof-edge Tiles 135
A. Pairings of Roof-edge Tiles 135
B. Periods of Manufacture 136
C. Uses of Large and Small Type Round Roof-edge Tiles 138
2. Pottery 139
A. General Classification into Heijo Imperial Palace Stage I-VII 139
B. Pottery Clusters 146
C. Differentiation and Standardization of Bowl and Dishes 147
D. Conclusion 149
3. Sites 150
A. Location Changes of Government Offices within the Dairi North Exterior Precinct 150
B. Absolute Chronology for Government Offices in the Dairi North Exterior Precinct 156
C. Government Offices on the Periphery of the Heian Palace Dairi 159
4. Conclusions 164
A. Nature of Government Offices in Dairi North Exterior Precinct 164
B. Terminal Occupation of Government Offices in Dairi North Exterior Precinct 167
Summary 168
Supplementary Tables 172
English Summary 193
Plans and Sections
Plates
Index
NARA (HEIJO) IMPERIAL PALACE SITE EXCAVATION REPORT VII
Nara, located 40 kilometers south of Kyoto, was Japan’s capital from 710 to 780 A.D. and was generally called Heijo-kyo （平城京）. The exact pronunciation given at that time to the characters designating the capital is, as in the case of all words in this period, un known.
Somewhat earlier, the ruling class, centering around a male or female sovereign, had made efforts to establish a legally regulated, centralized government and had, during the second half of the 7th century, constructed capital cities at the site of present-day Osaka (viz. Naniwa-kyo) and at a site 20 kilometers south of Nara (viz. Fujiwara-kyo). However, in order to establish a still stronger political system, the ruling class decided to build a new capital city at the northern edge of the Nara Plain. This new capital was Heijo-kyo.
Heijo-kyo measured 4 kilometers from east to west and 4.5 kilometers from north to south. It was a city according to the Chinese pattern, its east-west “avenues” and north-south “streets” forming a grid. The part of the city to the west of the 84-meter-wide central street, or Suzaku-oji （朱雀大路）, was called the “right city” (ukyo, 右京) while the part of the city lying to the east was called the “left city” (sakyo, 左京). Attached to part of the eastern side of the “right city” was an area known as the “outer” or “peripheral city” (gekyo, 外京). The presentday city of Nara has developed in such a way that its central part lies within the site of the ancient “peripheral city.”
At the southern end of the Suzaku-oji was the principal gate to the city, the Rajomon （羅城門）, and at the northern end was the Heijo Imperial Palace (Heijo-kyu, 平城宮 ).The palace precincts measured 1000 meters from east to west and 1000 meters from north to south, plus an adjoining rectangular area to the east which measured 250 meters from east to west and 750 meters from north to south. The total area of the palace grounds thus measured approximately 1.19 square kilometers. The palace grounds consisted of the Imperial Domicile (dairi, 内裏 ); the Halls of State (chodoin, 朝堂院); and the various government offices and bureaus (kanga, 官衙). It formed the political nucleus of the Japanese nation-state and provided direct employment for about ten thousand officials and public employees.
The Nara National Cultural Properties Research Institute has since 1959 carried out a continuing series of excavation studies at the site of the Heijo Imperial Palace, which is designated a "special historical site." Results of these studies have been pulished in Heijo-kyu hakkutsu chosa hokoku, Nos. I-VI. The present report brings together the results of Surveys 10, 11, 13, and 20 (see Table 1 and Figs. 3-6), carried out between 1962 and 1964, as well as the results of two small-scale surveys (Figs. 7-8) carried out in conjunction with later alterations at the sites in question. (See Chapter I).
The combined area covered in Surveys 10, 11, 13, and 20 is a rectangle measuring 300 meters from east to west and 90 meters from north to south. It lies near the northern edge of the eastern half of the palace grounds, within that part of the grounds having the highest elevation (73 to 74 meters). The area of the present study is bordered on the south by the walled corridor forming the north side of the so-called “second dairi” and is thought to be a part of the palace closely connected in some way with this “second dairi.” Thus, the area of the present study is called the “dairi north exterior precinct,” in contradistinction to “dairi interior precinct” surrounded by walled corridors. (See Ch. I; Ch. II, Sect. 1.)
The full report in Japanese is prefaced by a brief outline of the progress of the work carried out each individual survey (Ch. II/2-3).
Before discussing individual remains, the report states the various types and total number of the remains, and a few technical terms are discussed. The principal types of remains are: buildings of hottate-bashira type, i.e., with posts deeply set in holes in the pround (52); buildings with foundation stones (6); buildings originally of hottate-bashira construction but later remodeled with the use of foundation stones (1); outside partitions (20); corridors (1); ditches (25); wells (3); and disposal pits (22). (See Ch. III/1A). Brief outlines are given of the scale of the buildings and the methods of ascertaining the distances between posts (Ch. I11/1B), and of stratification characteristics (Ch. I11/1C).
The area of the present study can, by means of the north-south earth walls SA 505 and SA 838, be (Fig. 20). The Central Sector can, by means of the north-south partition SA 630, be further divided into an eastern (Fig. 17), and a western half (Fig. 15). The policy is adopted of describing the remains in each sector and sub-sector in such away as to combine the results of the various surveys, rather than to describe the results of each survey separately. First of all, there is a discussion of SA 505 and the other walls and partitions subdividing the surveyed area, together with a discussion of the earth-walled corridor SC 060 at the southern edge of the surveyed area (forming the northern edge of the dairi) and of the east-west partition SA 486 (Ch. III / 2A). This is followed by a discussion of remains in the Western Sector (Ch. III / 2B), the western half of the Central Sector (Ch. III / 2C), the eastern half of the Central Sector (Ch. III / 2D), and the Eastern Sector (Ch. III / 2E), followed in turn by the attempt to set forth a relative chronology for the various remains, as elucidated for example, through the relationships among overlapping elements. (Ch. III/2F.)
Before the constructs of the Heijo Imperial Palace, there existed on the site the present survey area the anterior part of a 250-meter-long keyhole-shaped tumulus with surrounding moat (viz. the “Ichiniwa Kofun 市庭古墳,” dating from the 5th century). The posterior part of this tumulus remains today as the so-called “Heizei Tenno-ryo 平城天皇陵.” (See PL. 33). At the time of the construction of the palace, the anterior part of the tumulus was leveled the corresponding portion of the surrounding moat was filled in and various palace structures were built in their place (See PL. 33; Figs. 9-10). During the present surveys, the area in question was partially excavated, and studies were made of the base of the tumulus along its eastern edge and southeastern corner, of a portion of what appears to have been an outer embankment, and of sites once covered with paving stones. (See Ch. III / 2G; and PL. 34-35.)
Unearthed remains are discussed in the following order; wooden tablets (mokkan 木簡); tiles and bricks; earthenware; wooden manufactures; fibers; and charcoal remains (Ch. IV). Among the wooden tablets, more than 1800 were unearthed from pit SK 820 alone in the Eastern Sector. Although more than 80% of these are fragments or shavings, they provide a wealth of notations in the form of food orders for the meals of guards and militia, shipping labels attached to produce offerings and taxes in kind, aids for learning Chinese classical writings, and various notations concerned with construction work, etc. (PL. 36-40). Since these mokkan have been described separately, they are discussed in the present report only summarily. It was pointed out that certain of the wooden tablets (viz. Nos. 696-764 and 689-1170) have, with the additional light shed by subsequent surveys, been determined to be commentaries by Li Shan (李善) on the collection of Chinese classics known as the Wen xuan (文選;Sino-Jap. pron.; Monzen). (See Ch. IV, Sect. 1.)
There were unearthed 3399 pieces of roof-edge tile. These consist of 31 types and 70 sub-types of round roofedge tiles and 25 types and 50 sub-types of curved root-edge tiles (See PL. 41-43; Supplementary Tables 2-4.). The most typical finds indicate pairings of round Type 6311A/B with curved Type 6664D/F and similar pairings of round Types 6313 and 6314 with curved Types 6666A and 6685A-D. These types account for 50% of the total number of round and curved roof-edge tiles unearthed. It is worthy of note that the above-mentioned patterns of tile pairing are also typical for the so-called “second dairi” (Ch. IV/2A). In addition, there are examples of three-color “demon-faced demon tiles” (kimen onigawara 鬼面鬼瓦), “demon-body demon tiles” (kishin onigawara 鬼身鬼瓦), and “phoenix pattern demon tiles” (hoomon onigawara 鳳凰文鬼瓦) (PL. 44), as well as a few examples of green glazed tiles and bricks bearing Chinese characters. (See Ch. IV/2C; Fig. 34.)
The earthenware is for the most part red pottery (hajiki 土師器) and grey pottery (sueki 須恵器). At the present time, seven Heijo Palace pottery “stages” can be distinguished on the basis of manufacturing technique, form, size, etc. And, with the help of accompanying wooden tablets and comparisons with earthenware from the Fujiwara (694-710 A.D.) and Nagaoka (784-794 A.D.) Palaces, it is possible to estimate an absolute date at least one point in each stage. (Tab. 23). Among .the earthenware discussed in this report, those unearthed from pit SK 320 in the Eastern Sector (PL. 45-50) are representative of Stage III; those from pits SK 2113 and SK 870 in the eastern half of the Central Sector (PL. 51-53) belong to Stage V; and those from well SE 715 in the Eastern Sector belong to Stage VII (See Ch. IV/3A.).
Glazed pottery wares are few in number, but some examples of three-color ware (PL. 56/1) and green glazed ware (PL. 56/4-7) do exist (See Ch. IV/3B.). There are various types of inkstones: some are round and flat-surfaced (resting on a cricular stand a ring of many flat supports or projecting supports of triangular section); some bird-shaped, and some are of the foliate type (PL. 55-56), (See Ch. IV/3C.). There is a rich variety of earthenware bearing inked Chinese characters. While the meaning of some of the notations is obscure, a few examples are given as follows: 内裏盛所 “dairi pantry (?)” (PL. 60/63); 鸚鵡鳥坏 “dish for feeding parrots” (PL. 58/4); 鳥食入器二口 (PL. 58/3) dish for feeding birds; （鳥坏口） (PL. 58/3). A dish for feeding birds; and. （醴太郎 炊女取不得 若取者苔五十）, whose humorous intent might perhaps be translatable as “Reitaro, Our Most Prodigious Master Sake ‐ Not to be Taken by Kitchen Maid ‐ 50 Cane Swats if She Dare” (PL. 61/71), (See Ch. IV/3K.).
Among wooden manufactures, there are a large variety of cypress fans, combs, and spools (PL. 63-75). After classifying and arranging these various manufactures according to type of processing (Ch. IV/4A), those items unearthed from pit SK 820 (Ch. IV/4B) and those unearthed from pit SK 2101 and elsewhere (Ch. IV/4C) are discussed separately. Various fiber manufactures such as rope (PL. 76) are mentioned (Ch. IV/4D), and the results of research by Yoshibumi Hirata on charcoal remains are introduced (Ch. IV/5E.).
After recording the above factual information, the report attempts some interpretations (Ch. V). It is pointed out that, judging from the fact that certain wooden tablets unearthed from pit SK 2102 in the eastern half of the Central Sector were in close association with round roof-edge tiles of Type 6311A/B and curved roof-edge tiles of Type 6664D/F (both representative of those roof-edge tiles unearthed in the area of the present study), the possibility becomes rather great that these tiles are older than previously thought, and may date back as far as the year 721. (See Ch. V, Sect. 1.)
The foundations are then given for suggesting a hypothetical absolute chronology for Stages I-VII of Heijo Imperial Palace earthenware, and outlines of the changes between stages are discussed. (See Ch. V/Sect. 2.)
Periods -of construction of official buildings in the “dairi north exterior precinct” are divided into three. The second of these periods may be further divided into three sub-periods. During Period I there were only a small number of buildings (Fig. 53). During Period II earth walls were erected to surround the Central Sector, The western half of this sector became occupied by a central building with aisles and by warehouse like buildings, while the eastern half came to play a subordinate role as the site of structures for various types of manual work, of a garbage dump, etc. (Figs. 55, 56, and 57 refer to subperiods III, II-2, and II-3, respectively.) In the course of Period II, part of the buildings were taken down and rebuilt, and some buildings were partially remodeled, but there is no evidence of any large-scale rebuilding in Period II that would cause us to think that there was within the period any physical transfer of government offices or major changes in their composition. In the first sub-period, areas marked for construction were divided into square sections 10 shaku (R) on aside, and the buildings were then erected in a very orderly way in accordance with such a scheme (Fig. 54).
In the Western Sector, buildings remained few, and it would appear that this sector did not form any independently constituted office compound. On the other hand, in the Eastern Sector, a large number of buildings stood close together in a small area, and there is evidence that buildings were on repeated occasions rebuilt or remodeled. One may hypothesize the existence there of an office compound distinct from that existing in the Central Sector.
The situation changes on entering Period III. The earth walls which flanked the Central Sector are removed, and there exist within the area of the “dairi north exterior precinct” three groups of buildings, consisting of two or three buildings each. The principal structure in each group is a large building with aisles. The three groups are located, respectively, in the western, central, and eastern parts of the precinct, and each group was provided with a well and a liberal amount of open space (Ch. V/3A). The so-called “second dairi” was formerly thought to date from after the end of the Tempyo 天平 Period (794 A.D.); however, recently it has come to be thought that its construction began around 721 A.D. preparation for the ascension to the throne of Emperor Shomu, 聖武天皇. Accordingly, one may hypothesize that Period I for the area of the present survey begins around 708; that Period III begins around 721; that Period II-2 begins around 745; and that Period II-3 begins around 761. Period III may well correspond to the years of retired Emperor Heizei, viz. 809-824. (See Ch. V/3B.)
At the Heian Imperial Palace in Kyoto, there existed within the area corresponding to the “dairi north exterior precinct” of the Heijo Imperial Palace three adjacent sectors (bo, 坊) known as the Ranrin-bo （蘭林坊）, Keiho-bo （桂芳坊）, and Kaho-bo （華芳坊）. From a study of 10th century and later written materials which mention these three bo, it is evident that they served as reserve or supplementary facilities of the dairi itself. It does not appear, however, that the nature of the “dairi north exterior precinct” of the Heijo Imperial Palace, as described in the present report, bears any direct relation to the above mentioned bo of the later palace in Kyoto.
Possible candidates for government offices which may have been located in the area of the present study are: a station of the Left Palace Guards (sa-hyoefu, 左兵衛府); those bureaus of the Department of the Palace Interior (kunaisho, 宮内省) that were in charge of food preparation (naizenshi, 内膳司), landscaping (enchishi, 園池司) or female servants (uneme no tsukasa, 采女司); those bureaus within the Department of Central Administration (nakatsukasasho, 中務省) that were concerned with the warehousing of precious articles (kuraryo, 内蔵寮) or with superintending the production of imperial wardrobes (nuidonoryo , 縫殿寮).
The Eastern Sector being quite small, it is difficult to suppose the existence there of government offices on a very large scale. In view of the excavation from pit SK 820 (in the Eastern Sector) of wooden tablets which relate to guard or militia affairs, we may hypothesize that in this sector there was posted a contingent of Left Palace Guards. Most probably, some of the wooden tablets from pit SK 820 relate to government offices which stood in the Central Sector.
The fact that these include numerous (viz. 46) labels which had been attached to Produce (nie , 贄) brought as offerings to the court, and also the fact that apiece of earthenware from pit SK 870 in the eastern half of the Central Sector bears the inked inscripition （内裏盛所） suggest the possibility that the naizenshi, in charge of imperial repasts, may have been located nearby. Some of the discarded tablets refer explicitly to the naizenshi （内膳司）, and on others there appear the names of offices such as the enchishi and uneme no tsukasa, the latter being functionally in close association with it. Thus, we may suggest the naizenshi as the most probable establishment to have occupied the Central Sector, further hypothesizing that it existed there together with one or more other bureaus which had related functions (Ch. V/4A).
In Period III it is possible that the three groups of buildings within the area of this study were not government offices but rather residences for in-laws and/or retainers of the retired Emperiod Heizei.
The area of the so-called “second dairi” had during Period III and afterward as many as 40 buildings, and as is evident from overlapping elements among their remains, these buildings were moved or rebuilt several times during the Heian Period. In contrast to the fact that this area has produced abundant remains of relatively new earthenware, there has not been found in the area of the present study any earthenware later than “Stage VII, and neither are there to be seen remains of any buildings postdating Period III. This suggests that the area in question was abandoned not long after the death of the retired Emperor Heizei, reverting to farmland (Ch. V/4B).
(translated by William R. Carter)
Chronological Table
EraName Year Emperor or Empress’s Name
Wado 和銅 1 708 Gemmei 元明
Reiki 霊亀 1 715 Gensei 元正
Yoro 養老 1 717 〃
Sinki 神亀 1 724 Shomu 聖武
Tempyo 天平 1 729 〃
Tempyo-Kampo 天平感宝 1 749 Koken 孝謙
Tempyo-Shoho 天平勝宝 1 749 〃
Tempyo-Hoji 天平宝字 1 757 〃
〃2 758 Junnin 淳仁
〃8 764 Shotoku 称徳
Tempyo-Jingo 天平神護 1 765 〃
Jingo-Keiun 神護景雲 1 767 〃
Hoki 宝亀 1 770 Konin 光仁
Ten‘o 天応 1 781 Kammu 桓武
Enryaku 延暦 1 782 〃
昭和五十一年三月三十一日 発行
平城宮発掘調査報告 Ⅶ
奈良国立文化財研究所学報 第二十六冊
