I INTRODUCTION

Goguryeo pottery is defined here as the pottery produced and used by the Goguryeo people who lived in the Goguryeo territory. According to the Samguksagi (三國史記: History of the Three Kingdoms), Goguryeo was established in Huanren (桓仁) around the Hunjiang basin in 37 BC and moved its capital twice, once to Ji'an and once to Pyeongyang (平壤), lasting for about 700 years until its collapse in 668 AD. As the early Goguryeo state matured, it expanded its territory both northward and southward, acquiring the Liaodong (遼東) area in the late fourth century and the Hangang (漢江) basin in the late fifth century. At the peak of its southernmost expansion, Goguryeo reached the areas of Gyeongsangbuk-do (慶尙北道) and Chungcheongnam-do (忠淸南道). Therefore, Goguryeo pottery encompasses the pottery that was produced and used by the people living in present-day northern and northeastern China, centering around Liaoning (遼寧) and Jilin (吉林) Provinces, and also in the south, the northern area of the Korean peninsula, as well as the Hangang area and regions of Gyeongsangbuk-do and Chungcheongnam-do, thus spanning some 700 years from the first century BC to 668.

Archaeological studies on Goguryeo began following the discovery of the stele of Gwanggaeto daewang (廣開土大王, r. 391-413: Gwanggaeto The Great) in the late nineteenth century, since which time they have flourished. Nevertheless, the study of Goguryeo has made slow progress compared with the archaeological findings on Baekje, Silla (新羅, 57 BC-654), and Gaya (42-532 [金官加耶]), due to the limited accessibility to the archaeological sites centered in northeastern China and North Korea, the geographical heartland of the Goguryeo state. For the same reason, the archaeological study of Goguryeo in South Korea relies heavily on secondary data such as research and reports by North Korean, Chinese, and Japanese scholars. However, as travel to China has become more convenient and communication and exchanges (albeit intermittent) with North Korea increasingly possible, so has direct access to Goguryeo's archaeological sites. Moreover, South Korea has recently begun to accumulate a significant amount of archaeological data of its own on Goguryeo from excavations of such areas as Achasan (峨嵯山: mountain) on the north bank of the Hangang, and the Imjingang (臨津江: river) and Geumgang (錦江: river) basins, including Daejeon (大田) and Cheongwon (靑苑).

Meanwhile, while it is true that archaeological studies on Goguryeo in the past centered on ancient tombs and focused on establishing a chronological system through analysis of the structures and types of the ancient tombs with few studies on actual artifacts including pottery, since the 1980s Chinese scholars have been trying to date Goguryeo pottery based on excavations in China, including Geng Tiehua (耿鐵華) and Lin Zhide (林至德),1 and there have been more studies on Goguryeo pottery in South Korea. It is nevertheless challenging to establish a definitive chronology of Goguryeo pottery as the chronology within the broader field of Goguryeo archaeological studies has yet to be settled. While there have been many advances in research in recent years, the criteria hitherto used to establish a definitive chronology have often been vague. At times, the dates of particular wares were estimated according to the age of certain ancient tomb types, the ages of which were in turn determined based on the dates of particular wares, creating a cycle of circular reasoning and contradiction. It is therefore imperative to establish a chronology of Goguryeo pottery through actual analysis of the pottery itself. Based on this, it would also seem possible to reconsider and refine the overall chronological system within the field of archaeological studies of Goguryeo.

II A BRIEF HISTORY OF GOGURYEO POTTERY STUDIES

Studies on Goguryeo pottery before the 1980s were largely confined to descriptions of pottery based on case-by-case excavation reports or general introductions to Goguryeo pottery by way of illustrations or introductory texts.2 In other words, these papers could not be considered as fully fledged academic papers but as descriptions of Goguryeo pottery in rather general terms.

In 1973 Jeong Chanyeong (鄭燦永) tried to form a comprehensive analysis of wares excavated from ancient tombs in his chronological study on Goguryeo tombs up to the fourth century.3 According to Jeong, between the second century BC and around the beginning of the first century BC gangdol dolgakdam mudeom (stone mound tombs in the shape of a rough circle or square) were prevalent. From such tombs various pottery artifacts such as brown polished bowls, black polished bowls made of sandy clay, and other bowls with added talc have been excavated. Moreover, from the dolgidan tombs (stone mound tombs with a solid square base), constructed from about the first to the fourth century, brown polished bowls made on a potter's wheel, and black, brown, and gray vessels were excavated, some of which were hand modeled. This paper conducts a chronological analysis of Goguryeo tombs according to their structural features and, in turn, dates the pottery pieces excavated from those ancient tombs in line with the approach recently taken by Chinese and Japanese scholars who have studied Goguryeo pottery.

In the 1980s China initiated comprehensive research projects on Goguryeo pottery, drawing on the data accumulated up to that time. In 1984 Geng and Lin conducted analysis on Goguryeo pottery excavated from Ji'an, China, and presented a theory of the process of change.4 The two researchers thought that the previous lack of studies of Goguryeo pottery was due to the absence of data and the limited dispersal of the wares. In the 1980s the two researchers presented a typological and chronological analysis of 96 pottery vessels excavated from the vicinity of Ji'an. They classified the excavated wares according to their form and material, and came up with three stages (groups) of pottery shapes in line with the chronology of ancient tombs. They found that the ho (壺: jar) and the guan (罐: jar with wide mouth and no perceivable neck) in particular had undergone much alteration, and divided these two vessel shapes into types I-V, further dividing each of these into three sub-groups.

In 1985, Wei Cuncheng (魏存成) devised a typology of the saijanggyeong-ong (四耳長頸甕: tall jar with a long neck and four handles), a typical Goguryeo vessel, as well as conducting a chronological analysis.5 Wei classified saijanggyeong-ong into two types: one with a neck that flares out straight toward the mouth from the body (Type I), and the other that stands upright before flaring out toward the mouth (Type II). He noticed that saijanggyeong-ong had in most cases been excavated from bongtoseoksilmyo (封土石室墓: earthen mound tomb with stone chamber) and attempted a chronological analysis of the wares based on the age of the stone chambered tombs in which they had been found. Wei concluded that saijanggyeong-ong had been prevalent in the mid-Goguryeo period around the fifth century; among them Type I had changed from having a longer neck to having a shorter one, while Type II had changed from having a shorter neck to having a longer one. Meanwhile, the maximum diameter of the body gradually rose to reach the shoulder level of the vessel, an area subject to considerable change in shape in both types.

In 1985, a comprehensive study of Goguryeo pottery took place in Japan. At the same time Okada Izumi (緒方泉) attempted a typology of saijanggyeong-ong and tried to devise a chronological study of Goguryeo tombs in the Ji'an area based on comparative analyses with other Goguryeo artifacts.6

Also in 1985, a study on Goguryeo pottery was presented in South Korea by Kim Giwung (金基雄).7 This paper, as the author himself indicated at the time, was not a comprehensive study of the chronology or change in the form of Goguryeo pottery but a very general presentation of Goguryeo pottery with basic facts, illustrations, and annotations. Nonetheless, Kim's paper was significant in that it was the first study on Goguryeo pottery to be carried out in South Korea.

In 1988, while undertaking a study on a chronology of Goguryeo artifacts, Azuma Ushio (東潮) introduced a typology and a chronology of Goguryeo pottery with a focus on saijanggyeong-ong. Azuma accepted the findings by Wei and Okada in general, classifying saijanggyeong-ong into four types depending on the shape of the vessel's mouth and body and providing descriptions of the special features and a chronology for each of the four types in addition.8 Azuma believed that the earliest saijanggyeong-ong was found in Tomb No. 332, Shanchengxia (山城下). He argued that they were made in the second half of the fourth century, and that the shape of the wares gradually changed over time tending towards ones with elongated neck and body with a wider mouth.

In the 1990s, several studies on certain specific features of Goguryeo pottery were published. Among them are two papers by a North Korean researcher, Li Gwanghi: one on Goguryeo glazed pottery (施釉土器, siyutogi) and the other on the decorative patterns of Goguryeo pottery.9 In the former paper, Li chronologically placed the tombs according to their structure having presented his findings on the glazed potteries excavated from Tomb No. 78 of Wanbaoting (萬寳汀), Jilin; Tomb No. 96 of Qixingshan (七星山), Jilin; Tomb No. 1 of Munak-ri (文岳里), Pyeongannam-do (平安南道); Tomb No. 7 of Pungcheong-ri (豊淸里), Pyeonganbuk-do (平安北道); Tomb No. 2 of Taeseong-ri (台城里), Pyeongannam-do; Tomb No. 3 of Anak (安岳), Hwanghae-do (黃海道); and Tomb No. 1 of Maxian'gou (麻線溝), Jilin. Li's main claim was that Goguryeo glazed pottery was extant before the third century. In the latter paper, Li classified 50 or so Goguryeo pottery vessels with engravings in accordance with their types and engraving techniques and asserted that most of the patterns on the wares were in use between the mid-fourth century and the late sixth century. It would seem that both papers take a rather simplistic approach in their respective analyses of the wares and the chronological classification of the tombs, although it is evident that the main focus of study was on the production techniques of the Goguryeo wares.

In 1993, in a paper on the international aspects of Goguryeo pottery, Gang Gyeongsuk (姜敬淑) of South Korea outlined the history of research on Goguryeo pottery both within and without Korea.10 Gang also presented some of the special characteristics of Goguryeo pottery by summarizing the papers by Kim Won-yong and Ahn Hwi-Joon, Geng Tiehua and Lin Zhide, and Jeong Chanyeong. Moreover, through a comparative study of the shape and glazing methods of Goguryeo pottery, Gang drew comparisons with Chinese wares. Gang's research paper on Goguryeo pottery is not comprehensive but it was significant at the time as it was the first comparative study in Korea on the production techniques and shapes of certain Goguryeo pottery vessels. As a research paper, nonetheless, her analysis is limited because it only compares the shapes of the vessels.

In the wake of the excavation of Goguryeo ruins in the Achasan area as well as Mongchontoseong (夢村土城: earthen fortress) in the Hangang basin in recent years, research on Goguryeo pottery has become increasingly active within South Korea.11 For instance, in 1995 an analysis on the Goguryeo pottery excavated from the Hangang basin was made by Choi Jongtaik of South Korea, which to some degree led to the revelation of the general distribution of Goguryeo pottery along the Hangang.12 However, as this paper almost exclusively focused on relics from the Hangang area, it fell short of providing a comprehensive analysis of Goguryeo pottery over a broader spectrum. The study was hampered, in its attempt at a comparative study of the shapes of the Goguryeo pottery from the Hangang area with those excavated in North Korea and China, by the dearth of research on the subject in both North Korea and China.

In 1999, another extensive analysis of Goguryeo pottery was undertaken by this author.13 By focusing on the analysis of Goguryeo pottery itself, this research paper aimed to transcend the skewed direction of previous research on Goguryeo pottery undertaken mainly for the purposes of chronological study of ancient Goguryeo tombs and to correct the inherent contradictions in such an approach. The result was a broader analysis of the types and processes of change in the form of Goguryeo pottery, based on an examination of 419 Goguryeo wares that were intact and which had been excavated in China, and in North and South Korea. However, as this paper still could not overcome the lack of data needed to establish a complete chronology, its methodology in the end depended much on the chronology of ancient Goguryeo tombs and further detailed research was necessary to establish a complete chronology of Goguryeo pottery.

In the same year a different study on the process of the formation of Goguryeo pottery was presented by the South Korean scholar Park Sunbal (朴淳發).14 Park assumed the firing of fine clay at a low temperature and a polished finish to be the primary features of Goguryeo pottery and focused on the formation process of these two features. He analyzed the pottery excavated from the ruins of the Hunjiang, the cradle of Goguryeo, and the middle reaches of the Amnokgang, between the Bronze Age and the early Iron Age. He went on to analyze the various characteristics of Nonam-ri (魯南里) type pottery from Pyeonganbuk-do, generally considered to be the precursor of Goguryeo pottery. Through this analysis Park concluded that, in terms of its shape, Goguryeo pottery had originated from the wares of the Bronze Age found around the Hunjiang and the middle reaches of the Amnokgang, and that features like the fine clay pottery fired at a low temperature were related to the production technique of hoedo (灰陶: gray pottery; Chinese: huitao) produced between the late Warring States Period and the early Han dynasty. Moreover, Park estimated the first Goguryeo pottery to have been made circa 200 BC, about the same time as the first production of Nonam-ri type pottery, and coeval with the establishment of an allied Goguryeo kingdom.

In 2001, upon analysis of glazed pottery vessels excavated from the Ji'an region, Geng Tiehua put forth the view that the glazed vessels were tomb ornaments buried together with the deceased, in view of the fact that these wares had been excavated mainly from the tombs of noblemen, namely tombs with mural paintings. Geng concluded that the glazed pottery was extant at the latest in the early fourth century.15

In 2003, Yang Sieun (梁時恩) of South Korea conducted an analysis on the production technique of Goguryeo wares excavated from Tomb No. 4 of Achasanboru (峨嵯山堡壘: mountain fortress) and the fortresses of Sirubong, Guui-dong (九宜洞), and Mongchontoseong in the Hangang basin, and determined that the differences in the production technique of the wares reflected different periods of production.16

In the same year, this author presented a synthesis of past discussions on Goguryeo pottery and an updated chronology of the subject through a presentation at the Korean Archaeological Congress.17 Subsequently, in 2004, this author analyzed Goguryeo pottery wares excavated from South Korea and put forth partially revised views. This last paper presented a revised chronology of Goguryeo pottery from the Hangang area by dating the artifacts found in Mongchontoseong, south of the river, sometime between 475 and 500 and those from Achasanboru, north of the river, between 500 and 551.18

As discussed in this brief overview of the history of research on Goguryeo pottery in South and North Korea, China, and Japan since the 1980s, few comprehensive analyses of the subject exist. The main reason for this dearth of analysis is primarily the lack of data, as the main historical area of the Goguryeo state is now in the Ji'an region in China and in North Korea. In fact, many Goguryeo tombs from these regions have been excavated but there have been few detailed reports or studies on the excavated artifacts, on pottery in particular. The excavation reports on these tombs typically consist of a few diagrams and a brief mention of the wares in just a few lines. Therefore, obtaining detailed information on the wares themselves has indeed been quite challenging.

However, in recent years, especially since the 1990s, the pace and scope of research has gained momentum with the publication of a comprehensive report for each field of Goguryeo archaeology of the Ji'an region. In addition to these new data, studies on Goguryeo pottery excavated in South Korea have also been gaining speed lately.

III TYPES OF GOGURYEO POTTERY VESSELS AND THEIR CHANGES

Goguryeo pottery can be largely classified into 24 types. Among them, types of wan (盌: small bowls), jongji (smaller bowls than wan), ttukkeong (lids), guhyeongho (球形壺: jars with a spherical body), ban (盤: plates and dishes), simbal (深鉢: beaker-like vessels), jangdongho (長胴壺: jars with a longer body than guhyeongho), and dong-i (urn-like vessels with a wide mouth and short body) constitute the largest share (Plate 1). Each Goguryeo vessel type developed into diverse forms over time and the following are typical changes presented by types.

( Plate 1 ) Configuration of Goguryeo pottery types

01 SAIJANGGYEONGHO AND SAIJANGGYEONG-ONG

The most distinctive types of Goguryeo pottery include jars with a wide mouth and a long neck such as saijanggyeongho (四耳長頸壺: jar with a long neck and four handles), saijanggyeong-ong (四耳長頸甕: tall jar with a long neck and four handles), saiong (四耳甕: tall jar with four handles), and janggyeongho (長頸壺: jar with a long neck). These four vessel types are different from one another but share one or more of the following attributes: a trumpet-shaped mouth, a long neck, and four daesangpasu. For this reason, in this paper they are studied as a single group. Considering that these four types are found in all Goguryeo tombs and that many of them were glazed with incised patterns, it is certain that they were mainly used as tomb ornaments, buried together with the deceased.

The saijanggyeongho type has a round body with four handles and a long, narrow neck. It was made only in the third century and is believed to be the prototype of the other three vessels, among which saijanggyeong-ong and janggyeongho continued to be used until the Balhae period. The saiong type has a wide mouth and neck and is characteristically larger than the other three types. Its large shape attests to its usability for everyday living and indeed saiong is quite frequently excavated from numerous dwelling sites. Janggyeongho is exceptional in that it has no handles but its shape developed in the same way as saijanggyeong-ong.

Of these four types saijanggyeong-ong is considered as the most typical Goguryeo vessel. As the changes in its shape over time have occurred in a relatively constant manner, it is usually used as the standard by which the dates of ancient tombs are measured. With regard to the origin of saijanggyeong-ong there are two different possibilities: one is the pankouhu (盤口壺: jar with wide, flat rim) of China19 and the other is the zhong (鍾: small jar originally used to contain wine) of the Han dynasty, which developed to become a particular Goguryeo vessel later in the fourth century.20 However, in view of the fact that saijanggyeongho was already extant in the third century and that the long neck and daesangpasu are features on other Bronze Age pottery vessels from the middle reaches of the Amnokgang, it is unlikely that saijanggyeong-ong was influenced solely by Han Chinese pottery.

As aforementioned, the chief features of saijanggyeong-ong are the trumpet-like, flared, long neck and the four strap handles (Plate 2). There have been various typological and chronological studies on saijanggyeong-ong over the years, most of which have focused on the shape of the vessel's neck. As already noted above, Wei Cuncheng categorized this vessel into two types based on the shape of the neck: one that flares out straight (Type I) and the other that rises upright before flaring out (Type II).21 Wei also observed that the neck of Type I grew shorter and that of Type II longer over time.

( Plate 2 ) Saijanggyeong-ong, excavated from Mongchontoseong, H: 59.0 cm, Seoul National University Museum

However, Wei's classification based on the neck shape is not without its weaknesses. Since the prominent attribute of this vessel type is the neck that flares out like a trumpet, small differences in the shape of the neck may be overlooked on account of the uniformity of this distinctive shape. Therefore, the criterion for analysis should include the point in the neck where it bends outward. Furthermore, there are differences and changes of the body shape to be considered in terms of roundness and slimness.

With such limitations in mind, this author has classified the shape of saijanggyeong-ong into Types I–V as shown in Plate 3 by using the mouth ratio (ratio of the length of mouth to the overall height: [mouth diameter/total height x 100]) and the circular body ratio (roundness of body part: maximum body diameter/body height x 100). Looking at the characteristics of each shape, Types I and II have similar mouth ratios but show differences in circular body ratios due to the differences in the length of the neck. The pieces belonging to Type II have a relatively shorter neck length than Type I. Type III has similar circular body ratio as Type II but has a bigger mouth ratio and this is also due to the differences in the neck length (Type III has a longer neck than Type II). Type V represents just one item excavated from Topo-ri daemyo (土浦里大墓: tomb site), and shows a smaller mouth ratio to the neck as well as a body more slender in shape.

( Plate 3 ) Correlation between body roundness and vessel slimness in saijanggyeong-ong

Based on these classifications saijanggyeong-ong may be recorded chronologically from round to slender body type. Type I vessels excavated from Maxian'gou Tomb No. 1 were assigned to the early to mid-Goguryeo period by Wei and to the late fourth century by Azuma, who made his estimation on the basis of the structure of the burial mound.22 Type II vessels from Yushan (禹山) Tomb No. 41 and Type III vessels from Zhangchuan (長川) Tomb No. 2 can be dated to the mid-fifth century.23 The chronology of the burial mounds coincides with the evolution of saijanggyeong-ong from Type I to Type V as well demonstrated by the differences in the circular body ratios. In other words, the circular body ratio of saijanggyeong-ong changes from large to small, and its body shape from round to slender (Plate 4).

( Plate 4 ) Diagram of changes in the shape of Saijanggyeong-ong

02 SIMBAL

Simbal, generally called guan in China, is a type of vessel that has been used from very early days (Plate 5). The chief features of this vessel type are a wide mouth with virtually no neck and a long body. Those with handles can be sub-divided as yang-isimbal (ܽ兩耳深鉢: simbal with two handles). Unlike other vessels, many simbal are made of a coarse paste and were built by hand rather than thrown on a potter's wheel. With respect to vessels with handles, there are some with bullhorn-type and some with sprig-type handles but most have the familiar four strap handles, daesangpasu. The daesangpasu are typically circular in section and attached in a vertical, upward direction. This kind of vertical daesangpasu or loop-type handle has been a common feature in pottery of the Amnokgang basin since the Bronze Age. In addition to the coarse body, this is evidence to demonstrate that simbal had been produced since early times.

( Plate 5 ) Diagram of changes in the shape of simbal

Simbal may be classified into three groups according to size with no relevance to the time period, although the slimness of the body does change over time. The early simbal came in a rather slim shape with a tall height relative to the maximum diameter of the vessel or the diameter of the mouth, but later it gradually took on a more rounded shape. As yang-isimbal is generally slimmer than simbal, it would seem that the former was produced earlier.

Plate 5 reflects these various attributes. In terms of shape, simbal may broadly be classified into three types: Type I with a slim body and a short neck length relative to the total height of the vessel; Type III with a short neck and a round, convex body; and Type II in between Types I and III. In the case of simbal excavated from Tomb No. 8 of Xiahuolongcun (下活龍村 JXM8), which falls under Type I, the ruins are estimated to be jeokseokchong (積石塚: stone mound tombs) of the earliest period – more specifically, before the third century.24 Furthermore, the simbal recently excavated from the third culture layer of Wunüshancheng (五女山城) in Huanren also featured a slim body and vertical daesangpasu, affirming the early provenance of this particular vessel.25 At the same time, the simbal excavated from the Guuidongboru ruins fall under Type III. It is therefore apparent that the type of yang-isimbal changed over time from Type I to II, and then to III, since the body shape of simbal changed from slimmer to rounder and at the same time the neck became shorter and more elaborate.

03 JANGDONGHO

Jangdongho has nearly the same mouth diameter as its base diameter with a relatively tall body. It was a vessel much used in practical everyday life (Plate 6). It can be classified into two types according to size, and into four types according to the shape of body and the location of the maximum body diameter. Type I, which has a convex, round body with the maximum body diameter below the mid-point of the overall height, is similar to simbal Type III, the latest type of simbal. This indicates Type I jangdongho to be the oldest of this vessel group. Its first production is estimated to be around the fifth century and, over time, jangdongho featured changes in its body shape particularly around the neck and shoulder. The neck gradually became longer and the location of the maximum body diameter moved up towards the shoulder. This vessel was in continual use until the fall of the Goguryeo state.

( Plate 6 ) Jangdongho, excavated from Guuidongboru, H (right): 31.3 cm, Seoul National University Museum, Korea

04 IBUHO

Ibuho (耳附壺: jar with a spherical body and handles) is a vessel with two or four handles on a round body but a shorter neck than that of saijanggyeong-ong. Ibuho is divided into two types according to body shape: Type I has a round body, later changing to Type II with either a globular or a taller body. The earliest ibuho was found along with saijanggyeong-ong from the partially destroyed ruins of Sanchengxia (山城下) Tomb Mound, dated to the third century.

05 GUHYEONGHO

Guhyeongho is a vessel with a spherical body, a short and curved neck and no handles (Plate 7). It is classified into three types according to the shape of the neck (including the mouth) and the body (Plate 8). The body of Type I is almost spherical but the maximum body diameter is located at the shoulder and the most distinctive feature is that the neck and the mouth are almost perpendicular. The body of Type II is spheroidal like a compressed ball, while its short vertical neck flares outwards. The body of Type III is almost spherical although it is a little slimmer than the other two. The neck flares outwards but is even shorter than that of Type II.

( Plate 7 ) Type III Guhyeongho, excavated from No. 4 of Achasanboru, H: 22.8 cm, Seoul National University Museum

( Plate 8 ) Diagram of changes in the shape of guhyeongho

Type I guhyeongho were mainly excavated from the yeonjeopsik jeokseokchong (連接式積石塚: stone mound tombs with connecting vaults and mounds) of Yushan (禹山) Tomb Mound. The date of the tomb is estimated to be sometime between the fourth and fifth century.26 Pottery vessels with similar body shapes have also been excavated from the tomb with murals of Yuan Taizi (袁台子) in Chaoyang (朝陽) and the Jin (晋) tomb with murals in Benxi (本溪), Liaoning, which are dated to the fourth century.27 Two parallel lines are incised around the shoulder of Type I guhyeongho, and junghomun or concentric semicircular pattern is typically incised inside. This particular pattern is thought to be among the earliest found on Goguryeo pottery.

Type II guhyeongho account for the majority of the spherical vessels. The paste of the vessel is either coarse and sandy or clay, with a greater proportion being of the clay variety. Unlike Type I, only one Type II vessel with patterns has so far been found. The potteries excavated from the historic site of Minzhuliudui (民主六隊) of the Balhae period tend to have a shorter neck and a flatter, globular body. At the same time, the Type II guhyeongho from Tomb No. 195 of Shanchengxia Tombs dated back to the early Goguryeo period have a longer neck and a spherical body.28 Therefore, it is conceivable that Type II guhyeongho gradually changed in body shape from spherical to spheroidal and in neck shape from longer to shorter.

Type III guhyeongho has its maximum body diameter in the shoulder area. It can be further divided into two groups, one with a narrow neck and the other with the maximum body diameter located in the middle or upper-middle part of the body and an underdeveloped shoulder. In the case of the former, three vessels have been excavated from Tomb No. 3 of Anak, which all have junghomun incised around the shoulders. Pottery vessels excavated from the historic site of Juwol-ri (舟月里), Paju (坡州), south of the Imjingang, also have the same shape and junghomun incised around the shoulders. As for the wares found in Anak Tomb No. 3, these can be dated to the middle or late fourth century. Within this vessel type, the guhyeongho with the maximum body diameter located around the center tend to be greater in height. The ones excavated from Guuidongboru and Achasanboru are thought to be from the sixth century, while those from Mongchontoseong can be dated to the late fifth century.

06 ONG

Ong (甕) is a tall vessel with a height of 40 cm or more and appears to have been used for storage. All extant ong have been excavated from the ruins of ordinary everyday habitations (Plate 9). Ong can be classified into three types according to the spherical ratio of the body. Type I has the smallest curvature to maximum body diameter ratio. Consequently, it is the slimmest. Pottery excavated from the Hangang area typically come in this shape. Type III has a larger base diameter than mouth diameter, with a very long maximum body diameter. Vessels of this kind have been excavated from Dongjincheng (東京城) of Balhae, which suggests this type to be the latest form among the three types.

( Plate 9 ) Ong, excavated from Guuidongboru, H: 67.0 cm, Seoul National University Museum

07 JIKGUHO AND JIKGUONG

Both jikguho (直口壺: jars with a flat mouth) and jikguong (直口甕: tall jars with a flat mouth) have a very short and straight neck. Both vessels were probably used with a ttukkeong (matching cover) but few have been found with one. These vessels are classified according to their height, less than 40 cm (壺, ho) or more than 40 cm (甕, ong). In either form, some come with handles which would indicate use of a practical nature (Plate 10).

( Plate 10 ) Jikguho (left), excavated from Achasanboru No. 4, H: 22.5 cm; Jikguong (right), excavated from Sirubongboru, H: 70.6 cm, Seoul National University Museum

The shape of jikguho changed over time into globular body type. The earliest jikguho is the one excavated from Tomb No. 242 of Wanbaoting (JWM242), dated to the third century. It was used until the Balhae period. Few jikguong have been found thus far; they are thought to have been used widely since the mid-fifth century as a vessel for everyday use.

08 SIRU

Siru (甑: vessel for steaming) can be classified according to the widths of the body and the base, while some have feet on the bottom. The feature that best shows the change in siru over time is the shape of the holes on the base of the vessel (Plate 11). In terms of these distinct bases with holes, there are those with several small holes clustered together and those with a larger central hole and with four or six holes around it. With respect to the former, the one excavated from Tomb No. 2 of Shanghuolongcun (上活龍村) (JMS2) has a mouth that curves inward and junghomun around the shoulder of the vessel. According to conventional chronology the siru from this tomb is dated to the end of the second century.29 However, as junghomun generally makes its first appearance around the mid-Goguryeo period, it is probably difficult to date these vessels back so early.

( Plate 11 ) Siru, excavated from Mongchontoseongboru, H: 27.0 cm, Seoul National University Museum

In the case of the small types of siru excavated from Yushan Tomb, considering that the latest end date of jeokseokchong has generally been revealed to be circa the fifth century Ji'an,30 they can be dated to the end of the fourth century or the beginning of the fifth century. These siru excavated from Yushan Tomb have a base with a hole in the center surrounded by eight holes. As the siru excavated from Shanghuolongcun have a base with several small holes, they can be dated back to a slightly earlier period, more likely the end of the third century or the early fourth century. The siru excavated from the Nonam-ri dwelling site are also thought to be from the same period.

All the medium and large siru, with the exception of those from Daepyeong-ri (大坪里), Bukchang (北倉) and Tomb No. 1 of Jigyeong-dong (地境洞), Pyeongannam-do, have a base with holes in an orderly configuration. The siru excavated from the Hangang basin have a base with a central hole and six surrounding holes. Meanwhile, the siru from Guershancheng (古爾山城) and Guoneicheng (國內城), Jilin, have a central hole and four surrounding holes on the base. In particular, the wares from Guoneicheng feature four oval-shaped holes surrounding the central hole, the intended dynamic of which seems greater intake of steam during cooking. Therefore, the ones from the Hangang basin are considered earlier than those from Guershancheng and Guoneicheng. As the ones from Guoneicheng are dated to the later years of the Goguryeo state, the ones from Guershancheng and Guoneicheng can be dated to the late sixth century, and the ones from the Hangang region to the late fifth century or the early sixth century.

09 BUHYEONGTOGI

Buhyeongtogi (釜形土器) are pottery vessels in the shape of an iron pot excavated along with small siru from ancient tombs. As iron pots (鐵釜, cheolbu) were widely used in Goguryeo, such pot-shaped vessels also seem to have been used as tomb ornaments (Plate 12).

( Plate 12 ) Siru on an iron pot, excavated from Guuidongboru, H: 45.5 cm, Seoul National University Museum

Buhyeongtogi can be classified into those with and those without a mount. The vessels gradually changed from those with a mount to ones without, generally in line with the change in iron pots.31 The buhyeongtogi from Tomb No. 3105 of Yushan (JYM3105) were excavated together with siru and are approximately dated to the end of the fourth century and the early fifth century. Meanwhile, the buhyeongtogi from another Yushan tomb (JYM3501) have a flatter mount and spheroidal bodies and thus are dated to the first half of the fifth century. At the same time, the buhyeongtogi from Yushan Tomb Nos. 1897 and 1815 and Tomb No. 1 of Jigyeong-dong have no mount and are dated to the middle or the latter half of the fifth century together with the siru excavated from Tomb No. 1 of Jigyeong-dong. Finally, the buhyeongtogi from Yushan Tomb No. 1493 have no mount but a developed shoulder and a long, narrow body. They can be dated to the late sixth century, slightly later than those from Tomb No. 1 of Jigyeong-dong.

10 DONG-I

Dong-i is a large flower-pot-shaped unpolished vessel of low height relative to the diameter of the mouth. In terms of functional utility it is the most practical of all Goguryeo pottery. The ratio of the mouth diameter to the height of the vessel, which reflects the distinctive shape of the ware, is quite diverse ranging from 120 to 296.9 (Plate 13). In other words, dong-i come in various shapes and sizes. The dong-i type is, however, classified into three groups according to the ratio of the mouth diameter to the height of the body. Type I includes those with a mouth diameter to height ratio of less than 160; Type II consists of those ranging from 160 to 220; and Type III are those with a ratio of over 220.

( Plate 13 ) Dong-i, excavated from Guuidongboru, H: 19.0 cm, Seoul National University Museum

Small dong-i are found only among Types I and II, unlike large dong-i found across all three types. Among Type I and II, several have daesangpasu regardless of the size. Among the dong-i excavated from Dongjincheng of the Balhae period, small ones belong to Type I and large ones to either Type I or III. The dong-i from Tomb No. 3103 (JYM3103) of Yushan Tombs, from which small dong-i (Type II) have been unearthed, are believed to have been produced at the latest by the fifth century as the latest end period of jeokseokchong is the fifth century.32 Therefore, small dong-i gradually changed in body shape from flatter to taller. In the case of large dong-i, this kind of change is hardly perceptible, although a general trend toward a flatter body can be perceived.

Based on the sectional diagrams of the dong-i from the Hangang basin, this author classifies these wares into four different groups by the shape of the mouth: Type I whereby the mouth area of the vessel is rounded but finished plainly; Type II whereby the mouth edges out at a sharp angle; Type III whereby the mouth not only edges out in a straight horizontal line but also the lower end of mouth is lifted up to make a groove; and Type IV whereby the edge of the mouth is slanted or treated in an oval shape while the lower end of the mouth is lifted up.33 However, as most of the data presented in this paper are based on analyses of diagrams, it is impossible here to classify these wares in greater detail based on the actual differences in the shape of the mouth. Nevertheless, in the case of dong-i from the Hangang area, Types III and IV are thought to have been produced at later dates.34

11 WAN

Wan (宛) is a bowl resembling present-day sabal (typical bowls for rice and soup). It is thought to have been used mainly as a serving bowl to serve food to individuals. The most frequent type of wan is with lid or ttukkeong (Type III) (Plate 14). Jongji is in fact a small wan, thought to have been used as a small container for ingredients such as the present-day soy sauce. Due to the simple shape of wan, noteworthy changes in form are virtually imperceptible. Wan can be classified into two groups, namely those with shallow mount and those without. In the case of wan with mount, some differences in their production technique can be detected: those with platform mount; those that have the platform mount attached first, after which the inside is hollowed out; and those with relatively high mount. These differences are considered to reflect the different periods of production.35 The wan with lid have often been excavated together with their lids and are thought to have influenced the daebubal (臺附鉢: mounted sabal) of Baekje in the Sabi (泗批, 538-660) period.36 On the base of some wan, a character such as 井 (jeong: well), 小 (so), and 大 (dae) is inscribed or molded in relief, most of which seem to have been intended to drive away evil omens.

( Plate 14 ) Wan with ttukkeong, excavated from Guuidonboru, H: 12.0 cm, Seoul National University Museum

12 BAN

Ban (盤) is flat dish with a wide mouth and low body height in the form of the plate today. The ratios of the vessel's mouth diameter to its body height ranges from 3:1 to 7.5:1. In the early period, ban was frequently used as a tomb ornament, and some were glazed. In the middle and late periods the vessels were more widely used in everyday life. Some ban have patterns incised on their base and dongsimwonmun or concentric circle pattern have been found on the base as well as the interior, on which various designs such as geoch, sagak (四角: square) and sagyeokja (斜格子: oblique lattice) are incised.

13 IBAE

Ibae (耳杯: eared cups) are similar in shape to wan but some have the jeon (展: flat rim) that partly circles the mouth area (Plate 15). All ibae excavated from ancient tombs have been found to be glazed, with slight variations in shape among those excavated from the fortresses of Guui-dong and Achasan. Namely, the ones from the tombs have the jeon on the longer side whereas the ones from the Hangang area have the jeon on the shorter side. The latter are considered to be produced later. Ibae type has also been found from the potteries of the Sabi period of Baekje, supporting the view that Bakje pottery has been influenced by Goguryeo pottery since the mid-sixth century.37

( Plate 15 ) Ibae, excavated from Yongmasanboru, No. 2, H: 8.5 cm, Seoul National University Museum

14 TTUKKEONG

As ttukkeong (蓋, gae: lids or covers) are a complementary part of other wares, strictly speaking they should not be considered as an independent pottery type. Nevertheless, since there are only a few cases of ttukkeong excavated together with the matching vessel this paper deals with ttukkeong as an independent pottery type for the sake of convenience. Ttukkeong are classified into six groups according to shape and size.

Type I ttukkeong has a disk-like body and a ring-like handle on top with no flange. In view of its shape, this type seems to have been used to cover vessels with a straight vertical mouth. Lids of this type have mostly been excavated from as early as the third to fourth centuries.

Type II ttukkeong have a flat body and a boju (寳珠: cintamani or precious pearl) knob or handle on top. The flanges are either straight or turned a little outward. This kind of ttukkeong has been found together with the bronze saiong excavated from the Geumgwanchong (金冠塚) or Gold Crown Tomb of Silla, and a Type II ttukkeong from Topo-ri daemyo has been confirmed to be the cover for a saijanggyeong-ong of the period, suggesting that this ttukkeong type has been in use since the mid-Goguryeo period.

Type III ttukkeong has deeper flanges and a boju-shaped knob. In view of its size, this type seems to have been used to cover wan and cylindrical tripod vessels. This type of cover is dated after the fifth century.

Type IV ttukkeong are the same as Types II and III in that they also have a boju-shaped handle but are different in that the flanges are either curved inward or lifted upward. Based on the shape of the flange, this type seems to have been used for vessels with a semicircular mouth and for small bottles, particularly considering its diameter. These ttukkeong are believed to have been in use mainly since the sixth century.

Type V ttukkeong have no handle and the flanges are short and straight or bent once. In view of their shape and size, these ttukkeong seem to have been used for hap (盒: small boxes). They are dated to sometime after the mid-Goguryeo period.

Type VI ttukkeong are the most common type among those excavated. When reversed, they have the same shape as ban and thus are called ban-type ttukkeong. This kind of ttukkeong usually has no handle, but some have been found with a strap handle at the center of the body or a pair of boju-shaped lugs on opposite sides of the flange (Plate 16). These ban-type ttukkeong are further classified into two subgroups according to their size, small and large, which have little or no correlation with the time period of production but more with the mouth size of the vessels with which these ttukkeong were matched. If the wares that these ttukkeong covered are assumed to have been yugaehap (有蓋盒: lidded box), ho, ong, and jikguong, then the small ttukkeong seem to have been used mainly for ho or yugaehap, and the large ttukkeong for ong or jikguong.

( Plate 16 ) Type VI ttukkeong, excavated from Guuidongboru, H: 11.7 cm, Seoul National University Museum

15 JEOPSI

Jeopsi (蝶匙 :dishes) are practical vessels that have mainly been excavated from the ruins of everyday dwellings. At the same time, as in the jeopsi found in Tomb No. 3 of Anak, some were also used as tomb ornaments. The simple shape of jeopsi makes it difficult to note changes in shape over time but in general jeopsi are classified into two types according to whether or not the foot is attached and into three types according to whether small, medium or large in size. A type of jeopsi like wan have some characters engraved or molded into relief on the base. Most of these characters were carved before the baking process and some wan and jeopsi contain even signatures on them perhaps to indicate the craftsman, the production site, or the patron, providing us with valuable information on the status of the owner or the ware itself, or the name of the production site (Plate 17).

( Plate 17 ) Engraved letters on the bases of various jeopsi, excavated from Achasanboru No. 4, Seoul National University Museum

IV THE FORMATION AND DEVELOPMENT OF GOGURYEO POTTERY

01 THE ORIGINS AND FORMATIVE PROCESS OF GOGURYEO POTTERY

Discussions on the origins and formative process of Goguryeo pottery have thus far been only tentative due to a persistent lack of archaeological data. From an archaeological standpoint, the formation of the Goguryeo state is believed to have taken place sometime around 100 BC,38 but archaeological evidence of pottery prior to the third century AD is extremely rare. Studies of burial sites in the early years of the Goguryeo state have made advances over the years but studies on the pottery still remain insufficient. Moreover, in order to establish the origins of Goguryeo pottery, archaeological evidence from the Bronze Age and the early Iron Age should be taken into consideration, although such archaeological studies also suffer from the lack of reliable data.

Considering that the geographical origin of the state of Goguryeo is the Hunjiang basin and the middle reaches of the Amnokgang in the Ji'an region, it is possible to ascertain the formative process of Goguryeo pottery through archaeological studies of the culture of these regions. Recent studies on the subject show that the Bronze Age culture of the Hunjiang and Amnokgang areas are understood to be a combination of features from the Miaohoushan culture (廟後山文化) in the Benxi, Liaoning Province of the pre-Bronze-Age Dagger Period (先銅劍期) and from the Gonggwiri culture (݀公貴里文化) in Ganggye (江界), Jagang-do (慈江道) of the Lute-Shaped Dagger Period (琵琶形銅劍期). This is followed by the Lalatun (拉拉屯) late Bronze Age culture in Changchun (長春) area and Dadianzi (大甸子) early Iron Age culture in the Liaoning area. The Dadianzi culture is regarded as of the early Iron Age, around 200 BC. Among them, the Lalatun culture exhibits features like the vertical daesangpasu, one of the most distinctive characteristics of Goguryeo pottery. This, in turn, is a feature found in Nonam-ri type pottery, the early prototype of Goguryeo pottery.39

The features of Nonam-ri type pottery are determined by studies on vessels excavated from the upper layer of Nonam-ri ruins.40 They are characterized by the daesangpasu attached either horizontally or vertically to the body and widely regarded as the prototype of Goguryeo pottery. According to available reports, the body of these vessels still contains fine sand, though generally in lesser quantity and in some cases the clay is of the same quality as in pottery vessels from the Bronze Age. Most that have been excavated are either brown or black polished vessels, and the characteristic is seen as another prominent feature of Goguryeo pottery.

According to Park, the principal characteristics of Goguryeo pottery are the daesangpasu, the fine clay paste, and the polished surface. Park states that the polishing technique and the daesangpasu are found in the pottery from Misong-ri (美松里), Pyeonganbuk-do and Gonggwiri, Jagang-do, which are regarded as distinct traditions of these respective regions. He sees, however, that the fine clay paste seems to have been influenced by the gray pottery of China from the late Warring States Period to the Han dynasty.41

Such views on the general characteristics of Goguryeo pottery seem reasonable, but this author finds that some Goguryeo pottery vessels are made of coarse earth paste containing sand grains. Therefore, the generalization that all Goguryeo pottery is made of fine clay paste seems not to be entirely appropriate but to be modified as a transitional characteristic. In any event, the origins of Goguryeo pottery are to be found in Nonam-ri, and will be illuminated by further comparative studies in relation to relics excavated from the early jeokseokchong from Ji'an and the middle reaches of the Amnokgang, as well as pottery of the early Iron Age.

Unfortunately, the task is quite challenging as there is only scanty material on pottery from this period and much uncertainty as to the chronology of Nonam-ri type pottery. Some scholars estimate the earliest date for these wares as 200 BC and the latest date around the beginning of the Christian era,42 while others date the wares as far back as 300 BC.43 The estimate of 200 or 300 BC are based on the fact that some Nonam-ri type potteries have been found along with myeongdojeon (明刀錢: knife-shaped coins used in China in the late Eastern Zhou period, circa 400–300 BC). However, since other artifacts like the yellowish white pottery native to the Pyeongyang region have also been found together at the Nonam-ri archaeological site,44 the fact that knife-coins were found does not necessarily mean that all the pottery from this site are from before the Common Era. As the Nonam-ri site was used over a long period, from the Bronze Age of the lower layers to the Goguryeo era of the upper layers, the earliest date for the lower layers of this site may reasonably be estimated at from 200 to 300 BC but the dates for the upper layers require further scrutiny.

In fact, the simbal type excavated from the upper layers of the Nonam-ri dwelling site is estimated at 300 based on the dates of jeokseokchong of the Ji'an region from which identical wares have been found.45 This much later date is a substantial departure from conventional estimates of 200 BC but the date seems reasonable considering the fact that the complexity in the chronology of the Nonam-ri site. This author views that, based on the shapes of the actual wares found at the site, the earliest date of pottery from the Nonam-ri site is hardly dated as far back to 200 BC. This view poses a problem in studying its pottery from the time of the establishment of Goguryeo to 300, but the problem is unavoidable due to lack of material evidences in the time period from 100 BC to around 200. Taken together all the available resource, it is only by the late third or the early fourth century that wares with typical Goguryeo pottery features have been found and the time span interestingly coincides with the formation period of Goguryeo as a governmental state.

02 THE EVOLUTION OF GOGURYEO POTTERY

To explain the evolution of Goguryeo pottery, Geng and Lin divide the entire time span into three periods: early, middle, and late,46 while Azuma Ushio divides it into five: incipient (200-300), early (300-400), middle (400-500), late (500-600), and final period (post-600 AD).47 However, there is simply insufficient evidence to support such divisions and, even more importantly, information on Goguryeo pottery before 200 AD is scant. Moreover, since the study of Goguryeo pottery is still in its infancy and lacks reliable sources, such a detailed chronological specification may only cause further confusion. This author thus explains the evolution of Goguryeo pottery in three stages: early (prior to 300), middle (300 to 500), and late (after 500), which coincide with the developmental phases of the Goguryeo dynasty. Scholars of ancient history generally agree with the view that Goguryeo was established as a state sometime between the late third century and the early fourth century48 and they regard the middle of the sixth century as the dividing line between the middle stage and the late stage of Goguryeo.49

Table 1 (see page 31) shows the characteristics of Goguryeo pottery with respect to the three different time periods. The material of Goguryeo pottery is largely divided into coarse earth paste or fine clay paste and then glazed or not glazed (undetermined). In the early stage there was not a single piece of glazed pottery, with coarse earth paste accounting for 91% and fine clay paste 9%. In the middle stage, fine clay paste increased to 57% while coarse earth paste decreased to 29%, with glazed pottery making up the rest, 14%. In the late stage, fine clay paste, glazed pottery, and coarse earth paste accounted for 89%, 8%, and 3% of the total, respectively. Since glazed pottery should be made from fine clay paste, it is safe to say that almost all of the pottery of the late stage was made of fine clay paste.

As the ratio of raw materials indicates, Goguryeo pottery was made mainly of coarse earth before the fourth century, marked by a steady increase in the use of fine clay, and by the late stage almost all pottery was made of fine clay. In the middle stage, glazing technique began to be used and was applied at a constant rate throughout the rest of the Goguryeo period. This glazing technique is a rather distinctive feature of Goguryeo vessels as glaze is not found in contemporaneous pottery wares of the Baekje or Silla–Gaya regions. Furthermore, as glazed wares have mainly been found in ancient tombs, they served the function of tomb wares. As for the origin of glazed potteries, there are three different opinions: prior to the third century,50 the late third century,51 and the early fourth century.52 The glazed pottery of Goguryeo is the product of low-temperature firing and light glazing. In China, such wares were made from the Han dynasty onwards; in Goguryeo such pottery of greenish brown color fired at a low temperature was made from the late third century onwards or, at the latest, from the early part of the fourth century.

The surface color of Goguryeo pottery is largely classified into yellow, black, and gray, of which brown, reddish-brown, and even red are classified under the color yellow. In terms of the surface color throughout the entire duration of the Goguryeo state, yellow is the most common, while black is the rarest. From the middle stage on, the number of gray-colored wares increases as compared to the early stage. In view of the fact that many gray, hard pottery vessels have been found among pottery from the Balhae period, one may speculate that the style of Goguryeo pottery eventually evolved toward a hard surface.

As for the manufacturing technique, two methods were mainly used: hand-made and wheel-thrown. In the case of hand-built pottery vessels, the general method of production would take the following steps: first, the base plate would be laid, after which the walls and rim of the ware would be built up by hand, then the ware would be smoothed and finished on a turntable. In the early stage, about 50% of Goguryeo pottery was hand-made. This proportion falls sharply to 7% in the middle stage and to only 2% in the late stage, which means that almost all Goguryeo pottery were made on a turntable plate or the potter's wheel with only a small number made by hand in the early stage.

In previous studies, Goguryeo pottery has generally been characterized as having a polished surface. However, among all Goguryeo pottery vessels, only 29 (a mere 7%), have been found to have a polished surface. Therefore, the general characterization of Goguryeo pottery as having a polished surface is unfounded. In addition, this small proportion of 7% is about the same as that of glazed pottery, indicating that glazed pottery vessels constitute most of polished pottery. Among the potteries excavated from the Hangang basin that are available for examination, most were polished with the potter's wheel or on a turntable. Among some types like jangdongho, there are examples which were first made on the potter's wheel or turntable then the foot and lower parts were smoothed by delicate carving or rubbing. This technique may be considered as an extension of the polishing method but it is still different from those that were smoothed and polished entirely by hand without the use of a turntable or wheel.

Another feature of Goguryeo pottery is that most do not have decorative patterns. In the early stage there were no patterns at all, whereas in the middle and the late stages a certain proportion of pottery vessels were made with decorative designs. Furthermore, only very few types of vessels featured patterns and the patterns themselves were also quite simple and limited. The techniques used for applying patterns on Goguryeo pottery include pressing, carving, and rubbing. The patterns include jeomnyeolmun, yeonsoksagakmun, geochimun, eogolmun (魚骨文: herringbone pattern), gyeokjamun, sagyeokjamun, dongsimwonmun, pasangmun, and junghomun.

The techniques used to apply these patterns varied from pattern to pattern. For instance, jeomnyeolmun and yeonsoksagakmun were made by impressing. Geochimun, eogolmun, gyeokjamun, sagyeokjamun, dongsimwonmun, pasangmun, and junghomun were made by carving. Other methods include ammun technique used mainly for generating irregular oblique lines, lattices, and continuous ring chains. This ammun technique in particular is a characteristic feature of Goguryeo pottery. The method involves pressing and rubbing the surface of the ware and seems to have originated from the rubbing of the surface with a hard tool. This method is apparent in pottery vessels of the middle and the late stages, and in some from the Balhae period.

To sum up, the production techniques and characteristic features of Goguryeo pottery are classified according to the raw materials, production technique, surface color, surface treatment, and patterns. In the early stage coarse materials were used and hand-made wares were made only in small quantities, but in the middle to the late stages all wares were made of fine clay paste with use of either the turntable or the potter's wheel. Also, from the middle to the late stage, some ceramics decorated with low-temperature glazes can be seen, the majority of which were used as funerary gifts. In terms of the surface color, yellow was the most common across the entire spectrum of Goguryeo history, with a progressively increasing share of gray-colored wares in production. Patterns on Goguryeo pottery were applied from the middle stage on until the end, but only on a limited number of vessel types and with limited variations.

Goguryeo pottery developed through a variety of forms depending on type and period. There are few vessel types that can be dated to the early stage. Only in the middle and late stages do all types of Goguryeo pottery vessels emerge.

Only a few pottery vessels from the early stage have thus far been excavated: they include: saijanggyeongho, simbal and yang-isimbal, ibuho, jikguho, siru, wan, ban, ibae, Type I lids, jeopsi, hap, and daebusaibal (臺附四耳鉢: mounted sabal [沙鉢: small bowls for rice] with four handles). A characteristic feature of the vessels from this early stage is that they have handles either of strap or knob type. Among the daesangpasu some are transverse, but most are vertical. As for the raw materials, coarse earth paste is the most common. Surfaces were often polished and no glazed pottery has been found in this early stage. No vessel at this stage has incised patterns and most were hand-made without the use of the potter's wheel or turntable.

By the middle stage, most of the known types of Goguryeo pottery were being made. But large vessels like daehyeong-ong and jikguong are found only in the later part of the middle stage. Gwangguho (廣口壺: jar with a wide mouth) were not made until even later. On the other hand, new types such as saiong, saijanggyeong-ong, and janggyeongho were produced in the middle stage. These new vessels are considered to be derivations of saijanggyeongho. As for the simbal type, a few yang-isimbal vessels with handles have been excavated among the remains of the middle stage while the total number of simbal increased over the same period. The shape of simbal gradually changed to a more bulging body type with bigger shoulder and neck, but decreased in quantity after the end of the middle stage. Jangdongho are a new type from the middle stage; gradually they developed a wider shoulder and a longer neck.

The guhyeongho type developed the most significantly in the middle stage, especially in terms of the variety of shape. Type I and Type III guhyeongho changed over time to assume a longer body and a larger shoulder, while Type II evolved from a spherical body with a long neck to a globular body with a short neck. The siru (steamer) gradually changed from those with one central hole surrounded by eight other holes at the base to those with one central hole surrounded by six other holes, with a corresponding increase in the size of the holes. From the early stage to the early part of the middle period, siru and buhyeongtogi (pottery vessels in the shape of iron pots) were used as funerary gifts. It is only in the second half of the middle stage that big siru were produced for practical use. Buhyeongtogi are another new type from the middle stage, some of which were glazed. The shape changed from those with feet to those without.

Greater numbers of dong-i, wan, ban, and ibae have been excavated from the middle stage. Among other new types first seen among the relics of this middle period are Type III lids with a rounded boju-shaped handle and (Type VI) with banhyeong (盤形: dish-shaped) ttukkeong or lids. The Type III lids were mainly used for covering wan and cylindrical three-legged vessels, while the banhyeong ttukkeong was for the large-sized or daehyeong (大形) pottery, as the emergence of such lids coincides with that of large-sized potteries. Hand-made dishes of the early stage evolved to the wheel-thrown type, becoming flatter with the mouth portion lower. Some were made with short feet at the base. A new type of jar similar to the hemispherical bottle with dish-shaped mouth (盤口瓶, bankouping) of the Chinese Six Dynasties was made for the first time, including a peculiar type such as the cylindrical three-legged vessel.

A small number of pottery vessels of the middle stage were made of coarse earth but most were made of fine clay. Glazed pottery was produced for the first time in this period and the wheel-thrown method became a dominant production technique of the time. Many vessels were decorated with concentric semicircles, wavy lines, dotted rows, saw-tooth, and herringbone patterns are regularly found on the shoulders of saijanggyeong-ong and the Type II spherical ho. Some wares bore ammun, a pattern made by pressing and rubbing technique that is seen only in the second half of the middle stage.

During the late stage, all the pottery types of the middle stage continued to be produced, with practical wares such as ho and ong increasing in quantity. In the case of saijanggyeong-ong, the body gradually became narrower, and by the Balhae period the neck had become so narrow and long that it had evolved into the shape of a bottle. In this stage, large wares such as jikguong, daehyeong-ong, and dong-i were produced in large numbers. Ban and ttukkeong accordingly grew larger during this stage. Such trends correlated with the practical function of the vessels of this stage, as evidenced by the preponderance of wares excavated from the ruins of ordinary dwellings. Pottery vessels in the late stage are mostly glazed and made of fine clay. As for patterns, the practice of ammun prevailed, while patterns like concentric semicircles, dotted rows, and wavy lines that had been used often in the middle stage became obsolete.

V CONCLUSION

This paper has so far examined the general characteristics of Goguryeo pottery, the developmental pattern, and the chronology of various types of vessels. As is the case with the general state of archaeological study of Goguryeo, the study of Goguryeo pottery remains stagnant mainly due to reasons beyond scholarly interests per se, namely the political environment. In fact, often in the past, research on this subject in China and North Korea - the hub of archaeological findings on Goguryeo pottery - has been varied in academic rigor and skewed in scope. Fortunately, however, some detailed and comprehensive studies on Goguryeo pottery have occasionally been published in China since the 1980s. Also in South Korea new investigations on Goguryeo historical sites have recently taken place, stimulating greater interest in research on Goguryeo pottery.

Much work lies ahead in overcoming the dearth of archaeological data and establishing the origins and the evolutionary process of Goguryeo pottery. However, recent research has shown that Goguryeo pottery was built upon the Bronze Age and early Iron Age pottery-making traditions of the Hunjiang and the middle reaches of the Amnokgang, supplemented by the introduction of pottery-making techniques from China during the Warring States Period to the Han dynasty. This theory is quite feasible from a historical perspective.

The production techniques of Goguryeo pottery have been clarified to some degree in terms of the developmental phases by analyses of raw materials, surface coloring, glazing, manufacturing method and surface treatment, and decorative patterns. However, such analyses should be complemented by further scientific research such as analysis of the composition of the raw materials and the temperature at which they become plastic.

In general, Goguryeo pottery is classified into about 24 different types and depending on size and physical attributes can also be divided into vessels for practical use and those without practical utility. The latter are mainly funerary articles, while the former are variously classified based on their functions for storing, cooking, serving, and transporting food. This kind of classification by function is an important factor in evaluating the vessels. But since such classifications are usually assumptions based on shape, they should further be complemented by analytic data related to the circumstances of excavation and their connection with other remains found together during excavation.

The chronology and associated different stages of the development of Goguryeo pottery are common subjects of interest to all researchers in the field, attracting numerous scholastic projects. However, as described above, most of the research on the subjects thus far have been conducted as part of a chronological study of old tombs, or as complementary data supporting the chronology of old tombs. Undoubtedly, ascertaining the dates of the old tombs and the historical relics from various ruins provide the most important information in establishing the chronology of pottery vessels. On the other hand, classification according to type and the analysis of evolutionary changes in regard to the different types should be conducted as a prerequisite to chronological research.

It is to be expected that there are substantial omissions and errors in this paper as the data used for this study were, with the exception of Goguryeo pottery vessels excavated in South Korea, based on indirect information in the form of diagrams, photographs, and other academic papers. Moreover, I have not made an independent interpretation of the ancient tombs, the main subject of archaeological studies on Goguryeo so far. In conclusion, more rigorous analyses on the manufacturing techniques behind Goguryeo pottery across the entire historical spectrum need to be undertaken, as well as verification regarding the developmental process and the chronology of Goguryeo pottery that I have presented herewith.