The Journal of Korean Art and Archaeology

Introduction

Excavations have recently taken place outside Wolseong Fortress at sites in the environs of the “Donggung Palace and Wolji Pond in Gyeongju” locale, as well as in the area of the nearby Balcheon Stream site. This has facilitated a full-fledged discussion on the expansion of the Silla palace complex into the outskirts of Wolseong Fortress during the late seventh century CE. It is well known that the grounds of the Silla palace complex, which had mainly been limited to the area within Wolseong Fortress, extended further beyond the fortress walls in the seventh century CE. The now commonly accepted view is that the Silla palace complex encompassed not only Donggung Palace and Wolji Pond but also reached Cheomseongdae Observatory. However, discussions have only recently begun to consider the changes that took place in the area that was newly incorporated into the palace complex, or discuss the types of facilities built in that area. This article therefore examines the transformations that occurred in the area newly incorporated into the palace complex in the latter half of the seventh century by tracing the spatial reorganization that took place, as is evidenced through the sites around Wolji Pond and the area around the Balcheon Stream site.

The Location of Dongji Pond (東池) and Its Relationship with Wolji Pond (月池)

An account from the fourteenth year of King Munmu (文武王)’s reign (674) in the Samguk Sagi (三國史記) records that “a pond was dug in the palace grounds and an island was constructed, where flowers and grasses were planted and rare birds and strange animals were kept.” Based on this, it was thought that Wolji Pond was first constructed in 674 CE, but evidence contrary to this statement has also been discovered.

During the excavation of Wolji Pond (1975-76), the cross-section of the pond’s west bank was also investigated in order to establish how the pond’s retaining wall (護岸石築, K. hoanseokchuk) had been built. As a result, it was possible to identify traces of earth that had previously been disturbed by human activity. This indicates that a pond had existed at the location of Wolji Pond prior to the construction of its stone retaining wall (Cho Yoojeon 1981). If that was indeed the case, what then did the pond look like prior to the construction of the stone retaining wall and the artificial island in 674 CE?

Wolji Pond is located where groundwater —trapped in the areas to the east and southeast of the pond— erupts through the ground surface (Hwang Sangil et al. 2010), and so the formation of a natural pond through water accumulation is inevitable. As such, it may be that the pond that existed prior to 674 was a natural, rather than artificially built, pond. The two following accounts in Samguk Sagi can be considered in relation to this possibility.

According to Historical accounts A-1 and A-2, a dragon is said to have appeared in the eastern pond of the palace. The palace is mentioned separately to Geumseong Fortress, making it possible to suggest that the “palace” refers to Wolseong Fortress. Therefore, it is logical to try to find the traces of this pond in the eastern part of Wolseong Fortress. In relation to this, the following accounts that appear in Samguk Yusa (三國遺事) should also be considered.

According to Historical accounts B-1 and B-2, a dragon resided in Dongji Pond, and in 754, the pond overflowed, sweeping away around fifty rooms units of the buildings forming the palace’s internal quarters. This indicates that Dongji Pond was located close to the internal quarters, and they were considerably large quarters. If this is the case, then where would the palace’s internal quarters have been located?

Up until the present, a pond dating to the Unified Silla Period has yet to be discovered within Wolseong Fortress. Given the fact that Wolseong Fortress is located in a hilly area, even if there was indeed a pond located within the fortress walls, its size is unlikely to have been big enough to contain water that could sweep away fifty rooms of buildings. In addition, although Dongji Pond is said to have overflowed in the summer of 754 (according to Historical account B-2), it is recorded in the Samguk Sagi that there was a drought in the eighth month of the year 754. It is therefore unlikely that such sudden rainfall in the summer of 754 would have led to the overflowing of Dongji Pond. As such, the only other likely possibility for explaining the overflowing of Dongji Pond is an eruption of a spring. However, the area within Wolseong Fortress is hilly and therefore not a place where springs erupt (Fig. 1). Due to this fact, it becomes difficult to find the location of Dongji Pond and, concomitantly, the location of the palace’s internal quarters within Wolseong Fortress.

Fig. 1. A cross-section of the center of the hilly landscape where Wolseong Fortress is located

It could be suggested, therefore, that the possible location of Dongji Pond could be established by determining the location outside Wolseong Fortress where the palace’s internal quarters could have been situated. Areas to the north or southeast of Wolsong Fortress are generally understood to have been the sites of government offices and their annex buildings, and therefore cannot be considered as likely settings for the palace’s internal quarters. In addition, building complexes comprising approximately fifty rooms or more of buildings have yet to be identified in the areas to the north or southeast of Wolsong Fortress. As a result, we cannot help but focus on the buildings remains identified around Wolji Pond.

It was recently suggested that the building complex located to the west of Wolji Pond was used not by the Crown Prince but by the Silla King (Yoon Mubyeong 2010; Choi Yeongseong 2014; Yoon Seontae 2019; Lee Hyuntae 2020a; 2020b; Lee Dong-joo 2020). In addition, the results of a geophysical survey revealed that Wolji Pond and the area to its north were not defined by using walls as spatial boundaries. Moreover, this area was found to have housed a large building complex aligned in an orderly matter along a North-South axis. This suggests that the spatial zone associated with Wolji Pond may have extended further to the north. In contrast to this, the areas to the south and east of Wolji Pond have yielded very little evidence of large building complexes, making it difficult to believe that the palace’s internal quarters may have been situated there. As such, it is proposed that the building complexes identified through geophysical survey in the areas to the west and north of Wolji Pond can be interpreted as the palace’s internal quarters mentioned in the Historical account B-2.

If this interpretation is accepted as being plausible, then it can be further used to establish the location of Dongji Pond. From the perspective of an individual standing in the area north of Wolji Pond (i.e. the presumed location of the palace’s internal quarters), Wolji Pond is located to the south. In addition, located to the northeast of Wolji Pond are several other ponds: the pond identified to the south of the Hwangnyongsa Temple History and Culture Center (K. Hwangnyongsa yeoksamunhwagwan), the pond located to the northwest of Hwangnyongsa Temple (皇龍寺址, K. Hwangnyongsaji), and the Guhwang-dong garden pond. Of these ponds, the only one to have been located close to the palace’s internal quarters in the eighth century and large enough to contain an amount of water sufficient to sweep away around fifty rooms of buildings is Wolji Pond. As such, Wolji Pond can be established as the “eastern pond” (i. e. Dongji Pond) that appears in the historical records.

Given that Wolji Pond is located to the east of Wolseong Fortress, it can indeed be referred to as the “eastern pond.” In addition, Wolji Pond drains out to the north; this fact is in concurrence with the historical account of the buildings of the palace’s internal quarters being swept away by the flooding of Dongji Pond. Noticeably, located north of Wolji Pond is Gaeseokgok Valley (開析谷), the largest valley of the capital city. Through this valley, Balcheon Stream flows from the east to the west. Due to this, the area north of Wolji Pond cannot help but be saturated with water (Fig. 2). It should be noted that, of the erupting springs located in the Gyeongju Basin, the one feeding Wolji Pond is the greatest in terms of water volume (Hwang Sangil et al. 2010); it is this water from Wolji Pond that flows into Balcheon Stream. The natural geography of Wolji Pond and the area to the north therefore provides the ideal conditions for an event such as the one described in Historical account B-2 (i.e. fifty rooms of buildings forming the palace’s internal quarters being swept away by the flooding of Dongji Pond) to take place. As for the mention of a dragon residing in Dongji Pond included in the Historical account B-1, it is worth pointing out that several pottery vessels featuring the inscription “Sinsim Yongwang” (辛審龍王) have been recovered from Wolji Pond. Based on the aforementioned evidence, it is possible to argue, with at least some degree of certainty, that Wolji Pond is indeed the same body of water referred to in historical texts as both “the eastern pond” and “Dongji Pond.”

Fig. 2. The distribution of streams fed by erupting springs and waterlogged areas in the Gyeongju Basin

It is therefore likely that Wolji Pond was originally referred to as the “Dongji Pond” but was provided with a new name, “Wolji,” following large-scale maintenance works undertaken at the pond in 674 CE, which involved the construction of a stone retention wall. This means that there were two names in use—“Dongji Pond” and “Wolji Pond”—referring to the same pond.

The Nature of Building Complex Situated to the West of Wolji Pond and the Background behind the Selection of Its Location

According to Samguk Sagi, in the second month of 679 CE, King Munmu renovated the palaces into a grander and more elaborate form, and in the eighth month built Donggung Palace (東宮). It is based on this account that the construction of the buildings around Wolji Pond is believed to have taken place since 679 CE. More recently, however, studies suggest that the buildings located to the south of Wolji Pond had already been constructed prior to 679 CE (Lee Sangjoon 2020). If this is the case, then it appears that the refurbishment of the pond in 674 CE and the large-scale construction of palaces in 679 CE took place when a pond, as well as some buildings, already existed in the area of Wolji Pond.

Previously, the building complex located to the west of the Wolji Pond (which likely represented the center of the building complexes around the pond) was linked to the “Donggung Palace” (東宮) mentioned in historical texts and interpreted as a palace complex used by the Crown Prince, rather than the Silla King. As mentioned above, however, it has recently been proposed that the space may have been used by the Silla King and not the Crown Prince. This view becomes more plausible when the layout of the building complex to the west of Wolji Pond is examined. Central to this building complex are Buildings A, B, and C, which face southwards and share the same central axis. In addition, these three buildings are situated within a central courtyard (中庭, K. jungjeong) formed by a surrounding corridor (回廊, K. hoerang) and transepts (翼廊, K. ikrang) (Figs. 3 and 4). This type of layout and the way in which Wolji Pond was renovated into a secret garden of sorts may have been influenced by the Three Courts (三朝, K. samjo) layout (Hanyuan Hall, Xuanzheng Hall, Zichen Hall) and Taiye Pond (太液池) of the Tang Dynasty’s Daming Palace (Yoon Mubyeong 2010; Kim Sangtae 2014; Lee Hyuntae 2020a). In particular, Building A is certain to have served as the throne hall (正殿, K. jeongjeon) since pillars are absent in an area measuring four rooms located in the center of the building. Stone features that appears to form the “king’s path” (踏道, K. dapdo) were also laid out in front of the pair of stairs located to the front of the building. Therefore, Building A and its associated central courtyard can be interpreted as having formed the Outer Court (外朝, K. oejo); Building B and its central courtyard – the Middle Court (中朝, K. jungjo); and Building C and its central courtyard – the Inner Court (內朝, K. naejo).

Fig. 3. A bird's-eye view of the building complex located to the west of Wolji Pond (reconstruction based on excavation)

Fig. 4. Layout of the reconstructed building complex located to the west of Wolji Pond

Building A and its central courtyard is likely to have been the place where joha (朝賀) ritual assemblies and state ceremonies, such as the meeting of foreign dignitaries, took place. Buildings B and C were where daily administrative affairs were dealt with, and also provided a place for banquets with officials and generals when needed (Lee Hyuntae 2020a; 2020b). In other words, the building complex located to the west of Wolji Pond functioned as a place where important state rituals and events took place, as well as a place of royal administration and control. The nature of the buildings located around Wolji Pond can be further interpreted in diverse ways according to how the palace’s “internal quarters” mentioned in Historical account B-2 are understood.

The palace’s internal quarters are generally thought to have been comprised of a pyeonjeon (便殿, the main office where the King undertook administrative duties), a chimjeon (寢殿, where the daily routines of the King took place), buildings where the Queen Consort and concubines resided, and buildings for those who served the King, Queen Consort, and concubines, as well as the court ladies (Jeon Deukjae 2009). Either Building B or C, located to the west of Wolji Pond, is highly likely to have served as the pyeonjeon. In this way, the throne hall, the main office, and the buildings central to the running of the kingdom were all located in the vicinity of Wolji Pond. However, if this is the case, then why did King Munmu decide to situate such an important building complex outside the walls of Wolseong Fortress? One possibility is that the area within Wolseong Fortress was already highly built up, making it difficult to find room for a plot of land on which large buildings sharing a common, north-south axis could be built. Then what may have been the reason that—of all the plots of land that could have been used following the expansion of the palace complex beyond the walls of Wolseong Fortress— such an important building complex was built on the plot of land located to the south of Wolji Pond? Particular reasons can be further explored in association with Historical accounts A-1, A-2, and B-1.

Prior to the construction of the building complex located to the west of Wolji Pond in the late seventh century, Dongji Pond (i.e. Wolji Pond) was known as a place where a dragon appeared (Historical accounts A-1 and A-2), and Historical account B-1 indicates that this belief persisted into the end of the eighth century. The dragon is widely known as a Water God (水神, K. susin), controller of water, as well as a symbol of the King. Therefore, the reason that the King’s main area of activity (represented by the Three Courts) was located next to Dongji Pond may have been due to the belief that a dragon, imbued with the spirt of kings, appeared and resided at that spot. In addition, the establishment of an arena for state rituals and activities, as well as a place where the administrative duties of the King were carried out, at a location where the spirt of kings was present in this manner may have functioned as a means of displaying the enhancement of royal authority and the might of the newly unified kingdom.

In this process of augmenting royal authority and laying down the foundations of the kingdom’s “state protection” (護國, K. hoguk) ideology, pre-existing perceptions of the “dragon of Dongji Pond” may have been actively utilized. This is reflected in Historical account B-1, in which the dragon residing in Dongji Pond is perceived as a “State Protector Dragon” (護國龍, K. hogukryong). It is along this same line of thought that the use of Wolji Pond as a ceremonial stage for the “Dragon King Ritual” (龍王祭, K. yongwangje) can be understood. This is evidenced by the discovery of several pottery vessels featuring the inscription “Sinsim Yongwang” (辛審龍王) at Wolji Pond (i.e. Dongji Pond) (Fig. 5) and the establishment of the “Office of the Dragon King” (龍王典, K. yongwangjeon) as one of the government offices belonging to Donggung Palace (東宮官, K. Donggungkwan). It must also be pointed out that such an arrangement would have conveyed to those participating in the Dragon King Ritual the message that Wolji Pond and its environs represented the heart of the Silla Kingdom, which was imbued with the spirit of kings and where the State Protector Dragon resided.

Fig. 5. A pottery vessel featuring the inscription “Sinsim Yongwang” (辛審龍王) recovered from Wolji Pond

The fact that the dragon was regarded as a Water God, controller of water, should also be borne in mind when making interpretations on the area around Wolji Pond. Analysis of the paleogeography of the Gyeongju Basin has revealed that a considerable portion of the land around Wolseong Fortress would have been waterlogged due to streams fed by erupting springs (Fig. 2). Therefore, the infilling of waterlogged land and the establishment of earthen foundations was required in order to utilize the area around Wolseong Fortress. The remains of such earthen foundations have been identified in the area east of Wolji Pond, as well as around Building A in the area to the west. The process of infilling waterlogged land and establishing earthen foundations represents, in a sense, the process of managing and controlling water, all of which was overseen by the Silla King at the pinnacle of the administrative pyramid.

The dragon is generally regarded as a being that tends to live in the water, a controller of wind, rain, lightening, and thunder, and a creator of water (風雨雷震, K. pungunoejin) (Cho Beopjong 2002). In addition, the dragon mentioned in the Historical accounts A-1, A-2, and B-2 can be classified as a “pond dragon” (池龍, K. jiryong), which was regarded as a powerful being associated with the construction of ponds and the management of water (Kang Yeonggyeong 2002). Taking these facts into consideration, it is possible to suggest that the act of refurbishing the pond (where the dragon resided) with a stone retention wall, as well as the establishment of an area around the pond where key activities of the King took place, can be regarded as a symbolic act of transferring the dragon’s quality of being a Water God, controller of the water, to the King. This is particularly likely, given that King Munmu is said to have relayed in everyday contexts his hopes “to become a ‘great dragon, protector of the state (護國大龍, K. hogukdaery-ong)’ after death, upholding Buddhist ways and guarding the kingdom” (as recorded in the Samguk Yusa).

The development of waterlogged lands around the “dragon palace” (龍宮, K. Yongkung) and the establishment of Hwangnyongsa Temple (皇龍寺) in the mid-sixth century was an act of great symbolism, intended to strengthen royal authority (Yang Jeongseok 2004). The establishment of the building complex to the west of Wolji Pond in the late seventh century, in contrast, can be understood as a proclamation, for all to see, of the King’s confidence in being able to control the waters as he saw fit. In addition, the construction of grand buildings—such as the throne room—in an area that had been reclaimed through control over water, can be interpreted as an act imbued with the intention to display the increasing royal authority of the time. King Munmu’s establishment of the building complex to the west of Wolji Pond can also be implied as an attempt to transfer the image of the dragon, as a symbol of the King and a Water God with control over water, to the King. To summarize, if Hwangnyongsa Temple was used to express the strengthening of royal authority in Silla’s Middle Ancient Period (中古期, K. Silla junggogi), then from the late seventh century, the building complex located to the west of Wolji Pond was used to express the newly established royal authority of Silla’s Middle Period (中代, K. Silla jungdae).

The Site of the Building with Front and Rear Toekan Excavated in the Balcheon Stream Area and the Nature of the Archaeological Features in Its Environs

As discussed above, the overall renovation of Wolji Pond in 674 CE and the building of a palace complex that followed the Tang Three Court (三朝, K. samjo) system next to Wolji Pond in 679 CE was a planned large-scale construction event. The spatial range of this planned construction was, however, not limited to the area around Wolji Pond, as can clearly be seen through the excavations that have recently taken place at the Balcheon Stream site (Fig. 6).

A stone waterway feature was unearthed during excavations at the Balcheon Stream site. It is believed to continue eastwards, eventually connecting to the water opening (出水口, K. chulsugu) of Wolji Pond (Fig. 6 and 7). A large-scale earthen foundation layer was identified around the base of the stone waterway feature, and the way in which the stones were stacked was found to be similar to that of the stone retention wall of Wolji Pond, as well as the stone moat of Wolseong Fortress. This indicates that the Balcheon Stream site is associated with the renovation works that took place in the area of Wolji Pond and its environs from 674 to 679 CE. This possibility is further backed by the discovery of stamped-design pottery sherds and assorted roof tiles within the earthen foundation layer of the site.

Fig. 6. The Balcheon Stream site under excavation

In addition to the stone waterway feature, there are several other archaeological features identified at the Balcheon Stream site that are worthy of attention. Of particular interest are the remains of a stone bridge that measured over 11.2 meters in length, located to the west of the presumed site of the North Gate (北門址, K. bukmunji) of Wolseong Fortress (where the Stone Ice Storage currently sits), and a road (paved with fine gravel) that leads off from both sides of the bridge in a north-south direction (Fig. 7). The width of this road measures approximately twenty meters in the area north of the stone bridge site. Excavated at a position located directly to the north of the stone bridge was a single foundation stone (round in cross-section) and two stone-packed timber supports which may possibly be interpreted as the remains of a gate. Until now, the North Gate of Wolseong Palace has been regarded to have functioned as the main entrance (Yeo Hogyu 2002). As such, it is hoped that the investigation of the remains of the stone bridge and the road extending in a north-south direction will help to answer questions regarding the main entrance of Wolseong Fortress.

Fig. 7. Remains of a stone bridge and stone waterway unearthed at the Balcheon Stream site

Another type of interesting archaeological feature excavated from the Balcheon Stream site is a building, represented by stone-packed timber supports, located in Area 10 (Figs. 8 and 9). This building is aligned along a north-south axis, and currently measures four rooms in width and three rooms in depth. However, given that this archaeological feature has been found to extend to the north (in the non-excavated area), it is likely that the true dimensions of the building extended beyond four rooms in width (Fig. 9). What is of particular interest is the fact that this building features both a front and rear toekan (前後退間). In the Gyeongju region, such front and rear toekan buildings have only been identified at the sites of temple buildings where Buddha statues were once placed (such as the site of the Main Hall of Gameunsa Temple or the sites of the Vairocana Buddha Hall [Birojeon] or Avalokitesvara’s Shrine [Gwaneumjeon] at Bulguksa Temple). In later periods, front and rear toekan buildings were frequently used as the Confucian shrine halls of hyanggyo (鄕校, government-run provincial Confucian schools) of the Joseon Dynasty. The Yeongnyeongjeon (永寧殿) building of Jongmyo, the Confucian shrine of Kings and Queens of the Joseon Dynasty, is also a front and rear toekan building. Such front and rear toekan buildings were usually designed so that they are greater in width than depth, resulting in a grand appearance when viewed from the front. This layout also provides greater space that can be used for worship or rituals.

Fig. 8. A building site with stone-packed timber supports and other surrounding archaeological features from Area 10 of the Balcheon Stream site

Fig. 9. Layout of the building with stone-packed timber supports and other surrounding archaeological features from Area 10 of the Balcheon Stream site

In front and rear toekan buildings, the front toekan section is mainly used as the area where bowing takes place, and the object of worship —be it a statue of Buddha or the tablet of the person being commemorated —is generally placed in the rear toekan section. It can be expected that a similar use of space took place at the front and rear toekan building excavated at the Balcheon Stream site. Given that no evidence linking the remains of the Balcheon Stream site to a temple complex has yet to be found, it appears more plausible to interpret the front and rear toekan building as a place where rituals took place. The nature of the archaeological features unearthed around the building provides support for this possibility.

Located directly to the south of the southeast stone filled timber support of the front and rear toekan building is a well (Fig. 10), the base of which was made using the sherds of large jar vessels. Such an example in which large jar vessel sherds were used to form the base of a well is rare. The well is not deep, measuring just over two meters. In addition, the well was located very close to one of the building’s stone-packed timber supports, which is a very unusual occurrence. These lines of evidence suggest that the well may have served a special function, rather than being a mere supplier of water. Both the Jongmyo Shrine (宗廟) and Sajikdan Altar (社稷壇) of the Joseon Dynasty feature wells that provided sacred water (明水, K. myeongsu) used during rituals or the water used to prepare the ritual offerings of food placed in front of ancestral tablets (神廚, K. sinju). Using this later practice as an example, it may be presumed that a well located in such a close proximity to the front and rear toekan building also served a similar function.

Fig. 10. The well unearthed at Area 10 of the Balcheon Stream site

A stone structure and a drain paved with tiles (Fig. 11) located close to the southeast section of the front and rear toekan building is also of interest. This stone structure consisted of a round stacked stone feature which had been built upon a rectangular stone foundation feature piled along the same axis as the tiled drain. As such, the stone structure may at first appear to be a well, but the presence of a hidden drainage facility located beneath the stone structure, which connects to a tiled drain, makes it possible to infer that water passing through the tiled drain would have reached the adjacent pond or have been fed into the stone waterway. It is therefore likely that the stone structure functioned as the entrance of an artesian spring, which acted as a water source (Jeong Taeyeol 2021). In addition, it appears that there was a tile roofed structure or house located above the stone structure and tiled drain based on the presence of three to four stone-packed timber supports placed in a symmetrical layout. These stone-packed timber supports were constructed by digging into the deliberately infilled earthen layer, which indicates that the tile roof structure was built around the same time (c. 674-79) as the stone structure and tiled drain that it was meant to protect.

Fig. 11. The stone structure and tiled waterway unearthed at Area 10 of the Balcheon Stream site

Interestingly enough, a similar structure was discovered during the excavation of a garden pond at the site of Asuka-kyo (the Capital of Asuka, 飛鳥京) in Japan. In 2019, a stone structure that functioned as the entrance of an artesian spring, a tiled waterway through which the spring water flowed, and a wide stone paved area flanking the tiled waterway, etc. were unearthed in the northeast corner area of the north pond (which formed part of a garden pond) of Asuka-kyo (Fig. 12). The fact that such a stone structure was installed at the location where spring water erupted, as well as the fact that this spring water was conveyed to the pond via a tiled waterway, are both similar in terms of outward appearance and function to the facilities identified at the Balcheon Stream site. Further investigation has revealed that the stone retention wall of the north pond and the stone structure were both constructed around the mid-seventh century, when the north pond was first built. The tiled waterway and wide stone paved area which flanks it, on the other hand, were laid out during the renovations that took place in the late seventh century. It is important to note that these archaeological features from Asuka-kyo have also been interpreted as being associated with water-related rituals (Archaeological Institute of Kashihara, Nara Prefecture 2019; Jeong Taeyeol 2021).

Fig. 12. The stone structure and tiled waterway associated with the north pond of Asuka-kyo

The fact that archaeological features similar to the “stone structure and tiled waterway” identified at the Balcheon Stream site have also been interpreted as having been used for water-related rituals adds support to the claim, proposed above, that the front and rear toekan building may have also been associated with rituals. If the front and rear toekan building can be seen as having served a ritual purpose, and if it can be identified as having been located within the newly expanded area of the palace complex, then it may be possible to suggest that the building functioned as a facility where state rituals took place. However, due to the fact that the front and rear toekan building faces to the east, it cannot be said that this single building alone would have formed the arena where state rituals took place. It is more likely that, as in the case of the building complex located to the north of Gyerim, there was originally a building complex consisting of buildings laid out in a ‘⊓’-shaped manner, of which only the western section has been partially revealed through excavation.

The results of the investigations undertaken at the Balcheon Stream site thus far suggest that the front and rear toekan building, along with the well, stone structure, and tiled waterway, etc. located nearby, represent state ritual facilities. If this is indeed the case, then King Munmu’s act of “renovating the palaces” as relayed in the Samguk Sagi, involved not only the large-scale maintenance of the palace buildings but also the construction of state ritual facilities. This can be seen to reflect the process of renewing the place of royal control and administration as well as establishing ritual institutions in the late seventh century following the active adoption of Tang (唐) culture.

Conclusion

The spatial rearrangement of the area surrounding Wolseong Fortress, as well as the area of Balcheon Stream, was not merely limited to renovating existing palace buildings to enhance their splendor and to constructing Donggung Palace. In the late seventh century, not only did King Munmu initiate the construction of large buildings displaying authority (including the Throne Room) along a north-south axis in the area around Wolji Pond, but he also instigated the expansion of the sphere of royal formal activity to areas beyond the walls of Wolseong Fortress, by establishing state ritual facilities in the northern part of the Balcheon Stream area. This was an act of great significance as it represented the removal of the central axis of government administration from the area within Wolseong Fortress to the area around Wolji Pond. The establishment of state ritual facilities in the area of Balcheon Stream is also of great significance. As is well known, the key ritual facilities of the state (such as the ancestral shrine for the Kings and Queens of the state and the altar where rituals for the Gods of the Earth and Crops took place) were also located outside the boundaries of the Imperial City (皇城, K. Hwangseong) in the Tang Dynasty. King Munmu, on the other hand, placed such state ritual facilities within the boundaries of Silla’s capital city (in the northern part of the Balcheon Stream site) when rearranging the area surrounding Wolseong Fortress. This indicates that King Munmu’s active adoption of Tang cultural elements and institutions in the late seventh century was based on an innovative adoption of the Chinese Capital City System in which changes were made so that this foreign system could fit better with the conditions of Silla.

Additionally, it is of interest to note, in association with the fact that the spatial reorganization of the area around Wolji Pond and Balcheon Stream took place in the late seventh century, that the fully-fledged development of the area to the south of Hwangnyongsa Temple took place from around the eighth century. Of the sites found in the area of Hwangnyongsa Temple and its environs, only Hwangnyongsa Temple itself, Bunhwangsa Temple, and the garden pond at the Guhwang-dong site can be dated to before the late seventh century, which is an interesting phenomenon. The plaza-like road (measuring fifty meters in width) located to the south of Hwangnyongsa Temple is believed to have been laid out as part of the same development plan which involved the construction of Building 26 of Area Ga (located to the east of Wolji Pond and interpreted as the East Gate of the “Donggung Palace and Wolji Pond” complex), but it has been dated to the turn of the eighth century. In other words, the time of its construction was around twenty years later than that of both Building 26 of Area Ga, located to the east of Wolji Pond, and the building complex located to the west of Wolji Pond. Artifacts that date to before the seventh century have also rarely been discovered in the area to the south of the plaza-like road. This indicates that fully-fledged development of the area to the south of Hwangnyongsa Temple into an urban landscape took place from the eighth century. The above sequences of development make it possible to suggest that the spatial rearrangement of the area around Wolji Pond also acted as an impetus for the urban transformation of the area south of Hwangnyongsa Temple. This, therefore, may also be understood as yet another aspect of the significance of the spatial rearrangement of the area surrounding Wolseong Fortress in the late seventh century.

Fig. 13. The area outside Wolseong Fortress reorganized in the late seventh century, including Donggung Palace, Wolji Pond, and the Balcheon Stream site