The Journal of Korean Art and Archaeology

I. Introduction

Silla headband crowns with tree-shaped uprights (sujihyeong daegwan, 樹枝形帶冠, Fig. 1) have been the subject of much debate in Korean archaeology. This paper examines fifty-five such crowns, including gold, gilt-bronze, and bronze examples dating from the Maripgan period (356-514 CE) and junggogi period (中古期, 514-654 CE). Not all of the crowns were examined firsthand, but all could be typologically studied based on information obtained through various media. Silla crowns can be regarded as a form of public ornamentation that served as an emblem of the wearer’s social status through their shape, metal type, and decoration, all of which were highly regulated according to institutionalized rules. Hence, each crown type (including the headband crown with tree-shaped uprights) is assumed to have embodied a certain symbolic system, and moreover, to have reached its quintessential form during the period when that symbolic system existed. Accordingly, the decline of each Silla crown type can be associated with the emergence of a new symbolic system. Therefore, the development of each crown type can be divided into the three stages of “prototype,” “standard,” and “regression.” Additionally, Silla crowns can be regarded as a material expression of the symbolic system to which the wearer adhered and reproduced prior to death. Such an interpretation assumes that all Silla crowns were prestige items worn by the living, a contentious point that remains the subject of lively debate. This, in turn, allows the wearer’s year of death to be established as the terminus ante quem of the crown’s manufacture. This study takes the above functions and characteristics into account in charting the development of Silla headband crowns with tree-shaped uprights.

Fig. 1. Parts of the headband crowns with tree-shaped uprights.

II. Chronology of Typological Groups of Silla Headband Crowns with tree-shaped Uprights

The dates in the chronological scheme of Silla headband crowns with tree-shaped uprights come from Silla tombs, particularly Geumgwanchong Tomb (Gold Crown Tomb), Cheonmachong Tomb (Heavenly Horse Tomb), and the south mound of Hwangnamdaechong Tomb. The dating of the south mound of Hwangnamdaechong Tomb is a central issue in Silla archaeology, as there has been disagreement regarding the identity of the occupant of this “royal tomb.” Different studies have argued that the tomb may be occupied by one of three possible Silla kings: Naemul Maripgan (d. 402) (Lee Heejoon 2010, 192-206); Silseong Maripgan (d. 417) (Ham Soonseop 2010, 226-45); or Nulji Maripgan (d. 458) (Kim Yongseong 2010, 207-25). I have argued that the south mound of the tomb is occupied by Silseong Maripgan, based on the relative abundance of Goguryeo objects found within, which would seem to date the tomb to the first quarter of the fifth century CE.1 Other studies have asserted that Namul Maripgan actually lies in Gyo-dong Tomb 119 in Gyeongju, as described in the “Chronology of the Kings” (wangryuk, 王曆) chapter of Samguk Yusa (Ham Soonseop 1996, 371-2; Kim Yongseong 1996, 119-21).

Objects demonstrating interactions between Silla and Baekje have also been useful in dating the tombs. For example, a Silla-style, three-leaf openwork belt ornament found at Songsan-ri Tomb 4 in Gongju was identical to one found in Geumgwanchong Tomb. This exchange of objects between the two kingdoms was originally believed to have taken place after the Baekje capital was moved from Hanseong to Ungjin in 475 CE (Lee Hansang 1998, 29-30). However, given that Silla and Baekje established a second alliance in 433 CE, they may conceivably have begun to conduct exchanges around that time. Objects reflecting such exchange have also been used to date Cheonmachong Tomb. Specifically, metal objects from Cheonmachong Tomb have been found to match those discovered in the Baekje Tomb of King Muryeong (d. 525), thus allowing Cheonmachong Tomb to be dated to the early sixth century (ibid., 30). Finally, the conclusion of the chronology of Silla headband crowns with tree-shaped uprights is provided by crowns from the “regression” stage that have been excavated around Yeoju and Paju in Gyeonggi Province. These crowns date to no earlier than 553 CE, when Silla began to expand into the Han River region (Ham Soonseop 2000, 81).

To use typological analysis to chart the development of headband crowns with tree-shaped uprights, both the “type” and “technical pattern” of the crowns must be considered. Type attributes were based on the outer appearance of the crowns, particularly the number of branches on the vertical uprights with symmetrical branches and the composition of the uprights (i.e., vertical with symmetrical branches[opposite branches] vs. angled with alternating branches[alternate branches]). As such, the Silla headband crowns with tree-shaped uprights can be divided into four types, presented in Fig. 2 and Table 1: vertical uprights with symmetrical angled branches (Type A); vertical uprights with 90-degree branches (Type B); vertical uprights with slightly angled branches and additional forms and embellishments (Type C); and vertical uprights with connected branches (Type D). Over time, the number of tiers of branches on the uprights generally increased, from one to three to four, with some coexistence between three- and four-tiered uprights. As such, the number of branches on the vertical uprights generally increased over time (Fig. 3). The composition of the upright ornaments developed in the following manner: first, three vertical uprights with symmetrical branches (sub-type “3/0”); followed by three vertical uprights with symmetrical branches and two angled uprights (“3/2”); and then vertical uprights with symmetrical branches only (“5/0,” “4/0,” “3/0”).

Fig. 2. Comparison of the number of vertical uprights with symmetrical branches.

Fig. 3. Comparison of vertical uprights with symmetrical branches of Silla headband crowns with tree-shaped uprights, (from the left) ❶ Gyo-dong (Gyeongju) Tomb ❷ North Mound of Hwangnamdaechong Tomb ❸ Geumgwanchong Tomb ❹ Seobongchong Tomb ❺ Cheonmachong Tomb ❻ Geumryeongchong Tomb.

In order to examine the development of the headband crowns with tree-shaped uprights more closely, some of the broader chronological phases were further subdivided based on technical patterns derived from production techniques. Diachronic changes in the technical patterns of the crowns can be most clearly observed in the surface patterns of the uprights. Uprights with the fewest branches have plain surfaces, but as the number of branches increases, so does the degree of decorative patterning of the surface. This may have been a deliberate technological development, as more patterning led to a corresponding increase in the curvature of the surface, which in turn helped the central upright sustain the increased weight of more branches. A total of nine different technical patterns were identified (“a” through “i”). The types and technical patterns of the Silla headband crowns with tree-shaped uprights were thus combined to form a total of six “typological groups” that were ordered chronologically. “Aberrant” examples of these typological groups were also identified. The results of this analysis are presented in Tables 1 and 2 and Figs. 4 and 5.

• Image Download
• Full Size

Table 2. Formation of typological groups through the combination of “types” and “technical patterns”.

Fig. 4. Chronology of Silla headband crowns with tree-shaped uprights.

“Typological group I” consists of “A1 (3/0)” and “A3 (3/0)” type crowns with “a,” “b,” and “c” technical patterns; its date can be established through comparisons with the crown from the south mound of Hwangnamdaechong Tomb. Crowns of the “a” technical pattern (from Gyo-dong Tomb in Gyeongju and Bokcheon-dong Tomb 11 in Busan) have distinctively varied styles, but a more prevalent common style soon emerged with “A3 (3/0)” type crowns that are associated with all three technical patterns (a, b, c). Since subsequent crowns were based on this common style, the crowns of this typological group can be regarded as the “prototypes” of the Silla headband crown with tree-shaped uprights. The appearance of “Typological group 1” can be dated to around the end of the fourth century, since artifacts found in association with the gold crown at Gyo-dong Tomb are contemporaneous with or slightly later than Wolseong-ro Tomb Ga-13 (Lee Hansang 2000, 102-5). The concluding date of “Typological group I” can be estimated by cross-dating between Imdang-dong Tomb 7 and the south mound of Hwangnamdaechong Tomb. The protective stones of the Imdang-dong tombs show that Tomb 7A was constructed first, followed by Tomb 7B, and then Tomb 7C. Also, mounted cups found inside Tomb 7B generally appear to be earlier than those from the south mound of Hwangnamdaechong Tomb (along with a few contemporaneous examples). Thus, the Imdang-dong crown is estimated to have been produced no later than the early first quarter of the fifth century, the period just prior to the construction of the south mound of Hwangnamdaechong Tomb. However, the time period of “Typological group I” crowns may extend a little beyond this point, to around the time of the construction of the south mound of Hwangnamdaechong Tomb. For example, pottery contemporaneous with the south mound of Hwangnamdaechong Tomb has been found in Hwango-dong 100 Beonji Tomb 1 and other tombs containing crowns of this typological group. “Typological group I” crowns are found clustered around Gyeongju, as well as in Dongrae in Busan to the south and Gyeongsan to the west.

“Typological group II” is represented by five gilt-bronze headband crowns with tree-shaped uprights from the south mound of Hwangnamdaechong Tomb. A time gap, albeit insignificant, can be observed in the production period of these crowns, with the crown worn by the deceased [“B3 (3/0)” type] having been manufactured after the “A3 (3/0)” type crowns that were placed as grave goods. The conclusion of the time period for “Typological group II” crowns can be estimated based on the identity of the deceased entombed within the south mound. These crowns appear to have been made around the same time or slightly later than the “A3 (3/0)” type crowns of “Typological group I.” The time period for these crowns ends slightly after the construction of the south mound of Hwangnamdaechong Tomb, based on pottery from the second and third cists of Hwango-dong Tomb 16. “Typological group II” crowns are found clustered in Gyeongju and also in Gyeongsan. They have also been found along the middle reaches of the Nakdong River, in Daegu and Dalseong, which was the western frontier of Silla territory at the time.

Although they were manufactured at slightly different times, the crowns of both “Typological group I” and “Typological group II” appear to have been used as grave goods around the same time. The crowns of these two typological groups can thus be considered together as representing the “prototype” stage of the Silla headband crowns with tree-shaped uprights, which spans from the end of the fourth century to the first quarter of the fifth century. A key feature of this stage is the rapid establishment of a single common style.

Fig. 5a. Gold crowns of Silla, Gyo-dong (Gyeongju) Tomb.

Fig. 5b. Gold crowns of Silla, North Mound of Hwangnamdaechong Tomb.

Fig. 5c. Gold crowns of Silla, Geumgwanchong Tomb.

Fig. 5d. Gold crowns of Silla, Seobongchong Tomb.

Fig. 5e. Gold crowns of Silla, Cheonmachong Tomb.

Fig. 5f. Gold crowns of Silla, Geumryeongchong Tomb.

“Typological group III” is represented by the gold headband crowns with tree-shaped uprights from Geumgwanchong Tomb and the north mound of Hwangnamdaechong Tomb. Typologically, the crown from the north mound of Hwangnamdaechong predates the Geumgwanchong crown, even though the Hwangnamdaechong crown has a “later” technical pattern than the Geumgwanchong crown. Notably, the Hwangnamdaechong crown may have required more patterning to help sustain its wider uprights with symmetrical branches (Fig. 6). In addition, the north mound of Hwangnamdaechong Tomb was found to contain Goguryeo ornaments, whereas only Silla ornaments were found in Geumgwanchong Tomb. “Typological group III” crowns are found throughout the distribution area of “Typological group II” crowns, with the addition of Changnyeong, along the Nakdong River. The beginning of the time period for this typological group can be estimated based on a three-leaf openwork belt buckle found inside Songsan-ri Tomb 4 in Gongju, which resembles one from Geumgwanchong Tomb. This tomb was constructed sometime after 475 CE, when the Baekje capital was moved to Ungjin (present-day Gongju), and this date has been used previously to cross-date the Geumgwanchong Tomb. However, the three-leaf openwork belt buckle could have reached Baekje when the second alliance between Silla and Baekje was established in 433 CE, which would push back the earliest date for these crowns. Therefore, the crown from the north mound of Hwangnamdaechong can be dated to the second quarter of the fifth century and the Geumgwanchong crown to the third quarter of the fifth century. “Typological group III” can be seen to represent the establishment of the “standard” phase of the Silla headband crowns with tree-shaped uprights.

Fig. 6. Comparison of gold headband crowns from Cheonmachong Tomb and Geumgwanchong Tomb (Cheonmachong crown in the foreground)

“Typological group IV” is represented by the gold headband crowns with tree-shaped uprights from Seobongchong Tomb, Cheonmachong Tomb, and Geumryeongchong Tomb. The Seobongchong gold crown is of the “B3 (3/2)” type, but is similar to the Cheonmachong crown in terms of technical pattern and the style of the associated pottery. In terms of chronology, the Seobongchong crown is the earliest of the three, followed by the Cheonmachong crown, and then the Geumryeongchong crown. This group also includes the gilt-bronze crown from Tomb 55 of the Dalseong Tumuli Group in Daegu and other examples from Yangsan. This typological group can be dated from the fourth quarter of the fifth century to the early first quarter of the sixth century, based on the cross-dating of similar ornaments from Cheonmachong and the Baekje tomb of King Muryeong (Lee Hansang 1998, 30). “Typological group IV” represents the stage in which the “standard” Silla headband crown with tree-shaped uprights developed, although some crowns from this group show divergent elements. In addition, this group also includes some regional imitations of the headband crowns with tree-shaped uprights, such as the crown from Chodang-dong Tomb B-16 in Gangneung and two crowns from Joil-ri Tomb 49-2 in Ulsan, which are characterized by technical pattern “i.” These three regional crowns deviate from the standard developmental trend of increased size and increased elaboration through the use of more spangles. Both Ulsan and Gangneung were occupied by Silla at an early date, so the appearance of imitations of Silla headband crowns in these regions would seem to indicate that both areas were quick to adopt and imitate Silla culture.

“Typological group V” is represented by two crowns from the Dongwon Collection and Byun Jongha Collection (respectively) of the National Museum of Korea. Both crowns were donated to the museum by private collectors, and the precise details of their origin and excavation are currently unknown. The chronology of this typological group can be ascertained from Eunryeongchong Tomb, Houchong Tomb, and Bubuchong Tomb (“Husband-and-Wife Tomb”) in Bomun-dong, all of which have yielded crown fragments that are similar to the crown from the Dongwon Collection.2 Thus far, it has not been possible to determine a relative chronology for this group, as only two of the crowns are completely intact. Notably, both of the intact crowns have side horns added to the peak of the central tree-shaped upright. However, the crown from the Byun Jongha Collection has five tree-shaped uprights and no angled uprights, which is similar to the “Typological group VI” crown from the Sanno Futakoyama Kofun Tomb in Maebasi, Gunma Prefecture, Japan. To date, only gilt-bronze crowns have been identified with this typological group, which includes the largest and most opulent of Silla crowns. These crowns have wide headbands, measuring 5-6 cm, that were affixed using rivets, possibly for ease of wear. The uprights overlap, giving the crowns a long cylindrical form, and in some cases, thin metal wires were used to connect the neighboring uprights. The large size and opulent nature of the “Typological group V” crowns cannot necessarily be regarded as a development, as it involved the erosion of the standard form of the Silla headband crowns with tree-shaped uprights. This erosion eventually resulted in the production of overly simplified crowns that maintained only the basic structure of the ceremonial headwear. As such, this typological group represents the beginning of the “regression” stage of the Silla headband crown with tree-shaped uprights. Typological analysis of pottery discovered within the tombs indicates that Eunryeongchong Tomb is the earliest of the three, followed by Houchong Tomb, and finally Bubuchong Tomb (Lee Juheon, Lee Yonghyun, Yoo Hyeseon 2006, 73-4; Gyeongju National Museum 2011, 65-82; Yoon Sangdeok 2011, 132-9). Based on the formal modifications, the “Typological group V” crowns likely postdate the crown from Geumryeongchong Tomb. However, the associated grave goods suggest that this typological group actually may be contemporaneous with Geumryeongchong Tomb. For example, the three-leaf openwork belt buckle that was discovered with the Dongwon crown is similar to one found in Geumryeongchong Tomb, and the pottery from Eunryeongchong Tomb may also belong to a similar phase as that from Geumryeongchong Tomb. The conclusion of the time period of this typological group can be established by the appearance of “Typological group VI” crowns in the middle and lower reaches of the Han River region. Therefore, “Typological group V” is estimated to have lasted from the late first quarter of the sixth century to the second quarter of the sixth century. This group also includes the crown from Gacheon-dong Tomb 168 in Daegu, a regional imitation and an aberrant example, which is of the “B3 (3/0)” type. This regional crown resembles one of the crowns (②) from Joilri Tomb 49-2 in Ulsan, except that it has inscribed angular patterns along the spine of the uprights with symmetrical branches. Although Gacheon-dong now belongs to the administrative district of Daegu, it was originally part of the Gyeongsan region, which was occupied by Silla at an early date.

“Typological group VI” demonstrates the deterioration of the Silla headband crowns with tree-shaped uprights, as only the basic structure of the crown is expressed. Only the vertical uprights with symmetrical branches remain, as the branches have been vertically merged together. In addition, the technical patterns of the crowns show a wide variance, with the only shared traits of technical pattern “i” being the presence of convex saw-tooth patterns and the use of metal rivets and staples for the headband. Therefore, this typological group represents the diminishment of typological developments of the Silla headband crowns with tree-shaped uprights.

Among this group, only the crown from Sanno Futakoyama Kofun Tomb in Maebasi, Gunma Prefecture, Japan, which is of the “D4 (5/0)” type, shows a continuation with the preceding typological group. However, since this crown was not found in former Silla territory, its production site and circumstances have yet to be determined. The crown may have accompanied Silla immigrants to Japan or may have been sent there after having been produced in the Silla capital or an outlying region. It may even have been produced in Japan, although that seems unlikely, given that it shows a continuation of the typological trajectory of the Silla headband crowns with tree-shaped uprights. For example, like its predecessors, the crown has five vertical uprights with symmetrical branches and a headband decorated with convex sawtooth patterns. Given these details, and the overall maintenance of standard shape and form, the crown was probably produced in the Silla center and then sent to Japan. Notably, even as the headband crowns with tree-shaped uprights lost their symbolic power in the Silla center, they would certainly have retained great significance for Silla people residing in the Japanese Archipelago.

This typological group also includes “D3 (4/0)” type crowns, which show regression in all aspects. Like the Sanno Futakoyama crown, the crown from Ji-dong Tomb 2 has peaks on the branch-ends of the uprights. Evidence indicates that the crown from Maeryong-ri Tomb 5 in Yeoju also resembles the Ji-dong crown, although only the headband of the former remains. The bronze crowns from Ha-ri in Danyang and Chuam-dong District B Tomb Ga-21, both of which are “D4 (4/0)” type, have uprights with triangular spaces cut along the sides to express the branches, much like the crown from the stone-chamber tomb of Seongdong-ri, Paju. In the case of the Ha-ri crown, the peaks have disappeared from the branchends of the uprights, indicating that it is relatively earlier than the other crowns of this type, but this initial assumption is contradicted by the presence of simple protrusions to express the branches. Based on this example, the regressive features of these crowns cannot be used to establish an accurate chronology, but should be considered case by case. A crown found in Ullungdo Island features uprights similar to those of the Ha-ri crown (Choi Monglyong, et al, 1998, 126-30). As the headband crowns with tree-shaped uprights do not adhere to any standardized form, this typological group represents the culmination of the “regression” stage. Apart from the Sanno Futakoyama crown, the gilt-bronze crowns of this typological group have all been found in frontier territories that were newly annexed by Silla during the reign of King Jinheung (r. 540-576). On the other hand, bronze crowns from this group have been found in regions that were already part of Silla territory. The gilt-bronze “D3 (4/0)” type crowns can be dated to the third to fourth quarter of the sixth century, while the bronze “D4 (4/0)” type crowns have been found in association with pottery dating to the early seventh century (Yoon Sangdeok 2010, 137; 2011, 137).

III. Development and Historical Context of Silla Headband Crowns with tree-shaped Uprights

Based on the estimated production dates, Silla headband crowns with tree-shaped uprights seem to have been used for a period of approximately 250 years. They can be divided into three stages (“prototype,” “standard,” “regression”), which can be further divided into the following five sub-stages: “prototype”; “establishing the standard”; “developing the standard”; “onset of regression”; and “culmination of regression.” The transition between these stages may have occurred in association with changes in the contemporaneous symbolic structure, which will now be examined.

The institutionalized use of specific objects to signify and maintain social stratification can be traced back to the Samhan Confederacy (early first century BCE-late fourth century CE). Records show that, around the second quarter of the third century, the rulers of Samhan received seals attached with sashes (印綬) from China through exchange, and that these seals were used as symbols of authority. Even members of agricultural villages (下戶) are said to have come up with their own seals with attached sashes and official hats and robes (印綬衣幘). In Samhan, the rulers of guk polities were divided into the “eup leader” (eupgun, 邑君) and the “eup head” (eupjang, 邑長), with the “peasants” (haho, 下戶) below them; these different ranks were associated with different types and styles of clothing. In both Buyeo and Goguryeo, clothing and personal ornaments made from precious metals were also used to denote rank. Thus, the regulations adopted in the Maripgan period regarding clothing and rank would seem to have been rooted in earlier practices. The use of Silla crowns, which were emblematic of a distinctive symbolic system, would have also been regulated.

According to historical records and archaeological data, both Goguryeo and Baekje used exclusively conical cap-shaped crowns (mogwan, 帽冠) from the fourth century onward. From that time, headband crowns were used only in Silla, Gaya, the southern territories of Baekje, and Japanese Wa (Fig. 7). The use of different types of crowns among these states can be explained by the varying processes for achieving statehood, the differences in the respective class and administrative rank systems, and the different stages by which each state adopted the new ideology of Buddhism. In the case of the Silla Kingdom, the establishment of a centralized government system and the adoption of Buddhism occurred almost simultaneously, followed immediately by changes in social rules and regulations. Accordingly, these transformations were accompanied by the decline of headband crowns with tree-shaped uprights (Ham Soonseop 2000, 81).

Fig. 7. Distribution of crown types (Re-drawn from Ham 1997: 96).

Some scholars have attempted to associate the headband crowns with tree-shaped uprights with women (Kim Byungmo 1998, 55; Kim Yongseong 2010, 121-3). However, although the crowns do seem to have been used to signify notions of social stratification, current evidence does not indicate that their use was restricted by gender or age. In the large tombs of Gyeongju, the former capital of Silla, deceased members of the royal family were buried wearing their crowns. The funerary rites, including the method of laying out the deceased, were in keeping with the distinctive religious practices of the time. However, the headband crowns with tree-shaped uprights are found less frequently in the tombs of outlying regions, perhaps indicating some conflict between the indigenous funerary practices and those newly introduced from the Silla center.

In addition to their role as emblems of social rank, headband crowns with tree-shaped uprights also may have carried a more general symbolic meaning related to the religious rituals of the Silla capital. In particular, the crowns may have been regarded as “sacred objects” (神物) within the Silla religious system, which had to be maintained and disseminated as the Silla state expanded. The prototype of these sacred objects would have been the “sacred tree” (神木) that stood in the hallowed sites of the Samhan period, called “sodo” (蘇塗). The leaders of the Silla Kingdom were the head priests who had presided over indigenous shamanistic rituals; male priests were given the title “Chachaung,” while females had names that contained the sound “ar” (Na Heela 1999, 95-7). A sacred place that housed the “shrine of the clan founder” (始祖廟) was also maintained within the Silla capital. Notably, the earliest known inception of the headband crowns with tree-shaped uprights was merely an image of the sacred tree attached to a headband. These crowns were but one of many components of ritual dress that were formalized in the Maripgan period, and their development through the three stages—“prototype,” “standard,” and “regression”—matches the trajectory of the symbolic system that they represented. The headband crowns with tree-shaped uprights may be regarded as material manifestations of the Silla system of religious rituals, making it possible to chart the overall development of Silla religious rituals, tracing back to rites carried out at the shrine of the Silla founder, through this single type of artifact.

The crowns of the “prototype” stage, which first appeared around the end of the fourth century, likely emerged through interaction with Goguryeo, where crowns with metal decorations were already in use. Foreign exchange was controlled by the central elite of Silla, who likely oversaw the creation of headband crowns with tree-shaped uprights, as suggested by the gold crown from Gyo-dong in Gyeongju. This crown has no decorative patterns along the edges of the uprights and its headband was not made from a single gold sheet, indicating that metalworking techniques had not yet been fully developed. Such techniques had become more advanced by the beginning of the fifth century, however, as evinced by the crowns from Bokcheon-dong Tomb 11 and Hwango-dong 100 Beonji Tomb 1, which were produced with delicate metalworking techniques that permitted their headbands to be made from a single sheet of gold. The headband crowns with tree-shaped uprights from this stage, made with a variety of individual styles, were produced during the reign of Namul Maripgan (r. 356-402). The fact that a prevalent style had not yet been established indicates that the Silla religious ritual system, embodied by the headband crowns with tree-shaped uprights, had yet to become dominant with the outlying regions of the kingdom.

The crowns of the “A3 (3/0)” and “B3 (3/0)” types with technical patterns “b,” “c,” “d,” and “e” are representative of the reign of Silseung Maripgan (r. 402- 417). In these crowns, a prevalent style emerged, characterized by vertical uprights with symmetrical branches. In addition to the Gyeongju area, crowns of these types have also been found in such key regional centers as Dongrae in Busan and Imdang in Gyeongsan. This distribution pattern reflects the spread of Silla’s religious ritual system into these key regional centers, along with the establishment of a symbolic system that utilized headband crowns with tree-shaped uprights as an important medium. This development can be attributed to the reign of Silseung Maripgan, a leader who was “wise with much sense and [who] had the vision to anticipate the future,” according to the Samguk Sagi. The crowns of this stage are also related to the existence of headband crowns with feather-shaped (羽毛形) uprights, which are believed to reflect the direct influence of the Goguryeo crown decorated with feathers (jowugwan, 鳥羽冠). Two such headband crowns with feather-shaped uprights were found in the south mound of Hwangnamdaechong Tomb, perhaps attesting to the deceased’s close connections with Goguryeo. Notably, Silseung Maripgan spent ten years as a hostage in Goguryeo, where he may have gained firsthand knowledge of regulations, including those concerning dress, which were later implemented in the Silla Kingdom.

The crowns of the “standard” stage represent the period from the reign of Nulji Maripgan (r. 417-458) to King Jijeung (r. 500-514). The “B3 (3/2)” and “B4 (3/2)” types represent the quintessential Silla headband crown with tree-shaped uprights; the majority of Silla gold crowns that are refined and well-proportioned belong to these types. The material culture of Goguryeo continued to have a heavy influence during the earlier part of the “establishing the standard” sub-stage, but this influence gradually decreased, until Silla designs were firmly established by the end of the sub-stage. This development of Silla designs occurred around the same time that Nulji Maripgan repaired the royal tombs of the previous rulers and that Soji Maripgan (r. 479-500) founded the Naeul Divine Palace (奈乙神宮), which occurred in 478. Significantly, these events took place after the second alliance between Silla and Baekje in 433, when Silla began to break away from Goguryeo influence. The repair of the royal tombs was not merely an act of refurbishing ancestral graves; it represented Nulji Maripgan’s attempt to re-organize the preceding royal lineage (Na Heela 1999, 104-5). The Naeul Divine Palace was intended for the worship of the “God of the Heavens,” the supreme deity who originally sent the founder of Silla down to earth. As such, this sacred shrine seems to have been an extension of the shrine of the clan founder, but its establishment does not appear to have led to the abandonment of the latter. Following the founding of the Naeul Divine Palace, the two shrines came to serve different functions, with the shrine of the clan founder being used only for ancestral rites for direct ancestors according to the royal genealogy (ibid., 126-7).

Thus, the Naeul Divine Palace was used for ancestral rites of the state, whereas the shrine of the clan founder was used for rituals of the clan. This differentiation is highly significant, as it seems to have been linked to the appearance of imitations of headband crowns with tree-shaped uprights, which occurred around the fourth quarter of the fifth century, according to the pottery chronology. That is, the imitation crowns may have been created when the clans of the regions annexed by Silla at an early date were given permission to hold their own ancestral rites. Notably, the deceased who were buried with headband crowns with tree-shaped uprights in small regional tombs were also entombed with iron bells and swords with ring-shaped pommels and three-leaf designs. Although this circumstance does not necessarily prove that the deceased were the heads of regional centers, it certainly demonstrates that the ruling elite of these regions actively adopted and reproduced these changes. The newly permitted rites would have followed Silla-style religious rituals.

Crowns of the “regression” stage belong to the Middle Ancient period (Ham Soonseop 2000, 81). The conditions that brought about the transition to this regression stage were set in place at the end of the preceding stage. During the reign of King Jijeung (r. 500-514), the authority of the Silla ruler expanded beyond that of the Maripgan rulers, as evidenced by the new use of Chinese-style king names (503), the establishment of formal guidelines on mourning clothes (504), and the dispatching of provincial governors (505). In the old system of the Silla Kingdom, the Maripgan rulers were akin to a “Khan” whose power and authority was relatively greater than that of his people. This power and authority was supported by Silla-style religious rituals (evidenced by the Naeul Divine Palace and the clan founder’s shrine), which were presided over by the Maripgan, while indirect rule was practiced according to the bu-system. However, the changes implemented by King Jijeung marked the dawn of a new system. The transition to the new system was completed with the establishment of a new centralized of government during the reign of King Beopheung (r. 514-540), who also oversaw the promulgation of state laws and decrees and the implementation of rules concerning official dress (520), and the acknowledgement of Buddhism as the state religion (527). As the result of all these changes, an absolute kingship was created (Jo Bodon 19889, 124-5). Therefore, the erosion of the standard form of the Silla headband crowns with tree-shaped uprights can be associated with the establishment of the new system during the reign of King Beopheung and the accompanying changes to Silla-style religious rituals. The royal family continued to preside over national ancestral rites at the Naeul Royal Shrine, but also actively delegated its religious duties to Buddhism. The increasing emphasis on the heavenly nature of the Naeul Royal Shrine and those lineages that were descended from the heavens led to greater authority for certain groups (in both the capital and provinces) that had not been fully incorporated into the state system, which became an obstacle to the royal family’s attempt to centralize the government system. Therefore, the royal family turned to Buddhism to obtain the transcendental authority it required to accelerate the centralization process (Na Heela 1999, 133-4). Religious rituals were carried out at three different locations: the Buddhist temple, the Naeul Royal Shrine, and the clan founder’s shrine. The imitation crowns that continued to be used in the provinces were still accompanied by iron bells, but not by swords with ring-shaped pommels and three-leaf designs. This change may not be attributed to the decline of the regional heads, but may instead reflect the wide dissemination of Silla-style religious rituals, which had taken root in these regions. As discussed, these imitation crowns have been found in the regions that were incorporated into the Silla state at an early date.

During the “culmination of regression” sub-stage, only the basic structure of the headband crowns with tree-shaped uprights was expressed. “D3 (4/0)” type crowns have been found around the peripheries of the regional centers in the Yeongnam area, as well as in the Han River region, which became part of Silla territory in 553, during the reign of King Jinheung (r. 540-576). The discovery of headband crowns with tree-shaped uprights representing Silla-style religious rituals in this area is especially interesting because that area had already been exposed to Buddhism as a former part of Goguryeo and Baekje territory. Moreover, by 553, Buddhism had already been introduced to Silla, so these crowns may reflect the process of “Silla-fication” undertaken by the communities of newly annexed regions. The “D3 (4/0)” type crowns were probably produced locally, as they are highly simplified and crudely made, unlike the “D4 (5/)” type crown from the Sanno Futakoyama Kofun Tomb in Japan. Like the Sanno Futakoyama crown, these crowns may possibly have been imitations of contemporaneous examples from the Silla capital, given the fact that the Naeul Royal Shrine was still in use. But this seems very unlikely, given that the royal family had adopted Buddhism as the state religion, which left little reason for the capital to continue producing headband crowns with tree-shaped uprights.

From this stage and beyond, the number of gold personal ornaments found in archaeological contexts decreased dramatically. This diminishment is not due to a decreased demand for such objects; rather, it is due to the regulation of their use according to rank. In addition, with the implementation of the official rank system, covering everyone from the village head to the peasantry, there was no longer a need to indicate rank through personal accessories (Kim Jaehong 2001, 165-177). These changes reflect the will of the Silla king, who also restricted the use of gold to the royal family and Buddhist temples (Ham Soonseop 2010, 190), an expected result of state centralization. The last of the headband crowns with tree-shaped uprights were the bronze “D4 (4/0)” type crowns manufactured in the seventh century. The use of certain precious metals was restricted by rank, so such metals could no longer be used for the crowns, even as gilt. The bronze crown from Ha-ri in Danyang did not come from a tomb but was found in a stone grotto, along with several ceramic vessels. Therefore, it is assumed that the Ha-ri crown would have been worn by a shaman who served the local community (Choi Monglyong, et al 1998, 129; Ham Soonseop 2000, 81). Some Silla-style religious rituals and ceremonies continued to take place in the Silla capital, including the coronation ceremony. By this time, however, in the local regions, the headband crowns with tree-shaped uprights had taken on a new meaning as a piece of religious paraphernalia worn by shamans serving the local community.

IV. Conclusion

This paper provides a diachronic analysis of Silla headband crowns with tree-shaped uprights. In previous papers published in Korea, I have examined the research history and iconography of Silla headband crowns with tree-shaped uprights, established typological categories and stages, and considered the overall development of the crowns within the historical context. This paper, intended for the English-language audience, analyzes the typological categories and developmental stages of the Silla headband crowns with tree-shaped uprights in more detail, and also discusses the historical context of their development from a fresh perspective.

In the future, the results of this study may be built upon by exploring the following the avenues of research. First, the development of the conical cap-shaped crown (mogwan), another type of Silla crown, should also be examined in detail. Such research would enhance our understanding of how crowns were worn in the Silla Kingdom from the Maripgan to Middle Ancient periods. Second, this research can help form the basis of future attempts to reconstruct Silla’s entire system of formal dress through archaeological material. Third, comparative studies with archaeological material from northern China, other regions of the Korean Peninsula, and the Japanese Archipelago will allow a more comprehensive understanding of the ancient systems of formal dress in East Asia.