Introduction
Since the end of the twentieth century, a number of archaeological sites related to Thăng Long capital (AD 1010–1802) of Đại Việt have been found and excavated in Hanoi.Footnote 1 In particular, the large-scale excavation of the site at number 18 Hoàng Diệu street found in 2002 at the very centre of Ba Đình district (adjacent to the National Assembly building) drew popular attention to the continuity of the central palace zone from the Lý (AD 1010–1226) through the Trần (AD 1226–1400) to the Lê (AD 1428–1789) dynasties. In order to commemorate the 1000th anniversary of the Thăng Long capital, the government decided to preserve the site along with the site of the ‘Central Axis’ of Thăng Long after the Lê period (including the Thăng Long-Hanoi citadel of the Nguyễn dynasty [AD 1802–1945]); a north-south axis which lies east of the 18 Hoàng Diệu site (Fig. 1).Footnote 2 In August 2010, these sites were registered on the UNESCO World Cultural Heritage sites list.
Figure 1. Early modern Thăng Long, c. AD 1490. (Adapted from Nishimura, 2011: 190).
For the purpose of research and preservation of the Thăng Long Imperial Citadel site, technical aid was requested from Japanese specialists, especially those who had worked on the preservation of Heijo (Nara) palace site. In response, a joint committee of Vietnamese-Japanese specialists was set up in 2007, and which has undertaken a three-year project (2010–2013) funded by UNESCO and the Japanese Trust Fund for Preservation of the World Cultural Heritage (JFIT). As a specialist in Vietnamese history during the Lý-Trần period, I was invited to join the committee. This article summarises part of the research outputs of this committee.Footnote 3
Due to the scarcity of historical records, research into Vietnamese history during the Lý-Trần period is generally underdeveloped.Footnote 4 Though comprehensive arguments about Thăng Long during the Lý-Trần period have been advanced by such scholars such as Lê Văn Lan (2004), Yao Takao (2007) and Phan Huy Lê (2006; 2009; 2010), issues remain to be discussed or readdressed. This article deals with such basic questions as to the position of major walls and palaces, the dispositions of major functions necessary for a capital, and the spatial concepts regarding the capital. To address these questions I review written sources, in particular two dynastic annals,Footnote 5 Đại Việt sử lược (hereafter SL) and Đại Việt sử ký toàn thư (hereafter TT).Footnote 6 It is my hope that this article will contribute not only empirically (in its direct sense) but also methodologically towards better exploitation of sources produced in the Sinic world (including Sino-Vietnamese records quoted in this paper) for the historical research of Southeast Asia, a field where East Asian sources have, so far, been exploited only superficially.
Construction of Palaces and Walls
‘Forbidden’ Palace Area in AD 1010/1011 and AD 1029/1030
The direct origin of Thăng Long is thought to have been Đại La citadel (Đại La thành) built in AD 866 by Cao Biền (Gao Pian), a Tang governor general of Annam.Footnote 7 According to Zizhi Tongjian (Suma Guang Reference Guang1986: 8117), the perimeter of Đại La citadel was 3000 bu (steps), approximating to 4.5 km.Footnote 8 The citadel was probably located at almost the same place as Thăng Long citadel built in AD 1010 by the Lý Dynasty (Lê Văn Lan 2004: 39–40, 42).
When Lý Công Uẩn (Thái Tổ, r. AD 1009–1028), the founder of the Lý dynasty, relocated his capital from Hoa Lư (in present-day Ninh Bình province) to Đại La and renamed it Thăng Long in the autumn of AD 1010, he ordered a number of palaces, temples and other buildings to be built in Thăng Long capital citadel. As described in the TT annals (seventh month, 1010):
“In front was built Càn Nguyên hall as the place of audience, on the left side was made Tập Hiền hall and on the right Giảng Võ hall. Then was opened Phi Long gate leading to Nghênh Xuân palace,Footnote 9 and Đan Phượng gate leading to Uy Viễn gate [palace?]. To the very south [of Càn Nguyên hall] was built Cao Minh hall, and a terrace called Long trì. Four sides of Long trì were roofed cloisters. Behind Càn Nguyên hall, two halls of Long An and Long Thụy were installed as the places of imperial residence. The Nhật Quang hall was constructed on the left side and Nguyệt Minh hall on the right. Behind them were built two palaces of Thúy Hoa and Long Thụy as the residences of court ladies and imperial concubines. Storehouses were built, walls constructed, and moats dug. On the four sides of the Citadel wall were opened four gates: the east was called Tường Phù gate, the west called Quảng Phúc gate, the south called Đại Hưng gate, and the north called Diệu Đức gate. Inside the Citadel wall (thành nội) were made the Hưng Thien imperial temple and the Ngữ Phượng tỉnh tower, outside the Citadel wall (thành ngoại) the Thắng Nghiêm temple was made to the south.”
The following year, the TT (1011) annals also record that, “Inside the Citadel wall, were built Đại Thanh palace [a Daoist temple] on the left and the Vạn Tuế temple on the right, and the Trấn Phúc storehouse was made. Outside the Citadel wall were the temples Tứ đại thiên vương, Cẩm Y, Long Hưng, and Thánh Thọ. On the Lô river (Red river) the Hàm Quang palace was built.”
In the third lunar month of AD 1028, Công Uẩn died. After suppressing an attempted coup by three princes in AD 1028, Lý Phật Mã (Thái Tông, r. AD 1028–1054) ascended the throne and remade the master plan of the forbidden area. According to the TT annals (sixth month, 1029), the emperor had removed Càn Nguyên hall, but a dragon appeared above its site, and therefore he:
“…had officials design a larger plan, fix the right direction, and rebuild [the hall], and renamed it Thiên An hall. On the left side was made Tuyên Đức hall, and on the right Diên Phúc hall. The terrace in front of the [Thiên An] hall was called Long trì. To the east of Long trì was installed Văn Minh hall, to the west was installed Quảng Võ hall. On the left and right sides of Long trì were installed bell towers facing each other, and whoever would appeal against a false charge was allowed to ring the bell. On the four sides of Long Trì, there were all cloisters and row houses, where officials gathered and army troops were stationed. In front [of Long Trì] Phụng Thiên hall was made, and on the palace built Chính Dương tower; the place taking charge of [the] timepiece. Behind [Thiên An hall] was built Trường Xuân hall, on the hall was made Long Đồ multi-storied house. [These buildings were] made as the place of residence and amusement. Outside was built a wall to surround [these buildings], called Long Thành (citadel of dragon).”
In the next year, Thái Tông: “Made Thiên Khánh hall in front of Trường Xuân hall to use as the place of imperial political works. The plan of the [Thiên Khánh] hall was octagonal-shaped. In front of and behind the hall were built Phượng Hoàng bridges” (TT 1030). The master plan of AD 1010/1011 and that of AD 1029/1030 (sketched roughly as Figures 2 Footnote 10 and 3) shared two features. First, the palaces and other buildings were arranged symmetrically on both sides of a central axis that ran in a north-south direction. Second, the palaces were arranged in rows from the front to the rear, with palaces for political affairs in the front and palaces for residence behind (Lê Văn Lan 2004: 45, 47).Footnote 11
Figure 2. Thăng Long citadel built in AD 1010/1011. (Adapted from Lê Văn Lan, 2004: 45)
Figure 3. Thăng Long citadel rebuilt in AD 1029/1030. (Adapted from Lê Văn Lan, 2004: 47)
Other Constructions
There are many other records of construction and repair of palace buildings and citadel walls. For instance, an earthen wall surrounding the four sides of Thăng Long capital was built in AD 1014 (TT). This earthen wall is usually thought to be identical with the Đại La citadel wall repaired in AD 1078 (TT). The latter’s location is thought to have been in close proximity to the Đại La citadel (also called La Thành) after the Lê period, namely, the outermost wall of early modern Thăng Long. In other words, the Đại La citadel wall after AD 1078 was a different wall from that built by Cao Biền in AD 866 (Lê Văn Lan 2004: 41; Phan Huy Lê 2006: 8). In AD 1049, another wall name appears in dynastic annals, which record that in that year was “dug the imperial canal outside [the wall of] Phượng Thành (citadel of phoenix)” (SL 1049). Its position and the relationship with Long Thành wall erected in AD 1029 will be discussed later.Footnote 12
Large-scale construction and reconstruction of palaces and modification of the master plan of the imperial citadel are also recorded, for example, in AD 1055, AD 1098, and AD 1156. In AD 1203, a new palace (tân cung) including many buildings was built to the west of the hall of imperial residence (tẩm điện). Thus, it is likely that the Central Axis of the imperial forbidden area was moved to the west. Some palaces were built outside the Thăng Long citadel, in addition to the famous outer palace (hành cung) of Giao Đàm or Dâm Đàm on present-day Hồ Tây. For instance, Long Đức palace was built for the crown prince Phật Mã “outside the Citadel” so that the crown prince could become familiar with popular affairs (TT 1012). When Thái Tổ died in AD 1028, Phật Mã entered Càn Nguyên hall through Tương Phù gate (the eastern gate of Thăng Long citadel). Therefore, Long Đức palace must have been located to the east of Thăng Long citadel.Footnote 13
Many halls and other buildings were recorded only once. These include the residential halls where successive emperors died. Except for Nhân Tông (r. AD 1072–1127) and Thần Tông (r. AD 1127–1137), both of whom died at Vĩnh Quang hall, every emperor reportedly died in his own residential hall.Footnote 14 This implies that many buildings were replaced or restored in a short time period, or at least renamed (after a new emperor ascended the throne, for instance).Footnote 15 There may be two explanations for the short life of buildings. First, wooden buildings could not withstand the humid tropical climate. Second, the structure (or rather, the idea) of power in Southeast Asia required frequent renovation rather than continuity.Footnote 16
The Segmentation of the Capital and the Citadel Wall
Doubts about the ‘Forbidden’ Area
As shown in Figure 1, the Thăng Long capital after AD 1490 was composed of three areas, namely, the Palace Citadel or Forbidden Citadel (cung thành or cấm thành), the Imperial Citadel (hoàng thành) and the Capital Citadel (kinh thành). This model differed in several respects from Chinese capital models after the Sui-Tang period (AD 581–907). For example, the three areas did not have a concentric composition as at Khaifeng of the Northern Song (AD 960–1127). Though the innermost citadel, namely, the Palace Citadel was square or rectangular, the middle (Imperial Citadel) and outermost (Capital Citadel) areas did not have a quadrilateral form as those in northern Chinese capitals like Chang’an during the Sui-Tang period. Administrative offices appear to have been mainly positioned within the Palace Citadel, rather than in Imperial Citadel, as they were after the Sui and the Tang (Fig. 4).
Figure 4. The “Map of Trung Đô” in Hồng Đức bản đồ (text A.2499), a collection of Lê maps originally drawn in the end of the fifteenth century and adapted in the eighteenth century.
The outermost wall of the Lý-Trần period (Đại La citadel wall) is thought to have run in close proximity to that after the Lê period. However, the location and function of inner wall(s) remains unclear. First, terms related to a ‘forbidden’ palace area (cung trung, cấm trung, cấm đình, cấm thành, khuyết đình, đại nội) are confusing. It is doubtful whether the ‘forbidden’ palace area during the Lý period covered the same area as the palace citadel of the Lê period. When Thái Tổ died, the princes Đông Chinh, Dực Thánh and Vũ Đức opposed the accession of Phạt Mã (Thái Tông) and, “taking command of soldiers, entered the forbidden citadel. Prince Đông Chinh lay in ambush within Long Thành, while princes Dực Thánh and Vũ Đức each lay in ambush within Quảng Phúc gate” (TT, third month, 1028). After suppressing this revolt, Thái Tông organised ten troops of imperial guard to protect “inside the Forbidden Citadel” (TT, before the sixth month, 1028). If the record of AD 1028 is to be believed,Footnote 17 Thăng Long citadel wall, in which was the western Quảng Phúc gate, was the very wall of the ‘forbidden’ palace area. However, if Long Thành was not simply an abbreviation of Thăng Long Thành, the former (if it existed before AD 1029) must have been located inside Thăng Long citadel wall; an apparent contradiction. Furthermore, according to SL (1137), “a golden dragon flew in the evening from Thái Thanh palace to the inside of the forbidden area (cấm nội).” If Thái Thanh palace is identical with Đại Thanh palace erected on the left side of the [Thăng Long] citadel in AD 1011, the ‘forbidden’ palace area was smaller (in the north?) than Thăng Long citadel. In AD 1127, when Nhân Tông died, Lê Bá Ngọc ordered officials to withdraw through Đại Hưng gate (the southern gate of Thăng Long citadel wall), and then enter Long Trì terrace through the “right side gate” to attend the ceremony of Dương Hoán (Thần Tông)’s accession to the throne (TT, twelfth month, 1127). If the “right side gate” did not mean the right side gate of Đại Hưng gate but rather meant the Right-Side gate of the forbidden area stipulated in article 51 of Quốc Triều hình luật (usually called “The Lê Penal Code” but including articles before the fourteenth century) as a gate inside Đại Hưng gate, which was the southern gate of the Imperial Citadel. Therefore, it is possible that the ‘prohibited’ palace area was also smaller than Thăng Long citadel in the south. If the forbidden area during the Lý period covered almost same area as that during the Lê period, the distance between its northern and southern sides (and possibly the distance between the eastern and western sides, as well) was approximately 700 meters (Phan Huy Lê 2006: 8-16, 18-19, 2010: 13-15).
Yet, the đại nội (Great Interior), which usually means the ‘forbidden area’ of Chinese capitals, also seems to have covered the whole area inside the Thăng Long citadel. According to TT (1024), the Chân Giáo Zen temple was built “inside the Citadel wall (thành nội)”. However, a decade later this temple is called “Chân Giáo Zen temple in the Great Interior” (TT, tenth month, 1224). Vạn Tuế temple, which was reportedly built “on the right side of [the area] inside the Citadel wall” (TT 1011), is also called “Vạn Tuệ temple in the Great Interior” in Thiền uyển tập anh (vol. 2: 10a), a collection of biographies of Buddhist monks compiled in the early fourteenth century.
Inside and Outside ‘The Citadel’
Next, we should examine the term thành nội (‘inside the Citadel wall’), and also the antonymous term thành ngoại (‘outside the Citadel wall’). Which wall did ‘the Citadel wall’ mean? In the case of constructions recorded in AD 1010/1011, it cannot have been a wall other than that of Thăng Long citadel, although the latter's location may have changed on the occasion of the repair recorded in the SL annals in AD 1024. In AD 1037, the shrine of the Hoằng Thánh đại vương deity was erected to the west of the southern gate of the Citadel (TT 1037). Judging from the present position of the shrine to the south of the Temple of Literature (Văn Miếu), the southern gate of the Citadel appears to have been Đại Hưng gate (the south gate of Thăng Long citadel wall erected in AD 1010), rather than the gates of Đại La citadel, for example the Trưởng Quang gate (TT 1033, 1048), or the south gate (TT 1154; see Fig. 1). In other words, ‘the Citadel wall’ did not mean the outermost wall (Đại La citadel wall), even after such a wall was erected in AD 1014.
Why the outermost wall could not be called ‘the Citadel wall’ can be explained with early Chinese concepts of the capital. Before the Sui period (AD 581–618), only the palace area (gong) and the administrative and popular residential area (cheng) were included in the capital citadel (jingcheng). The suburban area (guo) outside the cheng was included in the capital (jing) but not included into the capital citadel area. A wall was often erected around the guo area, but it was usually not as high and solid as the wall of gong and cheng (Toyoda 2008, 2010). During the Tang and Song periods, the suburban area was also included in the capital city, but it was again excluded after the Yuan period (Toyoda 2010). Thăng Long capital in its early phase was probably planned after the model before the Sui, when it was sufficient to erect an earthen wall to protect the suburban area.
Returning to the problem of the ‘forbidden’ palace area. There seem to have been two ideas. The first one covered a smaller area than Thăng Long citadel, and the second covered the entire Thăng Long citadel. Such ambiguity, which also appeared in Heian Japan,Footnote 18 may have reflected a political structure in which the imperial house and the administrative apparatus were not clearly differentiated from each other, as they were after the Trần period. It is also possible that the ‘prohibited’ palace area was originally smaller than the Thăng Long citadel, but later enlarged to cover the whole ‘inside the Citadel’ area, for instance on the occasion of the construction of new palace area in AD 1203. This may have influenced not only the descriptions of temples but also the description of the AD 1028 revolt by three princes, which seemingly describes Thăng Long citadel as a ‘Forbidden Citadel’.
Among the four gates of Thăng Long citadel opened in AD 1010, Tường Phù (east) and Đại Hưng (south) gates continue to be recorded until the Lê period.Footnote 19 However, Quảng Phúc (west) and Diệu Đức (north) gates never appear in sources after AD 1029. Concerning the north side of the citadel, there are no available records. For the west side, the ‘Citadel’ wall itself may have been moved in AD 1029 or later to the west, to a place close to the Imperial citadel wall of the Lê period. If so, the new space probably served as religious and garden spaces, because few archaeological remains from the Lý-Trần period have been found in the Lê imperial citadel area (Phan Huy Lê 2006: 12–14).Footnote 20 “The west gate of the Citadel”, through which the Trần emperor proceeded to Đồng Cổ shrine to take the New Year oath with officials following the Lý tradition, appears in dynastic TT annals of AD 1227. After the Lê period, the Đồng Cổ shrine was located in the north-western corner of Thăng Long capital.
Long Thành and Phượng Thành
Examination of Long Thành and Phượng Thành walls may shed light on the ambiguity of the ‘forbidden’ palace area. During the civil war in the last phase of the Lý period, Trần Tự Khánh, the military leader of the rising Trần family, sent troops to the capital because he got angry with a court official who fawned upon the emperor’s will. When they reached Long Thành wall, a palace-guard commander named Nguyễn Ngạnh led quan chức đô (palace policemen?) and entered the forbidden area (cấm trung) to accuse and arrest the official (SL, twelfth month, 1212). This apparently tells us that the forbidden area was enclosed by Long Thành wall. The master plan of the forbidden area in AD 1029/1030 mentions fewer palaces than the master plan in AD 1010/1011. This may prove that Long Thành wall was situated inside Thăng Long citadel wall, as the record of the revolt of three princes in TT annals of AD 1028 implies.
It is more likely, however, that Long Thành was an abbreviation of Thăng Long Thành.Footnote 21 If so, the record of SL (1212) quoted above may show there was some space between the Long Thành wall and the forbidden area (in the literal sense). A similar description can be found in TT (third day, tenth month, 1459): “The partisans of Lạng Sơn prince Nghi Dân, made ladders in the night, got over the Eastern Gate of the Citadel through three routes, stole into the forbidden palace. The Emperor (Nhân Tông) and the Empress Dowager Tuyên Từ were all killed.” The “Eastern Gate of the Citadel” apparently indicates the gate of Thăng Long (Imperial) citadel in Hồng Đức bản đồ (see Fig. 4), which was distant from the forbidden palace. Between the gate and the forbidden palace, there were spaces like the Eastern Palace and the Grand Ancestral Temple (Thái Miếu). Such records imply that, once getting over [Thăng] Long Thành, there was no serious obstacle before entering the forbidden area. In other words, the wall protecting the forbidden area in the literal sense was not as high and solid as the wall of ‘the Citadel’, not only during the Lý period but also in the fifteenth century. If so, this may have been in accord with the ambiguity between ‘the Great Interior’ and ‘Inside the Citadel’ during the Lý period.
It is difficult to clarify the relationship between Long Thành, Phượng Thành (AD 1049) and Long Phượng Thành (AD 1243) quoted above. The dynastic TT annals record that in AD 1243 the emperor had “constructed ‘thành nội’ and called it Long Phượng Thành.” “Thành nội” still appears to have meant ‘the area inside the Thăng Long citadel’ as was the case during the Lý period.Footnote 22 However, it is not clear whether the name Long Phượng Thành, means that Long Thành and Phượng Thành were originally the same. It is not impossible that the then Long Phượng thành wall connected the former walls of Long Thành and Phượng Thành.
However, it is reasonable to propose that during the Lý period, the names Long Thành and Phượng Thành referred to two different citadels. If Long Thành was situated inside ‘the Citadel,’ the latter may have been called Phượng Thành. While, if Long Thành was the abbreviation of Thăng Long Thành (which is considered more probable), Phượng Thành may have indicated the wall of the forbidden area. The TT annals in AD 1490 record that in that the emperor had “extended [the] Phượng Thành following the plan of the Lý-Trần. The Emperor learned from the case that Nhân Tông had been killed, so sent the men to erect the wall. And [the wall was situated] outside ‘Cửu giảo giao trường(?)’ was eight lý [c. 4 km] long. After eight months of construction, it was completed” (TT, eleventh month, 1490). The term Cửu giảo giao trường is translated as ‘martial arts practice ground’ in the Vietnamese translation of the TT annals,Footnote 23 and usually identified with Giảng Võ drill hall (Giảng Võ điện) in the southwest of the Lê Imperial Citadel.Footnote 24 This leads to the conclusion that the wall that was extended in AD 1490 was the south-western part of the Imperial Citadel on which Bảo Khánh gate opened (see Fig. 4). However, I could not find any concrete grounds for identifying cửu giảo giao trường with Giảng Võ điện. It is also possible to identify this wall with the straight wall in Figure 4 that runs westward from the west side of the Palace Citadel and is shaded in the same way as the Palace Citadel wall (Đỗ Văn Ninh 2004: 30–31). This straight wall may have served the imperial need to secure an escape route in case of emergency (remember that Lê Nhân Tông was killed in an attack from the east). If so, it can be inferred that Phượng Thành during the Lý period originally meant the wall surrounding the ‘forbidden’ palace area and later extended to the west.
Conclusion
It is still difficult to prove the exact location of halls, palaces and walls of Thăng Long during the Lý period. In a narrow sense, the ‘forbidden’ palace area seems to be smaller than Thăng Long citadel constructed in AD 1010. Later, however, the former term could also cover the whole area of Thăng Long citadel in AD 1010, while the latter (abbreviated as Long Thành?) may have been expanded to the west. Throughout this process, the Thăng Long citadel wall (the Citadel wall) was more important in both physical and ideological terms than any other wall. It was only in the late Lý or early Trần period that the area surrounded by the outermost Đại La citadel wall was regarded an integral part of the Capital Citadel (kinh thành).
Two features of Thăng Long deserve more attention. First, its plan may have changed drastically on more than one occasion. In AD 1203, the Central Axis may have moved when the ‘New Palace’ was constructed. Next came the change in AD 1230, when it is recorded that “Inside the Citadel palaces [were] erected multi-storied houses and cloisters and offices in the east and west. To the left was Thánh Từ palace (where the Senior Emperor resided), and to the right Quang Triều palace (where the present Emperor resides)” (TT 1230). Thus, it is impossible to assume continuity from the Lý to the Lê period in any simplistic way. Second, not all the functions of the capital were disposed within Đại La citadel wall. Aside from the residence of prince(s) and the parade ground,Footnote 25 Hoài Viễn station was opened on the river bank of Gia Lâm (on the opposite side of the Red River to Thăng Long) as a guesthouse for foreigners (TT, twelfth month, 1044). Villages around the capital, especially those directly subordinated to the throne, such as Cảo xã (present-day Nhật Tảo, where convicts were taken to cultivate state-owned rice fields), must have played large roles as an economic base of the emperor.Footnote 26 Cham villages around Thăng Long, the most famous being Bà Gia village, which was reportedly established by Cham prisoners caught by Lý Thánh Tông (who invaded Champa in AD 1069) (TT 1330), also appear to have played various roles.
Such features lead this study of Thăng Long to wider comparisons with other capitals in East and Southeast Asia. In the East Asian context, not only in Southern Chinese citadels (Nanjin, Hangzhou, Guangzhou and Chengdu, for instance) – the outer wall(s) of which were usually irregular-shaped – should be examined more closely. But also the ‘authentic’ models of Northern China, including those before the Sui-Tang era, ought to be re-examined regarding the positioning of buildings and the segmentation of the capital.Footnote 27 Citadels in Japan, Korea and Manchuria may show fresh insights in studying the multi-layered influence of varied Chinese models and deliberate assortment of it by the local ruler.
Though few Southeast Asian capitals outside Vietnam followed the Chinese model,Footnote 28 prosperous foreign trade (mainly conducted in the market town to the east of ‘the Citadel’) was not the only feature that Thăng Long had in common with other capitals in Southeast Asia. The bureaucratic apparatus was still primitive, and therefore the palace area and administrative area had not yet been differentiated clearly. Temples and shrines were often more important than secular administrative offices in the master plan of the capital.Footnote 29 The ethnic denomination ‘Kinh’ originally meant the people of the capital. During the Lý period, people of Capital (Kinh sư) (SL and TT 1016; SL 1059) or Capital Citadel (Kinh thành) (SL 1083; SL 1099; SL 1209) were probably distinguished from commoners (bách tính).Footnote 30 It was from the Trần period onwards that the term ‘Kinh’ began to indicate people from the Red River delta.Footnote 31 In other words, Thăng Long during the Lý period represented the whole state.Footnote 32 And in the capital, the Lý emperors not only assumed themselves to be the southern emperor of the Sinic world, one of them also proclaimed himself Buddha as well (Cao Tông, according to the comment of Lê Văn Hưu in TT 1034). Despite their ideological suzerainty, these emperors could not control peasants as uniformly as Chinese rulers could.Footnote 33 Rather, they relied on people directly subordinated to the throne,Footnote 34 probably including many non-Vietnamese people (Chinese, Cham, Thai-Lao and so forth),Footnote 35 who were scattered in various regions. Similar features may be found in Southeast Asian capitals including Angkor, Pagan and Ayutthaya.
