Literature Review

In 111 bce, the Han Dynasty (206 bce–220 ce) destroyed the Kingdom of Nanyue and annexed its territory. Inheriting the administrative system of Nanyue, the Han put the region of Nanyue under the Jiaozhi Commandery (交趾郡). The headquarters of the Jiaozhi Commandery was originally Miling (麋泠, Vietnamese: Mê Linh), but was later moved to Leilou (羸𨻻, Vietnamese: Luy Lâu) and then to Longbian (龍編, Vietnamese: Long Biên). So far, there have been several hypotheses concerning the location of the Leilou and Longbian citadels.

The first hypothesis (hereafter Hypothesis 1), proposed by Henri Maspero, holds that the Leilou Citadel was located south of Hanoi. Maspero cited a note in Yuanhe junxian tuzhi (元和郡縣圖志) that suggested that Leilou was located 75 li (里) northwest of the Tang’s Protectorate Citadel.¹ This hypothesis, however, has been criticized for its lack of evidence, as it relies on a single piece of evidence.²

The second hypothesis (hereafter Hypothesis 2, simplified as “LÅ©ng Khê = Leilou”), proposed by Claude Madrolle based on fieldwork by Henry Wintrebert, holds that the Leilou Citadel was located at the same place as the later LÅ©ng Khê Citadel (Thuận Thành, Bắc Ninh).³ This hypothesis has been backed and developed by many other scholars,⁴ based on the following sources. First, Việt sá»­ thông giám cÆ°Æ¡ng mục (越史通鑑綱目) and Đại Nam nhât thông chí (大南一統志), compiled under the Nguyễn (1802–1945), both affirm that the LÅ©ng Khê Citadel was the Leilou Citadel under Han rule. Second, Việt điện u linh (粵甸幽靈), compiled under the Trân (1225–1400), suggests that Shi Xie (士燮, Vietnamese: SÄ© Tiêp/Nhiêp, 137– 226), the head of Jiaozhi Commandery, based his headquarters in Leilou. Ngụy Việt ngoại ká»· (偽越外紀), cited in Annan zhi(安南志), which was compiled under the Ming (1368–1644), also states that Shi Xie built the Leilou Citadel. In fact, the inner and surrounding areas of the LÅ©ng Khê Citadel include many relics and legends related to Shi Xie. Also, as Sakurai Yumio observed, the view that Leilou was located at LÅ©ng Khê does not contradict the notes on rivers in Shuijing zhu (水經注).⁵

Based on Hypothesis 2, scholars such as Đỗ Văn Ninh argue that the Leilou Citadel was also the Longbian Citadel under the Han, and that both were the same as the LÅ©ng Khê Citadel (hereafter Hypothesis 3, simplified as “LÅ©ng Khê = Leilou = Longbian”). The equation of Leilou with Longbian is based on records in Đại Việt sá»­ ký toàn thÆ° (Toàn thÆ°) (大越史記全書).⁶ However, this hypothesis was criticized by Trân Quôc Vượng because in many historical documents from the Han to the Song period (960–1279), Leilou and Longbian are described as two different administrative units.⁷

In 2011, based on his excavation at LÅ©ng Khê, Nishimura Masanari proposed a new theory, arguing that the LÅ©ng Khê Citadel is the Longbian Citadel, and both are distinct from the Leilou citadel (hereafter Hypothesis 4, simplified as “LÅ©ng Khê = Longbian ≠ Leilou”). Nishimura’s hypothesis is based on three lines of reasoning. First, although he did not pursue an in-depth analysis, Nishimura did question the reliability of historical documents (the earliest being Việt điện u linh) that hold that LÅ©ng Khê was Leilou. Second, in reference to Jiaozhou department heads such as Bu Zhi (步騭), Nishimura recognized that these mandarins’ titles often included the names of the places where they established their headquarters. As Shi Xie received the title of “Longbian marquess” (龍編侯), Nishimura believed that Shi Xie’s headquarters was at Longbian. Therefore, the LÅ©ng Khê Citadel, with many relics and legends associated with Shi Xie, should be at Longbian, not Leilou. Third, results from archaeological excavations by Nishimura and Vietnamese archaeologists have also suggested that the LÅ©ng Khê Citadel was occupied in the period between the Late Han (25 ce–220 ce) and the Sui period (581–618), which coincided with the period when the Longbian Citadel was occupied according to historical records.⁸

A review of the literature on the Leilou–Longbian–Lũng Khê question highlights several shortcomings. First, when discussing the issue of Leilou and Longbian, it is crucial to note that Leilou and Longbian were the headquarters of the Jiaozhi Commandery. As will be discussed in the second section of this paper, under Han rule there were two parallel administrative units in the region: the Jiaozhi Commandery and the Jiaozhi Department, headed respectively by the Jiaozhi taishou (交趾太守) and the Jiaozhi cishi (交趾刺史). Differentiating between the two headquarters of these two administrative units, the author will analyze their relationship, changes in their location over time, and the relationship between Leilou, Longbian, and Shi Xie.

Second, although many works mention Leilou and Longbian, no study has offered a comprehensive and systematic analysis of the relevant historical records. Information about Leilou and Longbian is distributed across a tremendous amount of material dated between the early Common Era and the early twentieth century. Within this vast historical time span, it seems that at certain points, in addition to the original data, some misinformation about Leilou–Longbian emerged and was later taken to be reliable. The subsequent use and interpretation of the original data and the later misleading information has yielded sedimentary layers of material in conflict with each other. It is difficult to locate Leilou and Longbian without properly distinguishing these layers. Indeed, it was the accumulation of undifferentiated historical and nonhistorical data that gave rise to the diverse hypotheses on Leilou, Longbian, and Lũng Khê.

In this context, Nishimura Masanari proposed some important steps toward solving the problem by combining archaeological evidence with written records. Based on the existing scholarly literature and particularly the ideas advanced by Nishimura, the author will try to separate the accumulated data into the sedimentary layers that led to each hypothesis. The author thereby attempts to recover the original sources in order to shed further light on the issue of Leilou and Longbian.

The Headquarters of the Jiaozhi Commandery and the Jiaozhi Department

To clarify the relationship between Leilou and Longbian, it is important to distinguish between two administrative units, the Jiaozhi Commandery and the Jiaozhi Department, and their respective headquarters.

The name Jiaozhi originated from the name of one of the two commanderies established on the territory of Âu Lạc (甌雒) by the Nanyue Kingdom, namely, Jiaozhi (交趾) and Jiuzhen (九真). The Nanyue administrated the two commanderies via two representatives, known as the “two envoys” (二使者). After the conquest of the Nanyue in 111 bce, the Han Dynasty maintained the system of two commanderies, Jiaozhi and Jiuzhen, designating the heads of both commanderies as taishou (太守).

In parallel with the Jiaozhi Commandery, the Han established a new administrative unit with the same name, the Jiaozhi Department.⁹ Five years after the conquest of the Nanyue, in 106 bce, the Han divided the whole country into thirteen “departments” (刺史部), each headed by a cishi (刺史). The Jiaozhi Department included seven commanderies: Nanhai (南海), Cangwu (蒼梧), Hepu (合浦), Yulin (鬱林), Jiaozhi (交趾), Jiuzhen (九真), and Rinan (日南).¹⁰ It is noteworthy that out of the thirteen departments, apart from Jiaozhi and Shuofang (朔方), the other eleven departments had the suffix -zhou, such as the Qingzhou (青州) Department or the Xuzhou (徐州) Department. Therefore, the Jiaozhi Department was equal to a zhou, and the headquarters of the Jiaozhi Department was regarded as the center of a zhou (州治). In 203 ce, for the sake of consistency, the Jiaozhi Department was renamed Jiaozhou (交州) (also known as the Jiaozhou Department).

Thus, under Han rule, in the period between 106 bce and 203 ce, there were two administrative units with the name Jiaozhi: the Jiaozhi Commandery (交趾郡) and the Jiaozhi Department (交趾 [刺史]部). The Jiaozhi Department was the higher-level unit, supervising the Jiaozhi Commandery. The head of the department was the Jiaozhi cishi, and the head of the commandery was the Jiaozhi taishou. In 203 ce, Emperor Xian of the Han, on the advice of the Jiaozhi cishi, Zhang Jin (張津), and the Jiaozhi taishou, Shi Xie, decided to rename the Jiaozhi Department the Jiaozhou (Department).¹¹ This crucial issue has not received sufficient attention, and confusion has arisen over the headquarters of these two administrative units.

As highlighted above, the headquarters of the Jiaozhi Commandery was the commandery center (郡治 or 郡守) of the Jiaozhi taishou. The headquarters of the Jiaozhi Department, meanwhile, was the department center (州治) of the Jiaozhi cishi (and later the Jiaozhou cishi). As the Jiaozhi cishi ruled over all seven commanderies, the department center could have been located anywhere within the boundaries of the seven commanderies. The headquarters of the Jiaozhi taishou, however, had to be somewhere within the boundaries of the Jiaozhi Commandery. The following paragraphs trace the shifting location of the headquarters of the Jiaozhi Commandery and the Jiaozhi Department.

According to Jiaozhou waiyuji (交州外域記), cited in Shuijing zhu, initially the headquarters of both the Jiaozhi Commandery and the “original” headquarters of the Jiaozhi Department (州本治) were located in Miling County.¹² Wang Fan’s (王範) Jiaoguang chunqiu (交廣春秋), written under the Western Jin (265–420) and cited in Shuijing zhu, holds that in 106 bce, the headquarters of Jiaozhou (the Jiaozhi Department) was moved from Leilou to Guangxin (廣信) in Cangwu (蒼梧).¹³ That is, the headquarters of the Jiaozhi Department was in Miling for a very short period before relocating to Leilou, and was relocated again to Guangxin in 106 bce. However, the headquarters of the Jiaozhi Commandery (and the Jiaozhi taishou) had to remain in the territory of the Jiaozhi Commandery. Where was it, and was it still located in Miling?

Many texts mention that Zhou Chang (周敞) moved the headquarters from Leilou to Longbian in the first half of the second century. However, it is vital to clarify that what was moved was the headquarters of the Jiaozhi Commandery or the Jiaozhi Department. According to Jiu Tangshu (舊唐書) and Taiping huanyu ji (太平環宇記), Zhou Chang was the Jiaozhi taishou, and the relocation from Leilou to Longbian was the relocation of the commandery center (郡守).¹⁴ Therefore, this event was the relocation of the headquarters of the Jiaozhi Commandery. This evidence shows that by the time of Zhou Chang, the headquarters of the Jiaozhi Commandery was no longer in Miling, but Leilou. This may be confirmed by the remark in Hanshu (漢書) that Leilou was the most important county of the Jiaozhi Commandery.¹⁵ In addition, Zhou Chang relocated the headquarters of the Jiaozhi Commandery from Leilou to Longbian, and consequently in Hou Hanshu (後漢書) Longbian was listed as the leading county of the Jiaozhi Commandery.¹⁶

Furthermore, according to Jinshu (晋書), in 141 ce Zhou Chang again proposed that the royal court establish Jiaozhi as a zhou. Although the court did not approve his proposal, it made a concession and promoted Zhou Chang to Jiaozhi cishi, therefore making him head of the department.¹⁷ This meant that in 141 ce, the headquarters of the Jiaozhi Commandery and the Jiaozhi Department were combined in the same location, which was in Longbian. This concurs with what Songshu (宋書) records, namely that in 203 ce, the headquarters of the Jiaozhi Department was located in Longbian. When the Jiaozhi Department was renamed Jiaozhou, the headquarters was moved again from Longbian to Guangxin (Cangwu).¹⁸ In 211 ce, the headquarters of Jiaozhou was moved to Panyu (番禺). In 226 ce, Sun Quan (孫權) tried to separate Guangzhou from Jiaozhou, but in the same year, he had to give up the plan because of the rebellion in Jiaozhou. In 262 ce, under the reign of Sun Xiu (孫休), Jiaozhou was finally separated from Guangzhou, and the headquarters of Guangzhou was located in Panyu.¹⁹ Jiaozhou at the time only included counties from Jiaozhi southward, and the headquarters of Jiaozhou was moved back to Longbian.²⁰

Based on the texts cited above, Table 1 summarizes the changes in the location of the headquarters of the Jiaozhi Commandery and the Jiaozhi Department.

As seen in Table 1, prior to the tenth century, there is no historical evidence that after the term of Zhou Chang, the headquarters of the Jiaozhi Commandery was relocated back to Leilou. In other words, there is no indication that Shi Xie (137–226) made Leilou his headquarters.

Second, Leilou and Longbian could not have been the same county (縣) (Leilou ≠ Longbian). As illustrated in Table 1, in both Hanshu and Hou Hanshu, Leilou and Longbian were listed as two distinct counties. Shuijing zhu (compiled in the early sixth century ce) describes a river that ran through the north of Leilou County, connecting with the King River that originated from the high grounds (高山) of Longbian County. Thus, Shuijing zhu also differentiates Leilou from Longbian, seeing the two as separate units.²¹ Jiu Tangshu and Taiping huanyu ji (both compiled in the tenth century) clearly note the relocation of the headquarters of the Jiaozhi Commandery from Leilou to Longbian.

However, after Vietnam became independent from China in the tenth century, a new opinion emerged in Toàn thÆ°, according to which Leilou was the same as Longbian (“Leilou = Longbian”). In Toàn thÆ°, the chapter on Shi Xie states, “(Shi Xie) was appointed Jiaozhou taishou, received the title of Long Độ Đình hâu (龍度亭侯), and situated his headquarters in Leilou (also known as Longbian).” With the words “Leilou (also known as Longbian),” Toàn thÆ° established the equation “Leilou = Longbian.” As Toàn thÆ° was the official historical book of the state, the idea of “Leilou = Longbian” was officially and widely accepted from the fifteenth century onward. And in the twentieth century, some scholars, notably Đỗ Văn Ninh, building on Toàn thÆ°, put forth Hypothesis 3: “Leilou = Longbian.”²²

table 1

Relocations of the headquarters of the Jiaozhi Commandery and the Jiaozhi Department

View Table

The view that “Leilou = Longbian,” however, was evidently in conflict with the idea that Leilou ≠ Longbian, which is well established in earlier sources, including Hanshu, Hou Hanshu, Shuijing zhu, and Jiu Tangshu. Thus, the earliest sources provide no support for Hypothesis 3: “Leilou = Longbian.”

Leilou in Historical Records

Tested against materials written before the tenth century, Hypothesis 3 does not hold water. Leilou and Longbian under the Han were two different citadels. Before considering the relationship between Leilou, Longbian, and the Lũng Khê Citadel, it is necessary to review the locations of Leilou according to pre-tenth century literature.

After the Tenth Century: Leilou = Lũng Khê

From the beginning of the eleventh century, however, new sources appeared that suggested that Leilou was located at the position of the Lũng Khê Citadel (Leilou = Lũng Khê). The earliest sources that contain this view are Thiên uyên tập anh (禪苑集英) and Toàn thư.

First, the Pháp Hiên (法賢) chapter in Thiên uyên tập anh referred to Pháp Vân Pagoda as Pháp Vân Pagoda of Leilou (羸𨻻法雲寺), implying that the Pháp Vân Pagoda (which overlaps with the area of the LÅ©ng Khê Citadel) was part of Leilou County.³⁴ Although the chapter of Thông Biện (通辨) did not directly suggest that “Leilou = LÅ©ng Khê,” it mentioned that Khâu Đà La (丘陀羅), Ma La Kỳ Vá»±c (摩羅耆域), and Pháp Hiên, the famous Zen masters associated with Pháp Vân Pagoda, once practiced Buddhism in the Leilou area.³⁵ Thus the view that “Leilou = LÅ©ng Khê” already appeared in the two chapters of Thiên uyên tập anh.

However, in Thiên uyên tập anh, the chapters on Tỳ Ni Đa LÆ°u Chi (毘尼多流支) (VinÄ«taruci, the master of Pháp Hiên) and Sùng Phạm (崇範) referred to Pháp Vân Pagoda of Cô Châu Hamlet of Longbian (龍編古州郷法雲寺), implying that Pháp Vân Pagoda was situated in the Cô Châu hamlet of Longbian.³⁶ Although the chapter on Thiên Nham (禪岩) mentions the location of another pagoda, Trí Quả (致果), it shares the view that Cô Châu hamlet was situated in the Longbian area.³⁷ Therefore, there are two views that coexist in Thiên uyên tập anh: “LÅ©ng Khê = Leilou” (the chapters on Pháp Hiên and Thông Biện), and “LÅ©ng Khê = Longbian” (the chapters on Tỳ Ni Đa LÆ°u Chi, Sùng Phạm, and Thiên Nham).

Second, according to Toàn thÆ°, in 1188 ce, Emperor Lý Cao Tông went to Pháp Vân Pagoda of Duênh Bà ( 瀛婆法雲寺). Duênh Bà (瀛婆) was believed to be an error for Leilou (羸𨻻), so this text has been considered as evidence that “LÅ©ng Khê = Leilou.” However, in the same text, the note of Ngô SÄ© Liên, the main editor of Toàn thÆ°, reads: “at the beginning of our dynasty [i.e., the Lê], we still had this custom” (今朝國初猶乃舊俗). Meanwhile, in Đại Việt sá»­ lược, compiled during the Trân, Pháp Vân Pagoda was never called “Pháp Vân Pagoda of Duênh Bà (or Leilou)” but “Pháp Vân Pagoda of Cô Châu” (古州法雲寺). This implies that the text of Toàn thÆ° was edited by Ngô SÄ© Liên in the fifteenth century.³⁸

Third, in the Shi Xie chapter of Toàn thư, Ngô Sĩ Liên suggests that the Leilou Citadel was situated in Lũng Khê. First, in the account of the life of Shi Xie, Ngô Sĩ Liên states that Shi Xie located his headquarters in Leilou, noting that “(Leilou) means Longbian” (“Leilou = Longbian”). Then, in his comments at the end of the Shi Xie chapter, Ngô Sĩ Liên remarks that the shrine dedicated to Shi Xie (located in the area of Lũng Khê) was once part of “the former Longbian Citadel” (“Lũng Khê = Longbian”). In other words, by combining the two equations “Leilou = Longbian” and “Lũng Khê = Longbian,” Ngô Sĩ Liên indirectly suggests that Leilou was Lũng Khê (“Leilou = Lũng Khê”).

The question, however, is whether Shi Xie really based his headquarters in Leilou. As demonstrated in the previous section, pre-tenth century literature never mentioned Shi Xie’s headquarters in Leilou. So where did this idea come from?

The following table compares the notes in Việt điện u linh, Toàn thư, and Sanguo zhi in order to prove that the idea that “Shi Xie’s headquarters = Leilou” in Toàn thư originated from Việt điện u linh. What was written in Việt điện u linh, however, was largely unsubstantiated. Table 2 cites specific sentences in Việt điện u linh and compares them with parallel statements in Toàn thư and Sanguozhi.

table 2

Shi Xie in Sanguo zhi, Việt điện u linh, and Toàn thư

View Table

View Table

View Table

View Table

As seen in Table 2, except in the case of sentences 11 and 30, Việt điện u linh copied almost every word in the Shi Xie story in Sanguo zhi. In Việt điện u linh, only some characters were modified: for example, Jingshi (京師) in Sanguo zhi in sentence 5 was changed to Hanjing (漢京), and the character cha (察) in sentence 8 was changed to ju (挙). In Việt điện u linh, some content from Sanguo zhi was also removed—for example, the section about Liu Ziqi (劉子奇) in sentence 5. In the case of sentences 31 onward, the order of the sentence was changed. In Sanguo zhi, these sentences were in the first section, but in Việt điện u linh they were relocated to later sections.

Sentence 30, “wang zhi di you sanzi” (王之弟有三子), was inferred from sentence 31 in Sanguo zhi (Shi Xie had three younger siblings), but Việt điện u linh mistook “the three siblings” as “the three children of his younger brother.”

In Sentence 11, Việt điện u linh added a new sentence with no equivalent content in Sanguo zhi—“wang zhi leilou ji guangxin ersuo” (王治羸𨻻及廣信二所)—stating that Shi Xie located the headquarters in the counties of Leilou and Guangxin. Comparing the content of the Shi Xie chapter in Toàn thư with Việt điện u linh and Sanguo zhi in Table 2, it is clear that the idea of “Leilou = Longbian” in Toàn thư was based on Việt điện u linh.

In the Shi Xie chapter, although Toàn thÆ° makes reference to Sanguo zhi,³⁹ in general it largely copies Việt điện u linh. This is evidenced by the fact that Toàn thÆ° retains almost every modification that Việt điện u linh made when copying Sanguo zhi, notably Hanjing in sentence 5, or ju in sentence 8.

Toàn thư also copied from Việt điện u linh, as the comments by Ngô Sĩ Liên in the chapter on Shi Xie show. In the comments, Ngô Sĩ Liên mentioned the event of Linyi warriors unearthing Shi Xie’s tomb at the end of the Jin. When compared with the contents of Việt điện u linh (Table 3), it is apparent that Ngô Sĩ Liên copied almost the same contents from Báo cực truyện (報極傳), cited in Việt điện u linh.

When copying from Việt điện u linh, Ngô Sĩ Liên modified sentence 11 (王治羸𨻻及廣信二所) in Việt điện u linh into 都羸𨻻（𨻻一作樓，即龍編）, meaning “situate the headquarters in Leilou, which was Longbian.” In short, Toàn thư’s idea that “Shi Xie’s headquarters = Leilou” was inherited from Việt điện u linh.

“Shi Xie’s Headquarters = Leilou” in Việt điện u linh

The above sections have shown that the idea that “Shi Xie’s headquarters = Leilou,” a premise that the editors of Toàn thư including Ngô Sĩ Liên used to establish that “Leilou = Lũng Khê,” was actually adopted from Việt điện u linh. The question, then, is how reliable the information in Việt điện u linh is.

As shown above, regarding the life of Shi Xie, in addition to the idea that he established his headquarters in Leilou, Việt điện u linh took most of its details from Sanguo zhi. However, the sentence 王治羸𨻻及廣信二所 is in conflict with a historical fact: Shi Xie never situated his headquarters in Guangxin County.

All the historical records concerning Shi Xie in Hou hanshu or Sanguo zhi agree that Shi Xie was not the Jiaozhou cishi (Department Head), but only the Jiaozhi taishou (Commandery Head). Both Hou hanshu and Sanguo zhi indicate that the position of Jiaozhou cishi existed— it was held by Zhang Jin and Bu Zhi, for example. After the Jiaozhou cishi Zhang Jin died, to prevent Shi Xie from taking the side of Liu Biao (劉表), Emperor Xian of the Han—but in practice Cao Cao (曹操)—tried to gain the loyalty of Shi Xie by granting him the authority to “supervise seven commanderies” (董督七郡). The question is whether Shi Xie could possibly have situated the headquarters in Guangxin.

table 3

Comparison of Ngô Sĩ Liên’s comments and Việt điện u linh

View Table

First, although Emperor Xian granted Shi Xie the authority to “supervise seven commanderies,” Shi Xie was retained as “the Jiaozhi taishou as normal” (領交阯太守如故). Meanwhile, Liu Biao appointed Lai Gong (賴恭) as the Jiaozhou cishi in place of Zhang Jin, and Wu Ju (吳巨) as the Cangwu taishou in place of Shi Huang (史璜). This meant that even after the death of Zhang Jin, Shi Xie was not the Jiaozhou cishi and his power did not extend to the area of Guangxin (Cangwu).⁴⁰

In short, all historical records in Houhan shu affirm that Shi Xie was never appointed as a Jiaozhou cishi who ruled over all seven commanderies, and that his highest position was only as a Jiaozhi taishou who ruled over the Jiaozhi Commandery. Guangxin County was part of the Cangwu Commandery, the headquarters of the Jiaozhi (Jiaozhou) cishi. Shi Xie, as the Jiaozhi taishou, could not have established his headquarters in Guangxin. The misleading information in Việt điện u linh might originate from the fact that Shi Xie’s homeland was Guangxin, or because the author of Việt điện u linh mistook Shi Xie’s position of Jiaozhi taishou for Jiaozhi cishi.

In conclusion, the viewpoint that “Leilou = Lũng Khê” was established very early; passages in the Pháp Hiên and Thông Biện chapters in Thiên uyên tập anh and in Toàn thư provided the basis for the conclusion that “Leilou = Longbian = Lũng Khê.” However, considering Thiên uyên tập anh as a collection of chapters that were written in different periods, it is clear that that that claim that “Leilou = Lũng Khê” in the Pháp Hiên and Thông Biện chapters was not very popular. Other chapters offered a different opinion, namely, that the area of Lũng Khê was part of Longbian (“Longbian = Lũng Khê”).

In Toàn thư, Ngô Sĩ Liên inherited from Việt điện u linh the idea that “Shi Xie’s headquarters = Leilou,” and he later noted that “Leilou means Longbian (Leilou = Longbian).” So he finally drew on the idea that “the shrine dedicated to Shi Xie was located in the old Longbian Citadel (Lũng Khê)” to establish that “Leilou = Longbian = Lũng Khê.” However, Leilou could not be Longbian. Moreover, the idea that “Shi Xie’s headquarters = Leilou” in Việt điện u linh was inserted when the author of Việt điện u linh copied Sanguo zhi, and is thus not reliable. Apart from this piece of information, however, Việt điện u linh copied almost every word in Sanguo zhi. It means that Việt điện u linh was not informed by a different and new source of material about Shi Xie. Furthermore, the sentence that linked Shi Xie and Leilou in Việt điện u linh contained misleading information, particularly as it implied that Shi Xie situated his headquarters in Guangxin.

Although the idea that “Leilou = Lũng Khê” developed very early in Vietnam (possibly since the Trân with the compilation of Việt điện u linh), it was not based on reliable material, and it was in conflict with the idea widely held in pre-tenth century literature that Leilou was located west of Hanoi.

Meanwhile, if we assume that “Leilou = Lũng Khê,” we will have difficulty locating the Longbian Citadel. If “Leilou = Lũng Khê,” because “Leilou ≠ Longbian” there must have been another Longbian Citadel different from Lũng Khê. To solve this puzzle, some scholars have tried to locate Longbian in the provincial center of Bắc Ninh or Yên Phong. However, investigations and fieldwork from French colonial rule until the present day have not found any remains corresponding to the location of this hypothetical Longbian Citadel.

“Longbian = Lũng Khê”: Recently Discovered Epigraphic Materials

Compared to the view that “Leilou = Lũng Khê,” the view that “Longbian = Lũng Khê” is even more widely held in the existing literature.

As illustrated above, in Thiên uyên tập anh, apart from the two chapters on Pháp Hiên and Thông Biện, other chapters agree that the LÅ©ng Khê area and adjacent areas belonged to Longbian (“Longbian = LÅ©ng Khê”). Đại Việt sá»­ lược, compiled under the Trân, does not contain any information on Leilou in the notes for 1034 and 1161.⁴¹

Although Toàn thÆ° embraced the view that “Leilou = LÅ©ng Khê,” the manner in which Ngô SÄ© Liên phrased his notes and his comments indirectly implied that in the fifteenth century, the LÅ©ng Khê area was widely known as “the old Longbian Citadel” (龍編故城). First, it is important to clarify how Ngô SÄ© Liên described the area where the shrine dedicated to Shi Xie was situated (LÅ©ng Khê). He did not call the area “the old Leilou Citadel” but rather “the old Longbian Citadel.” When he mentioned Leilou, Ngô SÄ© Liên noted that “Leilou means Longbian.” This shows that in Ngô SÄ© Liên’s view, in the fifteenth century the location of Leilou was not well known. Hence, he had to use Longbian as a cross-reference, for Longbian (the old Longbian citadel) was widely known as the place where the shrine dedicated to Shi Xie was located.⁴² This fact is compatible with Nishimura Masanari’s view that “Longbian = LÅ©ng Khê,”⁴³ which has been confirmed by recently discovered stele inscriptions.

First, in 2004, an inscribed stele from the Jiaozhou Sarira Stupa (Figs. 1, 2), dated 601 ce, was discovered along with a stone casket in Trí Quả commune (Thuận Thành, Bắc Ninh). In 601 ce, Emperor Wen of the Sui Dynasty distributed sarira (relics of Buddhist monks) to all thirty zhou across the country, including Jiaozhou. In Jiaozhou a stupa was built to house the sarira, and a stone stele was raised to commemorate the burying of the sarira and the construction of the stupa. The Sarira Stupa inscription matches the notes in Guang hong ming ji (廣弘明集), compiled by Monk Daoxuan (道宣) in the Tang, and this suggests that the place where the stele and the stone casket (presumed to be the original container of the sarira) were discovered was 交州龍編縣禅衆寺, namely, the Thiên Chúng Pagoda (禅衆寺) in Longbian County.

However, my research on the stele suggests that the location where this stele was discovered in 2004 was not the original location of the Thiên Chúng Pagoda. Rather, based on the notes in Thiên uyên tập anh, the Thiên Chúng Pagoda in Longbian County under the Sui was located on 天福峯, namely, on Thiên Phúc Mount, one of the eight mounts of Tiên Du Mountain (仙遊山). Tiên Du Mountain was five km north of the area where the stele was discovered in 2004, and approximately 6.5 km northwest of the LÅ©ng Khê Citadel.⁴⁴

View Full Size

Area near the Lũng Khê Citadel, showing the original location of the stele bearing the Sarira Stupa inscription and the place where it was discovered, and the location of the shrine built by Tao Huang.

Citation: Asian Review of World Histories 5, 2 (2017) ; 10.1163/22879811-12340004

Map Source: Google Maps.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Rubbing of the Sarira Stupa inscription discovered at Trí Quả commune (Thuận Thành, Bắc Ninh) in 2012.

Citation: Asian Review of World Histories 5, 2 (2017) ; 10.1163/22879811-12340004

Photo: Phạm Lê Huy.

• Download Figure
• Download figure as PowerPoint slide

This identification of the original location of the sarira stele suggests that the area of the LÅ©ng Khê Citadel belonged to Longbian County under the Northern and Southern Dynasties—i.e., in the Sui–Tang periods.⁴⁵ Because Longbian County under the Tang was part of Longbian County under the Han, the area of the citadel was also part of Longbian County under the Han.

Regarding the location of the Longbian Citadel under the Han, Yuanhe junxian tuzhi suggests that in Longbian County under the Tang there was a citadel called “the old Jiaozhou Citadel” (交州故城) built by the Jiaozhou cishi, Tao Huang (陶璜), during the Wu period, about 14 li (7 km) to the east of the Longbian County headquarters.⁴⁶ Although there are still different views on the location of Leilou, the majority of contemporary scholars, including Nishimura, agree that “the old Jiaozhou Citadel” built by Tao Huang was the LÅ©ng Khê Citadel.

This view is further confirmed by a stele inscription discovered in 2013. In the shrine of Tao Huang in Thanh Hoài hamlet (Thuận Thành, Bắc Ninh), near the LÅ©ng Khê Citadel (Fig. 1), Nguyễn Phạm Bằng, a researcher from Bắc Ninh museum, discovered an ancient stone stele. On the front and back of the stele are two inscriptions with different dates. Although the inscription on the front has faded and hence cannot provide much information, the inscription on the reverse has a clear date of 450 ce. It was composed by a person named Tao Zhenzhi (陶珍之), a descendant of Tao Huang, at Tao Huang’s shrine on December 25, 450 ce. Tao Zhen Zhi performed this task at the command of the Jiaozhou cishi Xiao Jing Xian (蕭景憲), announced by a person named Wang Fa Ling (王法齡) on November 10 of the same year.⁴⁷

The discovery of this stele indicates that Tao Huang’s shrine in Thanh Hoài hamlet was built and redecorated no later than the fifth century. Because the shrine was located only 2 km to the west of the LÅ©ng Khê Citadel, it is important evidence that the LÅ©ng Khê Citadel was the Longbian Citadel built by Tao Huang.⁴⁸

Conclusion

In Vietnam, it has long been held that Leilou and Longbian were the same citadel (“Leilou = Longbian”). However, building on ideas proposed by Trân Quôc Vượng and Nishimura Masanari,⁴⁹ this paper has examined historical documents and reached the conclusion that in pre-tenth century records, Leilou and Longbian were listed as two distinct administrative counties (“Leilou ≠ Longbian”). Early records such as Hanshu, Hou Hanshu, and Shuijing shu provide good evidence for rejecting the view that Leilou and Longbian were the same citadel.

Regarding the location of the Leilou Citadel under the Han, some materials in post-tenth century literature state that the Leilou Citadel and Leilou County under the Han were located west of modern Hanoi. However, as Vietnam entered the era of independence, a new idea emerged that Leilou was located in the area of the Lũng Khê Citadel and surrounding areas, notably in Toàn thư and the chapters on Pháp Hiên and Thông Biện in Thiên uyên tập anh.

In the fifteenth century, Ngô Sĩ Liên and Toàn thư played an important role in disseminating the idea that Leilou was located in Lũng Khê. The equation of Leilou with Lũng Khê in Toàn thư was largely based on the statement in Việt điện u linh that “Shi Xie situated his headquarters in Leilou and Guangxin.” However, this statement cannot be trusted. Throughout his lifetime, Shi Xie’s highest position was the head of Jiaozhi Commandery, not the head of Jiaozhi Department (Jiaozhou). Shi Xie therefore could not have situated his headquarters in the Guangxin Commandery as written in Việt điện u linh. The confusion in Việt điện u linh (and later Toàn thư) may have originated because Shi Xie’s homeland was Guangxin, or because the author of Việt điện u linh failed to recognize the difference between the positions of the Jiaozhi taishou and the Jiaozhou cishi.

When Ngô Sĩ Liên compiled the Shi Xie chapter of Toàn thư, he cited the statement in Việt điện u linh as evidence that Shi Xie had situated his headquarters in Leilou. Because Shi Xie’s shrine was located in the area of the Lũng Khê Citadel, Ngô Sĩ Liên concluded that Leilou was Lũng Khê. And a new body of literature in the later period—Toàn thư, stone stele inscriptions (the earliest being the stele 羸𨻻城重興聖廟碑, 1661 ce), the records in Việt sử thông giám cương mục, and Đại Nam nhât thông chí—adopted the idea that “Leilou = Longbian = Lũng Khê,” leading to the conclusion that “Lũng Khê = Leilou.”

By examining the texts in Thiên uyên tập anh and Toàn thÆ°, it can be shown that apart from the idea that “Leilou = LÅ©ng Khê,” there is another, more widely held idea that “Longbian = LÅ©ng Khê.” In Thiên uyên tập anh, the view that “Leilou = LÅ©ng Khê” existed only in the Pháp Hiên and Thông Biện chapters,⁵⁰ while other chapters such as Tỳ Ni Đa LÆ°u Chi, Sùng Phạm, and Thiên Nham all agree that the area of LÅ©ng Khê and the surrounding areas were the same as Longbian. Although Ngô SÄ© Liên adopted the view that “Leilou = LÅ©ng Khê,” he phrased his notes in Toàn thÆ° in a manner that demonstrates that by the fifteenth century, the area of Shi Xie’s shrine (LÅ©ng Khê Citadel) was widely known as “the old Longbian Citadel,” not Leilou. Toàn thÆ° noted that Leilou “means Longbian,” further supporting the above view.

Given the existing evidence, the view that the Lũng Khê citadel was the Longbian Citadel (“Lũng Khê = Longbian”), represented by Nishimura Masanari, is the most plausible opinion and is consistent with what is written in both pre-tenth century texts and recently discovered stele inscriptions. The Sarira Stupa stele inscription of 601 ce clearly shows that the area of Tiên Du Mountain and the surrounding regions (including the Lũng Khê) belonged to Longbian County during the Han, Northern and Southern Dynasties, Sui, and Tang periods. During the Wu period, the Longbian Citadel was redecorated by Tao Huang, the Jiaozhou cishi. Although not of decisive importance, the inscription dated to 450 ce discovered to the east of Lũng Khê in 2013 supports this view. Archaeological findings show that the Longbian – Lũng Khê Citadel was used until the Northern and Southern Northern Dynasties and the Sui. Under the Tang, the Longbian – Lũng Khê Citadel was no longer used as the headquarters of Longbian County, the latter being 14 li east of the Lũng Khê Citadel.

If we accept the hypothesis of Nishimura Masanari that the LÅ©ng Khê Citadel was the Longbian Citadel, the question then is how to pinpoint the location of the Leilou Citadel under the Han. If Yuanhe junxian tuzhi’s assertion that the Leilou Citadel under the Han was located 75 li to the west of the capital of the Protectorate under the Tang is correct, it can be hypothesized that the Quèn Citadel in the area of Tuyêt NghÄ©a commune (Quôc Oai, Hà Nội) had some connection with the Leilou Citadel under the Han. Indeed, fieldwork on the Quèn Citadel has shown that it was constructed during the Han period.⁵¹