The erected, the erector, and those for whom the pillars were erected

The two pillars were excavated in 1962 in Hanzhuang Township, Baoding City. The archaeological site was reportedly the former location of a monastery with a white stupa, which had been destroyed in the early twentieth century.Footnote ⁹ Each pillar consists of three octagonal parts: a top cover, a body with eight panels, and a base. Tangut inscriptions are carved onto the panels of each pillar body. Pillar One, standing at 263 cm, is slightly taller than Pillar Two, which measures 228 cm. Both pillars are now housed in Lianchi Park in Baoding (see Figure 1).

Figure 1.

Tangut Dhāraṇī Pillars (right: Pillar One; left: Pillar Two), 1502. Lianchi Park, Baoding. Photo by Andrew West. September 19, 2017.

The inscription on each pillar consists of three main parts: a commemorative inscription, the Uṣṇīṣavijayā dhāraṇī in a Tangut phonetic transcription, and a list of patrons. For scholars documenting the pillars’ historical background, the commemorative inscriptions are the most significant parts. The commemorative inscription on Pillar One states the following:

Pillar Two states the following:

On the leftmost side of the first panel of Pillar Two, there is also a line of Chinese script:

The dates recorded in all three passages are consistent. Both pillars were erected in the fifteenth year of the Hongzhi reign, which corresponds to 1502. The erector is also the same—Phjij śjo Tśia śji rjijr dźji̱j, as the Tangut inscriptions state. Phjij śjo is a well-known Tangut surname, while Tśia śji rjijr dźji̱j is his religious name. Zhashi Lingzhan is evidently the Chinese rendering of this Tangut religious name. The Chinese passage further states that he was an abbot, presumably of the Xji śji sə monastery mentioned in the Tangut inscriptions. Both pillars were dedicated to deceased monastics: Pillar One to a śia mji (Ch. sha mi 沙彌, Skt. śrāmaṇera) named Pja dja̱ dow dźjij, and Pillar Two to a “phji khjiw [Ch. bi qiu 比丘, Skt. bhikṣu] master” whose actual name is not given. Notably, the “stupa courtyard” mentioned in Pillar Two’s inscription appears to correspond exactly with the excavation site of both pillars, in light of reports that a white stupa once stood there.

Xji śji sə Monastery evidently followed the Tibetan Buddhist tradition at least partially. Tśia śji rjijr dźji̱j is a Tangut phonetic rendering of the Tibetan religious name Trashi Rinchen (Tib. Bkra shis rin chen), while Pja dja̱ dow dźjij corresponds to Penden Dorje (Tib. Dpal ldan rdo rje). The white stupa, a distinctive feature of Tibetan Buddhist architecture, became especially prominent after the establishment of the Yuan dynasty, so this detail further reinforces the monastery’s connection to Tibetan Buddhism.

Xingshan Monastery and Tibetan Buddhism

As previous scholars have rightly noted,Footnote ¹⁵ Xji śji sə monastery is most plausibly identified with Xingshan Monastery, with xji śji sə being a phonetic transcription of the Chinese xing shan si 興善寺.Footnote ¹⁶ The only known historical source providing information about this monastery is a short text titled A Brief Account of the Renovation of Xingshan Monastery” (Chongxiu xingshan si jilue” 重修興善寺記略, henceforth the Account). Written in 1671 by Guo Fen 郭棻 (1622–1690), a Baoding native and a Qing official, it offers valuable insights into the monastery’s history. The full text is as follows:

None of the steles that Guo saw in the early Qing have survived, making Guo’s second-hand report the only source that allows us a glimpse of the monastery’s history. The last paragraph is particularly relevant to our discussion. The names of the three monastics it mentions—Bandan duanzhu, Shilai jianzan, and Bandan duo’erzhi—are clearly Chinese phonetic transcriptions of Tibetan religious names: Penden Döndrup (Tib. Dpal ldan don grub), Shérap Gyentsen (Tib. Shes rab rgyal mtshan), and Penden Dorje. Nie identifies the Penden Dorje in this account as none other than the deceased Penden Dorje mentioned in Pillar One’s inscription, suggesting that the renovation these three monastics led must have taken place between 1489 and 1501, the year of Penden Dorje’s death.Footnote ²⁰

This identification is further supported by the dates of Chai Wu, the benefactor, who was active in the late fifteenth century (see discussion below). However, one lingering inconsistency remains: Guo Fen explicitly states at the beginning of his Account that “stele inscriptions from the Yongle and Xuande eras of the Ming all record its renovations.” If the renovation the latter part of the text describes occurred in the late fifteenth century, it must have been distinct from the Yongle (1403–1424) and Xuande (1425–1435) renovations. This raises a question about the internal consistency of Guo’s Account, making his narrative appear somewhat unnatural or ambiguous.

Nevertheless, there is little doubt that the Xji śji sə Monastery mentioned in the pillar inscriptions is Xingshan Monastery, which thus links the two sources. By the time Penden Döndrup oversaw its renovation, Xingshan Monastery was already following the Tibetan Buddhist tradition; this connection most likely dates back to the Yuan dynasty, when the monastery was probably first established. The renovations during the Yongle (1403–1424) and Xuande (1425–1435) eras were also not coincidental, as both emperors were known supporters of Tibetan Buddhism. One remaining question concerns the ethnicity of the three monastics mentioned in Guo Fen’s account. While they could certainly be ethnic Tibetans, they could also be Tanguts, as Nie suggests, since Tangut monastics often adopted Tibetan religious names.Footnote ²¹

Dhāraṇī and Tangut Culture

Tibetan Buddhism began to exert a significant influence on Tangut religious culture in the early twelfth century. However, the earliest datable Tangut translation of a Tibetan Buddhist scripture, the Aparimitāyurjñānasūtra, was published earlier, in 1094. Despite being titled a sūtra, this text primarily consists of dhāraṇīs associated with Buddha Amitāyus. In fact, various dhāraṇī scriptures were among the first to be translated as Tibetan Buddhism gained prominence within the Tangut state;Footnote ²² notable examples include the Pañcarakṣā and the Sitātapatrā. The Uṣṇīṣavijayā, which is inscribed on both pillars, was translated from Tibetan sometime before 1149,Footnote ²³ when the Tangut court convened a grand religious assembly to distribute officially printed copies to the public as an act of merit accumulation.

It is no coincidence that the Tangut state decided to prioritize the translation of dhāraṇīs which were available in Tibetan. A defining feature of Tangut Buddhism during the Xia period was the state’s strong involvement in religious affairs. The government was particularly concerned with how Buddhist practices could contribute to state welfare. Reciting dhāraṇīs, in particular, was believed to have powerful transformative effects: good fortune, protection from disasters, extended lifespans, and the warding off of evil spirits. To safeguard both the royal family and the broader population, it was therefore logical for the Tangut state to prioritize the translation and institutionalization of dhāraṇī practices at the state level.

The Tangut state’s inclination toward dhāraṇīs is reflected in the Legal Code of the Tiansheng Era that Revises the Old and Ascertains the New (Ŋwər ljịj kjwi ljij sjiw djɨj kie dzjɨ̱ ), issued in 1150. One article requires monastics to recite a specific set of Buddhist scriptures, the majority of which were dhāraṇīs. Notably, certain scriptures were designated exclusively either for Tangut and Tibetan monastics or for Chinese monastics. The Uṣṇīṣavijayā was among the few required scriptures for all three ethnic groups, underscoring its significance.Footnote ²⁴ The state’s sponsorship of the Uṣṇīṣavijayā is also evident in its material culture, as Linrothe has shown.Footnote ²⁵ Its fascination with the Uṣṇīṣavijayā persisted into the Yuan period. In 1345 the dhāraṇī was inscribed in six scripts, including Tangut, on the walls of Juyong Pass, a strategically vital section of the Great Wall.

While the Tangut state played a central role in promoting dhāraṇī practices at the national level, they must also have been widespread at local and personal levels within the state, although historical documentation on this aspect is limited. Postscripts in printed dhāraṇīs from the Xia period suggest that certain editions were privately produced rather than government-issued. Some of them also indicate that the dhāraṇīs were so popular that the original print blocks quickly wore out, necessitating new copies.Footnote ²⁶

The Uṣṇīṣavijayā was no exception. A Tangut manuscript of the Uṣṇīṣavijayā from Kharakhoto (Or.12380/323; see Figure 2) contains the following postscript:

Figure 2.

Manuscript of the Tangut Uṣṇīṣavijayā dhāraṇī, 1165. Kharakhoto. British Library, Or.12380/323.

This date corresponds to the year 1165, not long after the Uṣṇīṣavijayā was translated into Tangut and published by the government. The words of the Tangut author, Tśhjiw ˑo Sjwɨ̱ tjɨ̣j, reflect his deep reverence for this this dhāraṇī and how profoundly impressed he was by its extraordinary efficacy. His printed edition is thus a valuable testament to its widespread acceptance among the Tangut people. Notably, the version under examination is not Tśhjiw ˑo’s original print but a manuscript copy of it. This further attests to the popularity of his private print, which evidently became the version most readily available to the scribe who produced our manuscript.

The Tanguts’ reverence for the Uṣṇīṣavijayā at both local and personal levels is also evident in their art. Two images of the Uṣṇīṣavijayā maṇḍala, painted on wood, were discovered in Kharakhoto (see Figures 3 and 4). Goddess Uṣṇīṣavijayā, the personification of the dhāraṇī, appears at the center of each painting, flanked by her two attendants and surrounded by various deities associated with her. Though somewhat difficult to discern, the complete Uṣṇīṣavijayā dhāraṇī in Tangut is inscribed in the background. Both paintings depict a benefactor in the lower right-hand corner—one male and one female. Their appearance indicates that they were likely Tangut aristocrats and possibly a husband and wife. Both the artistic style and the Buddhist ritual framework the maṇḍala suggests are clearly Indo-Tibetan rather than Chinese.Footnote ³⁰ This demonstrates how Tibetan Buddhism’s underlying influence helped to sustain Uṣṇīṣavijayā worship within the Tangut state.

Figure 3.

Uṣṇīṣavijayā Maṇḍala, ca. late twelfth century. Kharakhoto. Hermitage Museum, XX-2406.

Figure 4.

Uṣṇīṣavijayā Maṇḍala, ca. late twelfth century. Kharakhoto. Hermitage Museum, XX-2407.

A distinctive function of the Uṣṇīṣavijayā dhāraṇī in Chinese Buddhism is its use in delivering the deceased from the risk of falling into hell. This tradition of employing this dhāraṇī in funerary rites dates back to the early Tang dynasty, following the scripture’s translation into Chinese. The practice of inscribing this dhāraṇī on pillars dedicated to the dead emerged shortly thereafter; the earliest known case is from 732.Footnote ³¹

While this practice became widespread in medieval China, what remains unclear is how widely the Tanguts embraced this tradition and to what degree they modeled their practices on the Chinese precedent. A woodcut from Kharakhoto depicts a Uṣṇīṣavijayā dhāraṇī pillar (see Figure 5) whose goddess is iconographically almost identical to that of the maṇḍala paintings shown above. However, whether this pillar was specifically dedicated to the deceased remains uncertain. A more relevant example, discovered in 1977, is a miniature wooden stupa (Figure 6) from a Xia-period tomb in Wuwei, which was then known as Liangzhou. This stupa has four parts: base, body, top cover, and tip. Like the 1502 dhāraṇī pillars, its body consists of eight panels, each 35 cm tall. These panels are inscribed with dhāraṇīs in Pāla script. These inscriptions include several dhāraṇīs from different scriptures, and Wang Rongfei recently identified a mantra at the end of the text: oṃ bhrūṃ svāhā,Footnote ³² which is the distinct inceptive mantra of the Uṣṇīṣavijayā dhāraṇī in the Tangut tradition.Footnote ³³ The stupa is hollow, and the inside of its top cover features the following inscription (see Figure 7):

Figure 5.

Woodcut of Uṣṇīṣavijayā dhāraṇī Pillar, ca. late twelfth century. Kharakhoto. Hermitage Museum, XX-2536.

Figure 6.

Wooden stupa, 1200. late twelfth century. Wuwei. Wuwei Museum, G32·002. After Ningxia daxue xixiaxue yanjiu zhongxin et al., Zhongguo cang Xixia wenxian, 260.

Figure 7.

Inscription inside the top cover of the wooden stupa.

This inscription reveals that the stupa was dedicated to the tomb’s occupant, Liu Deren, and that he passed away in 1198. While this object is not a stone pillar, its function is identical to that of the 1502 dhāraṇī pillars. Notably, Liu Deren was not a monastic but an Inspector-General (du’an 都案), an official responsible for handling government documents. His name also suggests that he was Han Chinese rather than Tangut. This indicates that, though their promotion was initially a governmental effort, by this time dhāraṇīs in general, and the Uṣṇīṣavijayā dhāraṇī in particular, were widely practiced by different social and ethnic groups within the Tangut state.

In many ways, the 1502 dhāraṇī pillars represent a continuation of Tangut social and religious culture from the Xia period. The local community that produced them, much like the people of the Tangut state, demonstrated reverence for the Uṣṇīṣavijayā dhāraṇī, the Tibetan Buddhist tradition that sustained it, and the funerary practice of erecting pillars. The dhāraṇī text inscribed on the pillars closely follows the 1149 edition, with only minor errors; in terms of textual accuracy, it is of even higher quality than the Uṣṇīṣavijayā inscription on the walls of Juyong Pass,Footnote ³⁵ a project sponsored by the Yuan government. While the Juyong Pass inscription is more elegantly engraved, it contains significantly more textual errors and inappropriate renderings of the dhāraṇī than this local, mid-Ming production. This suggests that local Tangut religious traditions may have had a more stable foundation, and greater resilience against changing political and cultural forces at the national level.

Patrons of the Pillars

But who were the individuals supporting this tradition? The names recorded at the end of each pillar provide an invaluable source of answers. However, interpreting these names presents significant challenges. Many characters are illegible due to wear and erosion. The small size of the patrons’ names compared to the dhāraṇī text further complicates the task, and it is sometimes unclear how characters should be grouped to form a single name. This issue is particularly critical for Tangut names; their structure did not follow a fixed convention, as in other naming systems. To date, only Shi Jinbo has made a comprehensive effort to transcribe all the names,Footnote ³⁶ and even that study did not include a detailed analysis. My renewed attempt to engage with this portion of the pillar inscriptions is therefore far from final or definitive. Nevertheless, it contributes to the ongoing exploration of the social history surrounding these inscriptions.

The Appendix provides a complete list of the patrons’ names. Pillar One contains ninety-six names, while Pillar Two has forty-nine. Notably, no name appears on both pillars. This could mean that different patrons were connected to each of the two deceased monastics, but it is hard to imagine that all those who supported the building of Pillar One had no interest in supporting Pillar Two, and vice versa. An alternative possibility, given that both pillars were erected at almost the same time, is that each patron was considered a supporter of both pillars, not just the one on which their name appears.

Each name is analyzed according to a “style,” a classification based entirely on how the name is presented, without implying anything about the individual’s actual ethnicity. For instance, the “Tangut style” generally refers to names that include a Tangut surname.Footnote ³⁷ Similarly, Chinese-style names typically consist of Chinese surnames, phonetically transcribed into Tangut. Tibetan names, such as Sönam Gyentsen (Tib. Bsod nams rgyal mtshan; Pillar One, no. 2), are recognizable due to their phonetic transcription. Additionally, there are several Indic-style names, including Padmaśrī (Pillar One, no. 9), as well as one clearly Mongol name, Dašmad (Pillar One, no. 92). For non-Tangut-style names, I have provided reconstructions to aid in interpretation. This approach does have inherent limitations, particularly for Chinese names, as a single Tangut syllable can correspond to multiple Chinese syllables and various Chinese characters. Despite this, I believe the method remains empirically useful for readers who wish to engage with the names. The final category, “nature,” distinguishes between religious and secular names.

Table 1 presents a statistical breakdown of the names. One category not included in the appendix is gender. The approach to determining gender is necessarily approximate: any surname suffixed with zjɨ̣ (Ch. shi 氏) is classified as female.

Table 1.

Statistical breakdown of the patrons’ names

Since the style of a name does not directly indicate its bearer’s ethnicity, what insights can we still uncover about their ethnic backgrounds? First, it seems highly reasonable to suppose that individuals with Tangut-style names were indeed Tangut, as there is no known historical case of non-Tanguts’ adopting Tangut names.Footnote ³⁸ Conversely, many Tanguts did take Chinese names, a practice already observed during the Xia period. For example, Pillar One contains five Chinese-style names with the surname Liang 梁. Though a Chinese surname, it was also associated with a powerful and aristocratic Tangut family during the Xia period. Similarly, the Luo 羅 family, whose surname appears sixteen times, was influential in the Xia period. While this family may originally have been Chinese, intermarriage with Tangut aristocrats led to their gradual Tangutization. Other Chinese surnames, such as Ye 葉 (five instances) and He 賀 (six instances), were also adopted by Tanguts in the Xia period.

Individuals bearing Chinese-style religious names (thirteen instances) could be either Chinese or Tangut. As previously discussed, some Tanguts also adopted Tibetan religious names. Likewise, the one recorded Mongol name may have belonged to a Tangut who adopted a Mongol name, a common practice during the Yuan period. Taking all these factors into account, the percentage of ethnic Tanguts among the pillar inscriptions’ name-bearers is likely close to fifty percent.

The presence of four Indic names does not necessarily indicate that all these individuals originated from the Indic world. However, at least one seems to have had a direct connection: the first figure listed on Pillar Two, who is explicitly identified with the title “Indic master from the Western Heaven” (). Some of these figures may have been Uyghurs, who often adopted Indic religious names. It would have been unusual for a Chinese individual of this period to take a Tibetan name, even in a monastic context. However, Mongols, like Tanguts, may have adopted Tibetan names as well. At the other extreme, it is difficult to believe that no ethnic Tibetans were present at all. As the previous section demonstrated, Xingshan Monastery clearly adhered to Tibetan Buddhist traditions, and the presence of at least one Indic master shows that the monastic community did not have to be entirely local. Hence, it is reasonable to assume the presence of some ethnic Tibetans.

A religious name does not necessarily indicate a monastic identity. For example, the name ˑJɨ low dow dźjij (Pillar Two, no. 18) is religious in nature, as dow dźjij is a phonetic transcription of the Tibetan rdo rje (“vajra”). However, a monastic’s religious name usually has more than one semantic component. Since ˑJɨ low is a Tangut surname, this individual may have been a layperson whose given name was religious in form but not an indication of monastic status. Nevertheless, there are strong reasons to believe that most individuals with religious names were indeed monastics. The majority of these names are grouped together at the beginning of the lists, suggesting a hierarchical ordering that reflects their superior monastic status in a religious context. Pillar Two features another cluster of religious names at the end, possibly a later addition, yet still arranged as a distinct group. In contrast, ˑJɨ low dow dźjij appears as a single religious name within a sequence of secular names, further supporting the interpretation that this individual was a layperson.

It is difficult to identify female names within the secular category, apart from those containing zjɨ̣. Women who entered the monastic order under Chinese-style names also cannot be identified with certainty. Female names are predominantly Tangut in style, with only three exceptions. Two of these include the surname Liang (Pillar One, nos. 76 and 90), while the other (Pillar One, no. 39) is of uncertain reconstruction, making it unclear whether it was a Chinese name. The presence of female Tangut names in this context is not surprising, as women were already active participants in religious activities during the Xia period. Several prominent women, such as Empress Dowager Liang (d. 1099), were known for their strong patronage of religious assemblies.Footnote ³⁹ Buddhist paintings from the Xia period often depict female worshippers (see Figure 4), further attesting to their involvement in religious life.

One significant piece of information about the patrons comes from the lists’ use of official titles, affixed to personal names. On Pillar Two, three Tangut names—ˑJɨ low lo śjɨj (no. 16), Bji̱j mji □□□ (no. 21), and Sjɨ pji xu śjɨj (no. 27)—are preceded by a title, inscribed in a relatively larger size: źju lo thej xu (see Figure 8, upper circle). This title is a phonetic transcription of the Chinese ronglu dafu 榮祿大夫, “Grand Master of Honor and Emolument,” an honorary title (Ch. sanguan 散官) of the junior first rank. Below these names, another title appears: tu tśjĩ xwej , (Figure 8, lower circle), a phonetic transcription of the Chinese du zhihui 都指揮, “Commandant.” This was a functional official title (Ch. zhiguan 職官), granted to head officers of the Regional Military Command Office (Ch. du zhihuishi si 都指揮使司)—a second- or senior third-rank military position. This suggests that these three individuals simultaneously held both an honorary title and an actual official title,Footnote ⁴⁰ and that they served at a Regional Military Command Office—possibly that of Daning 大寧, whose headquarters at that time were in Baoding.

Figure 8.

Rubbing of Pillar Two (Detail). After Du and Deng, Dangxiang yu Xixia beike tiji, 2:411.

A Tangut’s holding a high-ranking military position was not unprecedented in Ming history. A Tangut version of the Scripture of the High King Avalokiteśvara (Ch. Gaowang Guanshiyin jing 高王觀世音經), likely printed in 1432,Footnote ⁴¹ records its principal patron as Zjwị ˑo ˑJi-r[jɨr] kja̱ tha-ˑjij . His surname clearly identifies him as Tangut, yet his given name, which can be reconstructed as Irgetei,Footnote ⁴² is Mongol, reflecting a partially Mongolized identity under the influence of the previous dynasty. His name is followed by the title tu tu , a phonetic transcription of the Chinese du du 都督, “Military Commissioner.” This was an even higher-ranking title than Commandant, corresponding to a military position of the first or senior second rank. His status within the Ming military hierarchy is consistent with that of the Commandants who supported the 1502 dhāraṇī pillars.

Pillar One provides another important clue: the title tśjĩ xwej, which straddles two lines of names. It corresponds to the Chinese zhihui 指挥, “Garrison Commander” (see Figure 9). A Garrison Commander was one of the head officers of a garrison (Ch. wei 衛)—the largest military unit under a Regional Military Command Office. However, it remains unclear to which specific names this title applies. Since it is positioned precisely between two lines of names, it seems likely that the four names below it (nos. 64, 65, 78, and 79) should be prefixed with this title—or at the very least, the top two names (nos. 64 and 78). I have been able to identify the individual who bore the first name, Tśjiw ˑji. In the register of the Baoding Left Garrison (Baoding zuo wei 保定左衛), which was under the jurisdiction of the Regional Military Command Office of Daning, the following record appears:

Figure 9.

Rubbing of Pillar One (Detail). After Du and Deng, Dangxiang yu Xixia beike tiji, 2:413.

On the same page, there is also the following record:

It seems highly reasonable to identify Tśjiw ˑji with Zhou Ying—and not merely due to their similar pronunciations.Footnote ⁴⁵ Zhou Ying was appointed Assistant Garrison Commander in 1473, and was stationed in Baoding. He likely passed away in 1504, at which point his son, Zhou Tang 周堂, inherited his position. These details align perfectly with the information recorded on the 1502 dhāraṇī pillars.

The Zhou family was not originally from Baoding. Their ancestor, Zhou Asun, began his military service in southeastern China, as a soldier in Zhu Yuanzhang’s army. However, by Zhou Cheng’s time the family was already stationed at the Regional Military Command Office of Daning, whose headquarters were then located not in Baoding but outside the Great Wall, in present-day Inner Mongolia. Zhou Cheng apparently joined Zhu Di’s army during the Jingnan Campaign, and was promoted as a result of its victories. In 1403, following Zhu Di’s ascension as the Yongle Emperor, the Command Office was relocated to Baoding; this was likely the beginning of the Zhou family’s history in the prefecture. Zhou Yao earned significant military merit in 1449 during the defense against the Oirat Mongols, and was ultimately promoted to Assistant Garrison Commander. He did not inherit his father’s position immediately after his father’s death; this suggests that before taking office he still had to reach adulthood, normally sixteen years of age. This means Zhou Yao likely came of age in 1426, which in turn suggests that he was already a Baoding native, possibly born around 1411. Zhou Ying, then, already represented the second generation of the Baoding-based Zhou family.

Unlike the head officers in a Regional Military Command Office, who obtained their positions by appointment, the head officers in a garrison could inherit their posts. These positions were passed down within military households, as the case of the Zhou family exemplifies. In particular, since the Zhou family stood with Zhu Di in the Jingnan Campaign and served meritoriously, they were regarded thereafter as the “New Officials” (Ch. xin guan 新官) within the garrison system,Footnote ⁴⁶ entitling them to a rich hereditary emolument. This further solidified their continuing local presence. In fact, their position of Assistant Garrison Commander in Baoding Left Garrison remained within the Zhou family until the death of Zhou Shangde 周尚德 in 1636. His son, Zhou Zhizhen 周之禎, was then only four years old but continued to receive the emolument until 1647, even after the Ming dynasty fell. This system of fixed stationing and hereditary succession made military households like the Zhou family a significant social force in areas such as Baoding, where multiple garrisons were concentrated.

A similar lineage is that of the Chai family, whose members appear eight times in total across both of the 1502 dhāraṇī pillars. Xu Cheng has examined the Chai family in Baoding and identified Chai Wu, the benefactor who helped renovate Xingshan Monastery (discussed above), in the register of Baoding Left Garrison.Footnote ⁴⁷ Chai Wu became Deputy Garrison Commander (Ch. zhihui tongzhi 指揮同知) in 1476, inheriting the position from his nephew, who had died prematurely. Chai Wu passed away in 1501, just before the construction of the dhāraṇī pillars, which explains the absence of his name from the list of patrons. His career closely overlapped with Zhou Ying’s; the two served together as colleagues and head officers in the Baoding Left Garrison for nearly three decades.

Unlike the Zhou family, the Chai family was granted the additional title “High Official.” However, the Chinese term da guan 達官 could also mean “Tatar Official” (鞑官), a title frequently bestowed upon nomads—primarily Mongols—who became officers of the Ming military. Indeed, according to the military register, the Chai family originated from Ta Tan 塔灘; this geographical designation generally refers to the northern part of the Hetao region. The family’s early patriarchs retained Mongol names as part of their given names, further indicating their nomadic ancestry. Therefore, the Chai family members listed on the dhāraṇī pillars were most likely from the same lineage as Chai Wu, and their Mongol heritage remained an integral part of their identity.

Now, let us piece the puzzle together and reconstruct a picture of the local community that sustained Xingshan Monastery’s activities in Baoding. This was a multiethnic community: approximately half Tanguts, some Chinese, a few Mongols, several Tibetans and Indic figures, and possibly one or two Uyghurs. Monastics made up only about a fifth of the population, but within this religious setting they held a higher status. Most laywomen were Tanguts, while all men with official titles were military officers; these included Tanguts, Chinese, and Mongols. Some of these officers belonged to hereditary military households, permanently stationed in Baoding due to their garrison assignments. This fixed military presence ensured their long-lasting local influence, shaping both military and religious affairs in the region.