The multiple siyin half seals: reconsidering the Dianli jicha si (1373-1384) argument.
Over the past three centuries, discrepancies have arisen in academic studies of the siyin half seal. Several late Ming and early Qing dynasty (1644-1911) court officials deciphered the half seal as representing a jingli si (registry office). This recognition would later fall into disfavor, as Qing private collectors launched a new reading of the characters on the seal as "jicha si," an office which did not historically exist but would loosely translate to "Office of Regulations and Investigations." The Qing understanding is echoed by twentieth-century scholars of Chinese art, who have developed an alternate explanation of the first character and thereby assign the siyin seal to the eunuch-run Dianli jicha si (Office of Regulations and Investigations) between 1373-1384, during the reign of the first Ming Emperor Hongwu (r. 1368-98).
Through a forensic tracing of the siyin art pieces, Ming court diaries, official memorials, legal statutes, and other governmental seals ending with the characters si and yin. I demonstrate that the siyin seal was not used by the Dianli jicha si office. In contrast to twentieth-century scholarship, I argue that late Ming officials in fact correctly identified the jingli si registry office (1368-1644) as the seal user. This unstudied office was in charge of inventorying items; its branch offices were distributed empire-wide throughout the Ming era. Different branch offices marked their own jingli siyin seals on the art seized from multiple ledger on the right. Today, the ledgers have all vanished and only the left halves of the seal impressions remain, leaving two full and two partial characters arranged in two columns. While the existing half of the seal impressions clearly reads siyin, which literally means "office seal," the identity of the office remains unknown because of the missing characters. In addition, the governmental office's duration, function, operational mechanism, and relationship to the imperial court are all obscure. Decoding the missing characters should reveal significant information regarding the legal system regulating the half seal and the imperial agency administering this system. By deconstructing the xi yin historiography, this paper raises new socio-political questions about ownership, censorship, and imperial competition over the siyin-marked art pieces. regions, resulting in nuanced differences in seal dimensions and stroke configurations. This finding challenges the binary scholarly categorization of siyin impressions as either genuine or counterfeit. Instead of assuming that only a single siyin seal existed throughout the Ming dynasty, I contend that three types of seals can be considered authentic. By critically reviewing the siyin historiography, this paper builds the foundation for a forthcoming paper which will determine the alternative origins of the siyin seals and reconstruct the jingli Si branch offices through tracing intricate networks of art ownership.
JINGLI SIYIN: MING INTERPRETATIONS OF THE HALF SEAL
Siyin historiography begins around 1616, when the late Ming historian Shen Defu ' (1578-1642) encountered various types of siyin impressions on paintings and calligraphies, including two genuine and numerous fake types. According to Shen, the two authentic types of siyin seals were used by two prefectural jingli si offices on imperially confiscated art: thus the full names of these seals are Yuanzhou fit jingli siyin (Seal of the jingli office of Yuanzhou prefecture) and Jingzhou fu jingli siyin (Seal of the jingli office of Jingzhou prefecture).2 Shen's contemporary Wen Zhenhen (15851645), Secretariat Drafter and Supervising Secretary at the Wuying Palace Hall, gained special access to Ming court diaries and the imperial collection. Upon viewing Wu Daozi's (689-after 755) painting, Sandalwood Image of Deities (Zhantan shenxiang), Wen wrote a colophon to pinpoint the half seal impression on this piece as a jingli siyin. (3)
Although Shen did not specify which artworks displayed the half seals, and the Sandalwood Image recorded by Wen is now lost, we can still reconstruct the half seal's appearance. Crucial evidence can be found in Huang Tingjian's calligraphy, Transcribing Du Fu's Poem to Helan Xian, which survives and bears the siyin seal in the lower right corner (Fig. 1). When Emperor Qianlong's (r. 1735-95) officials compiled the Qing imperial catalogues in 1744, they defined the half seal on Huang's calligraphy as jingli siyin.4 These entries foreshadow later descriptions by Li E (1692-1752) and Chen Weizhen (18th c.). The two Qing historians reiterate that the siyin seals on art were the result of confiscating artworks from the residences of two deposed Ming ministers by two prefectural jingli si registry offices.5 Since several Ming-Qing historians acknowledge the half seal as jingli siyin, we can exclude the possibility of typos or misprints.
Jingli Si literally means "registry office." This office attracted little attention in twentieth-century scholarship, yet played an essential role in the clerical systems of the Yuan (1279-1368), Ming, and Qing dynasties. Starting from the Yuan era, the palace library, six ministries, Bureau of Military Affairs, and myriad guard units all supervised their own jingli si offices that were responsible for cataloging official documents and property. (6) For example, the jingli si office of the Yuan palace library registered the imperial art collection and classical literature; its seal (jingli sun) was cast in 1285. (7)
Just as with the preceding Yuan clerical system, the first Ming emperor installed various jingli si offices after he established the dynasty. (8) Manned by trained clerks that registered the arriving items, rejuvenated Ming jingli si offices spread across the country. The six ministries, the Court of Judicial Review, the Censorate, the Embroidered-Uniform Guard, the Five Chief Military Commissions, the Office of Transmission, the Court of the Imperial Clan, the censorial departments, and many prefectures each opened their own jingli si offices. (9) Eachjingli si office possessed its own seal with varied first characters and dimensions according to their official ranks. (10) For instance, the Collected Statutes of the Great Ming stipulated that the jingli siyin seals used by the military guards measure 2.1 cun long, whereas a prefectural jingli siyin seal measures 2 cun. (11) Although half a century ago there existed debates on how to convert cun to centimeters, recent scholarship generally approximates 1 cun to 3.26-3.3 cm. (12) This conversion is not only consistent with the existing siyin impressions, which measure 6.6-6.9 cm, but also echoes Shen Defu and Wen Zhenheng's ascription of the siyin to the jingli siyin seal. Unfortunately, this identification was obscured by Qing private collectors, as presented below.
JICHA SIYIN: THE ALTERNATIVE INTERPRETATION OF QING MERCHANTS
Early in the Qing era, two theories concerning the fun names of the siyin seal held sway. The first, jingli siyin, promulgated by the Qing imperial court and Li E, was a holdover from Shen and Wen. The second, jicha siyin, was created by private collectors in 1692 and was advanced by art historians for three centuries. I argue in this paper against the second theory and in favor of the first.
In their catalogues published in 1692 and 1693, the two wealthy art aficionados Gu Fu gni (ca. 1662-1692) (13) and Gao Shiqi (1645-1704) observed that several artworks in their collections bore the same half seal. (14) Due to the paucity of precise information, Gu misattributed the siyin seal to a nonexistent jicha si office of the Song dynasty (960-1279). (15) However, twentieth-century scholars modified Gu's conjecture and argued that the siyin seal hailed not from the Song, but from the Ming dynasty, and that the so-called jicha si office never existed. Furthermore, Gu interpreted the characters to be ji and cha (both literally mean "surveillance"), but in fact the phrase does not designate any particular office, but rather refers to any office charged with censorship, policing, or adjudicating. Both Ming- and Qing-era legislators customarily referred to such offices as jicha, (16) jicha. jicha, (17) jicha, jicha, or similar terms. (18) In Hongwu's edicts, the characters that make up jicha are interchangeable with other homophonic synonyms, such as jicha.(19) Similarly, the renowned Ming novel, The Plum in the Golden Vase (1580s-1600s), employs homophones of jicha to ridicule the ubiquitous spy agents, who terrorized officials and commoners under the emperors' mandates. The novel's author sarcastically relates jicha to secret services like the Embroidered-Uniform Guard and Eastern Depot.
The Qing period also saw the frequent use of jicha and its homophones; private collectors such as Gu Fu, Gao Shiqi, and An Qi all made ample use of the term. (20) At the time when these collectors encountered the siyin seal, various jicha stations existed. The Qing government not only institutionalized the jicha headquarters (jicha fang) to monitor the six ministries, but also placed jicha si offices in county governments to conduct criminal trials. Officials stretched the term jicha to designate an array of government offices, so that any event, case, or agency subject to government inspection could be tagged as jicha Si. In accordance with Ming-Qing practice, Gu read the siyin half seal as jicha siyin, but did not specify the date or source of the seal. Subsequently, Qing merchants and publishers espoused Gu's vague reference to jicha siyin. In other Qing editions, texts denoting siyin were altered to jicha siyin or other homophones. Certainly Gu would not have anticipated that, after three centuries, his vague annotation would spur twentieth-century researchers to hunt for an office to fit his ambiguous jicha si designation.
DIANLI JICHA SI: TWENT1ETH-CENTURY INTERPRETATIONS
More recent art historians, who realize that jicha si does not match any real office's name, have endeavored to solve this puzzle: Ma Heng (1949), (21) Chen (1956), (22) Max Loehr (1961). (23) Na Zhiliang mizLL (1970), (24) Zhuang Shen (1972), (25) Liu Jiu'an 11.0L14 (1972), (26) Cheng Qi (1973), (27) Hironobu Kohara (1975), (28) Jiang Zhaoshen (1977), (29) Jiang Yihan (1980), (30) Fu Shen (1982), (31) Suzuki Kei (1988), (32) James Cahill (1999), (33) and Ding Xiyuan (2001; 2005), (34) among others. Unaware of the variety of jicha si nomenclature in the Ming-Qing dynasties, they have scoured the records in search of the elusive office named jicha Si.
How can it be, then, that scholars have located a specific office as the seal user? The first misstep was to latch onto homophones. Throughout the Ming era, the eunuch office jicha si altered its name four times as it transformed into new incarnations: Neizheng si (Justice office of inner palace eunuchs; only 1373/10), Dianli si (1373/11), Jicha si (1374/11-1375/3), Dianli jicha si (1373/11-1374/11; 1375-1384/4), and finally Sill jian (Directorate of Ceremonial, 1384/4-1644). (35) Since the jicha si could represent either the jicha si office or an abbreviation of the Dianli jicha si office, further controversy surrounds how many characters constituted the siyin seal. Assuming that the four characters were ji cha si yin. Suzuki Kei surmises that the seal existed for only two months, from 1382/1 to 1382/3. (36) Others have opted for the six characters Dian ji cha si yin.
Regarding the acquisition of the siyin art pieces, the aforementioned scholars extract from the Ming shi (1739) that in 1368/8 the Ming general Xu Da (1332-1385) marched into the Yuan palaces to remove treasures. They estimate that the loot upon which the Dianli jicha si office imprinted the siyin seal was delivered from Dadu (modern Beijing) to the Ming capital Nanjing. (37) But as Zhuang Shen admits, only twelve works bearing the siyin seal also display Yuan royal seals (e.g., Emperor Wenzong or Princess Sengge Ragi) or the Yuan official seal Dusheng shuhua zhiyin (Seal of calligraphy and painting inspected by the Secretariat Department). (38) In addition, two late Yuan paintings carry the siyin seal: Tang Di's (1287-1355) Fishermen Returning on a Frosty Bank (Shuangpu guiyu zhou, dated 1338; NPM) and Ma Wan's (ca. 1310-1378) Secluded Dwelling amid Lofty Peaks (Qiaoxiu youju zhou dated 1349; NPM). Fu Shen argues that, given the completion of the paintings at the end of the Yuan dynasty, it is unlikely that these paintings would have entered the collection of the already deteriorating Yuan court, which collapsed in 1368. Fu consequently offers the possibility that a few siyin artworks might have been levied from the southern prefectures during Hongwu's early reign. (39)
As for the issue of why the siyin impressions are mostly vertical but occasionally horizontal, Zhuang Shen speculates that the registrar workforce--either marauding soldiers of Xu Da or eunuchs from the Dianli jicha si office--stamped the siyin seal in random orientation due to their illiteracy and ignorance of the internal content of the art. Misreading a different seal as an upside-down siyin (on Auspicious Pines in Spring Mountains attributed to Mi Fu), Zhuang further concludes that the siyin was stamped in a rush. (40)
Following the thread that connects the siyin seal to the Dianli jicha si office, Zhuang, Liu, and Jiang Yihan base their approximation of the seal's date on two clues. First, this office operated ca. 1373-1384. Second, several artworks bearing the siyin also carry the seals of Hongwu's sons, such as the third son, Prince Zhu Gang (1358-1398), and the tenth son, Prince au Tan (1370-1389). (41) Three paintings bearing the siyin were excavated from Zhu Tan's tomb in Shandong. The Shandong Provincial Museum curator, among others, opines that Zhu Tan's enfeoffment in 1385 signifies the terminus ante quem for the siyin seal on the three buried paintings. Because, the theory runs, a prince's enfeoffment symbolized his permanent exodus from the capital, Zhu Tan must have acquired the siyin art pieces from Hongwu prior to his emigration. (42) Nevertheless, as proclaimed by Hongwu, his sons still took turns visiting the capital Nanjing after their departure. (43) With the emperor's special permission, the princes could temporarily enter the court according to seniority. (44) Hongwu's edicts from 1379, 1380, 1381, 1382, 1393, 1394, 1396, and 1397 reveal that he sent messages and goods to Prince Zhu Gang many times and summoned the prince to court. (45) Thus enfeoffments cease to clearly indicate dates for acquiring artworks from Hongwu's collection.
Zhuang, Liu, and Jiang proffer additional hypotheses about which storeroom supervised by the Dianli jicha si office preserved the siyin art pieces. They quote three seventeenth-to-eighteenth-century sources to propose that the siyin artworks were amassed in the Shuji minghua ku (Storeroom of Classic Literature and Famous Paintings). (46) However, no known Ming document describes a stockroom belonging to the Dianli jicha si office. The storeroom actually emerged after Hongwu abrogated the Dianli jicha si office in 1384, when the storeroom fell under the jurisdiction of the Directorate of Ceremonial.
Notably, the term jicha si (not a precise office's name) occurs six times in Gu Fu's catalogue, but the characters denoting the eunuch-run jicha si office do not appear. Twentieth-century scholars have overlooked the low status of the eunuchs and the legal restrictions inflicted on them under Hongwu's tyrannical reign, particularly from ca. 1368 to 1384, a time when the ruler stifled the eunuchs. Moreover, the siyin seal strokes and dimensions actually vary among good samples (see below), but the Dianli jicha si office ran for only eleven years.(47) The most apparent lacuna in the scholarship is that, as Suzuki Kei acknowledges, no record substantiates the claim that this eunuch office had its own seal. (48) Ding Xiyuan admits that only a small percentage of art pieces sport both the siyin seal and the Yuan imperial seals, which undermines the theory that the siyin artworks derived from the Yuan palace. (49)
The three interpretations of the siyin seal that we have reviewed (jingli siyin, jicha siyin, and Dianli jicha siyin) direct us to multiple offices, one of which (jicha si office) did not actually exist. To resolve the academic discrepancy, I will now scrutinize the siyin seal impressions in order to determine whether any of the interpretations accord with the character strokes.
CONNOISSEURIAL SCRUTINY: SEAL STROKES DO NOT MATCH THE CHARACTERS JI AND CHA
Visual analysis of the seal strokes challenges the assumption that the Dianli jicha siyin was the full name of the siyin seal. A few siyin impressions retain fractions of two partial characters in the middle column, which Ma Heng and subsequent scholars read as the components and of the characters ji and cha V (Fig. 2a). (50) Based both on this identification and on the characters' homophonic relationship to the jicha si mentioned by Gu, modern scholars have concluded that the eunuch-run jicha si office employed the siyin sea1. (51) Should the academic reading be correct, the character cha would be comprised of three components: (which should include a dot at its top), (whose center should consist of two long, unbroken strokes), and However, careful scrutiny of the siyin images overturns the jicha si ascription. As Figure 2b demonstrates, the first radical lacks a dot, and the second radical contains short and broken strokes. Such visible discrepancies recur in other siyin impressions, contradicting the accepted office name jicha si. Since the characters do not support the jicha si reading, the siyin seal must have derived from another office.
Because the top component does not contain the critical dot, I read the marking as rather than. The middle component should be read as (instead of), since features short vertical lines and, in conjunction with and other components, forms various characters. The bottom component should then be. The resultant missing second character must be li (rather than cha). As my drawing in Figure 2c shows, the seal can consequently be reconstructed as jing Ii siyin, which buttresses the identifications by the Ming-Qing court officials and repudiates the twentieth-century assertion of the jicha siyin seal.
MODULAR SYSTEM: MULTIPLE SEALS FROM THE MASTER MODEL
Why, then, did the Ming state cast assorted types of siyin seals for the jingli si registry office, and how did these types correlate to that office? The subtle variations in siyin seal strokes discussed below corroborate that multiple authentic siyin seals existed and were manipulated by more than one si-level office. The empire-wide jingli si office, with its multiple branches, is a better candidate to have wielded multiple genuine seals than the Dianli jicha si office, as the latter only operated for a scant eleven years.
Interestingly, many other surviving official copper seals end with the same siyin characters, though researchers have not previously noted this parallel. Figure 3 itemizes assorted official seals and impressions, dating from 1130 to 1620, in chronological order. The two characters si and yin of the half seal on art (hereafter our siyin seal) stylistically resemble those used by offices with equivalent hierarchical status, but deviate from prior Song, Jurchen Jim (1115-1234), and Yuan dynasty seals. (52) Typically, Ming governmental seals feature straightforward lines, sharp turns, and an elegant courtly style, which contrast with the round, soft, and free compositions of the Song, Jin, and Yuan seals. Ming and pre-Ming seals differ in compositional structure, seal dimensions, and border thickness--differences that reflect how the Ming government codified their own aesthetic and stroke structures.
The Seal Service of the Ministry of Rites employed seal-drafters both to compose new designs for empire-wide official seals and to convert them into master models so that further casts could be made. The Seal Service poured molten copper or silver into molds through a formalized seal-casting procedure. Governmental seals reflected hierarchical categorization: ministry (bu), directorate (jian), office (si), and bureau (ju). Seals for institutions of comparable authority shared comparable dimensions and stroke con-figurations. All si-level seals end with the same characters si and yin, such as can be found in the seal impressions of the Office of the Superintendent of Paper Currency (Baochao tiju siyin, Fig. 3b) from 1375, of the Pacification Office in Jintong (Jintong unfit siyin, Fig. 3c) from 1407, of the Pacification Office in Dayuan (Dayuan xuanwei siyin) from 1606, and others. I have discovered ten more examples of si-level seals that hail from various regions such as Nanjing, Tibet, Hubei, Jilin, and Fujian. (53) Throughout the samples, the layout of the characters si and yin is generally consistent after the advent of the Ming dynasty, even though the seal dimensions deviate. The consistent composition of these si-level seals is reflected in our siyin half seal.
Figure 3a shows a representative of our siyin seal juxtaposed with a Baochao tiju siyin seal impression from 1375 (Fig. 3b). (54) Figure 3c is a detail of the left half of the Jintong anfir siyin seal impression dated 1407, measuring 7.3 x 7.3 cm, slightly larger than our siyin seals. Strikingly, our siyin seal has a similar layout to other si-level office seals dated 1375, 1406, 1407, 1408, 1606, and 1645. (55) These remarkable stylistic resemblances suggest that our siyin seal was created between 1375 and 1645.
In order to verify the authenticity of official seals, the Seal Service of the Ministry of Rites cast copper seals from a single master model. Mirroring the movable type printing process, this copper casting procedure afforded the Seal Service the ability to alter the first few characters while retaining the last characters. Therefore, instead of merely referring to the ficha si or Dianli jicha si office, our siyin seal could also have been left over from another si-level office. The aforementioned Baochao tiju siyin and Jinwng anfu siyin seals have analogous character layouts and numbers. The elongated shape of the characters si and yin makes clear that our siyin impressions represent only one third of a three-columned seal. As shown in Figures 3a-c, each column accommodates two characters. The middle column denotes the office's name and the right column would have indicated which superior department oversaw the office. However, one cannot rule out the possibility that the right column of our siyin seal fit three characters to denote a long prefectural name, as exemplified in 7itduhewei zhi-huishi siyin (see n. 53) and other si-level seals. If this is the case, our seal could have been comprised of six or seven characters, depending on the region and superior department with which it was affiliated.
SIMILARITY AND DIFFERENCE: MULTIPLE SIYIN SEALS ON ART
Since multiple si-level office seals originated from a master model, the shared characters si and yin featured analogous but not identical stroke outlines, The master model was fashioned after a design disseminated by the imperial court. For both reasons, the casting process resulted in subtle distinctions in the stroke configurations. This helps to explain why existing siyin impressions on art display many more discrepancies in composition and dimension than might be expected.
In order to explore deviations in impressions made from the same copper seals, I assemble two impressions printed from the same seal, Tuduhewei zhihuishi siyin. Though the two impressions were stamped on different materials using different pressing strengths and ink pastes. these conditions impact stroke thickness rather than structural configuration or turning angles. On the other hand, post-impression distortion of the warp and weft of the silk could change the stroke structure, a possibility that I took into account in my analysis.
Considering the reasons for variations in the seals is pivotal to authenticating atypical siyin impressions. Thus, while art historians customarily divide the various siyin impressions into genuine and counterfeit, the evidence I present here contradicts this binary approach. The subtle distinction of seal structures must derive from separate casts, meaning that multiple siyin seal objects existed. This deduction also resolves the academic confusion about why the dimensions of seemingly authentic siyin would measure differently: measuring the siyin impressions on eighteen artworks, Zhuang Shen found seal dimensions varying from 6.5 to 6.9 cm.56 The NPM curator measured the siyin on the Song painting Calico under Peonies (Fuguihua ii zhou) at 6.8 cm. (57) Ding Xiyuan determined 6.8 and 7.0 cm. (58)
Of all 199 existing or now-lost siyin impressions, 148 specimens survive. (59) At least forty-eight of these appear on forged paintings and calligraphies from the fifteenth to the twentieth century or exhibit dubious features such as wrong characters misshapen structures, crooked strokes, or smaller dimensions. (60) Examples include assorted fake siyin seals on Feng Chengsu's (ca. 627-650) Copy after Preface to the Orchid Pavilion Gathering (Mo Lanting xu, PPM), on Zhang Daqian' (1899-1983) forgery of Ju Ran's, (ca. 960-980) Dense Forests and Layered Peaks (Maolin diezhang zhou, British Museum), and on Flowers and Birds Sketch (Xiesheng tujuan, NPM) attributed to Mu Xi (ca. 1210-1269). The siyin impression on the Flowers and Birds Sketch measures only 6.1 cm in height, which indicates that this seal was a counterfeit. As witnessed by the late Ming historian Shen Defu in 1616, in his time unscrupulous art dealers affixed forged siyin seals on fabricated artworks. I group all poor quality seals as Type A--a collective term for forty-eight doubtful siyin impressions--and will exclude them from my discussion.
Disregarding the now-lost examples and the dubious Type A siyin, the remaining 100 siyin examples constitute the heart of my investigation. They show refined spatial arrangement, well-formed composition, and appear on ancient masterpieces that ordinary Ming collectors would not have had the chance to acquire. The more complete transmission history of these canonical siyin artworks adds to the authenticity of their siyin impressions. (61) I classify these 100 impressions into three different types based on stylistic comparison.
Despite the shared features of the seal impressions, some deviations cannot be accounted for by printing irregularities, and must instead stem from differences in the three seal objects themselves. Designating the three types as X, Y, and Z, I photocopied each type of seal image onto a transparency maintaining its original dimension and proportion, and layered the transparencies on top of the siyin impressions. Since I do not have access to all 100 siyin artworks due to museum restrictions, the ratio for the Types X, Y and Z below are approximations that integrate my own and other scholars' measurements, as well as catalogues which specify the enlargement. (62) Roughly 1/10 of the siyin samples match Type X, including Qian Xuan's (1235-1305) painting White Lotus Scroll (Bailian tujuan, Fig. 4a) and two twelfth-century painters' Autumn Mallows (Kuihua jiadie tuanshan) and Golden Landscape (Song jinbi shanshui, Shandong Provincial Museum). About half of the siyin impressions fit Type Y, including those on Wang Wei's (701-761) Fu Sheng Expounding the Classics (Fu Sheng shoujing tujuan Fig. 4b), and on Zhao Gan's (10th c.) painting Early Snow along the River (Jiang xing chuxue tujuan, Li NPM). Finally, the remaining forty or so siyin samples match Type Z, including those on Ma Lin's (ca. 1180-after 1256) painting Layers of Icy Silk (Cengdie bingxiao, Fig. 4c), on Jia Shigu's (ca. 1131-1162) Ancient Temple in a Mountain Pass (Yanguan gusi tuce, Fig. 5a), and on a Song album, Ape and Heron (Yuan lu tuce, Shanghai Museum).
As Figures 4a-c demonstrate, Types X, Y, and Z siyin impressions diverge in line angles, border thickness, and seal dimensions. Their heights vary slightly from 6.6 cm (Type Y) to 6.8 cm (Type Z) and 6.9 cm (Type X) according to my own measurements. These different heights attest that the impressions came from multiple seal objects, which explains the different dimensions among the siyin impressions assumed to be authentic by scholars.
The metallurgical process of replicating a standard governmental seal further confirms that the three types of seals were cast at different times. Each seal type features analogous but not exactly identical solid bulges on their upper left edge (Figs. 4a-c). Since these seals were cast in liquid copper, the bulges suggest empty spaces between the inner and outer molds. Types X, Y, and Z seals all carry a large bulge at the same corner (which the fake impressions do not have). These bulges may represent the spot where molten metal emerged from a tube used to direct it onto what would become the seal surface. When the molten copper was poured through the tube, air hidden inside the joint between the tube and the seal surface would prevent it from fully filling the outer mold, resulting in an imperfection in the surface of the seal when the liquid copper solidified. The bulges in the three types of siyin take different shapes, indicating that the seals were cast separately.
Types X, Y, and Z siyin impressions also feature different linear angles and compositions, as labeled in Figures 4a-c. Taking the character si as an example, the two short horizontal strokes in the middle of each of the seals vary in their respective heights. Type X has a higher left shoulder and Type Y has a higher right shoulder. The left shoulder of Type Z is similar to that of Type X, but Type Z tilts the right shoulder outwards while Type X tilts the right shoulder inwards. Another deviation can be detected in the lower right corner of the first character. Type X shifts from vertical to horizontal lines smoothly with a curve, in contrast to Types Y and Z, which turn at a 90-degree angle. Likewise, at the very bottom strokes of the same character, Type X shows a straight horizontal line turning downward at a right angle on the left side, but Types Y and Z display a slight curve in the same horizontal line, ensuring that the lines' intersection with the left vertical is not at a perfect right angle.
Nonetheless, these three types all exhibit courtly, elegant layout, with evenly spaced lines, whose neat, precise execution testifies to their courtly origin. Each type of siyin appears on several exquisite sixth to fourteenth century artworks with genuine Song-Yuan seals, which corroborates the authenticity of the siyin seals.
All of my preceding arguments suggest that the Ming central government had created at least three siyin seal objects for three si-level offices over the course of the dynasty. These offices were in charge of similar duties on similar hierarchical levels, like prefectural branch offices. The Dianli jicha si office alone was unlikely to use multiple siyin seals. This eunuch office, which only existed from 1373 to 1384, had no reason to create so many seals during its brief lifespan. Moreover, the alleged jicha si characters do not match the siyin strokes. Compounding the visual conflicts are textual discrepancies between the scholarly hypothesis and reliable Hongwu-era accounts. For one, Ma Heng and subsequent scholars often cite eighteenth-century sources, such as Ming shi, to conflate the much more influential Directorate of Ceremonial with the earlier and weaker Dianli jicha si office despite the temporal gap in their operations. While historians of the Ming period Hiroshi Danjo (1978) and Sarah Schneewind (2002) have traced how Hongwu frequently abolished and reinstituted the administrative system, (63) this analysis is seldom incorporated into art-historical scholarship, which rarely addresses the instability and evolutionary pattern of the Ming institutions. Modern art historians have relied on historiographical misconceptions embedded in post-1644 sources, leading to widespread misapprehension of the siyin seal. The following sections contrast Hongwu-era accounts with the siyin story rendered by art historians.
THE YUAN-MING TRANSITION: THE SEAL THAT GENERAL XU DA STAMPED ON PIECES IN HONGWU'S COLLECTION
Zhuang Shen and Jiang Zhaoshen aver that general Xu Da affixed the siyin seal on art taken from the Yuan palace Dadu in the turmoil of regime changes. An analysis of statistical data contravenes this possibility. As Ding Xiyuan contends, only a few artworks, which later were imprinted with the Slyill seal, came from the Yuan palace. (64) In fact, Xu commanded his army to obey a rigid rule which segregated the Yuan assets. (65) At least five Ming accounts articulate this episode, but none of them mentions the siyin seal. (66) The captured treasures passed undamaged into Hongwu's hands, and arrived at the Ming capital between 1370 and 1382/8 under the escort of military guard commanders. This severe sequestration process invalidates the conjecture that Xu's illiterate soldiers impressed the siyin seal in random orientation. In reality, Xu did not even inventory the Yuan items.
Significantly, instead of the siyin seal, new evidence indicates that Xu stamped another seal on the Yuan assets. (67) The siyin seal was cast in copper, but the late Ming official Wu Weiye (1609-1672), who once served as Hanlin Academician and had access to imperial storehouses at the end of the dynasty, testified that Xu actually stamped the Yuan collection with his personal jade seal. In Wu's epic lament for the fall of the Ming in 1644:
From the Jin and Yuan dynasties until now [early Qing], half of the imperial collection was acquired from Emperor Huizong's (r. 1100-1126) court of the Northern Song Xuanhe period. The jade seal Zhongshan [Xu Da's zi] imprinted on the Yuan imperial collection physically remained, but the collection mostly perished in the flames of war.
Wu went on to remark that all of the impressions of Xu's seal had gone lost during the Ming-Qing dynastic transition:
The Zhishen ku Storehouse (Auditing Office) contained millions of books and documents, which formerly belonged to the late Northern Song Emperor Huizong and were then passed down from the Jin, Yuan, and early Ming eras. Impressions of Xu Da's jade seal survived during the early Ming, but nowadays they have vanished. (68)
Wu's contemporaries, the Ming historians Tan (1594-1657), Sun Chengze, and Yao Zhiyin (jinshi 1721), continued to recount that Xu stamped his own jade seal, which carried the characters zhong and shan, on Yuan items. (69)
These depositions support the view that Xu's own seal, rather than the siyin half seal, was displayed on Hongwu's collection. Xu, the veteran general and junior grand councilor, had earned the right to place his own seal on the sequestered Yuan collection due to his contributions to the founding of the Ming dynasty. Subsequent to Xu's demise in 1385, Hongwu preserved Xu's jade seal to commemorate his merits. During the Ming-Qing transition, several officials stepped into the Auditing Office and witnessed Xu's seal object.
1373-1384: THE DIANLI JICHA SI OFFICE DID NOT POSSESS ITS OWN SEAL OR STOREROOM
A complete perusal of Hongwu's treasury system disproves the claims that the siyin seal emerged at this time and that the so-called siyin art pieces were stored at the Storeroom of Classic Literature and Famous Paintings of the Dianli jicha si office. Between 1368 and 1384, Hongwu installed ten imperial storehouses, as detailed in the The Great Ming Official System, the Record of the August Ming Ancestral Injunctions (1384), and the Veritable Records of Ming Taizu (1418). (70) The supposed storeroom cannot be verified by early Ming accounts. References to this storeroom all date from the late Ming era (1590, 1638, (71) 1640, (72) 1739 (73)) and denote the Directorate of Ceremonial. This indicates that such a storeroom did not come into existence during Hongwu's reign, and was under the jurisdiction of the Directorate of Ceremonial. If this storeroom of the Directorate of Ceremonial (Sili jian) had its own seal, the seal would end with characters like kuyin or jianyin, and would not contain the siyin characters that signify the seal's derivation from a si-level office.
The Han, Sui, Tang, Song, Jurchen Jin, and Yuan dynasties (ca. 140 B.C. to 1368 A.D.) stored the imperial art collections in the palace libraries (known as mishit flan RAM or mi-related terms), and Hongwu continued their tradition. (74) He emulated the Yuan precedent, directing Mishit flan (1370/3-1382/7), the palace library in Nanjing, to superintend imperial art and book collections. (75) The palace library in Nanjing supervised a mounting room with bookshelves and compartments to store brushes, ink sticks, papers, and office supplies. As Hongwu recalled in his anthologies, he frequented this mounting room between 1373 and 1376 to oversee the remounting of ancient paintings and calligraphies that he had recently acquired. The emperor assigned the artisan Sheng Shuzhang (14th c.) to remount imperial edicts, writings, and damaged art scrolls. (76)
The exact number of art pieces and their import is unknown, but in the years 1373-1380 alone Hongwu chronicled at least twenty-two artworks in his anthologies. As far as can be ascertained, none of them displays the siyin seal. (77) At this juncture, Hongwu's art holdings were cataloged with a different inventory system in full character marks, rather than with half marks. In 1372/6, the imperial tutor Song Lian (1310-1381) viewed the imperial collection, and composed a colophon to Li Bai's (701-762) calligraphy Under the Moon (Yuexia tiejuan ). According to Song Liam the Li scroll was stored on a bookshelf labeled "yin-series, no. 5" The scroll itself bears a "yin-series" seal at the end of the scroll. (78) All of this evidence suggests that before 1384 Hongwu's art collection was cataloged with full character marks, rather than with half marks.
THE LEDGER: THE COMPLEXITY OF THE HALF-SEAL SYSTEM VS. ILLITERATE EUNUCHS
The legal framework that regulated the half seal system was the karzhe system. (79) Kanhe literally means a legal certificate cut into two halves. A comprehensive kanhe system consisted of two security features: an inventory code number (X-series, X-number) and an official seal, both marked across the seams of an inventory object and a page of a ledger to bridge the two items. (80) Sewn in a booklet format, a ledger booklet normally contained a multiple of 100 pages under a designated code series as tracking numbers for treasury audits. If the entire inventory procedure was carried out thoroughly, the registered objects would bear the left halves of what I call "double credentials" : a registry seal and an inventory code inscription. Accordingly, each matching ledger would carry the right halves of the credentials. If an item did not go through an entire registration procedure, only one mark would be applied.
The double credentials prove that the imperial court tracked the artworks before ana atter transfer to their final destination, the court. The inventory process included two steps: marking the siyin seal and inscribing an inventory code. In the first step, the siyin seal was bridged on each artwork and a ledger page at a local registry office before its departure to the capital. The name of the siyin seal denoted the remitting office and the art pieces, when stamped, became state-owned property. The ledger containing the right-side siyin impressions was sent along with the artworks, which bear the left-side siyin, since en route there were opportunities for the art to be secreted away. In the second step, another clerk would inscribe the inventory code on the delivered artwork and a ledger for confirmation upon the artworks' arrival at the imperial treasury.
The Ming dual ledger system for initial and final verification had an earlier precedent in Chinese history: it echoes the documentation of imperial offerings for the Famen temple in the Tang dynasty (618-906). A stone stele, known as Inventory of Imperially Bestowed Gold, Silver, Precious Jewels, and Clothes (Famensi enci jinyin baoqi yiwu zhang bei) from 874, is currently preserved at the Famen temple in Shaanxi province. (81) The linguistic structure of the inscription is consistent with that of contemporary Tang ledgers. (82) The text of the stele comprises several sub-inventories from different donors, which signifies that this checklist was not compiled for the treasures' departure, but for their arrival. The checklist would help determine whether treasures had gone missing on the trip from the capital at Changan to the Famen temple. This delivery procedure can be seen as a historical precedent that illuminates the Ming counterpart. The Ming court's control over inventory exceeded that of the Tang's, with two extra layers of authentication on the transferred goods via bridging seals and codes on the goods and ledgers. The siyin seal not only served an archival function; it also acted as insurance that nothing was stolen during transit. If the siyin mark resulted from the Ming court's initial check, there must have been a final check, which would produce another type of inventory mark, upon the arrival of the stamped artworks and their matching ledgers. Since twenty-nine inspectors from central and local governments jointly checked the transfer of the treasures in the Tang era, it seems unlikely that the same process in the Ming involved just one clerk. Rather, the art pieces imprinted with the siyin seal were overseen by several clerks from assorted departments.
Therefore, theoretically, each siyin seal impression (the initial check) would accompany a half-code inscription (the final check). However, for two reasons, more seal impressions than inscriptions survive: either because artwork was stolen during its journey to the imperial court, or, more prosaically, because the codes could also be sliced out during later remounting due to their placement on mounting textiles. These scenarios help to explain why we only sometimes find a siyin impression juxtaposed with a half-code inscription. At least six surviving siyin art pieces display double credentials: two paintings from Prince Zhu Tan's tomb carry the siyin seal and the inventory code in full character format (ri-series, no. 119; and ri-series, no. 148). (83) Four other paintings and calligraphies bear both the siyin and the inventory-code inscription in half characters: Fu Sheng Expounding the Classics (juan-series, no. 1), Breezes along a Mountain Stream (Xifeng tujuan, juan-series, no. 3). Butterflies and Flowers (Jiajie tujuan, juan-series, no. 10), and Xianyu Shu's (1257-1302) calligraphy Autumn Poems in Running Script (Xingshu qiuxing shice, juan-series, no. 60). (84)
Specifically, Xianyu Shu's calligraphy belongs to the album Precious Paintings from the Song and Yuan Eras (PPM). Five other leaves from the same album were collected by the same Ming owners and carry either the siyin half seal or half inventory code. Leaf 7 (Huang Tingjian's calligraphy; Fig. 1) contains the siyin that was identified by the Qing imperial catalogue as a jingli siyin registry seal. The same seal appears on leaves 12 and 13, which show Mi Fu's Large Character Running Script (Xing cao da shu). Leaves 16 and 17--Wang Tingyun's calligraphies Poem to Fahua (Fahua shi) and Poem to Daolin (Daolin shi)--bear minister Lu Wan's (1458-1526) seals, whereas leaf 25 (Xianyu Shu's calligraphy) carries the half inventory code as noted. (85) This corroborates that the siyin practice abided by a comprehensive kanhe procedure, resulting in the double credentials on each artwork.
The siyin clerks were trained specialists who could draft and file ledgers following kanhe law, procedures that illiterate eunuchs or soldiers could not have performed. After years of literacy training, the siyin clerk needed to spend another few years learning legal practice and calligraphy. Systematic decade-long training strengthened his practical knowledge such that he could read seal scripts and create legal archives that described inventory objects. The complexity of these clerical tasks indicates that scribes who processed the siyin art pieces must have been proficient in bookkeeping. No record elucidates how such scribes were selected by Ming central and prefectural governments, but they were likely chosen in the same way as licensed scribes in Qing counties, who had to pass the clerical exam or be nominated by several clerks. (86) A group of such experienced scribes, if drawn from Dianli jicha si eunuchs between 1373 and 1384 (a time during which Hongwu institutionally prohibited eunuchs from gaining knowledge), would have severely violated Hongwu's own anti-eunuch politics.
The imprinting process of the siyin seal involved familiarity with legal formats and calligraphic skill. Because the kanhe system was technically difficult, it was not yet used in Hongwu's art collection when the Dianli jicha si office was in operation. It should be noted that at this time the kanhe system was performed only for limited purposes, such as autopsy reports (enacted in 1368), household registers (in 1370), grain audits (in 1376), six ministry documents (in 1382), and trade permits between China and Siam (in 1383). (87) Before 1384, the half-character kanhe law did not apply to art collections.
After 1384, various kanhe marks emerged for security verification. Several governmental divisions supervised registry offices that marked government property with half- or full-character kanhe signs, depending on the provenance of the inventoried goods. Variations of such marks appear on waist-badges, legal documents, land contracts, military household registries in 1387, imperial treasuries in 1387, and confiscated booty in 1393. The use of the siyin half seal was not an isolated phenomenon in the post-1384 era. It blended with a complex, well-designed information control network, with effective coordination between the central and regional governments, the registry offices, and experienced bookkeeping clerks.
How many literate hands wrote out the ledgers and stamped the siyin seals? To ensure that the transferred treasures would not themselves be secreted away by local officials before their remittance to the court, the Ming emperor would customarily dispatch a cluster of professional clerks to mutually monitor the inventory process. A single clerk working alone could not perform the siyin procedure. The siyin procedure highlights the institutional involvement and legal transportation of thousands of canonical paintings and calligraphies. During 13731384, could Hongwu authorize the Dianli jicha Si office to publicize a eunuch office's seal on the imperial art collection? Enmeshed in suspected conspiracy and subversion, the emperor deterred eunuchs from gaining education and periodically promulgated laws to curtail their power. Based on his negative personal experiences under the previous dynasty, the emperor suppressed eunuchs at the time, as evident in his decrees from 1368, 1369, 1370, 1371, 1373, 1376, 1377, and 1384. (88) In Hongwu's diagnosis:
Those people [eunuchs] can solely be given sprinkling and sweeping jobs. They should not be given responsibility and their number should not be large. ... Of these people, not one or two out of thousands are good. (89)
This controverts the predominant theory that Hongwu appointed the Dianli jicha si eunuch office to register the imperial art collection with the siyin seal. Prior to 1384 most eunuch offices engaged in manual labor. Institutionally, only the Neishi jian (Directorate of Palace Attendants; 1368-1384) and the jishi si (Office of Recording; 1373/78-1384) were designated to perform work requiring literacy. (90) The former belonged to the directorate (jian) level and its seal would carry the characters jian yin. The latter functioned as Hongwu's private secretary; eunuchs from this office were able to read, write, and handle document transmission and boolckeeping. (91) If this office had its own seal, the seal would include the four characters jishi siyin. Of these four characters, the second one--shi--does not match the partial character in the middle column of the siyin seal, indicating that the Jishi si office could not have wielded this seal.
Prior to 1384, Hongwu engaged in inaugurating proper protocols and ancient customs to distinguish his rule from Mongolian barbarism. In his vision, the Dianli jicha si office and its previous incarnations--Neizheng si (1373), Dianli si (1373), Ji cha si (1374-1375)--were designated to rectify inner court etiquette. Documents from 1375, (92) 1382, (93) and 138494 regarding the Dianli jicha si office all refer to the same duties. As proclaimed in the Record of the August Ming Ancestral Injunctions, the emperor juxtaposed the Dianli jicha si office with three other si-level eunuch offices: the offices of Imperial Horses, of Military Weapons, and of Punishing Mischievous Eunuchs. Above the si-level offices stood seven superior jian-level directorates.95 This outlines how Hongwu circumscribed the power of the Dianli jicha si office up to 1384.
Certainly, Hongwu occasionally empowered an individual eunuch as a one-time messenger to run errands for him. But such court activities involving literate eunuchs are reported in great detail for this period, as they fascinated contemporary Ming historians from the bureaucratic elite. (96) In 1375, a palace servant asked the head eunuch Li Binyan of the Dianli jicha si office to deliver a copy of Hongwu's own anthology to the official Song Lian. (97) In 1382, Li's successor Tang Shou discussed ritual attire with a Minister of Rites. (98) In 1377, 1378, 1379, and 1381, Hongwu mandated not the Dianli jicha si eunuchs, but eunuchs from other offices--Li Jing, Wu Cheng, Chen Jing, and Xu Bao, respectively--as temporary messengers to run errands to military frontiers. (99) Even such trivial events involving eunuchs were recorded by different court officials. Such commissions were individual, irregular, and traceable. The siyin seal on thousands of ancient art pieces reflects the archival practices of a legal institution. It was a significant undertaking. A project of this magnitude could not have gone undocumented, as assumed.
AFTER 1384: SIYIN FOLLOWING THE LIBU PING YAN SHUHUA GUANFANG SEAL
The siyin seal's placement on the artwork indicates that the seal must have been imprinted later than the rectangular Ming official seal, the Libu pingyan shuhua guanfang (Security inspection of calligraphy and paintings from Ministry of Rites) seal (hereafter Libu seal). As Figures 5a-b illustrate, both seals appear on Jia Shigu's painting Ancient Temple in a Mountain Pass and on Xianyu Shu's calligraphy Transcribing Du Fu's Poem on General Wei. Nevertheless, the Libu seal occupies the predominant location, one of the four corners of the painting. Because only the left half of the siyin seal could be displayed, the imprinter had two options for stamping: either the right side or the bottom edge of the painting. Since the bottom and the upper- and lower-right corners of Jia's painting had already been imprinted with the Yuan official seal Dusheng shuhua zhiyin and the oblong Libu seal, and the bottom of the painting is pigmented with very dark ink, there was no choice but to squeeze the siyin seal into the middle of the right side (Fig. 5a). The same situation occurs in Figure 5b. in which both the Libu seal and siyin appear on Xianyu's calligraphy. The siyin seal, however, due to its later impression, was forced to yield to the Libu seal which already occupied the lower right corner. The siyin must therefore have been stamped after the Libu seal.
When, then, was the Libu seal created? While most scholars believe that it was used by the Ming to mark the Yuan collection, new evidence proves that this rectangular seal came into being after 1384, when the Diann jicha si office ceased to operate. Of all twenty-one traceable Libu seal impressions on art (including those recorded but now lost), seven covered the full character inventory codes, which must thus have been inscribed before the Libu seal. The seven inventory codes comprise ren-series, no. 1 (100) wen-series, no. 1 wen-series, no. 6; wen-series, no. 7; (101) wen-series, no. 17; wen-series. no. 19; (102) and wen-series, no. 20. Only wen- and ren-series are found.
Although the two characters wen and ren are in the Thousand Character Essay, the codes did not derive from its numbering system. The Essay consists of 1000 characters, among which wen and ren correspond to numbers 263 and 369, but no other code series appears beneath the Libu seal. This indicates that the wen- or ren-series codes must have been drawn from another numbering system. Moreover, each code was inscribed by different hands with different calligraphic styles, suggesting that the wen- or ren-series codes were neither from the Essay system nor from any other single system. Artworks with these codes must have been assembled by the court from diverse locations and were registered with different codifying systems.
The ren-series code appears on the painting Six Arhats attributed to Lu Lengjia (see n. 100). According to the Collected Statutes of the Great Ming:
The Storehouse of the Imperial Treasury employed the qing- and ren-series as kanhe inventory codes. ... After describing the content of the treasure on the ledger and codifying both with inventory codes, the clerk together with the Directorate of Palace Seals (Shangbao jian) presented a request [to superior agents] for permission to use the seal. After registration, the ledgers would be submitted to the Directorate of Ceremonial for preservation. (103)
This passage attests to the fact that artworks bearing the ren-series inventory codes fell under the jurisdiction of the Directorate of Ceremonial. Because this directorate only began its work in 1384, this is the earliest possible date for the ren-series code. The siyin seal on Jia's and Xianyu's works all postdate the Libu seal, so the siyin seal must have appeared after 1384, by which time the Directorate of Ceremonial had superseded the shuttered Dianli jicha si office.
POST-I384: THE DIVISION BETWEEN THE DIANLI JICHA Si OFFICE AND THE DIRECTORATE OF CEREMONIAL
Beginning in 1384, the emperor overhauled all eunuch institutions, including the Dianli jicha si office:
Staffed with one director and one deputy, the Dianli jicha si office was in charge of handling inner court etiquette, taking memos of imperial instructions, and reporting transgressions of palace eunuchs and runners. Their other duties comprised manufacturing brushes, ink sticks, and mounting scrolls. (104)
In 1384/4 the Dianli jicha si office began to take notes from the emperor, monitoring inner court ritual and etiquette, in addition to producing studio objects for practical purposes. However, this office was immediately disbanded in the same month and transformed into the Directorate of Ceremonia1. (105) In contrast to its previous incarnations, the Directorate of Ceremonial was given authority to inscribe inventory codes on imperial writings bestowed upon meritorious officials, and on rank-badges for palace functionaries to pass through gates. (106) Similarly, two other eunuch directorates issued inventory codes--the Directorate of Palace Seals and the Directorate for Credentials (Yinshou jian) (1384-1644). But a seal from a directorate (jian) would end with the characters jianyin.
While the Dianli jicha si office exercised only limited functions, the power of its later incarnation--the Directorate of Ceremonial--was drastically augmented. As the emperor's delegate, this directorate in its dynasty's final state infiltrated the central bureaucracy, dominating all other eunuch offices of the dynasty. This elevated directorate intervened in judicial verdicts, controlling the twelve eunuch directorates, four offices, eight bureaus, and a few palace storerooms. But this later stage of the Directorate of Ceremonial should not be injected into the equation of the Dianli jicha si office. Since the siyin half seal could not have been wielded by the Directorate of Ceremonial or any other directorates, it must have originated with another office.
This paper refutes the dominant twentieth-century theory about the siyin half seal by arguing that this seal could not have originated from the eunuch-run Dianli jicha si office in 1373-1384, nor could it have been used by the Ming general Xu Da on pieces of the Yuan palace collection as previously hypothesized. New evidence suggests that the alleged Dianli jicha si characters do not match the siyin seal strokes, and Xu actually stamped his own Zhongshan seal on the Yuan collection, rather than the siyin half. The siyin in fact appeared later than the Libu seal, which came into existence after 1384, by which time the Directorate of Ceremonial had superseded the shuttered Dianli jicha Si office.
Instead of simply dividing the various siyin seal impressions into binary genuine/forged categories as in past studies, I demonstrate that three types of siyin seals can be considered authentic, all of which were used by different branches of the jingli Si registry office. This finding can be further substantiated by remarks of several Ming-Qing historians, such as Shen Defu and Wen Zhenheng, as well as Qing imperial cataloguers. The jingli si office was in charge of bookkeeping, and part of the characters spelling out jing and II can be reconstructed from the remnant of the half seal. The fact that a half seal was employed suggests that the jingli si office followed the kanhe inventory system, which imposed an official seal on the seam of both an inventoried object and its complementary ledger.
Much remains to be uncovered. The precise dates, duration, and full names of the jingli Si branch offices demand further investigation. It is equally essential to untangle the cultural biographies of the 199 siyin art pieces, in order to trace the overlapping art owners and those who targeted their art holdings. The imperial motives that prompted the cast of the three types of siyin seals will be the topic of a future article. Here it may suffice to determine that all were cast later than 1384, after the closure of the Dianli jicha si office. The current paper takes an initial but significant step toward penetrating the intricate siyin historiography. The research allows a rethinking of Ming-era transmission histories of the canonical siyin artworks that lie at the heart of Chinese art. Including a wide spatial and temporal range, from the fourteenth-century Ming capital of Nanjing to a fourteenth-to-sixteenth-century empire-wide registration network, this research links the siyin seals to increasingly regimented numerical inventory systems and the institutions that developed them. In this way I show how social processes and bureaucratic history can be derived from the material culture of art. Further pursuing this theme, my forthcoming essay will explore the role inventory systems played in mapping, defining, and legitimizing the royal power of the Ming dynasty.
An abridged version of this paper was presented at the International Conference on Middle Period China, 800-1400 (Harvard University. June 6th, 2014). I am deeply indebted to Professors Richard Vinograd, Ronald Egan, Albert Dien, Jeffery Moser, Alfreda Murck, the anonymous reviewers of the Journal of the American Oriental Society as well as Annalisa Bolin and Chen Kaijun for their productive suggestions. Thanks also to the Stanford Humanities Center Geballe Dissertation Prize Fellowship and Mellon Foundation Dissertation Fellowship, which supported my research. Dates follow the Chinese lunar calendar in the following form: "year/month/date."
(1.) In 1972, Zhuang Shen listed 120 existing or now-lost siyin seal impressions, two of which were in tact not siyin seals: one on Huai Su's (737-799) calligraphy Autobiography (Zixu tie, National Palace Museum. Taipei; hereafter NPM), and one on Auspicious Pines in Spring Mountains (Chun shan rui song tu NPM) attributed to Mi Fu (1051-1107). See Zhuang Shen, "Gugong shuhua suojian Mingdai banguan-yin kao", in Zhongguo huashi yanjiu xuji (Taipei: Zhengzhong shuju, 1972), 1-46. In 1980, Jiang Yihan. assembled 142 siyin seal impressions; see his "Yuannei fuzhi shuhua shoucang" Gugong jikan 14.3 (1980): 18-28. In 2001, Ding Xiyuan discussed 89 siyin artworks; see his articles "Dianli jicha siyin kao" Gogong wenwu yuekan 214 (2001): 69, and "Siyin yu Tang Song Yuan Ming huaji" in Guobao jiandu (Shanghai: Renmin meishu chubanshe, 2005), 241. Based on previous scholarship and my new findings. I examine 199 siyin art pieces that comprise 148 extant and 51 now-lost but recorded pieces.
(2.) Shen Defu, Wan yehuo bian (Harvested in the wild during the Wanli period, 1619) (Beijing: Zhonghua shuju, 1980), 8.211. Wang Yaoting notices Shen's account on the two types of jingli siyin seals but does not pursue this line of investigation further. See "Beyond the Admonitions Scroll: A Study of Its Mounting, Seals, and Inscriptions." in Gu Kaizhi and the Admonitions Scroll, ed. Shane McCauseland (London: British Museum, 2003), 216.
(3.) See Zhang Zhao (1691-1745) et al., comps., Midian zhulin (Pearl Forest in the Secret Hall), in Shiqu baoji Midian zhulin hebian (1744; Shanghai: Shanghai Shudian chubanshe, 2011), 9.105 Below Wen's colophon, the Midian zhulin editor added an annotation in 1744, explaining that the j ingli siyin seal had gone missing.
(4.) See Zhang Zhao et al., comps., Shiqu baoji (Treasured boxes of the Stone Moat) (1744; Shanghai: Shanghai shudian chubanshe, 2011), 1.465.
(5.) Both Li 's and Chen 's accounts were transcribed in Wu Qian (1733-1813), Lunyin-jueju (Quatrains on ancient seals) (Shanghai: Baogu zhai, 1922), 5-6. Zhuang Shen notices Li's passage, but brushes aside the connection between siyin and the seals of registry offices. See Zhuang, "Gugong shuhua," 18.
(6.) Wang Shidian (d. 1359). Mishit jianzhi, (Records of the Palace Library) (1342; Hangzhou: Zhejiang guji chubanshe, 1992), 3.59.
(7.) Mishu jianzhi 1.25; 9.180; 3.49; 3.52.
(8.) Hu Shi'e ed., Ming Taizu ji (Hefei: Huangshan shushe, 1991), 86; Huangming zuxun lu (Record of august Ming ancestral injunctions; 1384; hereafter HMZXL), in Mingchao kaiguo wenxian ed. Wu Xiangxiang (Taipei: Taiwan xuesheng shuju, 1966), 3:1752.
(9.) Da Ming guanzhi (The great Ming official system), in Mingchao kaiguo wenxian, 1V:2419; 2254; Zhang Tingyu (1.672-1755). et al., Ming shi IIJJ2 (History of the Ming; hereafter MS) (Taipei: Dingwen shuju, 1975), 75.1747; 75.1849; 76.1856; and Huatigming zuxun (August Ming ancestral injunctions; 1395 edition; hereafter HMZX), in Mingchao kaiguo wenxian 111, 38.1654; Liu Tingji (1654-after 1715), Zaiyuan zazhi AltilliCit (Miscellanies of the Zai Garden) (1715; Beijing: Zhonghua shuju, 2005), 1.8.
(10.) MS 75.1849; 76.1857; 76.1858; 76.1860.
(11. Li Dongyang ( 1 447-15 16) et al., comps., Zhengde huidian (1511; SKQS 617), 78.754.
(12. Qiu Long Em, Ming Qing shiqi de duliangheng, in Zhongguo gudai duliangheng wenji (Henan: Zhongzhou guji chubanshe, 1990), 349; Qiu Guangrning, Zhongguo lidai duliang hengikao (Beijing: Kexue chubanshe, 1992), 104. Wang Yaoting cites an outdated paper from 1957 to convert 2 cun to 6.4 cm, which obliterates the connection of the siyin seal with the jingli siyin seal. See Wang, Chuan Gu Kaizhi niishizhen tuhua wai de jige wenti, Mei-shushi yanjiu jikan 17 (2004): 30.
(13. Gu Fu, Pingsheng zhuangguan (The great sights in my life; hereafter PSZG) (1692; Shanghai: Guji chubanshe, 2011), 1.19; 2.49; 3.79; 6.239; 7.250; and 8.292.
(14. Gao Shiqi, Jiangcun xiaoxia lu (Record of works of art seen by Gao Shiqi) (1693; Shanghai: Guji chubanshe, 2011), 2.291; 3.332.
(15.) PSZG 3.79.
(16.) Long Wenbin (1824-1893). Ming huiyao (Collected institutions of the Ming) (Taipei: Shijie shuju, 1963). 56.1074.
(17.) Sun Chengze (1592-1676). Tianfu guangji (Record of buildings and institutions of the capital) (after 1672; Beijing: Guji ehubanshe, 1982), 10.119: 10.123.
(18.) Da Ming guanzhi 2254; 2419.
(19.) Ming Taizu ji 140.
(20.) An Qi (1683-1744), Moyuan huiguan (Collected records of works in ink) (1743; Qingxuan tongyuan Man, 1909), calligraphy 1.15; 1.35; 1.51; 1.53; and painting 1.4-8; 1.12-17; 1.21-22; 2.3; 2.25; 2.70-7-1.
(21.) Ma Heng (1881-1955), Ma Heng riji (1949; Beijing: The Forbidden City, 2006), 105; 126 Thanks to Sun Bo 4114 of the National Museum of China for helping me to access this book.
(22.) Chen Rentao (1906-1968), Jingui canghua ji m (Kyoto: Benrido, 1956), 12; and Chen, Jingui lunhua (Hong Kong: Dongnan shuju, 1956), 53-55.
(23.) Max Loehr, "Chinese Paintings with Sung Dated Inscriptions," Ars Orietztalis 4 (1961): 224.
(24.) Na Zhiliang, Nieyin tongshi (Taipei: Shangwu yinshu guan, 1970), 87.
(25.) Zhuang, "Gugong shuhua," 25.
(26.) Liu Jiu'an, "Zhu Tan mu chutu huajuan de jige wenti" 195 (1972): 64-66.
(27.) Cheng Qi, Xuanhui tang shuhua lu (Hong Kong: Xuanhui tang, 1972), painting 6.
(28.) Hironobu Kohara, "O I ga to sono densho sakuhin," in Bunjinga suihen: O I (Tokyo: Ch56 Koronsha, 1975), 145.
(29.) Jiang Zhaoshen, "Shanzhe jique zaochun yu wen hui: Tan gugong sanzhang songhua" Gugong jikan 11.4(1977): 13-21.
(30.) Jiang, "Yuannei fu," 18-28.
(31.) Fu Shen, Yuandai huangshi shuhua shoucang shilue (Taipei: Gugong bowuyuan, 1981), 93-99.
(32.) Suzuki Kei, "Kenkyu yoroku: Shiin sanko. Kokka 1117(1988): 27-29.
(33.) Cahill, "The Case against Riverbank: An Indictment in Fourteen Counts," in Issues of Authenticity in Chinese Painting, ed. Judith G. Smith and Wen C. Fong (New York: Metropolitan Museum of Art, 1999). 13-61.
(34.) Ding, "Dianli," 69; and Ding, "Siyin," 244.
(35.) MTZSL 85.1520; 86.1526; 94.1639; 161.2502; Tianfu guangji 15.199; and MS 304.1823; 304.1824.
(36.) Kei, "Kenkyu," 29.
(37.) Fu, Yuandai huangshi, 97. and "Tianxia diyi Su Dongpo hanshi tie" Gugong wenwu yuekan 2.7 (1984): 80.
(38.) Zhuang, "Gugong shuhua," 34-46; Fu, Yuandai huangshi, 97-100.
(39.) Fu, Yuandai huangshi, 99. See also Ma Heng riji, 105; Zhang Guangbin, "Ma Wan de shanshui hua" Gugong wenwu yuekan 17 (1984): 96.
(40.) Zhuang, "Gugong shuhua," 16.
(41.) Zhuang, "Gugong shuhua," 24; Liu, "Zhu Tan mu," 66; Jiang, -Yuannei fu." 18. On the other hand, Ding Xiyuan recently argued for a longer duration of the siyin seal from 1373-95. See Ding, "Siyin," 245.
(42.) Zhuang, "Gugong shuhua." 24; Liu, "Zhu Tan mu," 66; Fu, Yuandai huangshi, 98; and Jiang, "Yuannei fu," 29.
(43.) HMZXL 1765; Edward L. Farmer, Zhu Yuanzhang and Early Ming Legislation: The Reordering of Chinese Society Following the Era of Mongol Rule (Leiden: Brill, 1995), 108.
(44.) Hok-Lam Chan, "Ming Taizu's Problem with His Sons: Prince Qin's Criminality and Early-Ming Politics," Asia Major 20.1 (2007): 48: and Ray Huang, 1587, A Year of No Significance (New Haven: Yale Univ. Press, 1981), 16.
(45.) Ming Taizu huangdi qinlu (A compendium of Hongwu's edicts to his princes), in Chang Bide, "Ming Taizu huangdi qinlu." Gugong zushu jikan .1.4 (1971), 75; 76; 84; 88; 90; 99; and 104.
(46.) Liu, "Zhu Tan mu," 66; Zhuang, "Gugong shuhua," 26; and Jiang, "Yuannei fu," 18.
(47.) Zhuang measures eighteen siyin samples and points out that their dimensions vary even though these seals have been accepted as authentic. See Zhuang, "Gugong shuhua." 33.
(48.) Kei, "Kenkyu," 28.
Duoganwei duzhihui shi siyin (Seal of the Commander in Duoganwei Region), dated 1407, silver, 6.4 x 6.4 cm, Cultural Relics Administration Committee, Tibet: 4) Tuduhewei zhihuishi siyin (Seal of the Commander in Tuduhe Region), dated 1408, copper, 7.3 x 7.3 cm, PPM; 5) Hetun jiwei zhihui-shi siyin (Seal of the Commander in Hetun jiwei Region), dated 1409, copper, Jilin Museum: 6) Libu wenxuan qingli si zhiyin J.EFi (Seal of the Bureau of Appointments in the Ministry of Personnel), Southern Ming period, dated 1.645, copper, 7.5 x 7.5 cm: see Zhongguo xiyin zhuanke quanji (Shanghai: Shuhua chubanshe, 1999), plate 1033.
(49.) Ding, "Dianli," 72.
(50.) Ma Heng riji. 126; Zhuang, "Gugong shuhua," 15. (51.) Liu. "Zhu Tan mu," 66; Zhuang. "Gugong shuhua," 29.
(52.) For a discussion of Song dynasty governmental copper seals and their authenticity, see Huiping Pang, "Shangshusheng yinzhi yanjiu huigu ji wuxiang shangque", Gugong bowuyuan yuankan 1 (2009): 44-59; and Pang, "Cunshi shuhua suoqian Shangshusheng yinkao" Wenwu 11(2008): 77-93.
(53.) Listed below are examples of si-level seals: 1) Nanjing tongzheng shisi jingli siyin (Seal of the Office of Transmission in Nanjing) on the painting Xiao Yi Stealing the Orchid Pavilion Preface (Xiao Yi zhuan Lanting tujuan), attributed to Yan Liben (601-673), 26.5 x 75.7 cm, Liaoning Provincial Museum; 2) Wusizang xuanwei si fen siyin (Seal of the Branch of Pacification Office in Wusizang), dated 1406, copper, 8.4 x 8.4 cm, Cultural Relics Administration Committee, Tibet; 3)
(54.) For paper money issued in 1375, see Richard Von Glahn, Fountain of Fortune: Money and Monetary Policy in China, 1000-1700 (Berkeley and Los Angeles: Univ. of California Press, 1996), 71-72.
(55.) See n. 53 above.
(56.) Zhuang. "Gugong shuhua," 1; 33.
(57.) National Palace Museum, ed., Daguan: Bei Song shuhua tezhan (Taipei: National Palace Museum, 2006), 207.
(58.) Ding, "Siyin," 241; 247; and Ding, "Dianli," 76.
(59.) See "Appendix I: Provenance of the Sun Art Pieces," in Huiping Pang's "The Emperors and Their Henchmen: Art Collecting, Inventory, and Criminality in Ming Imperial China" (Ph.D. Dissertation, Stanford Univ., forthcoming).
(60.) Because I do not have access to all 148 existing artworks, I cannot precisely evaluate how many siyin seal impressions are fake. The fake-to-authentic ratio is subject to future revision.
(61.) Pang, "The Emperors and Their Henchmen," chapters 2-3.
(62.) For example, Zhuang Shen measures dimensions of eighteen siyin impressions. I consider ten of these authentic and thus include his measurements. See Zhuang, "Gugong shuhua," 33.
(63.) Sarah Schneewind, "Visions and Revisions: Village Politics of the Ming Founder in Seven Phases," Toung Pao 87 (2002): 317; Hiroshi Danjo, "Minbaro cho seiritsu ki no kiseki: Kobucho no gigoku jiken to kei-shi mondaio megutte", Toyo shi kenkyu. 37.3 (1978): 23.
(64.) Ding. "Siyin." 243.
(65.) Fu, Yuandai huangshi,98.
(66.) Yao Guangxiao (1335-1418) et al., Ming. Taizu shilu (Veritable records of Ming Taizu; hereafter MIZSL) (1418; Taipei: Academia Sinica, 1962), 34.600; 34.1606; 147.2307; MHY 26.418; Yu Jideng (1544-ca. 1601), Diangu jiwen (Literary allusions and anecdotes) (1601; Beijing: Zhonghua shuju, 1997), 4.67.
(67.) MTZSL 34.600.
(68.) Wu Weiye, Wu Meicutz quanji' (Shanghai: Guji chubanshe, 1990), 19.508.
(69.) Yao Zhiyin, Yuan Ming shi leichaoTc (Collected and arranged writings on Yuan and Ming affairs) (SKQS 884), 21; and Tan Qian, Beiyou lu (Record of traveling in the North) (1656; Beijing: Zhonghua shuju, 1997), ji you, 66-67.
(70.) HMZXL 1732-34; HMZX 1641; MTZSL 83.1487; 98.1673; 161.2503-4; 124.1990; Da Ming guanzhi 2226.
(71.) Liu Ruoyu (1541-ca. 1642), Zhuo zhong zhi (A weighted and unbiased record) (ca. 1638; Beijing: Guji chubanshe, 1994), 16.94.
(72.) Liu Ruoyu, Minggong shi (History of the Ming court) (Beijing: Beijing guji chubanshe, 1980), 1.2.
(73.) MS 74.1819.
(74.) On establishing the Northern and Southern Song palace libraries, see Huiping Pang, "Liangsong gongting shuhua chucang zhidu zhibian: Yi Mige wei hexin zhi jiancang jizhi yanjiu", Palace Museum Journal 1 (2005): 12-40. On the Yuan palace library see Fu Shen, "Princess Sengge Ragi: Collector of Painting and Calligraphy," in Flowering in the Shadows, ed. Marsha Weidner (Honolulu: Univ. of Hawai'i Press, 1990), 57: Fu, Yuandai huattgshi 3.
(75.) Xu Xueju (jinshi 1583), Guochao dianhui (Memorials and legal statutes of our Ming dynasty), in SKQS cunmit congshu, ce 264 (1642; Taipei: Zhuangyan wenhua, 1996), 22.602; Diangu jiwen 1.10; 16.283.
(76.) MTZJ 290. Sheng was a nephew of the renowned Yuan painter Sheng Mao (14th c.).
(77.) Such as Liu Gongquan's (778-865) Stele of the Shence Imperial Army (Shence jun bei, National Library of China), Fan Quan's (ca. 990-1030s) painting Four Season Landscapes (Sishi jing ), Li Gonglin's (1049-1106) Copy of Wei Yan's Pasturing Horses (Lin Wei Yan fang mu tujuan, PPM), Zhao Xiyuan's (ca. 1124-1182) painting Wild Birds in Autumn Pond (Qiu tang yeqin tu ), Zhao Boju's (ca. 1120-1162) Rivers and Mountains in Autumn Color (Jiangshan qiuse tujuan PPM), Liang Shimin's (12th c.) Dense Snow on Reedy Sandbanks (Lu ting mixue tujuan, PPM). Li Song's (ca. 1170-1255) West Lake (Xihu tu ), and an ancient painting titled Orchid Pavilion Gathering (Gu Laming qushui liushang tujuan ). MTZJ 16.354; 16.388-93; Yuthi wenji 18.220-32; 18.316-26; Yuzhi wenji. bu 325.
(78.) For Song Lian's colophon, see Yang Enshou (1835-1891), Yanfu bian (A treat for the eyes) (1885; Changsha: Tanyuan conggao, 1885), 4.15.
(79.) See Ray Huang, Taxation and Governmental Finance in Sixteenth-century Ming China (London: Cambridge Univ. Press, 1974), 14; Kei, " Kenkyu, " 28; Huiping Pang, "Stolen Art and Lost Inscriber: Inventory Codes on Artworks in the Tumultuous Ming Wanli Period (r. 1573-1619)," Artibus Asiae 73.2 (2012): 411.
(80.) Da Ming guanzhi 2400; MTZSL 58,1143; 141.2223; Fu Fengxiang (ca. 1486-1551), Huangtning zhaoling (Edicts of the Ming) (1548; Taipei, Chengwen chubanshe, 1967), 2.115; Zhengde huidian 104.995; Wanti huadian 41.763; 116.1694; Yuzhi dagao xubian (Grand pronouncement; 1386 edition), in Hongwu yuzhi quanshu (A complete collection of the imperial commands of Hongwu) (Anhui: Huangshan shushe), 839.
(81.) Han Wei, "Famensi digong Tangdai sui zhenshen yiwu zhangkao" Wenwu 5 (1991) : 29; Wang Cangxi "Famensi gongwu zhangbei shiyi ", Wenbo 4 (1989) : 30-33. Thanks to Prof. Albert Dien for this reference and for bringing the Tang-dynasty example to my attention. See also Patricia Eichenbaum Karetzky, "Esoteric Buddhism and the Famensi Finds," Archives of Asian Art 47 (1994) : 78-85.
(82.) Wang Cangxi, "Cong famensi digong chum jinyin qitan Tangdai hengzhi", Wenbo 1 (1992) : 51-79; Hong Yifang, "Lun famensi Tangdai yiwu zhangzhong de geti liangci", Hanxue yanfiu 24.2 (2006) : 136; and Zheng Shaolin, "Cong Tang shike kan liangci shide shengchan jiqi fazhan yanbian", Huabei shuili shuidian xuebao 29.1 (2013) : 126-27.
(83.) Shandong Provincial Museum, "Fajue Ming Zhutan mu jishi", Wenwu 5 (1972) : 29-30.
(84.) Wang, "Beyond the Admonitions," 214; Pang. "Stolen Art," 413.
(85.) For the underlying logic that tied minister Lu Wan's art collection to the siyin seal and the inventory code, see Pang, "Stolen Art," 404.
(86.) Bradly Ward Reed, "Scoundrels and Civil Servants, Clerks, Runners and Local Administration in Late Imperial China" (Ph.D. Dissertation, Univ. of California, Los Angeles, 1994), 56; 74.
(87). Da Mingling (The great Ming commandment) (1368; Beijing: Falu chubanshe, 1999). 262; Shen Shixin (1535-1614) et al., comps., Wanli huidian (Collected statutes of the great Ming) (1587; Beijing: Zhonghua shuju, 1989), 108.585.
(88.) MTZSL 110.1285; 112.1850; (1526-1590), Yanshan tang bieji (Yanshan's essays about Ming history; hereafter YSTBJ) (1590; Beijing: Zhonghua shuju, 1985), 91.1739-40; MS 2.27; 304.7766; MHY 39.699; Baoxun (Precious instructions of Hongwu, 1374), in Hongwu yuzhi quanshu 487; Shihshan Heney Tsai, The Eunuchs in the Ming Dynasty (Albany: State Univ. of New York Press, 1996), 13.
(89.) MTZSL 44.861; Diangu jiwen 2.31; 4.70; YSTBJ 91.1739; MS 74.1826; and MHY 39.697.
(90.) HMZXL 1726-28.
(91.) MTZSL 84.1495; 161.2501; Luan Chengxian, "Hongwu shigi huanguan kaolue" in Ming shi yanjiu luncong 2 (Suzhou: Jiangsu guji chubanshe, 1983), 106.
(92.) Song Lian, Wenxianji (Collected works of Song Lian) (SKQS 1223), 13.628.
(93.) Yu Ruji (17th c.) et al., Libu zhigao (Draft monograph of the Ministry of Rites) (1620; SKQS 597), 86.
(94.) YSTBJ 87.1666; and 90.1723.
(95.) HMZXL 1726-30.
(96.) Wang Chunyu., Mingchuo huanguan (Beijing: Zijincheng chubanshe, 1989), 234.
(97.) Wenxian ji 13.628; Huang Zuo (1490-1566), Hanlin ji (Records of the Hanlin Academy) (Shanghai: Shangwu yinshuguan. 1936), 16.213.
(98.) MTZSL 143.2253.
(99.) MTZSL 120.1949; YS7'BJ 87.1661; and Ming Taizu huangdi qiniu 72; 74.
(100.) See the Libu seal and ren-series. no. 1 code on the painting Six A rhats (Liu zunzhe xiang ce) attributed to Lu Lengjia (ca. 730-760). Album leaf, ink and color on silk, 30 x 53 cm, PPM.
(101.) See the Libu seal and wen-series, no. 7 code on Xie Yuan's (13th c.) painting Broken Branch of Blossoming Emerald Peach (Zhezhi bitao tujuan). Handscroll, ink and color on silk, 26 x 64 cm, private collection, Taipei.
(102.) See the Libu seal and wen-series code on Xianyu Shu's Transcribing Du Fu's Poem on General Wei.
(103.) Wanli huidian 213.1065.
(104.) HMZXL 1729; Luan, "Hongwu," 104.
(105.) MTZSL 161.2502.
(106.) HMZX 31.1689; YSTBJ 90.1726; Wanli huidian 41.292; Farmer, Early Ming Legislation, 137.
|Printer friendly Cite/link Email Feedback|
|Publication:||The Journal of the American Oriental Society|
|Date:||Jul 1, 2014|
|Previous Article:||Anterican Oriental Akcietp.|
|Next Article:||Breaking Bread with the Dead: Katumuwa's Stele, Hosea 9:4, and the Early History of the Soul.|