The Relationship Between the Tangut Art and Textual Heritage : Some Observations *

The paper examines several cases where the Tangut textual materials can be juxtaposed against visual materials and iconography. The problem of the relationship between the two types of sources is discussed, and textual examples are provided to support the author’s point of view. The author argues for the priority of texts before images and intends to show that the ‘systematic nature of Tangut Buddhism’ reveals itself through the uniformity between texts and images. Hualin International Journal of Buddhist Studies, 1.2 (2018): 183–201 * Images used in the paper, are available in public domain available from the website: hermitagemuseum.org; images of the Tangut texts are courtesy of the Institute of Oriental Manuscripts, Russian Academy of Sciences. The Relationship Between the Tangut Art and Textual Heritage:

F rom a historical perspective, the system of Tangut Buddhist imagery supports the traditional division of the Tangut Buddhism into the Sinitic and Tibetan domains, whereas alleged synthetic forms provisionally identified as 'Sino-Tibetan art' are supposedly representative of the indigenous Tangut Buddhism. This division is somewhat parallel to the threefold division of the Tangut Buddhist language, which operates either within the scope of Sinitic, or Tibetan terminology, or tends to combine both sets of terminology within the same discourse. These specific developments are provisionally identified as the 'Sino-Tibetan Buddhism', and 'Sino-Tibetan Buddhist art' respectively. 1 There is only limited correspondence in subject matter between these two rubrics, however, their existence is a reality, which one must encounter in their research.
The problem of the relationship between texts and visual representations is crucial for the reconstruction of the Tangut Buddhist art, but is even more crucial if the art is envisaged as a component of the Tangut Buddhist history in a broader sense. The attempts to reconcile the Buddhist images and the textual data have a long history from the initial studies of the Tangut art by S. Oldenburg in the 1930s, were continued by N. Nevskij, and still later by K. Samosyuk and other scholars including R. Linrothe, Xie Jisheng, Liao Yang, and M. McCoy. One observation presented in this paper, in a rather ex cathedra fashion, is that the art objects (images) in general do not belong to the official mainstream Tangut Buddhism, represented by the teachings of Huayan, Prajñāpāramitā, Madhyamika (Dbu ma), Vijñānavāda, 'Hundred dharmas' and 'generating the Bodhicitta' as determined by the famous Tangut Tiansheng Law Code. 2 This set of teachings is by no means exhaustive, and the actual excavated corpus reveals a greater variety of textual materials and exceeds the legal stipulations, but a clearly identifiable set of texts in both Tangut and Chinese is fully compatible with the above rubrics. One major 1 The term 'Sino-Tibetan Buddhism' (Hanzang Fojiao 漢藏佛教) was originally suggested by Shen Weirong 沈衛榮. Publications on the topic are not as numerous, but one example is: Suoluoning, 'Jin'gang boruo jing'. 2 Details in: Solonin,'Tangut Buddhist Literature', Suoluoning, 'Xixia Fojiao',[22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38] Nevskij, 'Kul't Nebesnykh Svetil', 52-73. Generally, the approach developed by Nevskij was maintained on a larger scale by K. Samosyuk  addition is the recently identified collection of the atiyoga writings. On the other hand, while the official layer of the Buddhist teachings in Xixia was represented by the abovementioned texts, the images of various Tantric deities correspond to the layer of the Tangut Buddhism, which is currently identified as Tangut 'popular Buddhism '. 3 The criteria for demarcating between 'official' versus 'popular layers' of Buddhism remain vague, however, the hypothesis of the existence of these layers does not seem counterintuitive, and some arguments in favor of this hypothesis have already been presented.
Congruence between the surviving Tangut imagery and the extant textual corpus of Tangut esoteric Buddhism is commonplace in Tangut studies: this means that the available Tangut tantric ritual manuals are distributed among the cycles of Vajravārāhī, Avalokiteśvara, Hevajra, Cakrasaṃvara, etc., and the repertoire of the surviving images is again represented by the images of these deities. In other words, one set of sādhanā texts corresponds with one set of images, which allows one to draw correspondence lines between what is written in the texts and what is represented in the images. The same is true for the Tangut cult of the celestial bodies: Tangut paintings of the planet maṇḍalas have close correspondences in the texts, as was demonstrated by N. Nevskij early on. 4 These two cases are relatively easy to solve. Below, I present several studies which tend to treat images historically, in terms of extracting some information on the transmission of various Buddhist lineages in Xixia. Some cases are easy to solve, whereas for others the connection between texts and images is not as straightforward, and for some we can only proceed from guesswork. 5 Huayan Haiyin daochang chanyi 華嚴海印道場懺儀, CBETA, X 74 no. 1470, 356a24-b18. 6 See: This engraving has long since been in the focus of the Tangut studies, its standard interpretation belongs to Shi, 'Xixia Yijing tu jie', 208-19.

Example 1. Tangut Translation of the Avataṃsaka Sūtra
Unlike its Tibetan counterpart, the way in which the Sinitic Buddhism made its way into the Tangut realm is not clear. That is, the texts of the Tibetan subject matter often include transmission lineages, which are traceable to the well-known Tibetan lineages, and contain indicators for the transmissions of the lineages among the Tanguts. The Sinitic texts present much less historical information concerning transmission of specific doctrines and the initial translation of the Buddhist scriptures. That is, although there is little doubt that the first texts to have been translated into Tangut were the Mahāyāna sūtras, and the translations were based on the Chinese versions, we do not have information on which text was the first to have translated. One useful clue is provided by the Tangut engravings attached to the covers of doctrinal compositions. One such example, which can be interpreted historically is the famous 'Picture of the Translation Session' (yijing tu 譯經圖), which can be partially connected with the 'Lineage of the Huayan Masters of the Great State of Xia' (大夏國弘揚華嚴諸師), preserved in the Repentance on the Platform of the Ocean Seal Samādhi of the Avataṃsaka Sūtra (Dafang guang Fo Huayan jing Haiyin daochang shizhong xingyuan changbian lichan yi 大方廣佛華嚴經海印道場十重行願常 徧禮懺儀), 5 a Yuan dynasty composition by Yixing Huijue 一行慧覺 (early-mid thirteenth century), a Yuan monk of Tangut descent. 6 It was long before scholarship noticed, that corroboration of the data from the engraving with the name list provided by Yixing Huijue allows the identification of one Tangut master lu 2 pụ 2 sjịj 2 djɨj 2 (i.e. Lubu Zhiyun 魯布智雲) from the engraving with the 'Tripiṭaka master Liberation, State Preceptor Lubu Zhiyun (Wisdom Cloud)' 救脫三藏魯布智雲國師, mentioned by Yixing Huijue in his Tangut Huayan lineage. This further allows a suggestion of the 7 Solonin, 'Sinitic Buddhism in the Tangut State'. 8 Wang, Songdai Huayan, 271-92. very early translation of the Avataṃsaka sūtra into Tangut: at least to the period of the third Emperor Huizong (惠宗, Bingchang 秉 常, 1068-1085), but probably even earlier. This observation is based on indications by Yixing Huijue: the first in the lineage the State Preceptor Xianbei Zhenyi 鮮卑真義國師 (probably the same person is depicted in the mural from Yulin cave 29) is mentioned as the 'one who retranslated many sūtras' (重譯諸經), and his successor Lubu Zhiyun as the 'one who transmitted the translation' (傳譯經).

Example 2. Tangut Chan Buddhism
The origins of the Tangut  Solonin,'Chan Contemplation in the Tangut Buddhism', 203-45. 10 Suoluoning, 'Baiyun Shizi Sanguan jiu men chutan ', 9-22. 11 Suoluoning, 'Xixia wen Suiyuan ji ', 285-93. 864), the lower level shows the so-called 'Buddha son White Cloud' (djɨ̣ j 2 phiow 2 śjɨ 2 gji 2 , Baiyn Shizi 白雲釋子) and Chan Master Zhang (tśjow 1 śjã 1 dzji̱ j 2 , Zhang chanshi 張禪師). Suggested reading of the engraving is left to right and from top to bottom, and thus the lineage is represented as follows: Zongmi-Pei Xiu-Baiyun Shizi-Chan master Zhang. Pei Xiu is the author of the Quan fa putixin wen 勸發菩提心文, available in the Tangut translation. 9 Baiyun Shizi is represented by a composition known as The Key to the Three Gates and Nine Contemplations, i.e. 'the three contemplations of Tiantai', 'the three contemplations of the Dharma Realm' and 'three contemplations of the Scripture of Perfect Enlightenment'. The same person emerges in the Tangut composition known under the title sọ 2 śjij 1 dźjwɨ 1 swew 1 ŋwu̱ 1 śio̱ 1 ·jwɨr 2 [The Collection of Mutually Illuminating Words from Three Generations, Tang 27 #109], which is a collection of Buddhist poetry, probably erroneously attributed to the Baiyun sect 白雲宗 of Sinitic Buddhism. 10 Chan Master Zhang was probably the main person to facilitate the publication of a collection of Tangut and Chinese Chan Buddhist lore known under the title tsji 1 na 1 phji 2 xiəj 2 śjã 1 dzji̱ j 2 jiw 1 śjɨ 1 śio̱ 1 [A Collection on Various Occasions by the Chan Master *Biyan from China, Tang 398, #2609, 2610]. 11 Despite obvious differences in their subject matter, the texts mentioned above equally belong to the tradition generally identified as Huayan Chan, and are thus congruent with the image on the engraving discussed above. Despite being anachronistic (Pei Xiu and Baiyun Shizi are separated by at least three hundred years), the Tangut image from the Chart of the Transmission of the Chan Teaching is representative of the reality of the Chan transmission in Xixia, and indicative of the fact that the Tangut Buddhists themselves were aware of the uniformity of their tradition.

Example 3. New Text on Astral Worship
The most rewarding aspect of the inquiry into the relationship between the texts and images was the study of the images of the astral deities popular in the Tangut Buddhism: this study in fact did reveal the symmetry between the texts and the images. The original research was initiated by N. Nevskij, who studied the texts of the 'Laudations of the Celestial Bodies' and the descriptions of the Tangut astral maṇḍalas with the special notice of both Chinese and Indian astronomical systems. Nevskij was able to connect the textual data with the images available from the Khara-Khoto collections. Recently this work was reproduced by K. Samosyuk who identified the visual materials supposedly used by Nevskij and demonstrated that there existed a relationship between the textual and visual materials, at least in the realm of the cult of celestial deities as it circulated in the Tangut state. However, the Tangut textual materials concerning the astral deities are not limited to the ones studied by Nevskij and Samosyuk, and new texts continue to emerge. Again, as in general, the texts devoted to the astral worship, revolving around the cult of Tejaprābhakumāra, are not among the texts of the official Tangut Buddhist curriculum; however, these are the texts where relations between the textual traditions and images can be established.
Below I would like to introduce a text which is currently under study. It is the Tangut text known as 4737 (courtesy of the Institute of Oriental Manuscripts, Russian Academy of Sciences), which is a fragment possibly connected to the text studied by Nevskij, which now is unavailable. The text contains laudations of the deities together with their descriptions. The surviving part contains the entries for the celestial astronomic maṇḍala of the zodiacal planets (Muxing 木星, Yuebo 月孛, Ziqi 紫氣, Yuexing 月星, Rahu 羅睺, Ketu 計度, etc), and the Four Heavenly kings together with their descriptions and votive prayers. 12 For this discussion, I limit myself to two textual fragments, juxtaposed with the images from Khara-Khoto in the for the publication of the complete manuscript is underway; see: Wei, Suoluoning, and Xie, 'Xixia wen xingyao xinyang wenxian'.
Hermitage Museum, available now in the public domain. The text presents several minor reading problems, but in general the translation is transparent and reliable. One example below: The Son of Earth, Yuebo, with wrathful appearance, whose body is red, with one face and two arms, […] screaming with high voice (lit. translation), who covers three thousand [worlds] in one step, we revere the great Yang Lord Yuebo, who was transformed by Tathāgata for the protection of the sentient beings, [we revere?] The images presented below are available on the website of the State Hermitage Museum at www.hermitagemuseum.org/wps/portal. Figure 2, attributed as the image of Yuebo from Khara-Khoto, 13 demonstrate almost exact congruence between the image of Yuebo as he is represented in the texts and in the picture. One can presume that a manual not much different from the one translated above was the background for the currently available pictorial representation of Yuebo. Another example is the textual paragraph and the relevant image of the imaginary star Ketu. The entry for Ketu reads as follows: Laudation for Ketu, his body is dark green and issues radiance, one face two arms, holds iron hook, and reveals wrathful image, and demonstrates courage and ferocity. Praise for Ketu and 14 Figure 3 image available at https://www.hermitagemuseum.org/wps/ portal/hermitage/digital-collection/25.%20archaeological%20artifacts/477258/; the website identifies the deity as Yuebo. Figure 4 image available at https://www. hermitagemuseum.org/wps/portal/hermitage/digital-collection/25.%20archaeo-logical%20artifacts/477259/. revere the great Yang, Lord Ketu, who was transformed by the Tathagata for the protection of the sentient beings.
Currently, two Tangut images of Ketu from Khara-Khoto are available (Figure 3, Figure 4), 14 and both reveal similarity to the description presented in the Tangut text. The same situation is observed in the following parts of the text, which allows us to confirm that the astral imagery in the Tangut iconography was standardized and followed the guidelines in the received texts.

Example 4. Images of Mahāsiddhas
One can suspect that the Tibetan Buddhism, to which a substantial part of the Tangut art object belongs, was not an important factor in the Tangut Buddhist system until the early twelfth century, and its spread in Xixia gained momentum after the 1130s, when the areas of the Qinghai Lake were incorporated into the Tangut domain. As the textual evidence confirms, the major source of the Tangut familiarity with the Tibetan Buddhism was not so much Tibet proper, known to the Tanguts as 'Madhyadeśa of the Bod people' ( 西番中國), but the Amdo area, known as the 'Bod of the Minyag people' (mji 2 wə 1 phə 1 ). This is despite the well-known fact that the famous Indo-Tibetan masters, such as Jayānanda, made their way to the Tangut Kingdom. However, shortly thereafter, the Tibetan Buddhism in Xixia gained momentum, and became a major form of religion among the Tanguts, which was reflected even in the Tangut New Laws, dating from the early thirteenth century. While the nature of Tibetan Buddhism in Xixia remains debated, the traces of both 'tantric' and 'revealed' teachings of Tibetan Buddhism are  15 The high resolution image is not available to me, the image used here is borrowed from https://xw.qq.com/cmsid/20181108B1HXAD00. discovered among the Khara-Khoto findings. Reasonably, the surviving imagery revolves around tantric deities and Indian mahāsiddhas. These personalities were well known among the Tanguts. However, juxtaposition of textual and visual materials does not often return palpable results in terms of positive identifications.
Images of the Tibetan tantric personages including the mahāsiddhas present more problems in the identification. This is especially true for the images of the mahāsiddhas: apparently, the images followed the Tangut textual tradition, such as the one represented in the collection on the lives of 85 mahāsiddhas (八十五成就禱 頌) by Vajrāsana, included in the Yuan-Ming period composition Dacheng yao dao miji (大乘要道密集). Currently, another two versions of the Tangut mahāsiddha lore were identified in the Tangut holdings in the British Library and in the private collection in Beijing. Still another was discovered in the covers of the Tangut Mahāmudrā composition, known as the Essential Collection of Ultimate Seal tha 2 tjɨ̣ j 2 zji 2 dźjwa 1 tshji 1 śio̱ 1 . These traditions are alternative both to each other and to the version by Rtsami lotsāwa, and again alternative to the Seven lineages of Tāranātha and the version by Abhayadatta Śri. Juxtaposing the textual data with the available imagery poses much more trouble than the case with the celestial deities. One positive example is the images of the eight mahāsiddhas of the Vajravārāhī maṇḍala from the Hermitage museum (I am thankful to Christian Luczanits for this comment) ( Figure 5) 15 . Here the eight mahāsiddhas are represented according to their standard descriptions, found in both Indo-Tibetan and the Tangut texts. However, it is not always the case, and in most of the Tangut paintings the mahāsiddhas are impossible to identify with certainty.
Another alternative version of the mahāsiddha identified lineage only in Tangut is as follows: the text identified by Gao Shanshan 高山杉, who was the first to introduce it to the scholarly public, contains seventeen entries with the names, not all completely identifiable. Correlating the data from the above list with the iconographic data might reveal important correspondences. However, this is impeded by a substantial number of unstandardized transcriptions in the Tangut texts, as well as by the non-standard versions of the mahāsiddha's lives. However, this works remains to be done. The situation is made easier by the fact that most of the personalities mentioned in the above laudation, emerge in the colophons of the Tangut texts, and in the transmission lineages included into the Tangut recensions of the Tibetan sādhanā and upadeśas.
Another important figure, responsible for several lineages was Kṛṣnapāda whose name is transcribed into Chinese as Nagchong 捺 乙鐘 identified as historical Pha Dampa Sangs rgyas (Dam pa rgya gar nag chung). He emerges as number ten in the above list under his Tangut name. Given his overall importance in the transmission of the Tibetan Buddhism into the Tangut state, including the system of Vajravārāhī, one would expect to find his image among the Tangut collections. However, no standard (i.e. naked) image of Kṛṣnapāda was yet discovered. One unlikely candidate is the foreign monk with 16 Image available at https://www.hermitagemuseum.org/wps/portal/hermitage/ digital-collection/25.%20archaeological%20artifacts/477198/. black skin in the bottom of another Vajravārāhī maṇḍala ( Figure 6). 16 That is, the identification here cannot be based on the image alone, but needs textual confirmation. That is, the identification of new textual material, or correct interpretation of the existing one, is imperative for the study of the Tangut art.

Conclusive Remarks
The relative importance of the Tangut textual heritage over the visual materials is a fundamental fact of the Tangut studies. That is, the Tangut texts reveal much greater variety of the Buddhist cults and their versions as compared to the ones demonstrated by the visual materials. Although the observations concerning identifications of the iconographical materials are not always exact, the newly emerging texts as well as the careful reading of the previously available materials allows reassessment of the visual materials. A vice-versa process is not possible: when iconographic data is interpreted outside of the textual scope of Tangut textual heritage, the results of such identifications are even less reliable.
On the other hand, juxtaposing Tangut visual and textual materials reveals an important feature of Tangut Buddhism, i.e. its systematic nature: one can observe direct correspondence between what is written and what is represented in the images. Study of the Tangut images combined with the research of the in-depth inquiry into the contents of the Tangut texts is able to return important information or substantiate previous hypothesis.