HOMA IN CHINESE TRANSLATIONS AND MANUALS FROM THE SIXTH THROUGH EIGHTH CENTURIES
Charles D. Orzech
THE EARLY NINTH-CENTURY monk Huilin 慧琳1®describes homa as follows: “In Tang language this Indic term is called ‘fire offering.’ Substances used to make food offerings for the worthies and sages are incinerated in a fire, just as in the four suburban and five marchmount sacrifices, etc.”2
As a disciple of the esoteric master Amoghavajra (Bukong jin’gang 不空金剛, 704–774), Huilin was well aware of the importance of homa not only in esoteric ritual but also as a key instrument in esoteric Buddhism’s support of the imperium. From the very beginning of his rise to power, Amoghavajra had promoted homa as the instrument for supporting the imperial house.3 One of the first pieces of correspondence we have from Amoghavajra’s hand is dated 758, not long after Emperor Suzong retook the capital Chang’an from An Lushan’s forces. Suzong had sent a gift of incense to Amoghavajra to thank him for his efforts on behalf of the imperial cause. In his reply Amoghavajra wrote,
In the tenth month [you] purified the palace in order to set up a convocation to smash the demons … it is fitting that at the three times I bathe the images and at the half month perform homa, so that the thirty-seven worthies protect the Bright King’s territory and so the sixteen [bodhisattvas] nurture and increase the sage emperor’s awesome spirit. (T. 2120.52:828a10–16)
During the next decade altars for homa and consecration (abhiṣeka) were established in the imperial palace, in several imperially sponsored monasteries, and at the great mountain pilgrimage site Wutaishan.4 Given this development, what are we to make of Huilin’s analogy between imperial rites and homa? Is he simply making a comparison of rituals offered on a periodic schedule that involve formal food offering to deities by the method of incineration? Certainly this is the case. But there appears to be more to Huilin’s comment. His choice of indigenous terms in making the comparison is jisi 祭祀—a term used for highly formal sacrifice, and the indigenous rites he chooses are rites performed by the emperor that have the most august, hoary, and orthodox pedigree.5 Huilin seems to be saying that homa is the equivalent of some of the most powerful rituals performed by the Son of Heaven.
Huilin’s comment underscores the importance of the fire sacrifice in the ritual practice of esoteric Buddhism. Homa is used to evoke deities, to eliminate obstructions, to finalize rituals, and even to dispatch ones’ enemies. Indeed, the appropriation of the fire sacrifice was one of the most remarkable developments in late Mahāyāna Buddhism. Fire sacrifice was the core technology of Brahmanic traditions, and homa is detailed in the Taittiriya Aranyaka of the Black Yajur Veda. There, and in later Vedic-influenced traditions, we find homa employed for a wide variety of ends, including ease in childbirth, production of wealth, averting disease or illness, and so on. In these works, homa anchors a vast semiotic system—classifications of plants, ritual actions, and the etiology of fortune and misfortune.
The practice of homa did not cease with the spread of Buddhism, even among Buddhists. Although Buddhism criticized some Vedic practices, notably animal sacrifice, in a wide variety of early Buddhist scriptures including the Kūtaḍanta Sutta of the Dighanikāya, Suttanipāta, and elsewhere the Buddha is depicted as endorsing or reinterpreting the Vedic practice of fire offering for householders and even claiming to be the original teacher of the Vedas in past existences.6 Nonetheless, for the better part of a millennium homa apparently remained a marginal ritual adjunct practiced by some Buddhist householders. But sometime around the fifth century C.E. homa and other elements of the Vedic system such as abhiṣeka began to assume prominent roles in ritual manuals and dhāraṇī scriptures.7 Within a century homa had become central to the new technology of esoteric Buddhism.8 Along with its patron deity Agni, homa is a case in point in the late Mahāyāna and esoteric Buddhist redeployment of traditional Indian deities and ritual practices.
Abundant evidence for the Buddhist appropriation of homa, most of it long ago lost in Indic languages, has been preserved in Chinese translations and ritual manuals.9 The first part of this essay outlines the Buddhist appropriation of homa as reflected in sixth- through eighth-century Chinese translations.10 In the second part, I advance an understanding of homa based on recent studies of the liturgical construction of public subjectivity. Through such an approach, I argue that we can understand homa as producing a form of subjectivity based on culturally constructed “types,” which is then extended to create an “interior” ritual space.
HOMA IN EAST ASIA
The impact of homa in East Asia was decidedly uneven. Although Michel Strickmann’s “Homa in East Asia”11 and his Mantras et Mandarins: Le Bouddhisme tantrique en Chine12 provide a durable overview of the topic, reading his works can lead one to conclude that homa had similar distribution and influence in both China and Japan. Homa was imported to Japan first in a variety of late Mahāyāna translations, then by returning pilgrims such as Saichō and Kūkai as a core technology of esoteric forms of Buddhist practice. As Strickmann and others have shown, over time homa and its most prominent deity Acala (Jp. Fudō 不動) came to have a wide ritual and semiotic impact on Japanese religion and culture. To this day one can go to a Shingon temple and engage a Buddhist priest to perform a *homa.*13
In contrast to the broad cultural distribution of its offspring in Japan, homa in China was largely confined to a highly trained elite and appears to have had little impact beyond that milieu.14 Despite widespread appearance in translations and despite the foundation in the eighth century under imperial patronage of numerous altars for homa and abhiṣeka in the Tang capital and beyond, homa was for the most part confined to its role as a powerful technology for producing imperial buddhas. Paired with abhiṣeka and supported through a variety of specialized ritual paraphernalia, its performance remained the preserve of a few exclusive lineages in the service of imperial courts and elite patrons.15
BUDDHIST APPROPRIATION OF HOMA: THE TRANSLATIONS OF YAŪOGUPTA, ATIKŪṬA, AND BODHIRUCI
Ritual can be understood as a means for shaping the body, mind, and world. As we shall see, the performance of homa has been used to create and reinforce a variety of subjectivities, and these are rationalized in a taxonomic logic. Indeed, we might very well think of the practice of homa as evolving from simple analogical practice into the ritual embodiment, performance, and extension of an entire taxonomic system.
Although homa is mentioned in Chinese documents of the fifth and sixth centuries, detailed descriptions of homa begin to appear near the end of the sixth century. Most early references to homa are cursory and appear to be what Michel Strickmann wryly (and perhaps misleadingly) termed “pious fumigation.”16 For example, Saṅghabhara’s early sixth-century translation of the Mahāmāyūrīvidyārajñī-sūtra (Kongquewang zhou jing 孔雀王咒經, T. 984) involves precisely this sort of “fumigation” to expel demons: “When the mustard seeds are incinerated in the fire the bodies of the evil demons catch fire” (T. 984.19:459a). No other details are provided. What is clear is that the ritual relies on an analogy with incineration. Fire burns and consumes, and the acrid burning of mustard seed, by analogy, burns and consumes the bodies of demons.
A different ritual logic appears in what is arguably the earliest unambiguous record of a votive homa. The rite, in which offerings are made to a deity, is found in the sixth-century Avalokiteśvaraikādaśamukha-dhāraṇī-sūtra (Shiyimian guanshiyin shenzhou jing 十一面觀世音神呪經, T. 1070) of Yaśogupta 耶舍崛多, dating from 561 to 578.17 This scripture describes a ten-inch tall image of the eleven-headed Avalokiteśvara fashioned from wood, to which the practitioner presents various offerings while chanting a dhāraṇī over the course of the first fourteen days of the month, and a fire altar is set up in front of it (T. 1070.20:150c22–151a19). Beginning on the fifteenth day of the month the practitioner is to take 1,008 pieces of incense, dip them in soma oil (sumo yu 蘇摩油),18 and offer them into a sandalwood fire. If the ritual is properly done, on the evening of the fifteenth day Avalokiteśvara enters the altar; the image shakes, and a voice praises the practitioner and offers to grant four supernormal boons (151a20–151b2). While an exchange with the deity that characterizes later homa is present, the focus here is on the physical image, the recitation of the spell, and on the discipline of performing a lengthy and detailed rite over many days.
Yet, in comparison to later examples of homa, certain elements are absent. There is no invocation of Agni nor is there a summoning and enthronement of deities in a mandala. These elements first appear almost a century later in Atikūṭa’s (阿地瞿多, fl. 650s) Tuoluoni ji jing 陀羅尼集經, the Collection of Coded Instructions (Dhāraṇī-saṃgraha-sūtra, T. 901) of 654.19 Homa appears repeatedly in the Tuoluoni ji jing. Some of the homas resemble those found in Yaśogupta’s translation of the Avalokiteśvaraikādaśamukha-dhāraṇī-sūtra produced a century earlier. For instance, “Vajragarbha’s Reception into the Dharma Altar” again employs an image before which offerings are made, and once again the image speaks to the practitioner (T. 901.18:851a23–851c04).
Elsewhere in the Tuoluoni ji jing, however, we find homa’s Vedic lord Agni playing a key role, and the rite is closely integrated with abhiṣeka (consecration/coronation) and the logic of a mandala.20 In a long section once again focused on the Eleven-headed Avalokiteśvara we find disciples given abhiṣeka, and this ritual is followed by a homa. Here, in contrast to other early descriptions of homa, the scripture says that the ācārya is to sit in the western gate of the altar, facing east, with body properly arranged and with correct mindfulness first call Agni and seat him in the fire altar. Next, using honey, sesame, and so on, seven times make offerings into the fire while intoning the spell. Next the ācārya says the spell mentally and Agni goes to the outside of the fire altar and is seated. Next call Horse-headed Avalokiteśvara to take a seat on the lotus throne in the fire altar. Next the ācārya grasps the vajra and one by one summons the disciples to approach (T. 901.18:816a27–816b5).
After further offerings Horse-head is dismissed, and Eleven-headed Avalokiteśvara is invited to take up the seat in the homa fire. A series of other deities is likewise summoned. It is notable that an image is no longer the focus of the rite—instead, the fire as the seat of the deities becomes the focus. The text describes the ācārya summoning and dismissing a series of deities, beginning with Agni. While these divinities are summoned seriatim they are now deployed according to the logic of the comprehensive mandala that structures the collection, the “all-gathering mandala” (douhui daochang 都會道場)—and in certain respects (the hearth as lotus throne, for instance) the logic of the mandala now structures the performance of *homa.*21
Yet the homa here is laid out without any overt mention of an interiorized mental performance that is found in eighth-century manuals—there is no discussion of “contemplation” (guan 觀) or “visualization” (xiang 想).22 The practitioner summons deities one after another as though they are visible nearby. But indications of interiorized mental performance of ritual are found elsewhere in this scriptural collection.23 Indeed, Koichi Shinohara argues that the insertion of an “interior” visualized dimension into otherwise typical dhāraṇī scriptures led to the emergence of a comprehensive mandala structure.24 Be that as it may, in the Tuoluoni ji jing the inner-versus-outer homa discourse that structures much later discussion is largely absent.25 I will return to this distinction between outer versus inner homa below.
Further, in the Tuoluoni ji jing, the fire throne makes the deities and their powers immediately accessible and actualized, starting with the Vedic lord of fire himself. With the fire/Agni as mediator of divine presence, the disciples can be introduced directly to each divinity.26 The summoning of the deities of the mandala is directly paralleled by the summoning of humans to interact with them. Indeed, the list of possible humans who may be introduced to the deities begins with the king and government ministers and ends with disciples. Clearly the intended audience for the performance, at least ideally, is people in power. In the Tuoluoni ji jing, then, homa served not only to extirpate obstacles or “seal” the results of a rite but also as an interface, a device to harness divine power for human purposes.
Less than half a century later, in a group of texts translated by Bodhiruci 菩提流支 (?–727) under the auspices of Empress Wu’s imperial patronage, homa is once again central to the development of esoteric ritual. Bodhiruci’s translation of the Scripture of the Cakravartin of the Single Syllable of the Buddha’s Crown (Ekākṣara-uṣṇīṣa-cakravartin; Yizi fo ding lunwang jing 一字佛頂輪王經, T. 951, 709 C.E.) contains a long segment titled “Homa Altar” (humo tan 護摩壇), which forms the final section of this extensive compendium.27 The text sets out the differently shaped altars suitable for each type of rite and appears to be the earliest use of what becomes the canonical threefold taxonomy of rites: śāntika for pacification (anyin fa 安隱法, T. 951.19:262a13), pauṣṭika for prosperity (qiu dafengrao zhu zhong shan fa 求大豐饒諸眾善法, T.951.19:262b3), and abhicāruka for subjugation (diaofu ta fa 調伏他法, T.951.19:262b21).28 Here we also see the appearance of a fully formed system evident in types of ritual and details of performance. Each rite has a designated time of execution, type of wood, color of garments, and so on.
An emerging ritual synthesis is evident in another of Bodhiruci’s translations, the Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations (Bukong juansuo shenbian zhenyan jing 不空罥索神變真言經, T. 1092) produced in 707 c.e. In this text mudrā, mantra, mandala, abhiṣeka, and homa now appear as a fully integrated system.29 There are richly detailed descriptions giving extensive instruction on performance and accoutrements. In the section titled “Homa Augmentation” (humo zengyi 護摩增益)30 details of the altars are minutely specified; the placement of deities forming a mandala are indicated by their mudrās; instructions are given to expel evil influences, protect the body, cordon off the ritual area, and invite the deities (some of which had already appeared in the Tuoluoni ji jing). The text specifies offerings, including vases of argya water, incense, and flowers, and then goes on to detail the offerings of wood into the fire.
Whereas the homa rites of the Tuoluoni ji jing either omit interior mental operations or indicate them in the barest terms, the Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations sometimes includes detailed instructions for visualization. For instance, the practitioner is instructed to contemplate the golden-colored flames of the fire becoming a ra (囉) bija, or seed syllable, which then changes into Agni, whose body, color, implements, faces, eyes, and so on, are described.31
By this point in the early eighth century, homa has a range of functions—from the expulsion of demons to the rectification of ritual shortcomings and the ritual expression and actualization of relationships between practitioners and deities of the mandala. In these texts the role of homa depends on its placement in the larger ritual and its relationship to other ritual operations. Further, by the early eighth century homa texts are beginning to include sophisticated instruction concerning mental operations to be executed in conjunction with ritual performance.
Canonizing Homa
Seventeen years after Bodhiruci’s translation of the Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations, Śubhākarasiṃha’s (善無畏, 637–735) translation of the Mahāvairocana-sūtra (Mahāvairocanābhisaṃbodhi-sūtra, Da piluzhena cheng fo shenbian jiachi jing 大毘盧遮那成佛神變加持經, T. 848, 724 C.E.) and the production of its Commentary with the aid of Yixing (Da piluzhena cheng fo jing shu 大毘盧遮那成佛經疏, T. 1796;) makes clear that homa was now at the heart of a comprehensive ritual praxis. The Mahāvairocana-sūtra, its Commentary, and Śubhākarasiṃha’s 726 translation of the Susiddhikāra-sūtra (Susiddhikāramahātantra-saddhanopāyikā-paṭala, Suxidi jieluo jing 蘇悉地揭羅經, T. 893) describe three types of homa with encyclopedic information on ritual times, implements, supplies, and performance. They set the basic pattern for all subsequent homa ritual in East Asia.32
Although Bodhiruci’s translation of the Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations (707 C.E.) and those of Śubhākarasiṃha not quite two decades later appear to reflect the same developments in South Asia, the newly translated works add important elements to Bodhiruci’s Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations. For its part, the Susiddhikāra, something akin to a ritual reference work, reflects the rationalization and codification of these developments and allows us to view the system in all of its details. The Mahāvairocana-sūtra provides another form of rationalization: it offers a theological justification for the use of such rituals by monastic practitioners.
The Susiddhikāra codifies the growing mass of specialized late Mahāyāna ritual technology. The radical reduction of possible types of homa, apparently in line with the focus on three buddha families, is only the most obvious sign of its work in rationalizing ritual practice. Producing such a work and restricting access to it through rigorous training and initiation indicates a further regimen of control. The notion of an all-encompassing fully rationalized system is evident everywhere in the Susiddhikāra. All things come under the sway of the taxonomic order. The following passage from the section on food offerings is typical:
Cakes of rice flour and so forth are for the Buddha Family, śāntika [rites], and higher accomplishments; all cakes of wheat flour are for the Lotus Family, pauṣṭika [rites], and middling accomplishments; and cakes made from sesamum or beans are for the Vajra Family, abhicāruka [rites], and lower accomplishments.33
In other words, the taxonomic logic of the threefold homa is potentially extended to all things in the world. Everything can be classified according to its use in rites of pacification, increase, or subjugation.
While the Mahāvairocana-sūtra exhibits the same kind of rationalizing force, another form of rationalization is also evident. Unlike other ritual procedures (mantra, dhāraṇī, abhiṣeka, mudrā, etc.) the Mahāvairocanā-sūtra goes out of its way to wrap the appropriation of homa in a special theological justification, as if to deflect the opprobrium of those who would charge Buddhism with being a pale version of the Veda. Chapter 27 of the sixth fascicle of the Mahāvairocanā-sūtra recounts the genealogy of Agni through forty-four “generations,” many of which are “fires” for specific kinds of events, such as conception, weddings, and so on.34 In contrast to other texts on homa we have encountered, the discourse here is rendered almost entirely in gāthās (verses). For his new post-enlightenment dispensation, the Buddha replaces the Brahmanic fires with twelve fires for the accomplishment of all aims:
Having become perfectly
Awakened to Enlightenment,
I have taught twelve kinds of Fire.
What are those twelve?
The first is the Fire of Awareness
Who is called Mahendra….35
Here the fires have gained metaphorical extension. But the Buddhist appropriation of homa is not yet complete, for these twelve fires, even when carried out with punctilious piety, are merely “exterior.” True homa has a simultaneous interior aspect “which eradicates all karma.”36 In the words of the Commentary, “Homa is the fire of the Tathāgata’s wisdom. It is able to incinerate the karmic connections that produce all misfortunes.”37 Thus, the Commentary reaches back to the fundamental analogy of incineration only to recast it in a higher metaphysical key. Just as the Upaniṣads had appropriated and internalized the Vedic fire as the tapas of the yogi, now Buddhism, while preserving the external rite, claimed an inner or higher understanding.38 This appears to be the first explicit use of the inner versus outer hermeneutic applied to homa. In short, as homa was appropriated it underwent a process that subordinated the exterior practice of ritual to the process of interior contemplation and visualization connected with the mandala. The subjectivity of the performer of homa apparently needed ideological shaping, lest the practitioner fall into the heresy of merely “exterior” performance.
THE MANUALS OF AMOGHAVAJRA AND FAQUAN
The system embodied in the Mahāvairocana-sūtra and the Susiddhikara was not the final one, and it is clear that the practice of homa continued to be refined and developed in South Asia. Vajrabodhi (Jin’gangzhi 金剛智, 671–741) and his young disciple Amoghavajra arrived in the Tang court in 720, a mere four years after the arrival of Śubhākarasiṃha. Vajrabodhi rapidly produced an abbreviated form of the Sarvatathāgata-tattva-saṃgraha, which is based on a five-buddha family model rather than the three-buddha family model found in the Mahāvairocana-sūtra and the Susiddhikara. In line with this fivefold taxonomy, some of the newly translated texts expand the taxonomy of homa. For instance, the Scripture of the Vajra Summit Pavilion of All Yogas and Yogis (Jin’gangfeng louge yiqie yujia yuqi jing 金剛峰樓閣一切瑜伽瑜祇經, T. 867) attributed to Vajrabodhi expands the list to five kinds of homa: protection (xizai 息災), prosperity (zengyi 增益), gaining love (aijing 愛敬), subjugation (xiangfu 降伏), and captivation (gouzhao 鉤召).39
Up to this point homa has been found solely in Chinese translations of Indic texts.40 By the late eighth century, indigenously composed Chinese homa manuals make their appearance. Although there are numerous descriptions of homa in Japanese collections, two manuals appear to be Chinese compositions: the Regulations for Homa for Vajra Pinnacle Yoga (Jin’gang ding yuqie humo yigui 金剛頂瑜伽護摩儀軌, T. 908) attributed to Amoghavajra and the Regulations for Establishing a Mandala for the Performance of Homa (Jianli mantuoluo humo yigui (建立曼荼羅護摩儀軌, T. 912), dating to the mid-ninth century and attributed to Faquan (法全).41
Amoghavajra’s Regulations for Homa for Vajra Pinnacle Yoga presents five types of homa. On first glance, it is immediately apparent that this text is different from the descriptions of homa found in most translations of Indic works. Indeed, Amoghavajra’s text opens not with prose instructions but with verses (gāthās). In this regard, the form of Amoghavajra’s Regulations for Homa emulates the five-word couplets discussing homa in the Mahāvairocana-sūtra, where the Buddha launches into a description of the genealogy of Agni and the forty-four Vedic fires.42 Similarly, Amoghavajra begins by extolling the role of homa (T. 908.18:916a10–a12):
I will now speak of homa—because of it [one] swiftly [obtains] siddhi
Owing to homa’s acts and rules, union is available
Thus, all matters are consequently clarified and done as they should be
Performing homa according to type is the act of highest accomplishment.
In contrast to the translations of Atikūṭa and Bodhiruci, not simply the first few lines but the entire first half of the text and its conclusion are rendered in five-word couplets. The balance of the text consists of prose ritual instructions focusing on the details of performance of the sort encountered in the translations of Atikūṭa and Bodhiruci. The verses render a common feature of Sanskrit Mahāyāna literature in which gāthās often present a summary of a story or points of the teaching. This is exactly their function in Amoghavajra’s manual, where the first half of the text rehearses the taxonomic outline of the homa system. Thus we find the following topics summarized:
Five types of homa
Shapes of altar for each type
Preferred time to perform each type of ritual
Direction to face while performing each type
Type of mudrā
Type of wood to be consumed for each kind of rite
Type of esoteric emblem (wheel, three-pointed vajra, etc.)
Size of hearth
Layout of surrounding altar/mandala (disposition of the families) according to type of rite
Description of deities (Agni, Yama, Indra, Vayu, Vaiśravaṇa, etc.)43
The order of the first long section is typological and structured by the fivefold taxonomy of rites. It could not be used as a manual for a particular performance of, say, a homa for subjugation. It seems clear, therefore, that the first half of Amoghavajra’s manual was to be chanted, either alone or in unison, but this chanting practice must be distinct from the performance of a particular homa, and the material presented in the gāthā half of the text contrasts sharply with the prose instructions that follow.
The prose instructions, like those found in the manuals of Atikūṭa and Bodhiruci, present a series of mantras accompanied by mudrās and ritual actions—the actual mudrā-by-mudrā actions involved in performing each sort of homa. Here, for instance, the color of garments and the demeanor and posture for each type of homa are specified. As the instruction proceeds, the manual presents the iconographic details for the evocation, visualization, and worship of Agni. Depending on the type of rite, different main deities are used. Thus, for rites of increase Mahāvairocana is the focal deity; for subjugation, Acala or Trailokyavijaya. The text ends with another radical break, shifting back to a final sixteen couplets in verse form describing and extolling the characteristics of the two ladles used for *homa.*44 In short, Amoghavajra’s text is half group recitation and half ritual instruction.
Faquan’s manual, in contrast to Amoghavajra’s, begins with a very brief prose list of seven steps for preparing for homa, commencing with propitiation of the earth god and ending with instructing disciples. Most of the remaining text is in five-word couplet form. The few exceptions are instances where mantras are inserted and where the form changes from five-word couplets to seven-word couplets, specifically designated “gāthā” (ji 偈, T. 912.18:929c2) or “praise” (zanyong 讚詠, 934a5). Although the gāthās appear superficially similar to those in Amoghavajra’s manual their tone is quite different and their cadence seems more natural. Faquan’s manual is more closely tied to ritual performance than Amoghavajra’s manual. Indeed, while the text was obviously intended for chanting, it could also structure a performance in the order presented.
Although widely used in Chinese translations of Indic scriptures, it is notable that the use of the gāthā form is dominant in these key treatments of homa—the Mahāvairocana-sūtra and the manuals of Amoghavajra and Faquan. It may not be a coincidence that the Mahāvairocana-sūtra, in setting the style for subsequent treatments, is also the most overtly “doctrinal” of the so-called esoteric scriptures. So we should ask, how does the use of the gāthā form in these instances structure the practice of the initiate? What work does the gāthā perform? The gāthās in the Mahāvairocana-sūtra and in Amoghavajra’s manual cannot have been a part of the actual performance of homa. I would suggest, rather, that they were designed to inculcate a particular taxonomic ideology of homa and thus to preemptively shape the subjectivity of the potential performer. This chanting, performed individually or in a group, would be ancillary to the performance of the homa rite itself. The chanting then embodies and conditions the subjectivity of the performer—it “subjugates” him and shapes the mental attitude with which he undertakes performance of the ritual. Although Faquan’s text, in contrast, appears to be compatible with a chanted, liturgical performance of the homa ritual, it too appears to be aimed at shaping the performer’s subjectivity.
The Subject of Homa
Ritual practice in esoteric Buddhism is not an individual pursuit. Not only must key rites such as consecration (abhiṣeka, guanding 灌頂) be performed with others, but enlightenment itself is inextricably social—it is a spectacle produced through ritual practice. Even when one performs a ritual “alone” one generates a mandala and populates it with deities; the ritual subject is socially conceived. Although disciples swear oaths of secrecy, the rites themselves are performances intended for an audience, sometimes an audience of a few disciples, sometimes an imperial audience, and sometimes an even wider audience. For instance, Zhaoqian’s biography of Amoghavajra says that in 755 he performed abhiṣeka for the military commander Geshu Han and his subordinates and that “nobles and the like, an assembly of some one thousand persons ascended the ritual arena.”45 Such rites are designed to interpellate both the initiate and those observing it into a social practice.46 Homa rituals are equally social, although by the eighth century they are discussed within an interpretive framework of “inner and outer” that might lead one to imagine that the former is of more import than the latter.
The impression of the solitary quest can be traced back to the foundational myth of Siddhartha’s renunciation—itself a story shaped by Upaniṣadic discourses that distinguish exterior from interior reality, promoting the latter over the former. Esoteric texts are often structured around an opposition between “exterior” and “interior” performance.47 Beginning with the Mahāvairocana-sūtra in the early eighth century, discussions of homa are also structured around this opposition. As such, our attention is channeled by an ontology that separates the self into subject and object. From such a perspective we can only observe outward “signs” while the interior “experience” remains obscure. The social becomes, by definition, secondary and derivative. Esoteric texts regularly include sections on discerning the signs (lakṣana, xiang 相) of inner qualities—a good disciple, a good teacher, success in ritual practice. Indeed, the easy fit between this traditional taxonomy of inner and outer and much contemporary discourse on religious “experience” can divert our gaze from the social production of subjectivity or self.48
Recent work on the creation of ritual subjects or selves, both in the present and in antiquity, affords us an alternative.49 In this view, “subjects” are socially produced ritual and discursive objects. They are codes that are produced, propagated, and shared through institutional means. In this reading the scriptures and ritual manuals for the performance of abhiṣeka and homa detail the process for producing and displaying a subject constructed in ritual and liturgy. This subject then can be understood as an institutional construct, which is typical, rather than unique and autonomous—a subject produced socially for institutional ends.50 Unlike the interior self, the subject of abhiṣeka or homa is a socially accessible subject, produced for religious manipulation and available for study. Indeed, its utility is precisely the fact of its social accessibility. Thus, although traditional discourses privilege the interior self, a social approach to these rites allows us to invert the usual hierarchy of interior and exterior to view the socially produced subject of ritual as the primary fact.
This social production and display of the subject is demonstrably the case, for instance, when we examine manuals concerning the process of homa. Homa is the enactment of a ritual taxonomy, and it produces and displays a typical subject. Unlike the Lenten liturgies of Byzantine Christianity examined by Derek Krueger or the rite of abhiṣeka, homa does not construct a self through the worshiper’s channeling of a religious persona.51 Rather, homa constructs a typical subject in the same way that a system of conventional iconography does. The types are not restricted to unique stories but function as elements in a semiotic code, with the practitioner taking on elements of the code as appropriate to the type of homa being performed (for pacification, increase, subjugation, etc.). In performing homa one instantiates a set of relationships that define one as a subject.
Let us return to the “Homa Altar” (humo tan 護摩壇) section of Bodhiruci’s translation of the Scripture of the Cakravartin of the Single Syllable of the Buddha’s Crown (T. 951). As noted above, the differently shaped altars to be used for each type of rite (śāntika for pacification, pauṣṭika for prosperity, and abhicāruka for subjugation) are indices of an entire taxonomic system. For instance, in describing abhicāruka rites for subjugation the text specifies not only that the fire altar be triangular in shape but also what direction the corners should be oriented to, specifics about its size and construction, what part of the month is best for performance of the rite, that one must wear black or red garments, where to sit and what posture to assume, how one should look when chanting (furious), what sort of wood to use (jujube wood, kudong wood, both sour/bitter), and so forth (T. 951.19:262b6–13). The same taxonomic logic is set out in in the Susiddhikara, and is clearly manifest in Amoghavajra’s Regulations (summarized in Appendix I).
Inner and Outer
By the beginning of the eighth century, as we have seen, reiterations of the homa in ritual manuals were frequently accompanied by instructions for interior mental practice. Thus, one “visualizes” deities evoked in the homa rite, including Agni, Acala, and so on.52 Bodhiruci’s 707 c.e. translation of the Scripture of the Mantra of Amoghapāśa’s Miraculous Transformations (Bukong juansuo shenbian zhenyan jing 不空罥索神變真言經, T. 1092) gives specific directions for the mental evocation of Agni:
One should take the incense water and sprinkle it onto the flames snapping the fingers three times, one should visualize the golden flames making a ra character and transforming into Agni with a single face, three eyes, a white body with four arms, and yellow hair as a crowning topknot. Of the left two hands, one grasps a pitcher, one a jeweled staff. Of the right two hands, one displays the mudrā of fearlessness, one clutches a rosary. [He is in] half-lotus position emitting great flames. (T. 1092.20:260b4–7)53
Although earlier esoteric texts such as the Tuoluoniji jing employed a hermeneutic of inner and outer and privileged the ontological status of the former over the latter, they did not overtly do so with respect to homa. Only with the Mahāvairocana-sūtra do we see the same hermeneutic applied in discussions of *homa.*54 Thus, according to the Mahāvairocana-sutra:
There are two kinds of homa, referred to as inner and outer.
One attains liberation from karma and rebirth and also the production of sprouts and seeds [of enlightenment]
[And] because it is able to incinerate karma we speak of it as inner homa.
For outer use there are three positions (practitioner, deity, and hearth)
And each of the three abides in the others
Accomplishing the way of the three karmas (body, speech, mind)
Is the surpassing mundane homa (T. 848.18:32c23–25).55
Incorporating homa within the binary framework of inner versus outer, the Mahāvairocana-sūtra creates a hierarchical distinction in a process hitherto unmarked by one.
The textual descriptions of inner mental ritual paralleled to outer action are a crucial part of the ideology of esoteric Buddhism from the eighth century onward. In its fully developed form this ideology goes beyond linking external performance with internal mental performance. It is designed to produce a living divinity through inner visualization and union with the deity coupled to outer ritual action. The Commentary on the Mahāvairocana sūtra by Yixing and Śubhākarasiṃha repeatedly emphasizes this act of union,56 saying that “if practitioners practice intensively they will cause their three karmas (body, speech, mind) to be identical with those of the focal deity”57 and that “contemplating one’s own body as identical with the focal deity causes one to understand inner and outer.”58
This is the final move through which the Vedic rite was fully domesticated. In turn, by inverting the received taxonomic hierarchy of inner and outer we can see textual instructions concerning interior mental rituals as an extension of a ritually created, iconographically conventional, and socially shared subject. In this light, then, the ritual process of homa is affirmed and extended through textual descriptions of inner vision, and these descriptions contribute to the ritual production of a socially constructed subject.
APPENDIX I
SOME RITUAL CORRESPONDENCES IN REGULATIONS FOR HOMA FOR VAJRA PINNACLE YOGA (T. 908)
This chart omits the more complex directives concerning the layout of the types of altars.
[APPENDIX I.
OME RITUAL CORRESPONDENCES IN REGULATIONS FOR HOMA FOR VAJRA PINNACLE YOGA (T. 908)]
NOTES
REFERENCES
Althusser, Louis. “Ideology and Ideological State Apparatuses.” In Hazard Adams and Leroy Searle, eds., Critical Theory Since 1965, pp. 239–251. Gainesville: University of Florida, 1986.
Barrett, Timothy Hugh. “Stūpa, Sūtra and Śarīra in China, c. 656–706 CE.” Buddhist Studies Review 18 (2001): 1–64.
Brakke, David, Michael L. Satlow, and Steven Weitzman, eds. Religion and the Self in Antiquity. Bloomington: Indiana University Press, 2005.
Chen Jinhua. Crossfire: Shingon-Tendai Strife as Seen in Two Twelfth-century Polemics, with Special References to their Background in Tang China. Studia Philologica Buddhica Monograph Series 25. Tokyo: International Institute for Buddhist Studies in Tōkyō, 2010.
Copp, Paul. “Dhāraṇī Scriptures.” In Charles D. Orzech, Henrik H. Sørensen, and Richard K. Payne, eds., Esoteric Buddhism and the Tantras in East Asia, pp. 176–180. Leiden and Boston: E. J. Brill, 2011.
Davidson, Ronald M. “Abhiṣeka.” In Charles D. Orzech, Henrik H. Sørensen, and Richard K. Payne, eds., Esoteric Buddhism and the Tantras in East Asia, pp. 71–75. Leiden and Boston: E. J. Brill, 2011.
Davidson, Ronald M. Indian Esoteric Buddhism: A Social History of the Tantric Movement. New York: Columbia University Press, 2002.
Davidson, Ronald M. “Some Observations on the Uṣṇīṣa Abhiṣeka Rites in Atikūṭa’s Dhāraṇīsaṃgraha.” In István Keul, ed., Transformations and Transfer of Tantra: Tantrism in Asia and Beyond, pp. 77-97. Berlin and New York: Walter de Gruyter, 2011.
Forte, Antonino. “The South Indian Monk Bodhiruci (D. 727) Biographical Evidence.” In Antonino Forte and Federico Masini, eds., A Life Journey to the East: Sinological Studies in Memory of Giuliano Bertuccioli (1923–2001), pp. 77–116. Kyoto: Scuola Italiana di Studi sull’ Asia Orientale, 2002.
Giebel, Rolf, trans. Two Esoteric Sutras: The Adamantine Pinnacle Sutra/The Susiddhikara Sutra. Berkeley, CA: Numata Center for Buddhist Translation and Research, 2001.
Goble, Goeffrey C. “Chinese Esoteric Buddhism: Amoghavajra and the Ruling Elite.” PhD dissertation, Indiana University, 2012.
Grapard, Allan G. “Japan’s Ignored Cultural Revolution: The Separation of Shinto and Buddhist Divinities in Meiji (shimbutsu bunri) and a Case Study: Tonomine.” History of Religions 23 (1984): 240–265.
Hodge, Stephen. The Mahā-Vairocanā-Abhisambodhi Tantra with Buddhaguhya’s Commentary. London: Routledge, 2003.
Hume, Robert Ernest, trans. The Thirteen Principal Upanishads. Oxford: Oxford University Press, 1921.
Kamei Sōchū. Goma no rekishiteki kenkyū 護摩の歷史的研就. Tokyo: Sankibō Busshorin, 1967.
Krishnan, Y. “To What Extent Buddhism Repudiated Vedic Religion?” East and West 43/1–4 (December 1993): 237–240.
Krueger, Derek. “The Great Kanon of Andrew of Crete, the Penitential Bible, and the Liturgical Formation of the Self in the Byzantine Dark Age.” In Brouria Bitton-Ashkelony and Lorenzo Perrone, eds., Between Personal and Institutional Religion: Self, Doctrine, and Practice in Late Antique Eastern Christianity, pp. 57–97. Cultural Encounters in Late Antiquity and the Middle Ages 15. Turnhout: Brepols Publishers, 2013.
Krueger, Derek. “Romanos the Melodist and the Christian Self in Early Byzantium.” In Proceedings of the 21st International Congress of Byzantine Studies, vol. 1, pp. 255–274. London: Ashgate, 2006.
Krueger, Derek. Liturgical Subjects, Christian Ritual, Biblical Narrative, and the Formation of the Self in Byzantium. Philadelphia: The University of Pennsylvania Press, 2014.
Orzech, Charles D. “On the Subject of Abhiṣeka.” Pacific World 3/13 (2011): 113–128.
Orzech, Charles D., Richard K. Payne, and Henrik H. Sørensen. “Introduction: Esoteric Buddhism and the Tantras in East Asia: Some Methodological Considerations.” In Charles D. Orzech, Henrik H. Sørensen, and Richard K. Payne, eds., Esoteric Buddhism and the Tantras in East Asia, pp. 4–13. Leiden and Boston: E. J. Brill, 2011.
Osabe Kazuo. Tōdai mikkyōshi zakkō 唐代密教史雜考. Kobe: Kōbe Shōka daigaku gakujutsu kenkyūkai, 1971.
Payne, Richard K. “Introduction.” In Richard K. Payne, ed., Tantric Buddhism in East Asia, pp. 5–8. Boston: Wisdom Publications, 2006.
Robson, James. Power of Place: The Religious Landscape of the Southern Sacred Peak (Nanyue 南嶽) in Medieval China. Cambridge, MA: Harvard University Press, 2009.
Sharf, Robert. “Experience.” In Mark C. Taylor, ed., Critical Terms for Religious Studies, pp. 94–116. Chicago: University of Chicago Press, 1998.
Sharf, Robert. “Thinking through Shingon Ritual.” Journal of the International Association of Buddhist Studies 26/1 (2003): 51–96.
Shinohara, Koichi. “The All-Gathering Maṇḍala Initiation Ceremony in Atikūṭa’s Collected Dhāraṇī Scriptures: Reconstructing the Evolution of Esoteric Buddhist Ritual.” Journal Asiatique 298/2 (2010): 389–420.
Shinohara, Koichi. Spells, Images, and Maṇḍalas: Tracing the Evolution of Esoteric Buddhist Rituals. New York: Columbia University Press, 2014.
Strickmann, Michel. “Homa in East Asia.” In Frits Staal, ed., Agni: The Vedic Ritual of the Fire Altar, vol. 2, pt. 1, pp. 418–455. Berkeley, CA: Asian Humanities Press, 1983.
Strickmann, Michel. Mantras et Mandarins: Le Bouddhisme tantrique en Chine. Paris: Éditions Gallimard, 1996.