One Clue Fewer about Klu sgrub

One Clue Fewer about Klu sgrub.[1]

Not long ago, I read, with Ogyan Tanzin Rinpoche, a 13th century partly autobiographical work by Guru Chowang (Chos dbang), his Great Treasure History (gTer ‘byung chen mo).[2] It is clear from the narrative that nāgas (klu) are deeply involved in the guarding of his treasures. The gter is contained in a casket (sgrom bu) which takes the form of a nine-headed nāga, and various signs and omens of nāgas appear during the discovery processes. Local people are greatly alarmed to see Chowang digging for treasure, fearing his excavations will annoy their territorial deities. 

Scroll forward to the mid 20th century. The late Dilgo Khyentse Rinpoche (1910-1991) is standing by a lake in East Tibet, surrounded by a crowd of people observing the action. Everyone in the crowd believes they see a nāga emerge from the lake, and in their full view, hand the great treasure revealer (gter ston) a treasure (gter) comprising an Amitāyus scriptural tantra, together with its associated ritual practices.

At all points between the 13th century and the 20th century, as we can see from the literature, nāgas (= klu) continue to play a similarly major part in gter ma revelation. Nāgas are not the only gter ma protectors, although they are certainly quite prominent. Other territorial deities, often implied by Tibetan savants as local functional equivalents to India’s yakṣas, can be observed behaving similarly.

Public gter ma revelations (khrom gter) were not so unusual in Tibet. Everyone in the audience tended to believe they could see the nāgas in the waters, or see a gter ston magically opening a supernatural treasure portal (gter sgo) in a rock face. 

Such treasure portal openings were performed in Tibet in a manner with some strikingly granular resemblances to Indian underworld entry (bilasādhana) and treasure excavation (nidhivāda) texts, of which a substantial quantity was translated into Tibetan from the Imperial period onwards, and preserved in their tantric Buddhist canons (Kangyur and Tengyur). Some such texts were attributed to Nāgārjuna. 

Did anything akin to the above ever happen in Buddhist India, so rich in nāga lore and from where the innumerable magical texts originated? At least in the case of great scholars like Nāgārjuna, the verdict of many Buddhologists seems to be a resounding “No!” They seem convinced that nothing like this ever happened to Nāgārjuna, and that when Indian Buddhist texts describe such interactions of humans with nāgas and yakṣas to retrieve relics or scriptures, or the entry of a great Buddhist master like Bhāviveka into a magic cave, they are in a great many or even most cases enjoying literary flights of the imagination, or conducting marketing exercises, but not describing a lived, experienced, traditionally established, ritual and cultural reality. So far, I  have not yet found it altogether easy to share in their conviction. 

The Ineffable Nāgārjuna

Klu sgrub, or Nāgārjuna as he is better known outside of Tibetology, is famous for his matchless ability to undermine our most cherished certainties. Once we subscribe to his interpretation of emptiness (śūnyatā), as most Mahāyānists so enthusiastically do, we can no longer take very much for granted. According to Nāgārjuna, there is no ultimate substratum upon which our apparently real world is based, because no existents whatsoever can have svabhāva. In the course of arriving at this radical conclusion, Nāgārjuna systematically and with relentless reasoning refuted the inherent existence, or svabhāva, of everything most sane and normal people take for granted: not merely commonplace entities like tables or pots, but even fundamental principles, like causation, change, the self, knowledge, language, and even truth (Westerhoff 2022: “Conclusion”). 

Fittingly perhaps, Nāgārjuna’s own persona seems similarly ineffable. Despite being the most influential figure in Buddhist history after the Buddha himself, academic scholars still know very little about him with any certainty. For example, debates have long persisted about when he lived. As Ye Shaoyong puts it, “The chronology of Nāgārjuna has long been an issue of debate. Proposals made by scholars vary roughly between the 1st and 3rd centuries of the Common Era.” Likewise no one is exactly sure which of the many works attributed to him were actually written by Nāgārjuna. As Ye Shaoyong (2019: 341) puts it, “The authentic oeuvre of Nāgārjuna is a hotly debated issue.” Jan Westerhoff concurs: “It is not easy to come up with a precise list of texts Nāgārjuna composed.” (Westerhoff 2022: “Life and Works”).

Nāgārjuna’s personal life story seems, if anything, even less knowable. As we will see below, there appear to be only two details about Nāgārjuna’s biography upon which many academics agree: that he did live in South India, and that he did not dabble in magic or have any truck with nāgas—these latter despite the profuse accounts of his varied occult activities and his famous visit to nāgaloka that are extant in Sanskrit, Tibetan, and Chinese alike. The purpose of this blog is to reduce our already meagre certainty by a further 50%. I propose to demonstrate that it is not after all so certain that any ‘historical Nāgārjuna’ was not involved in magic, or with nāga beliefs.

Nāgārjuna’s many life stories

Mahāyānists have long revered Nāgārjuna as the greatest of sages, who successfully restored the Dharma in India from a low ebb. They further believe that he was the first to introduce hitherto unavailable Buddhist teachings of incalculable importance (most famously the Prajñāpāramitā scriptures, but others can be mentioned too), which he received in a nāgapalace. With such a reputation, it is not surprising that Nāgārjuna soon became a metaphor for subsequent tradition building. Perhaps this was particularly true in China. As Stuart Young puts it (2019: 727), “By the end of the Tang dynasty (618–907), Nāgārjuna had become a synecdoche for all things Mahāyāna, and was widely regarded as a founding father of almost every East Asian Buddhist tradition – including most notably Tiantai / Tendai (天) and the so-called Three Treatise school (Sanlun zong / Sanron shū [三論宗]).” One might add, he was equally attributed with recovering the Buddhāvataṃsaka from the nāga palace, cementing his importance to the Huayan  (華嚴) School, and likewise often claimed to have opened the famous iron stūpa in South India to retrieve the Vajraśekhara-sūtra, thus becoming the notional source of most East Asian tantrism. In Tibet, Nāgārjuna was said to be the founder of the Ārya school of Guhyasamāja exegesis (Wedemeyer, 2007:14). And so on.

The situation is further complicated by the proposed existence of several subsequent Nāgārjuna namesakes. Note however that nearly all Asian Buddhist traditions maintain there was only a single Nāgārjuna, many claiming that Nāgārjuna was supernormally long lived. For narratives across much of Asia extol his expertise in alchemy, medicine, and mantra, and how these afforded him great longevity. 

Over time, many varied Nāgārjuna biographical narratives began to proliferate. The challenge facing academic historians has therefore been to extract reliable knowledge about a historic Nāgārjuna from the many contrasting sectarian narratives that evolved over many centuries.

In pursuit of this historical goal, the predictable and reasonable strategy of academic scholars has been to seek out the earliest surviving biographical narratives about Nāgārjuna. As far as we currently know, the earliest of such are in Chinese, and they exist in two clusters. Since I am not a Sinologist, the following section is paraphrased from the publications of Chieko Yamano and Stuart Young, and the great many emails they so generously sent me.

[1] The very earliest extant biographical narratives about Nāgārjuna come down to us in a number of fragmentary accounts bequeathed by the direct Chinese colleagues of the great Kashmiri-educated Kuchean-born translator Kumārajīva (344-413), notably Sengrui (僧叡; c. 352–436), Sengzhao (僧肇; c. 374–414), Lushan Huiyuan (廬山慧遠; c. 334–416), and Tanying (曇; c. 348–418). Ye Shaoyong (2019: 335) estimates 300 CE as a reasonable terminus ante quemfor Nāgārjuna, hence this earliest and broadly concordant tranche of Chinese narratives about Nāgārjuna might have been written within one or two centuries of the man himself. Ye Shaoyang (2019: 337) proposes they “may be derived from an oral tradition from northwest India in the latter half of the 4th century.”

These earliest sources tell us that Nāgārjuna was born as a brahmin, and only later became a Buddhist monk (Ye Shaoyang 2019: 337). Three of them also describe Nāgārjuna’s dealings with nāgas. These three are Prefaces that were included within Sengyou’s (僧祐) Compilation of Notes on the Tripitaka (出三蔵記集) (Yamano 2009a: 186). Firstly, Sengrui in his Preface to the Mahā-prajñā-pāramitopadeśa 大智釈論序;[3] secondly Lushan Huiyuan in his Preface to the Brief Exposition of the Mahā-prajñā-pāramitopadeśa 大智論抄序;[4] and thirdly Tanying in his in his Preface to the Mūlamadhyamaka-kārikā 中論序,[5] These three sources, each of them contemporary with Kumārajīva, all agree that Nāgārjuna was invited down to visit a ‘Nāga Palace’ (龍宮, nāgaloka), where he was given previously inaccessible Buddhist scriptures. Stuart Young (2019: 728) concurs. Regarding other supernormal abilities, Yamano further mentions that Sengzhao (僧肇), in his commentary on the Vimalakīrtinirdeśa-sūtra describes an episode in which Nāgārjuna subdued heretics using his supernatural powers.[6]  In short, as Ye Shaoyang (2019: 337) quotes these early sources, “due to his great contribution to the revival of Buddhism, he was worshipped in India as a buddha” (and yet, as Shaoyang adds, other evidence suggests Nāgārjuna may not have gained wide popularity in India until the end of the 4th century). 

Young (2019: 728) explains further that these above-mentioned Chinese colleagues of Kumārajīva were scholarly monks dedicated to arduous meditational disciplines, which suggests their accounts of Nāgārjuna might have been shaped to conform with their own austere ideals. Yamano independently argues the same point in her PhD thesis. Describing this respectable spin, Young writes (2019: 728): 

“The most robust example is provided in Huiyuan’s preface to his abridged edition of the Da zhidu lun, in which he describes Nāgārjuna as a “high-minded gentleman” who “grieved at the benightedness of the masses” when the Dharma was “decadent and weak,” so took the tonsure and vows of a Buddhist monk. Thereupon, “he dwelt in seclusion in the woods and swamps, lived as a hermit and practiced dhyāna. He stilled his mind and studied the subtle, his thoughts penetrating the supra-normal.” He then “had an awakening” and “turned his steps towards the snowy mountains [of the Himālayas]” where he met a śramaṇa (renouncer) who introduced him to the Mahāyāna teachings. Nāgārjuna next proceeded to the famed dragon (nāga) palace of Buddhist lore, in which “there was no important canon or secret text that he did not master.” “When the roots of his impediments had been pulled up, his name crowned the stages of the [bodhisattva] path … Consequently, the non-Buddhists esteemed his manner and famous gentlemen submitted to his dictates. From that time, the enterprise of the Greater Vehicle flourished again”. 

[2] As Young explains (2019: 727), it was for many years wrongly thought that the earliest Nāgārjuna hagiographies were those found within our second cluster of early sources, the Longshu pusa zhuan (龍樹菩薩傳; T. 2047) and Fu fazang [yinyuan] zhuan (付法藏[因緣]傳; T. 2058). Most scholars of that era (e.g. Robinson 1977: 21-26) also assumed that the Longshu pusa zhuan was a direct transcription of Kumārajīva’s teachings by his disciples, but this is apparently not the case. It is now understood that the texts of this second tranche, which share similar narratives, were compiled several decades later than the first tranche described above (Young 2019). 

Yamano however adds a further perspective in suggesting that many elements in the Longshu pusa zhuan narrative nevertheless might reflect actual information from what the Chinese referred to as the ‘西域 (Western Regions)’. She explains, even though the Longshu pusa zhuan per se was only compiled into a work at a later date, some academic scholars believe there had been an oral tradition relating to Nāgārjuna’s life story, referred to as ‘the Indian Tradition (天竺傳),’ which was a shared source for the Longshu pusa zhuan and the Fu fazang [yinyuan] zhuan, and which dated from Kumārajīva’s time. Yamano herself believes this tradition was to some extent also circulated in written form, and that fragments of it still exist. Thus the contents of the Longshu pusa zhuan might indeed reflect some actual Indian data (Yamano 2009a: 220-1; Yamano 2010). 

Be that as it may, unlike the above first tranche of texts that were compiled slightly earlier, the Longshu pusa zhuan and other texts of this tranche were not written in scholarly academic Chinese, but in a simpler more accessible register of the language, and moreover their contents show little sign of possible redaction to conform with austere monastic ideals.

In these latter sources, just as we see in the Nāgārjuna narratives extant in Sanskrit and Tibetan, much more emphasis is placed on Nāgārjuna’s involvement with magic (Ye Shaoyang 2019: 341). Here he is said to have started out as a brilliant non-Buddhist brahmin youth from South India who was deeply learned in the Vedas and all branches of knowledge. He was also interested and skilled in the magical arts. Notoriously, as a youth, he once brilliantly reverse-engineered the secret medicinal recipe of a magician he had consulted, a complex concoction of seventy different ingredients to be placed on the eyelids to confer invisibility (which, as we shall discuss below, looks like a remarkably accurate description of a specific Indian magical category, añjana). As Roger Corless (1995: 527) translates the passage, the magician tells Nāgārjuna and his friends “..grind this up with water, then smear [the mixture] on your eyelids. Your form will become invisible..” The magician had hoped to exploit his young clients by keeping the mixture secret so that they would have to come back to him for more, but Nāgārjuna was able to ascertain each of the añjana’s seventy different ingredients merely by their smell. He and some friends had planned to abuse the power of invisibility in an audacious attempt to gain sexual favours from a king’s harem (note that in Indian magic, añjana can also be used to confer powers of attraction, although this is not spelled out in the Longshu pusa zhuan). They succeeded for a while, but after some misadventures, Nāgārjuna repented, and converted to Buddhism, dedicating himself to the practice of dharma. At first, his exceptional brilliance made him arrogant, and he tried to establish his own heretical Buddhist sect. However, a great nāga bodhisattva took pity on him, and took him down to his palace, where he shared with Nāgārjuna numerous Buddhist scriptures taught by the Buddha during his time on earth, but hitherto unavailable to humans. It was these texts that allowed Nāgārjuna to achieve an authentic realisation free of pride, upon which the nāga bodhisattva sent him back to South India to begin his teaching career. Thus he became the great Buddhist philosopher known to posterity, who also brought with him key Mahāyāna scriptures from the palace of a nāga king, and this is why he has the element ‘Nāga-’ in his name. 

Reducing our certainties by a further 50%.

What certainties do academic scholars believe they have extracted from these richly flavoured narratives? As mentioned above, one is the association with South India, which is not my primary interest here, and which I have no plans to challenge. The other ‘certainty’ is that all mentions of a great philosopher like Nāgārjuna being involved with nāgas and magic, must be later accretions invented for marketing purposes, that we can safely ignore in any attempt to understand a historical Nāgārjuna. It is this latter ‘certainty’ that I wish to undermine. (NB: in this endeavour, I will necessarily focus on the two magical feats attributed to Nāgārjuna before his Buddhist realisation, the invisibility magic and the visit to nāgaloka, since further miracles often attributed to him later in his career might be better explained in terms of more orthodox Buddhist categories like the ṛddhipāda).

This latter ‘certainty’, that all mentions of Nāgārjuna being involved with nāgas and magic are later accretions we can safely ignore, is quite prominently represented in the secondary literature. For example, Walser’s otherwise excellent monograph (2007) freely acknowledges that pretty much every story related to Nāgārjuna contains the recurring elements of nāgas and magic; yet he feels confident to assume a priori that these narratives are in their entirety apocryphal, invented at a later date to sell the philosopher to royal patrons in China and India. 

Yet Walser (2007: 75) has also falsely assumed, along with several other scholars (e.g. Young 2017: 178), that the seventy-substance invisibility añjana Nāgarjuna is described as replicating in the Longshu pusa zhuan (which in 2007 was still generally attributed to Kumarajīva) should somehow be understood as ‘alchemical’ and can thus be discounted as anachronistic (2007: 77), since alchemy was only first attested in India at a later date. I find that a bit of a stretch. 

  • Firstly, añjana is not classified as alchemy (rasa-rasāyāna) within Indian magic. Añjana is a separate category, including its uses as here described for invisibility.
  • Secondly, as Chieko Yamano points out (2004: 70), “The earliest existent text that narrates Nāgārjuna’s episodes involving immortality or alchemy is Xuanzang’s Da Tang Xiyu Ji (大唐西域記) [7th century]. The earlier biographies of Nāgārjuna do not mention his relationship with rasa-rasāyāna.”  Thus the Longshu pusa zhuan does not specifically mention alchemy (rasa-rasāyāna) either by name or by description; it merely describes añjana, a different category than alchemy (rasa or rasāyāna). 
  • Thirdly, an encyclopaedic (and elite!) education in plants, animal substances, minerals, and motley other ingredients, for medical or magical or other pragmatic purposes, is extremely ancient in India, and the shared foundation of a great many different aspects of India’s sciences and magic. Hence I find it problematic to classify Nāgārjuna’s purported expertise in such learning within a narrow category of ‘alchemy’ that is specifically associated with a later period of Indian history. 

Nevertheless, on the basis of this misunderstanding of Nāgārjuna’s seventy-substance based invisibility magic as specifically ‘alchemical’, and the consequent inclusion of this Longshu pusa zhuan narrative within his chapter on ‘Alchemy’, Walser writes (2007: 75): “Whether one is trying to sell the Nāgārjuna legend to an Indian audience or whether one is trying to export the legend to a Chinese audience, claiming that the saint is an alchemist would have ensured the audience’s attention.” He then makes the observation that no Indian sources describe alchemy (i.e. rasāyana) at such an early date, so that the alchemical reference is anachronistic in an Indian context, although not of course in a Chinese context. From this he concludes that the story of Nāgārjuna’s dabbling in ‘alchemical’ invisibility magic was most likely made up in China by Kumarajīva, to sell Nāgārjuna to an alchemy-crazed Chinese audience. In truth, as mentioned above, invisibility magic need have nothing to do with any later Indian science of alchemy. To reiterate, as Aleksandra Wenta wrote me (24th September 2023), “Invisibility magic need have nothing to do with [later] alchemy. This is an old magic practice, which is found even in the Arthaśāstra, Chapter 14.3.8-13.” So even if the mercury pills or guṭikās typical of much later ‘alchemical’ works do indeed bestow the siddhi of invisibility (which is perhaps what Walser had in mind), this siddhi was already widely attested by other means in much earlier brahmanic literature, notably in this case, through añjana

Walser has undoubtedly produced a great study of Nāgārjuna in Indian context as a Mahāyānist. Yet he tends to skirt around the inescapable fact that magic and nāga cults were likewise extremely prominent parts of the historical and cultural contexts in which any historical Nāgārjuna must have lived, not least within educated brahmanic circles and Buddhist monasteries.

Similarly, I think the very admirable scholar Max Deeg is probably articulating the opinion of many of his Buddhological colleagues when he writes:

“The Buddhist philosopher Nāgārjuna…certainly had nothing to do with a nāga-cult…the narrative of Nāgārjuna being given Buddhist texts (sūtras) by the nāgas in the netherworld is more likely an etiological post-ex-nomine means of making sense of the name in a hagiographical context than the reason for the ‘individual’ having been given the name in the first place” (Deeg 2021: 53). 

My contention is that these authors, and the many Buddhologists who seem tacitly to agree with them, might be mistaken. If we approach Indian magical beliefs and practices with a greater granularity, with a finer focus on their ritual and social-historical actualities, we might conclude that it is in fact not impossible that any historical Nāgārjuna might have (as he saw it) interacted with nāgas to gain dharma treasures from them, or in his youth practice añjana magic based on complex concoctions of substances to smear on his eyelids. 

Now, I am definitely not proposing to claim with any certainty that Nāgārjuna did engage in such activities; nor am I in any way whatsoever attempting to deny that there might have been several later Nāgārjuna namesakes, which I believe is altogether possible. I am merely bringing to the table further evidence and alternative perspectives that I believe cast significant doubts on any existing certainties that Nāgārjuna the Mahāyānist and madhyamaka philosopher was a cerebral intellectual who somehow managed to stand quite apart from his prevailing cultural contexts, and could not possibly have engaged either with the prevalent nāga beliefs and practices of his time, nor with the widely attested añjana magic.  

I wonder if our academic scholarship might have made a category mistake, a failure correctly to understand the relevant classifications of Indian magic and nāga lore, thus not appreciating the cultural cohesiveness of Nāgārjuna’s purported occult activities. By assuming inapplicable categories a priori, we might well have failed to notice the actual Indian magical categories to which his activities collectively seem to conform in the Longshu pusa zhuan narratives. For us, communications with nāgas to obtain treasures from them, and multi-substance invisibility magic, might seem like quite separate pursuits that were randomly attributed to Nāgārjuna to increase his fame, sharing little in common with each other beyond their supposed irrationality and sensationalism. But the separation of these practices into different silos is not really the way that Indian magic was understood. 

On the contrary, the behaviours attributed to Nāgārjuna point rather consistently to his engagement with specific and salient classifications within Indian magical thinking, the densely interrelated and overlapping fields that at some stage came to be called (i) khanyavāda and/or nidhivāda; (ii) pātālasiddhibilasiddhi, or bilasādhana; (iii) añjana

  • The two terms khanyavāda and nidhivāda were sometimes used interchangeably or synonymously, so it could be argued that they should be seen as aspects of the same extended discipline, rather than as separate disciplines (Balbir 1993: 19-22). Both of these terms refer to treasure discovery, the acquisition of various kinds of precious valuables, whether mundane or religious, from within or from under the earth. Magical practices of this sort long predate Nāgārjuna. As several scholars have already pointed out (Goudriaan 1976: 307; Einoo 2009: 34; Vasudeva 2012), they can be found in Vedic sources, such as the Sāmavidhāna Brāhmaṇa, the Atharvavedapariśiṣṭāṇi, and the Arthaśāstra. The most salient finding of Balbir 1993 is precisely to show that khanyavāda and nidhivāda were deeply dependent on a vast knowledge of plants and similar substances, typically for making the magic eyeliner, añjana (see (iii) below), to render the normally invisible magic treasure portals visible, or to evade the dangerous spirits that guarded the treasures by making oneself invisible (for a translated example, see Mayer 2022: 417-438). As Balbir points out (1993: 18), this kind of knowledge of substances and plants for magical reasons was often the preserve of the learned and powerful, of brahmins and kings. Somdev Vasudeva (2012) and Chieko Yamano (2008b: 353-60) describe the same general pattern, of complex potions concocted out of many ingredients, serving as a shared basis for many different forms of magic, often practiced by elites. In the Longshu pusa zhuan, Nāgārjuna displays an exceptional knowledge of the plants and substances used to make a specific añjana, which he can identify merely by smell.
  • Major subterranean treasures were believed almost inevitably to be guarded by nāgas in particular, but also by other powerful spirits, so that an associated and prominent sister science to khanyavāda and nidhivāda was the knowledge of how to directly interact with nāgas. The great advantages of direct familiarity with nāgas generated a copious parallel set of magical practices that eventually came to be called pātālasiddhibilavādabilasiddhi, or bilasādhana: the magical techniques for gaining entry to the luxurious and splendid underground domains of nāgas and asuras. In these subterranean abodes, many boons could be obtained, be they nāga or asura maidens to entertain the human visitor, or magnificent treasures gifted by the nāga kings who luxuriated in vast and wealthy palaces (Vasudeva 2012: 275). There could in fact be several different reasons for wanting to enter nāgaloka to commune directly with the nāgas living there (Vasudeva 2012: 275-278), but all of them circled around the central, pervasive, and extremely ancient cosmological idea of nāgas as controllers of the climate and its waters, and thus owners and masters of all the wealth and treasures of the natural environment. From this premise, nāgas became widely perceived as the traditional guardians of all kinds of treasures, including the many Buddhist relics discussed in Vogel 1926. More recently, Scheible (2016: 109-112) has analysed at length one of the many Pāli narratives about the human sangha retrieving priceless relics of the Buddha from the nāgas in nāgaloka. Given that nāgas were generally understood as controllers of the entire agricultural economy and dispensers of its vast wealth, it should be noted that the ability to visit them was typically associated with elite persons, the powerful and their educated assistants. Thus Vasudeva (2012: 276-7) describes various famous kings and the Mahāyāna Buddhist sage Bhāviveka as vaunted bilasiddhi adepts.
  • añjana, eye ointment or eyeliner, refers to a major category of magic in which various complex magical potions are concocted out of many plants and other substances to put on one’s eye lids. Depending on the recipes and associated rituals employed, these confer a number of varied magical siddhis, which Goudriaan (1976: 318) conveniently lists. They include (a) the power to see otherwise hidden objects, such as the hidden entrances to the nāga realms, or hidden treasures or nidhis as sought by the nidhivādin or khanyavādin (see (i) above). (b) Notably however, and here exactly as we read in the Longshu pusa zhuanañjana equally had the power of making oneself invisible: for an early example of añjana used to confer invisibility, probably from between 2nd century BCE and 3rd century CE, see the Arthaśāstra Chapter 14.3.8-13; for slightly later examples, see Garuḍa Purāṇa 178, 9, as cited in Goudriaan (1976: 318); Daṇḍin’s 7th century Kāvyādarśa at 2.151; and a host of tantric texts. It seems the use of añjana to confer invisibility is a long-standing stock theme in Indian magic. (c) Moreover, which might possibly also connect with the Longshu pusa zhuan narrative, añjana could also confer the power of attracting the devotion of those who glance upon its wearer, be they a king or his harem (Goudriaan 1976: 318).

Whichever of their slightly varied classificatory terms one might prefer to use, it does seem to be precisely these three interrelated and overlapping aspects of classic and specifically Indian magic with which Nāgārjuna is associated. Even though the Longshu pusa zhuan narrative describes that Nāgārjuna had to learn the invisibility magic from a (possibly disreputable?) practitioner who had not been a part of his regular brahmanic education, it is abundantly clear from this story that Nāgārjuna was already so deeply familiar with the plants and other substances used in Indian magic, that he was able very swiftly to reverse-engineer the secretive magician’s añjana recipes, merely by the power of smell, we are told. Indeed, such a knowledge of plants and substances was the shared common foundation of most of the arts and sciences that became attributed to Nāgārjuna, such as medicine, alchemy, treasure retrieval (nidhivāda), and the associated entry into the subterranean worlds where nāgas dwell (bilasādhana, pātālasiddhi). The invisibility añjana narrative in its entirety rings true of actual Indian magical practice in sufficient ways that I find Walser’s proposal that it pertains instead to Chinese alchemy rather less likely. Moreover, Chieko Yamano believes that such an eye liner to achieve invisibility is quite untypical of Chinese magical beliefs, and if she is correct, this would make an Indian origin for the story all the more likely.

It is true that in the Longshu pusa zhuan narrative Nāgārjuna is freely invited to the nāgaloka by the benevolent nāga king, and does not have to engineer his own entry into that otherwise inaccessible place by means of some kind of bilavāda or bilasiddhi. This non-magical approach is reasonably typical of earlier Buddhist narratives, since outright magical rituals to achieve an uninvited entrance into the luxurious underworlds of nāgas and asuras don’t become commonplace in Buddhist literature until the early tantric period gets underway from the 4th century CE, 100 years after Nāgārjuna’s terminus ante quem, at which point they become ubiquitous, with substantial Buddhist tantric texts dedicated to this topic. Thus in the Pāli Mahāvaṃsa we find the monk Soṇuttara using Buddhist iddhi and Buddhist meditative concentration (rather than brahminic pātālasiddhi or bilavāda or bilasiddhi) to make his way into nāgaloka (Scheible 2016: 109).

Nevertheless, it’s important to take note that within the wider South Asian culture, visits to nāgaloka were not merely a conventional literary conceit, to be deployed for dramatic effect in popular narratives. On the contrary, what we can learn from the Indian magical literature and its many ritual handbooks is that the practice of magically visiting the nāga realms was surely seen as a real accomplishment, something which was actually achievable by suitably trained persons. Certainly such perceived actual direct interactions with nāgas (klu) remain very much a living part the Tibetan Treasure (gter ma) tradition, for, as I described above, in contemporary Tibetan Buddhism, treasure revealers (gter ston) are still widely believed to be receiving treasure caskets (sgrom bu) directly from the hands of nāgas. In other cases, they might obtain treasures from out of subterranean magical treasure portals (gter sgo) often inhabited by supernatural beings, which they alone have the power to access. Both of these kinds of Tibetan treasure discovery can be performed in full public view of an attendant crowd, which underlines that they are understood culturally as real human behaviours, not mere literary conceits; and from my readings of Indian magical texts, I definitely believe a similar cultural understanding must have prevailed in ancient India too. It is interesting that many of the methods for locating and gaining access to magic treasure portals in Tibetan Buddhism remain substantially similar to those described in ancient Indian pātālasiddhi or bilasiddhiinstructions preserved in Tibetan translation, such as those included within a lengthy Kriyātantra translated in the Imperial period, the Āryavidyottamamahātantra D746 (Mayer 2022).

I consider it highly improbable that any of the above magical practices did not yet exist in any form in Nāgārjuna’s time, even if different terminology might have been used in some cases. Several contemporary specialists in Indian magic have pointed me towards so-called vidyādhara traditions of magical praxis which they say are early enough to pre-date Nāgārjuna (I am not yet aware of a recent monograph or article specifically addressing this important topic although there are several older studies). We have already seen how such magical practices are already attested in Vedic sources, but they are equally described in early Jain and Buddhist texts. Pāli texts in general have many references to nāga kings, who inhabit magnificent subterranean palaces full of incomparable valuables, where they are visited for important pragmatic reasons by usually elite humans. We find this prominently in one of the older strata of Pāli texts, the jātakas. In the Mahāpaduma-jātaka a nāga conceals a human prince in his underground domain to protect him from his murderous father. In the Champeyya-jātaka, a nāga king named Champaka entertained in his underwater palace a politically beleaguered human king of Magadha, gave him great treasures, and made treaties with him that eventually restored the human king’s dominion. In the Bhūridatta-jātaka, a nāga gives a wish-fulfilling gem as a gift to a human brahmin who visits him in his underground palace. And so on: one can learn much by reading the entirety of Vogel 1926, Chapter III. 

To borrow the words of Andrea Acri, while it is true that “the 4th-6th century zeitgeist was all about magic—see the Buddhist Kriyātantras and such Śaiva texts as the VīnāśikhaGuhyasūtraGāruḍatantras etc.”, nevertheless, these magical preoccupations from the 4th century onwards came only about one hundred years after Nāgārjuna’s putative time, and the overwhelming interest in magic of this slightly later period did not arise ex nihilo without precedents. So while it might be the case (we are not yet sure) that such formal magical classificatory terms as pātālasiddhibilasiddhi, or bilasādhana perhaps only became prevalent with the efflorescence of magical literature approximately 100 years after Nāgārjuna, it appears extremely probable that actual practices for visiting and receiving boons from nāgas were already established quite a lot earlier. 

I also feel it might be significant that Nāgārjuna is said to have started out as a brahmin, because the Buddhist Vinaya placed restrictions on some aspects of worldly magic in a manner not necessarily imposed upon brahmins. As Chieko Yamano neatly puts it (personal communication, 2 October 2023): “Historically, among the Vedic priests (hotṛadhvaryuudgātṛ and brahman), brahmans who possessed knowledge of the [conspicuously magical] Atharvaveda were the most (economically) successful. They were often employed as royal purohita due to their expertise in mantras and rituals described in the Atharvaveda. Therefore, there is no contradiction in the legend of Nāgārjuna, who is often portrayed as the teacher of the king of south India, being knowledgeable in magical arts.”

For a while, I was tempted by the idea that Nāgārjuna might have been a Pāśupata or proto-Pāśupata, since Sanskrit narrative literature tends quite often (for example Bāṇa’s Kādambarī) to specify nidhivādinskhanyavādins, and rasavādins, as Pāśupatas (Balbir 1993: 22). Moreover, Diwakar Acharya (2013, 2018) has argued for the existence of Pāśupatas by another name well before the 4th century, taking them within the usual range of dates estimated for Nāgārjuna. Conversely, Andrea Acri has pointed out that the narrative descriptions of Pāśupatas might be mere caricature, and should not be relied upon too much. 

Yet we can nevertheless be confident that visiting the nāgas was an elite behaviour, and a technique that many brahmins would take pride in mastering. So even if we can’t know which exact category of brahmins the young Nāgārjuna might have been educated amongst, we can nevertheless be sure that involvement with nāgas and visits to nāgaloka were consonant with an elite education. It is after all an anthropological commonplace that rulers of pre-modern agricultural societies typically needed to show a closeness with and influence over natural powers such as rainfall and fertility, which governed essential food production. Those who were not seen to have good relations with the spirit powers that fed the people might not remain kings for long. Hence visiting nāgas in nāgaloka and receiving gifts from them was something that could be admired in kings, or the brahmins acting for them (Vasudeva 2012: 276). The Pāli narratives allude to this in so often showing nāgas to interact with kings and brahmins, and likewise we can point to the claims of several historical South Indian royal houses to lineal descent from nāga ancestors (DeCaroli 2004: 92), as is also claimed for Nāgārjuna himself in several Jain sources (Walser 2007: 74) . 


Now, let us recall that it is generally agreed by most scholars that the above-mentioned earliest Chinese sources, in both the clusters we have discussed, portray Nāgārjuna as a brahmin before converting to Buddhism, and describe him as visiting the nāgaloka to receive Dharma treasures. It follows, then, that if Nāgārjuna had in fact belonged to some brahmin grouping educated in various magical arts, most of which were underpinned by the thorough knowledge of plants and substances fundamental to much Indian magic and medicine, it might be a bit rash to try to claim with such absolute certainty that he could never have engaged in these practices, despite the unanimity of all Asian traditions to the contrary. Might we be unconscious victims of the same kind of over-rationalising mindset that once attempted to deny the mathematician and physicist Sir Isaac Newton’s deep and life-long immersion in the study of alchemy, the Philosopher’s Stone, the Temple of Solomon, and other occult topics? 

If a super smart brahmin suitably conversant with such contemporary topics as khanyavāda and nidhivādaañjana, and visiting nāgaloka, were to convert to Buddhism, can we really be so absolutely certain that he would never feel the urge to deploy whatever know-how he had on behalf of his new Buddhist faith? And given his new faith’s already well-established involvement in nāga cults, might he not feel confident to describe a visit to a nāga palace, and the gift from a long-lived nāga king who might have been amongst the Buddha’s original audience, of some wondrous teachings of the Buddha that were otherwise lost to humans? 

It is necessary to recall that nāgas were considered quite exceptionally important as protective deities in early Buddhism. More specifically, South India was a major centre for nāga cults, even more so than other regions of India, and archaeologists have recovered a prolific quantity of nāga representations in the region, not least in Buddhist sites. Can we really be absolutely certain that Nāgārjuna the Mahāyānist and Madhyamaka philosopher remained aloof from his cultural background? If so, why? 

Much of the literature on nāgas in Indian Buddhist monasticism has focused on their rain-making and monastery-protecting duties. However, in addition, early Buddhist texts are similarly full of narratives about nāgas serving the dharma by guarding in their subterranean palaces a variety of specifically Buddhist treasures. For example, nāgas played exactly this role in an important Pāli Buddhist text from Nāgārjuna’s contemporaneous and neighbouring Sri Lanka, the Dīpavaṃsa, composed between the 1st and 3rd centuries CE. Kristin Scheible, (2016: 94–116) has thus argued with considerable cogency that the prime function of nāgas in the Pāli Buddhism of Sri Lanka during Nāgārjuna’s time was actually guarding the various types of sacred relics of the Buddha in their nāgaloka. There were powerful connections between the Buddhism of South India and nearby Sri Lanka at that time. As Scheible explains, it was believed these relics had mainly been hidden in nāgaloka during the Buddha’s own time on earth, to be guarded by nāgas until the moment arrived when they would be needed by humans; at which point, a suitably qualified human would have to visit nāgalokato retrieve them. Scheible suggests this relic-guarding function was the main reason for the great prominence of nāgas in Buddhist texts of that period.

In Mahāyāna literature of that period, the role of nāgas is similar, except that otherwise ‘forgotten’ Mahāyāna sacred texts could also count amongst such relics. For example, the Pratyutpanna-buddha-saṃmukhāvasthita-samādhi-sūtra that was translated into Chinese in 179 CE, near or shortly before Nāgārjuna’s time, talks of the Buddha causing his Mahāyāna sermons, including that very sūtra itself, to be hidden with the nāgas, as a precaution that would enable their future recovery by spiritually authorised humans. This was necessary because the Buddha could foresee that many of his sermons or sūtras would otherwise become lost to the careless human race (Ch.13 v.9).

Given the perceived proclivity of the short-lived human sangha to lose their scriptural transmissions, and given that nāgas were believed to have usually been amongst the original audience of a sūtra when the Buddha first preached it, and given that nāgas were believed to have such great longevity that original auditors of the Buddha’s preaching were routinely still alive among them, and given the perceived frequent deep commitment of such nāgas to serve the dharma as its protectors, surely this strategy for scriptural safe-keeping must have looked like a natural and reasonable arrangement. By substituting Padmasambhava for the Buddha, a parallel narrative is still able to persist in rNying ma gter ma, no doubt inspired by these earlier Indian models.

In this context, it might be worth noting that much has been said in secondary literature about Nāgārjuna as a name used by later alchemists in India and his prominence in later Indian alchemical texts; likewise in matters of Indian medicine and longevity. But less has so far been said about his similar role in the related Indian arts of treasure discovery or nidhivāda, and his connection with nidhiśāstra texts such as the Nidhipradīpika (Balbir 1993:25), which, indeed, reproduces the two chapters on treasure discovery from the Kakṣapuṭa or Siddhanāgārjuna-tantra, attributed to the authorial name of Nāgārjuna (Chapters 15 and 16, see Yamano 2013: 66-67). 

We seem to encounter little difficulty in accepting that an actual historical Nāgārjuna probably served as the synecdoche or role model for future doctrinal and textual innovators, some of whom were also his namesake. Jean Filliozat (1979) has previously tried to assert that Nāgārjuna was the editor of one of the foundational texts of Indian medicine, the Suśruta Saṃhitā, although his proposal is not reproduced in such standard sources as the two Brill Encyclopedia of Buddhism articles upon which this blog has largely relied, and was critically reviewed. My own position is altogether different: I make no claim whatsoever to prove that Nāgārjuna did one thing or another. I simply wonder if a current academic certainty, that a historical Nāgārjuna did not have any actual involvement with nāga beliefs or magic, is sufficiently well founded. I don’t think we have enough evidence to support any such certainties.

This is merely a blog, an informal space where scholars can feel free to explore and suggest.


Acharya, Diwakar, 2013. “How to Behave like a Bull? New Insight into the Origin and Religious Practices of Pāśupatas.” In Indo-Iranian Journal 56 (2013) pp. 101–131. 

Acharya, Diwakar, 2018. “Pāśupatas”, in: Brill’s Encyclopedia of Hinduism Online, editors Knut A. Jacobsen, Helene Basu, Angelika Malinar, Vasudha Narayanan. Consulted online on 28 December 2021 < 5019_BEH_COM_9000000056>
First published online: 2018 

Balbir, Nalini, 1993. “À la recherche des trésors souterrains.” In Studies in honour of Gerrit Jan Meulenbeld: Journal of the European Āyurvedic Society vol. 3, 1993, pp. 15– 55. 

Corless, Roger, 1995. “The Chinese Life of Nāgārjuna”, in Buddhism in Practice, ed. Donald Lopez. Princeton, New Jersey: Princeton University Press.

DeCaroli, Robert, 2004. Haunting the Buddha. Indian Popular Religions and the Formation of Buddhism. New York: Oxford University Press.

Deeg, Max, 2021. “Indian Regional nāga Cults and Individual nāga Stories in Chinese buddhist Travelogues.” In Acta Asiatica Varsoviensa,  No. 34, 2021. Pp. 51-78.

Einoo, Shingo, 2009. “From kāmas to siddhis. Tendencies in the Development of Ritual towards Tantrism.” In Genesis and Development of Tantrism, ed. Shingo Einoo. Tokyo: Institute of Oriental Culture, University of Tokyo. Pages 17-40. 

Filliozat, Jean, 1979. Yogasataka: Texte Médical, attribué a Nagarjuna. Pondicherry: Institut Francais De Pondichery.

Goudriaan, Teun, 1976. Māyā Divine and Human: A study of magic and its religious foundation in Sanskrit texts, with particular attention to a fragment on Viṣṇu’s Māyā preserved in Bali. Delhi: Motilal Banarsidass.

Mayer, Robert, “Indian nidhi, Tibetan gter ma, Guru Chos dbang, and a Kriyātantra on Treasure Doors: Rethinking Treasure (part two)”, Revue d’Etudes Tibétaines, no. 64, Juiller 2022, pp. 368-446. 

Robinson, Richard H., 1977. Early Mādhyamika in India and China. Madison: University of Wisconsin Press. 

Scheible, Kristin, 2016. Reading the Mahāvaṃsa: The Literary Aims of a Theravāda Buddhist History. New York: Columbia University Press. 

Vasudeva, Somadeva, 2012. “Powers and Identities: Yoga Powers and the Tantric Śaiva Traditions“, in Yoga powers: extraordinary capacities attained through meditation and concentration, ed. Knut A. Jacobsen. Leiden and Boston: Brill. pp. 265-302. 

Vogel, Jean Ph., 1926. Indian Serpent Lore or the Nāgas in Hindu Legend and Art. London: Arthur Probsthain. 

Walser, Joseph, 2005. Nāgārjuna in Context. Mahāyāna Buddhism and Early Indian Culture, New York: Columbia University Press.

Wedemeyer, C.K., 2007. Āryadeva’s Lamp that Integrates the Practice (Caryāmelāpakapradīpa): The Gradual Path of Vajrayāna Buddhism According to the Esoteric Community Noble Tradition, New York: The American Institute of Buddhist Studies at Columbia University. 

Westerhoff, Jan Christoph, “Nāgārjuna”, The Stanford Encyclopedia of Philosophy (Summer 2022 Edition), Edward N. Zalta (ed.), URL = <>.

White, David Gordon, 1996. The Alchemical Body. Siddha traditions in Medieval India. Chicago: The University of Chicago Press.

Yamano, Chieko (山野千恵子), 2013, “The Yakṣiṇī-sādhana in the Kakṣapuṭa-tantra: Introduction, Critical Edition, and Translation”.  Journal of the International College for Postgraduate Buddhist Studies Vol. XVII, 2013 

Yamano, Chieko (山野千恵子), 2010, 『龍樹菩薩伝』の成立問題,(Two Biographies of Nāgārjuna), in Sengokuyama Journal of Buddhist Studies, Vol. V, 2010. 

Yamano, Chieko (山野千恵子), 2009b, “Śrīparvata as Nāgārjuna’s Abode: At the Confluence of Legend, History, and Geography,” IBK 57/3, 1246−1252. 

Yamano, Chieko (山野千恵子), 2009a, “Early Biographies of Nāgārjuna”, in Journal of Rengeji Institute of Buddhist Studies, 2, 2009.

Yamano, Chieko (山野千恵子), 2008b. Siddha Nāgārjuna: The Legend of Buddhist Saints in the Tantric World (シッダ·ナーガールジュナ ―タントラ世界における聖者伝説の展開―) Dissertation presented to the faculty of International College for Postgraduate Buddhist Studies in candidacy for the degree of Doctor of Philosophy in 2008).

Yamano, Chieko (山野千恵子), 2008a. “Nāgārjuna and Sātavāhana,” IBK 56/3, 1157−1163. 

Yamano Chieko (山野千恵子), 2007. “Nāgārjuna to Sātavāhana” (ナーガールジュナとサータヴァーハ ナ), CG 56, 401−431. 

Yamano Chieko (山野千恵子), 2004. “Nāgārjuna to Rasāyana” (ナーガールジュナとラサーヤナ), CG 53, 63−82. 

Ye Shaoyong, 2019. “Nāgārjuna”, in Brill’s Encyclopedia of Buddhism, Volume II: Lives. pp. 335-347.

Young, Stuart H, 2019. “Nāgārjuna in China”, in Brill’s Encyclopedia of Buddhism, Volume II: Lives. pp. 728-734.

[1] In due course, parts of this blog will be refined and expanded into a chapter of a forthcoming book about gter ma  revelation. Other parts of the blog are contextual research towards that chapter. In the meantime, any constructive comments will be greatly welcomed. Chieko Yamano has so patiently and so generously contributed so much to my understanding that this blog could quite certainly not have existed without her. Stuart Young has likewise has offered me an immense amount of extremely valuable advice. Many thanks also to Aleksandra Wenta, Andrea Acri, Jan Westerhoff, and Somdev Vasudeva, for their priceless offerings of advice and references.

[2] We now have a DPhil student at Oxford, Ryan Jacobson, working on the same materials. 

[3] 出三蔵記集 T 55, no. 2145, p. 74.

[4] 出三蔵記集, T 55, no. 2145, p. 75.

[5] 出三蔵記集, T 55, no. 2145, p. 77.

[6] T no.1775『注維摩詰経』巻第二 38:339a.

This entry was posted in Mainly monthly postings. Bookmark the permalink.

Leave a Reply

Your email address will not be published. Required fields are marked *