As important as it is to examine the categories and taxonomies of bhūtavidyā, it is equally important to examine the healing process. This involves investigating the broader dynamics of exorcism: ritual, pharmacopia, healers, and the interaction between healer and client. Some of this has been addressed above, in anthropological and other ethnographic discussions of the healing process.212 We shall here look more closely at selected passages from the vast āyurvedic and tantric material, mostly the former.
It is impossible to address the exorcistic material in all the relevant texts. With the limitation of employing selected passages it is best to approach them chronologically, because this is the best way to convey an idea of how the practices and, indeed, the vidyā of bhūtavidyā evolved. This also helps illuminate key aspects of the morphology of the cultural, linguistic, and moral underpinnings of mental illness, especially of exogenously produced unmāda. A much longer study could be undertaken on this if more material is used. But the following should provide a few guidelines for such a study.
In discussion of these treatments, I must first summarize some of the exorcistic procedures found in Caraka, Suśruta, and Vāgbhaṭa, including the diagnostics, pharmacopeia and other “rational” modalities (yuktivyapāśrayacikitsā),213 as well as ritual and moral treatments (daivavyapāsrayacikitsā). Then I discuss in some detail the procedures found in the Cakradatta, an important āyurvedic text of the eleventh century that brings together āyurvedic and tantric therapies. It would be a mistake to assert that exorcisms are of a single style of practice or generated by indigenous scholars equally informed about the intricacies of medicine and ritual, especially after the turn of the first millennium C.E. Like all ritual in India, this was diversified even within specific performative traditions.214 The ethnography that concludes this chapter is drawn from fieldwork I conducted in Kerala and demonstrates how medical, tantric, and astrological traditions operate in tandem, in a seamlessly functioning and fairly ancient system of referral, how they strengthen rather than isolate one another.
Treatment in the Early Āyurvedic Compendia
The earliest compendia, Caraka and Susruta, of the early centuries of the Common Era, as well as those of the next chronological level four or five centuries later, Kāśyapa and Vāgbhaṭa’s AH and ASaṃ, do not mirror each other as closely as might be expected in their treatment of unmāda. Considerable differences may be found among the texts, including differences in opinion that go beyond the enumeration and designation of grahas. The most glaring instance of this is Caraka’s overall treatment of unmāda, which is less nuanced and more straightforward than Susruta’s. One aspect of this is Caraka’s more substantial treatment of the psychopharmacology.
Susruta defines the terms graha and bhūta more specifically than Caraka, but he adds details that were probably borrowings from other genres (e.g., the Sanskrit epics), namely, that grahas are powerful beings who achieved their present form through severe austerities, religious donations, performative and moral vows, and observation of appropriate socioreligious duties. These enabled them to achieve the “eight powers” according to their respective strength.215 However, adds Susruta, they do not themselves possess humans; it is their minions who do their bidding, who do the dirty work of possession, and, in a statement presumably directed at Caraka and others, Susruta states that “those who say that they possess are to be removed from the orbit of bhūtavidyā due to their ignorance.”216 This agrees with the general tenor of what may be found in the Mahābhārata, Rāmayana, and other major texts of the time, which mention myriad demons under the control of a few, an ideology of bhūtas that we still find today at Bālājī in Rajasthan, Chottanikkara in Kerala, and generally throughout India (as well as in many areas of Christendom, where the “evil one” is said to have innumerable agents).
After distinguishing the different bhūtas, Caraka (CaCi 6.9.23) declares that, if unmāda is caused by possession by spirits that demand sex or worship, the physician (bhiṣak) should determine the nature of the possession through the intent and behavior of the client and then prescribe suitable medicinal treatment and mantras, as well as sacrificial offerings and gifts of food.217 Caraka subsequently recommends a variety of therapies common to both exogenous (āgantukā) and endogenous (nija) unmāda—intended to cleanse the victim internally and restore his or her balance of doṣas. The therapies include laxatives, dietary modification, emetics, medicated oil applied to the head (sirovirecana), full-body oil massage, inhalation (of smoke from a dhūmavarti, a medicated cigar) and fumigation therapies, bloodletting, and, especially for exogenous unmāda, consumption of ghee. These therapies “enliven the mind, intellect, memory, and consciousness” (manobuddhismṛtisaṃjñāprabodhanam; CaCi 6.9.32cd).
Incurable insanity is also recognized, in which case treatment is not recommended. However, extreme insanity that is deemed curable may be treated with extreme therapies, including forms of shock therapy including beatings (tāḍanam), which might be useful in “stimulating the mind, intellect, and body” (manobuddhidehasaṃvejanam; CaCi 6.9.29cd). A person might also be tied up tightly, though gently, with cloth (rather than rope), and confined to a dark room from which dangerous objects like iron rods and wooden sticks have been removed. Other kinds of shock therapies include shouting angrily at the patient, terrorizing him with the help of the local police who threaten to arrest or even execute him, having snakes that have had their fangs removed bite him, and threatening the patient with tamed (if still frightening) lions and elephants.218 At the same time, the victim should be given presents, reassurance, and consolation by friends, and exposure to emotions opposite from the one on which the patient is obsessing—passion, grief, fear, anger, exhilaration, jealousy, and greed are mentioned—that might help turn the mind in a different direction. The principles of these largely exotic treatments appear to be, first, to rebuild the mind and body with medicines that address both the humoral imbalance and the personality affliction; second, to bring the patient out of his or her condition by exposure to positive emotions; third, to shock the patient out of his or her obsessions; and, fourth, to bring a constant undercurrent of support and consolation to the patient. All of these should “lead to a state of mental normalcy by countering memory loss.”219
Thus, a full arsenal of treatments is brought to bear on possession. CaSū 11.54 defines the three areas of treatment mentioned at the beginning of this chapter: daivavyapāśraya, yuktivyapāśraya, and sattvāvajaya: “Daivavyapāśraya indicates the use of mantras, herbs, gemstones, auspicious offerings, gifts of food, fire offerings, scriptural observance, expiatory rites, fasting, chanting of hymns, surrender to deities, and pilgrimage. Yuktivyapāśraya indicates proper food, medicine and [other] substances. Sattvāvajaya is the holding back of the mind from harmful objects.”220 The last of these, sattvāvajaya, is not mentioned beyond this, possibly because it was regarded as implicit in proper observation of daivavyapāśraya.221 Thus, the texts describe in detail only the first two, at least by name. Much later in his text, as a conclusion to his section on treatment for unmāda, Caraka expands on his description of daivavyapāśrayacikitsā. As a partial remedy for unmāda, he enjoins,
One should perform pujā, offer sacrifice [bali], give offerings of food, ritually apply collyrium sanctified by mantras, and perform pacification rituals, Vedic sacrifice, repetition of divine name, rituals for auspicious occasions, and expiatory ritual. Worshipping Īśvara [viz. Śiva], the lord of the universe, the leader of bhūtas, constantly and with effort, one becomes victorious over the fear born of unmāda. One should perform puja to Rudra’s attendants [gaṇaḥ] called pramathas, who roam about in the world, thereby releasing them from various forms of unmāda. Exogenous possession [unmāda] becomes pacified [prasamam] through sacrifice, auspicious actions, fire ritual, wearing amulets that are healing because they contain appropriate herbs, observing truthfulness and good conduct, performing penance [tapas], developing knowledge, becoming charitable, observance of scriptural limits and vows, through honoring deities, cows, brahmans, and gurus, and through the application of perfected mantras and medicines. (CaCi 6.9.89–94)
Caraka’s exposition of the “rational treatment” (yuktivyapāsrayacikitsā) of unmāda consists largely of descriptions of medicinal preparations used to treat multiple conditions, most of them vāta debilitations.222 Because, according to Āyurveda, vāta evolves into prāṇa, it is in many ways a more fundamental doṣa than pitta or kapha. This is why, according to many practitioners, treatment for vāta derangement is more capable of abetting a cure for pitta or kapha conditions than treatment for one of the latter two would be for healing vāta debilitation.
This may also explain why ghee-based herbal preparations are prescribed as a remedy for unmāda in general, for one whose “mind is struck by a bhūta” (bhutopahatacitta), for bad luck (alakṣmī), for transgression (papa), and for demonic seizure (rakṣoghna). These preparations are also said to destroy all grahas (sarvagrahavināśanam; CaCi 6.9.41d). Ghee is believed to act directly to alleviate vāta and pitta debilitation, while taken in moderation it does not exacerbate kapha debilitation. Because of the esoteric nature of the descriptions of ghee-based preparations, and the extensive lists of ingredients used in their preparation, I can here offer only a brief description. To begin with, most of the preparations consist of a large number of herbs mixed with ghee. However, a few special, relatively simple, recipes require “old ghee” (purāṇaṃ ghṛtam). Caraka states that ten-year-old ghee used as a base is capable of rectifying all the doṣas and eliminating grahas. Ghee older than ten years is called prapurāṇa (very old), and ghee which is a hundred years old can cure any disease: “just the sight, touch, or smell of this casts out all grahas.”223
This practice is still current. In the eighteenth century, when the British finally defeated the Peshwas in Pune, they are said to have found a large number of pots of prapurāṇa ghee in Shanwar Wada, the fortress in the center of the old city. With respect to its manufacture, I am informed that “you must make sure that it is totally cooked, meaning that all the water is evaporated, and then it must be kept very dry. Then it begins to age.”224 Other informants in India have told me that the production of purāṇa-ghṛtam, which includes the manufacture and vigilant protection of ghee, occurs elsewhere as well. It is stored in large clay pots lined on the outside with copper or bell metal (kāṃsya), and was once a relatively common āyurvedic practice.
In an ensuing discussion of treatments for apasmara, also said to be effective against demonic infestation and other kinds of unmāda, Caraka prescribes a large number of animal products to be mixed into the preparations. Most common are recipes containing pañcagavya, the “five products of the cow”: milk, sour yogurt, ghee, urine, and liquid squeezed from dung.225 Goat’s or cow’s urine is also frequently mentioned as an ingredient independent of pañcagavya. One recipe calls for a number of herbs to be mixed with cow’s or goat’s urine along with leech excrement, ash from incinerated goat’s hair, donkey bone, elephant’s toenail, and hair from a cow’s tail (CaCi 6.10.39–40). From a modern perspective such recipes appear to be exotic, to say the least. However, Caraka and others probably did not include them for shock value, except possibly to shock an offending graha; a patient suffering from convulsions or any kind of mental disorder quite likely was not able to identify the ingredients in these preparations. More likely, animal products were deemed more antithetical to the possessing spirits, thus indirectly influencing, and even rectifying, humoral imbalances.226
Suśruta offers greater detail on antidotes to specific grahas (SuUt 6.60.29–37), as well as a number of more generally applicable remedies. The latter are not as profuse as are found in Caraka, but the attention paid to individual grahas indicates that Susruta emerges from a different, if related, medical tradition than Caraka, an oft-noted fact. Susruta states that grahas are nourished by the color red; thus, they should be given gifts of red clothes and offerings of red flowers, meat, and blood. The eight grahas mentioned by Susruta should be pacified with these and other appropriate items in rituals held in a temple (for a devagraha), at a crossroad (for an asura), amid cows (for a gandharva), in a pleasant room (for a yakṣa), by a river (for a pitṛgraha or a nāga), at a crossroad or in a dense forest (for a rakṣas), and in an abandoned house (for a piśāca). If these offerings do not succeed in alleviating the possession, the next recourse is fumigation therapy. One of these fumigants consists of goat, bear, porcupine, and owl skin mixed with asafoetida and goat’s urine. Another consists of oil (taila, unspecified, but sesame is usually the default oil) mixed with feces of a donkey, mule, owl, camel, dog, vulture, crow, or pig. Most of the recipes, however, consist of a mixture of herbs and animal products, including bile from different animals, including fish and cow, and the nails of various feral beasts, including a variety of large and small felines, alligator, mongoose, and so on. Like Caraka, Susruta may have selected these creatures in part because of their perceived dominant doṣa, but probably largely because they were feral and could thus overpower the afflicting graha.
Susruta adds a chapter (6.62) on the diagnostics of unmāda according to humoral imbalance. It is not necessary to enter into the details here, but it suffices to note that the diagnoses comply with the characteristics of the grahas and that the details of the shock therapy are in essential agreement with those given by Caraka (a few details are added, such as whipping or tying the victim with chains), and a bit more dietary advice is given.
Kāsyapa Kalpasthānam, chapter 1, lists a large number of recipes to be used as demonifuges, for example, “A fumigant [dhūpa] named Uttama [Best] is recommended for people overcome by pretas and pūtanās [indicating all balagrahas]. It should be prepared from ghee, siddhārthaka [?], honey, sheep’s horn, goat’s milk, donkey’s hair and urine, and soma” (KSKa 1.11–12).227
The AH Uttarasthāna (AH-Ut) 6.3 discusses balagrahas and recommends an array of possible treatments for such invasion, including various fumigants, medicated ghee, decoctions for nocturnal baths (snapana), and ritual. Much of this is found in Caraka and Susruta, but the corpus is here expanded. AH-Ut 6.5, titled bhūtapratiṣedha (repulsing bhūtas), again covers a gamut of remedies, including recipes for pills (guḍa) to ward off insanity (6.5.15–17), again with an exotic mixture of vegetal and animal products, special offerings for most of the species of offending grahas listed in 6.4 (translated above), mantras invoking Dvādasabhuja (Twelve-Armed) Īśvara or Āryāvalokita Nātha, the mahāmāyūrī mantra (cf. discussion of the Bower manuscript above), and worship of various other multiforms of Śiva, including Bhūtesa, Sthānu, and the latter’s attendants, called pramathas (agitators, shakers).228
The Cakradatta: Pharmacopeia
As an example of the discourse of healing in the āyurvedic text passages on possession, I draw on the Cikitsāsaṃgraha, otherwise called Cakradatta, a therapeutic text of the mid-eleventh century, thus placing it chronologically in the third tier of āyurvedic textuality.229 The author, Cakrapāṇidatta, is best known for his commentary on Caraka. But he also wrote this independent treatise, which became widely followed in his native Bengal. Although much of the material is drawn from Caraka, it updates and situates it in a healing locus more definitively systematic, as Āyurveda had by then become a sastra with a long and distinguished tradition. This compendium of āyurvedic therapeutics is conspicuous for its rich pharmacopeia. Although it does not come out particularly strongly in the passage at hand, the Cakradatta surpassed its predecessors in integrating tantric medicine and practices, which include mercurial and other mineral-based preparations and a tantric repertoire of mantra and yantra, into the formulary and ritual recommendations of Caraka. This text probably came as close as any to representing a complete form of Āyurveda, in north India at least, at the time of its composition. Chapter 20 of the Cakradatta addresses unmāda, distinguishing psychological dysfunction with identifiable causes (nija) from exogenously induced (āgantuka) unmāda. In discussing the latter, Cakrapāṇidatta describes the pharmacopeia and mentions certain aspects of ritual treatment.
44. The practice of consuming ghee, etc., as well as the use of mantra, etc., are introduced for [treatment of] the exogenous variety [of unmāda]. The physician [bhiṣak], having undergone his own purification, should treat exogenous unmāda with ceremonies of worship [puja], normal sacrificial oblations [bali], nonsacrificial offerings [upahāra], vedic fire ritual [iṣṭi], non-vedic fire ritual [homa], mantra repetition, ritual application on the head or body of colored or scented salve or paste [añjana], etc., each according to its respective ritual injunctions.
45. A salve [añjana] prepared from long pepper [pippalī], black pepper [marica], black sulphurous salt [saindhava], honey, and bovine bile [gopitta] destroys extreme insanity caused by all kinds of spirits.
46. A pill [guḍika] made from dāruharidrā and honey in the lunar asterism called puṣya is [powdered and] made into a salve.230
47. Alternatively, marica mixed with bovine bile that has been placed in the sun for one month may be beneficial as a salve. This is used for distorted vision [hallucinations] when memory is lost due to imbalance of humors or spirit possession.
48. A fumigant [dhūpa] that wards off enchantment [ḍākinya], etc., made of neem leaves, vacā,231 asafoetida [hiṇgu], sloughed-off snake skin [sarpanirmoka], and mustard seed, may be used to destroy madness caused by spirit possession.
49. According to tradition, a fumigant [dhūpa] made of a ball packed together with cotton seeds, coppery eye of peacock feathers [mayūrapiccha], bṛhatī, and nirmālya, [added to a bundle consisting of] cinnamon [tvak], the soft extrusions at the joints of bamboo segments [vaṃśalocana], cat excrement [vṛṣadaṃsa-viṭ],232 grain husks, vacā, jatāmāṃsī,233 and sloughed-off snake skin [ahinirmoka], and equal quantities of [powdered] cow’s horn and dvipadanta, asafoetida [hiṇgu], and black pepper [marica], alleviates fever that arises due to possession [āvesa] caused by the skanda entities [skandonmāda], piśācas, rākṣasas, and suras.
1. A nasal preparation [nasyam] of roots and ripened fruit of aindrī [indravāruṇī] and cow urine234 allows victory over brahmarākṣasas. A nasal preparation of svetā and ricewater [jyeṣṭhāmbu] mixed with ghee wards off bhūtas.
2. In the case of madness caused by deities [deva], sages [ṛṣi], deceased ancestors [pitṛ], and gandharvas, an intelligent person should avoid irritant salves, etc., and other harsh measures.
3. The symptomatology of the departure of madness is clarity of the senses, intellect, self, and mind, as well as normalization of the dhātus.
We have seen many of the exotic substances recommended here in earlier texts. Among them, cat excrement (which may have been dried and powdered), sloughed-off snake skin (ahinirmoka), and the coppery eye of the peacock feather (mayūrapiccha) were probably classified, at least informally, as vāta (wind-reducing substances) and thus antispasmodics that at the same time rebuild afflicted systems.235 In present-day Kerala the vaidyas praise classical recipes like this one. They employ the herbal ingredients recommended, such as vacā, black pepper, asafoetida, neem leaves, and cinnamon, but use other substances in place of the animal products, including garlic, white śaṇkhapuṣpi, and chili.
Although unmāda could result from a specified imbalance of any of the three doṣas, most of the evidence encountered in both the āyurvedic and tantric texts suggests that externally induced (āgantuka) negative possession is vāta-based more often than it is pitta- or kapha-based. Among the symptoms of spirit possession described in these texts that are more characteristic of vāta debilitation are excessive facial gestures, uncontrolled body movements, stammering, cold limbs, and attention-deficit disorders. These symptoms, we should recall, align closely with Caraka’s general description of unmāda cited earlier, which, more than the other doṣas, is a description of vāta debilitation. Nevertheless, afflictions more often associated with other doṣas are also found in descriptions of the various grahas, as we have seen. Primary among these is extreme anger, more frequently associated with debilitated pitta. In all cases, however, it appears that antispasmodics, prescribed primarily to mitigate vāta debilitation, are recommended for alleviation of spirit possession.
Much of the following is based on research conducted in Kerala in 2002 and 2004. I undertook this research after hearing a rumor from a usually reliable source that a number of āyurvedic physicians and tāntrikas (or mantravādins, as they are called in Kerala) still follow the practices of possession and exorcism outlined in the ĪŚP. On close investigation, I learned that this is not true. Nevertheless, it is likely that some of the current practices are derivative of what is found there or at least from a complex heritage of texts and practices that gave rise to the ĪŚP.
My primary conclusions, after speaking extensively to more than a dozen highly placed vaidyas, mantravādins, and jyotiṣis (astrologers), are that: (1) notions of bhūtas and bhūtavidyā are substantially unchanged from what appear in the classical texts;236 (2) most of the clients are women, including a large number of Muslims; (3) an entrenched system of referral among the three groups of practitioners, which constitutes a mental health-care delivery system of considerable sophistication, is still very much in place;237 (4) the groups freely borrow both textuality and ritual apparatus from one another. A full ethnography of the possession events that I witnessed in Kerala is beyond the scope of the present work.238 Instead, I present a summary account of the points mentioned above, particularly the latter two, the system of referral and the shared textuality. These are most important for our purposes because they are crucial in assembling an epistemology that encourages the development of a complex and integrated healing tradition among the vaidyas, mantravādins, and astrologers of central and northern Kerala, particularly in districts of central Malabar, north of Cochin.
Two lineages of āyurvedic practice are currently found in these and the districts to the south, from Kollam (Quilon) in south-central Kerala to Kozhikode (Calicut) and beyond in the northern districts of the state and from the coast to Palakkad (Palghat) in the hills of eastern Kerala and beyond to Coimbatore, a city of over a million people in Tamilnadu, barely fifteen kilometers from the Kerala border. The two lineages are the aṣṭavaidya tradition, transmitted for centuries by Nambudiri brahmans, and the non-brahman tradition established by Vaidyaratnam P. S. Varier in 1902 in the town of Kottakkal in Mallapuram District. The latter has since spread throughout Kerala and is now creating a discernible presence on the āyurvedic map as far north as Delhi.239 Both lineages maintain the AH as their primary text.
My most valuable informants were Vaidyamadham Narayanan Nambudiri of Mezhathur and Kumaraswami Nambudiripad of Kāttumataṇ Mannā, both in Palakkad District. The latter is well known locally as an expert in mantravidyā to whom Vaidyamadham occasionally refers patients diagnosed with exogenous unmāda. Vaidyamadham, who learned Āyurveda from his father in the aṣṭavaidya tradition, is one of Kerala’s most-respected āyurvedic practitioners. His practices and opinions carry considerable weight.240 I spoke with him at length and observed some of his practices. As is the case with the other vaidyas in the predominantly Muslim districts of Malabar, most of Vaidyamadham’s patients are Muslims, including those who come to him for treatment of mental dysfunction or illness.
After observing a fair amount of practiced, but nevertheless unpredictable and captivating (to both ethnographer and participant), possession ritual at the Chottanikkara “Bhagavathy” temple, a well known goddess temple about thirty kilometers east of Cochin, where the use of puttalīs is common, it became clear to me that most of the victims were girls and young women from about fifteen to about twenty-five years old. This is in keeping with ethnographic reports throughout India. The women exhibited the classic symptoms of possession on having darshan of the goddess, exhibiting rhythmic rotating of the head with hair unbraided, wild body movements, shouting and thrashing about, and so on. Although this temple is noted for both possession and exorcism, what I observed was not entirely “negative” possession, even if there was undoubtedly a negative onus cast on the girls by their families. I say this because their apparent dissociation is in many cases accompanied by palpable states of exhilaration, rendering it more accurately a highly focused, if at a certain point uncontrollable, liminal state, albeit one without an apparent oracular component. Were it mere dissociation, as most scholars understand it, this complexity would be negated, particularly as the girls in question appear to maintain some awareness of their complex state of consciousness and the transformations (albeit temporary) wrought on their persons as a result of their possession. The temptation to express themselves through possession states appears to contain an element of choice, at least some of the time; even if many of the girls are “clinically” ill, the temptation to “act possessed,” to unleash this “spirit” within themselves to the point at which their self-imposed liminality overwhelms their socialized personalities, includes critical elements of excitement and fascination.
Similarly, their psychophysical attitude or comportment (bhava), regardless of whether a product of illness, is clearly liberating, at least in the short run. In either case, whether the possession fits into indigenous categories of ecstasy or derangement (or both), the experience is valorized, sanctified, and healed or exorcised by the goddess. In cases of negative possession, the evident attraction of this kind of ecstatic healing environment is that the uncontrollable meandering and eclipse of the socialized personality are given an opportunity to express themselves. This is exemplified in girls who appear to engage in both sides of a dialogue, in different voices and intonations, in which the goddess and the possessing spirits argue or even fight. If the possessed person is deemed by the healer, family members, or ritual officiant to be ready to expel the spirit, and if adequate administrative and ritual preparations are made, including payments and the acquisition of the proper material items (including a puttalī), a brief ceremony is then performed. Several regular attendees who had seen this ceremony (I did not witness it) described it to me. In it, I was told, the officiant drives the bhūtam, sometimes in the form of the puttalī, into the tree with a nail or railway spike in the presence of the afflicted individual and accompanying family members, after which the victim beats her (or, less frequently, his) fists or head on it. As expected, this act is less practical than symbolic and “ritual”; the human head cannot beat a nail or railway spike into a tree without causing irreparable damage. Thus, we must assume, the nail has been placed in a hole already drilled or otherwise hammered into the tree. The purpose of beating the nail with the head or fists, then, is to provide conviction and a sense of personal immediacy to the exorcism and, perhaps, to draw blood as an attractive offering to the goddess.
It is important to note that the girls (and a few men) at Chottanikkara are Hindus (and possibly a few Christians), and their conditions did not appear to be irreversable. By contrast, most of the victims of unmāda who attend the clinics of the local vaidyas are either Muslims who would not travel to Chottanikkara or members of other communities with serious psychological dysfunction. Vaidyas of both the aṣṭavaidya and Varier traditions, when repeatedly asked why most of their patients suffering from unmāda diagnosed as āgantuka were young women, gave one response with regularity: Young girls did not modernize as seamlessly as young men, and this opened them up to possession. This was not a satisfying answer, and it gave the impression that they were not telling the whole story. Finally, Vaidyamadham Nambudiri told me that these girls frequently are depressed, abused, and isolated, and suffer from drug addiction. Much of the abuse is sexual, he said, and occurs within the family; they have no outlet for expression in their families, and thus they are psychically weakened and become susceptible to possession, a situation greatly exacerbated by drug addiction. Regardless of the identity of the intoxicant, which from its description could be an opiate, he treats the victims with his arsenal of āyurvedic medicines, or refers them to Kumaraswami Nambudiripad for ritual treatment.
Kumaraswami was also very forthcoming with me, and I asked him what he does with these girls. On the whole, he counsels them and uses his position to establish a comforting and positive rapport with them. Very rarely does he perform tantric rituals for these girls. More than ritual, he says, they need someone to talk to, an admission that he has little ability or authority to deal with their actual problems. But Vaidyamadham, a physician, takes a more proactive āyurvedic approach. He prescribes fumigation (dhūma) with “foul-smelling substances,” three times a day for up to fifteen minutes, usually for several consecutive days. Although the vaidyas of Kerala continue to praise classical recipes for fumigants, such as the ones we saw in the Cakradatta, they employ only some of the herbal ingredients recommended, including vacā, black pepper, asafoetida, neem leaves, and cinnamon. They use other substances in place of the animal products, including garlic, white śaṇkhapuṣpi,241 and hot chili. Fumigation is administered on the premises of his clinic by his trained staff. He also prescribes emesis or induced vomiting (vamana) in which the patient expels the offending bhūta after drinking a concoction of goat’s milk, white śaṇkhapuṣpī, and a few other unnamed herbal substances (this concoction is used by several vaidyas in Kerala), oil massage (a specialty of Kerala āyurveda), and, if necessary, mantravidyā, in which case he refers the patient to Kumaraswami Nambudiripad.
If Mr. Nambudiripad or certain other members of his family who are privy to the meticulously protected family secrets deems mantravidyā to be advisable, they will counsel the victims and prescribe for them certain kinds of medicated ghee that help eliminate the bhūta. In the āyurvedic idiom employed by Vaidyamadham and Kumaraswami, drinking ghee (snehapānam) and inducing vomiting (vamana) are effective therapies for unmāda because they “boost buddhi [intelligence].” They manufacture their ghee-based medicines themselves and subsequently potentiate them by offering them to their deities, Bhagavatī (Durgā) and Gaṇapati, with mantras known only to them. Only then are they dispensed to their clients. In fact, irrespective of whatever ritualizing they might enact on the patient, their main form of mantravidyā is empowering these medicines with mantras.
I asked both Narayanan Nambudiri and Kumaraswami Nambudiripad whether their mantravidyā and practices of exorcism derive from the ĪŚP or other similar tantric compendia that were composed in Sanskrit long ago in Kerala. They had heard of this and other similar texts, including the (Viṣa-)Nārāyaṇīya, and knew, vaguely, that they contained mantras and instructions on possession and exorcism. However, they had not read any of them. Because of the secrecy of their traditions, I have no way of knowing whether their mantras resemble those found in these texts. I can only conjecture that the tradition of employing mantravidyā in exorcism remains unbroken in Kerala and that this tradition has always allowed considerable flexibility and variation, but with only a small number of survivals into the past few generations.
Kumaraswami Nambudiripad, as well as members of an unrelated family of Nambudiris from Poomkutil Mannā near Manjeri, in Mallapuram District, about fifty difficult kilometers away, state that they learned their mantravidyā and associated healing practices from their forefathers through direct observation and instruction, not from books or Sanskrit manuscripts.242 Kumaraswami, as suggested, is not the recipient of any immediate Sanskrit tradition. He regards himself as a mantravādin, not a tāntrika. In Kerala, he explained, tāntrikas perform temple ritual with the help of Tantras and Āgamas. He hails from one of four Nambudiri families of mantravādins ritually affiliated with the Mookambika goddess temple of Kannur, in northern Kerala. Not all four families, he says, actively practice mantravāda. Mantravādins specialize in treatments that require personal attention outside the temple. These treatments need not be based on known sastra. On Sundays, he says, more than a hundred people come to his Mannā for treatment of bhut-bādhā. The mannā itself is picturesque, consisting of a large family house at the bottom of a long driveway sloping down from the main road. Next to it is a well-maintained temple. These are surrounded by several acres of lush and well-tended fruit trees, with one prominent pipal tree in the front yard, which, says Kumaraswami, houses a chāttan.
This is a spirit indigenous to Kerala. All my informants agree that this is a class (gaṇa) of bhūtas, specifically a janglī bhūta (primitive, undomesticated jungle spirit).243 Among the symptoms of harassment by a chāttan are rocks falling inexplicably on one’s roof or large amounts of hair found in one’s food (specifically in ḍāl). A chāttan is also explained as “a force, something like a yakṣa.” It is not always malevolent and could even be protective if regularly and properly appeased, which, it is said, is accomplished by leaving votive offerings for it at a shrine constructed beneath certain large trees. In addition, certain people are called chāttans. They are most often depicted as members of backward classes who are short, dark, and demonic in a modest, nonfatal manner. They practice chāttan upāsana, meaning that they are possessed by a chāttan and practice chāttan jādu (black magic associated with a chāttan). However, even Nambudiris are capable of performing chāttan upāsana. In Thrissur District live several chāttan maṭhas or extended houses of people, mostly related, who are trained through family or lineage tradition in ritual appeasement of chāttans by making offerings at their shrines. Although leaving votive offerings beneath trees appears to be a simple matter, it is not so simple that any untrained person can do it. Another person informed me that his father once called in a Nambudiri from Kāttumatan Mannā to perform chāttan sevā, as the Nambudiris call it. His father believed that he was being bothered by a chāttan, as his businesses were going bad and illnesses were occurring in the family.
Kumaraswami’s clients report that they are afflicted by enemies, ancestral spirits (pitṛs), and chāttans. When they come to him, he recites simple mantras, most of them inaudibly, talks to the client, ties a black string around the wrist—a protective device that lasts several months before it weathers away, and, as mentioned above, gives medicated and ritually empowered medicinal ghee. Most referrals are through astrologers (jyotiṣī), but some come from vaidyas, such as Vaidyamadham. Kumaraswami says he knows only pratihāra (exorcism), mantra, and other procedures that “strike against” invasive beings; he does not claim to know medicine or astrology.
Vaidyamadham Nambudiri and Kumaraswami Nambudiripad also accept clients referred from astrologers. I therefore contacted a few highly recommended practitioners of this ubiquitous South Asian art form in Coimbatore and Kozhikode. Astrology and priesthood have had a long association in India,244 as have astrology and medicine, especially in South India. One of the convergences is textual. Perhaps the most important text of prasna or horary astrology (see Chapter 11), at least in Kerala, is the Praśnamārga (PrMā). The twelfth chapter of this complex text discusses the influence of the planets on health. Most of the chapter (verses 18–62) is dedicated to a discussion of unmāda and apasmāra. Verses 18–20 divide disease (roga) into the two familiar divisions of nija and āgantuka. Nija diseases may be either physical (sarirottha) or mental (cittottha), while those classified as āgantuka are unnatural, though they may be born of either visible or invisible causes (dṛṣṭādṛṣṭanimittajāḥ). The PrMā then describes the astrological configurations in a person’s natal horoscope (kuṇḍali) that indicate mental dysfunction. Space limitations prevent the discussion of the astrological significations here; I can only note that if a critical mass of discord exists between the planets, houses, and signs that govern one’s mind, emotions, general state of well-being, social capacities, and tendencies toward conflict, then mental or emotional instability of the visible (dṛṣṭa) type results. If these conditions are not met, and there is still instability, then the unmāda must be judged to be the work of obstructive spirits (bādhakagrahasambhavāḥ; 12.24d). The PrMā and its commentaries link planetary discord with emotional instability, untoward life-style choices, curses (sāpa-), black magic (abhicāra-), and other kinds of incidental decline, including social ostracism, into a fully evidentiary and cross-referenced whole, which is to say as consistent with the general presentation of astrology as a tightly woven epistemological or sāstraic system.245 According to my āyurvedic informants in Kerala and Coimbatore, the AH recommends astrology as an adjunct to Āyurveda. Their authority for this is AHSū 1.31c, which suggests that healing is facilitated when the planets are congenial (graheṣv anuguṇeṣu). This is widely quoted as a justification for employing astrology as a diagnostic tool.246
The PrMā (12:34a–c) describes the causes of unmāda in language familiar from the āyurvedic texts: excessive joy, desire, fear, grief, eating forbidden or unclean food, and the wrath of the gods and gurus. After breaking down the symptoms of unmāda according to the afflicted doṣa, the PrMā describes possession by a devagraha in familiar terms, but does not name other grahas. Instead, the text personifies as female grahas a dozen symptoms of seizure (apasmāra, 12.53–54): Svāsanā (Gasping), Malina (Dirtying), Nidrā (Sleep), Jṛmbhikā (Yawning), Anaśanā (Not Eating), Trāsinī (Fear), Mohinī (Enchantress), Rodanī (Weeping), Krodhanī (Anger), Tāpanī (Excitable), Śoṣaṇī (Dryness), and Dhvaṃsinī (Destructress). Throughout this section the PrMā prescribes different kinds of medicated ghee, including pañcagavya, and various kinds of sacrifice and homa. Most of these are to be administered only by qualified āyurvedic physicians and tantric ritual specialists.
One ritual, however, deserves special praise in the PrMā: the mṛtyuñjaya homa (13.36–39). This consists of a simple burnt offering of rice, black sesame seeds, and other uncooked grains, roots, fruits, and so on, with the so-called mahāmṛtyuñjaya mantra, the mantra that “conquers death.” This is perhaps the most commonly recited mantra for homa in India. The mantra, oṃ tryàmbakam yajāmahe sugandhím puṣṭivárdhanam | urvārukám iva bándhanān mṛtyór mukṣīya m
m
tāt| (We make offerings to Tryambaka (Rudra), the fragrant, increaser of prosperity; like a cucumber from its stem, from death may I be loosened, not from immortality) (ṚV 7.59.12, TS 1.8.6.2, etc.).247 The PrMā 13.36ab states that this ritual should be practiced to pacify all diseases, especially sharp fever (tivrajvara) and those brought on by black magic (abhicāra; 13.37), a category in which most spirit possession generally abides. This belief, as mentioned earlier, continues to the present day, as Dwyer’s survey at Bālājī attests. In Kerala, this takes the form of kaiviṣa, a Malayalam compound meaning “poison from someone’s hand” (Mal. kai [hand]; Skt. viṣa [poison]). In this form of abhicāra, a poison that is said to bring on unmāda or apasmāra is given surreptitiously to a person to drink or swallow, often with milk. The poison may be an ensorcelled hair, iron powder, aconite, or an insect’s body. Thus, the PrMā, standing outside the textuality of Āyurveda and Tantra, at least formally if not practically, actively invites the participation of experts in these two allied bodies of knowledge to treat what is to this text primarily astrological affliction.248
Beyond astrology, and beyond our concern with practices in Kerala, it is relevant to ask how is nija unmāda (mental illness with traceable physiological or psychological origins) distinguished from āgantuka unmāda (mental illness believed to be the result of possession)? In the case of nija unmāda, the vaidya takes a case history and undertakes a medical evaluation. In most cases, the cause is readily acknowledged: family problems, a death in the family, financial pressure, a life-threatening illness. The usual problems are depression and anxiety, as elsewhere in the world, and the unmāda is usually diagnosed as a vāta affliction and treated accordingly.
The tool most commonly employed for diagnosing āgantuka unmāda is observation. What vaidyas look for is abrupt and specifically defined behavioral change. However, this is not usually accepted as the final word, as most vaidyas try to verify these findings by taking a detailed case history and initiating a full medical evaluation. Never, to the best of my knowledge, except at last-resort places that specialize in exorcism, such as Bālājī, is there an assumption of possession. Besides observation, the diagnostic tools ordinarily used to check for āgantuka unmāda are pulse diagnosis, urinalysis, and the detection of physical signs of different families of spirits. For example, Clifford reports, “If the brightness of the eyes has deteriorated, if there are changes in the blood vessels in the whites of the eyes, if there are very small black dots on the whites of the eyes, or if the eyes look blurred or scratched, these are all signs of possession by a kingly or royal spirit (rgyal-po).”249 With respect to pulse, according to some of my informants in Pune, one should stare at the third eye of the patient, then take one’s own pulse. If the radial pulse is strongest, the unmāda is judged to be nija; if the ulnar pulse dominates, it is believed to be spirit possession (bhūtonmāda); if the medial artery pulse is dominant, then the unmāda is said to be caused by possession by an ancestral spirit (pitṛgraha). The urinalysis procedure reported by the same vaidyas is as follows. First, fill a cup with the patient’s urine. Then carefully place at the center of the urine one drop of sesame oil. If the oil flows to the south, then the unmāda is caused by a malevolent spirit, to the east by a devagraha, to the west by an animal spirit, but if it flows to the north then it is regarded as internal (nija).
This procedure, called tailabindu-parīkṣā (examination through a drop of oil), is related to a divinatory technique of urinalysis described in an eighteenth-century text called Yogaratnākara (YR), which is an excellent example of the integration of Āyurveda with tantric principles.250 The practice described to me in Pune is not found in the YR, but it appears to emerge out of a body of practice that is old and gave rise to the practices described in the YR. The passage in question is contained in the first chapter of the YR, in a section called mūtraparīkṣā (examination of urine). It will be instructive to give a near-verbatim account of the passage, because it illustrates the manner in which divination has been accepted in a medical work of the past few centuries. This is given neither to discount nor to accept the findings of the text and derivative practices, but to note that principles of a very different scientific paradigm were applied.
In the last hour of the last part [yāma] of the night, a vaidya should collect the urine and store it in a glass [kācamaya] vessel. He should then examine it after sunrise. The first part of the stream of urine should not be used, but the middle part of the stream should be collected. Only in this way can the vaidya determine the cause of the illness and its remedy. One should then take a drop of oil with a blade of grass and carefully place it in the middle of the vessel of urine. If the oil spreads, the disease is curable. If it does not spread, then the disease is difficult to cure. If it floats randomly on the surface, then the disease is incurable. If it flows to east, good health will be restored quickly. Movement to the south indicates a fever that will subside only gradually. If it flows to the north, a cure is definite. Movement of the oil to the west indicates happiness and health. If it flows to the northeast, then the patient will die in one month. If it flows to the southeast or southwest or if it breaks up, death is inevitable. If it flows to the northwest, then the patient will die even if treated with the best medicines [sudhā]. The same is true if it outlines the form of a plow, a tortoise, a buffalo, a beehive, a headless man, a part of an arm or leg, a weapon, an axe, a wooden pestle or spear, an arrow, a mace, or a crossroad with only three roads. In such cases it is better to do nothing at all. If it takes the shape of a goose, a duck, a pond [tāḍāga], a lotus, an elephant, a chowrie [cāmara], an umbrella, an arch, or a mansion, then good health will definitely be restored. If the drop of oil wanders aimlessly, there is possession by a family member [kuladoṣa] or a spirit [pretadoṣa]. If it resembles a human being or two skulls [mastakadvayam], there is possession by a bhūta [bhūtadosa] and bhūtavidyā should be applied. If it takes the shape of snake, the condition is due to vāta aggravation. If it takes the shape of an umbrella, the condition is due to pitta aggravation. If it takes the shape of pearls, the condition is due to kapha aggravation.251
In addition to the fumigants and varieties of medicated ghee prescribed for alleviation of bhutonmāda, some vaidyas in Maharashtra prescribe bhasma or ash from incinerated emerald (pannabhasma) or from the nine gems (navaratna) associated with the planets.252 Emerald is associated with the planet Mercury, which in astrology is associated with the mind. Thus, in astrological thought, an infusion of emerald in its subtlest form, alchemically produced bhasma, is believed to strengthen the influence of Mercury and, consequently, the mind.253 I am unable to find a prescription for this in any āyurvedic text. However, Indian alchemical literature of the last three hundred years has been insufficiently researched.254 I should add that these prescriptions do not appear to be followed in Kerala, where the physicians inform me that they do not use such bhasmas.
However, the most consistent statements on the application of pulse and urinalysis for diagnosis of possession come from Tibetan sources. Clifford states that these are not found in the Gyu-zhi. Her sources, rather, were contemporary practitioners, who probably learned these from other texts and from oral tradition. The most likely source is a series of thangkas that is used as a textbook, along with copious written and oral commentary, in Tibetan medical colleges.255 Three of the thangkas depict (among other things) divinatory diagrams that appear within the larger maṇḍalas that constitute the thangkas themselves. These smaller diagrams are circles trisected vertically and horizontally, thus yielding nine divisions, the central one being a square.256 To produce this maṇḍala or yantra, a circle is inscribed on a small mobile object, possibly a glass or copper plate. On this, the grid is created by placing four thin sticks one-third of the way into the circle from both sides, the top, and the bottom. The resulting nine sections are imagined as a turtle lying on its back, with its head toward the north, as a mirror of male urine, and as a turtle sitting right side up, with its head toward the south, as a mirror of female urine. The nine partitions represent the residence of sadak (sa-bdog; ground-owner demons), said to come from the Bön tradition.257 The center represents the self. Surrounding it, clockwise from the northeast corner are, in the case of the upside-down turtle, the ground-owner demons represented as gods (lHa), ancestors (pha-mes), cremation ground demons (dur-sa), house demons (khang), field demons (shing), progeny (bu-tsha), ghosts (gDon), and humans (mi). In the case of the right-side-up turtle, they are, from the northeast, cremation ground demons, ancestors, gods, humans, ghosts, progeny, field demons, and house demons. The category at the center, namely, self, requires explanation. According to the Fundamentals of Tibetan Medicine, a widely used handbook, this represents “a spirit associated with a person with whom the patient is emotionally close, although not necessarily friendly.”258 This yantra is then placed over a bowl of urine. If a bubble that looks like a fisheye appears in the urine and settles beneath one of the partitions, then the identity of the invading spirit becomes known.
The recognition of disease-producing spirit possession in South Asia has an epochal history of more than three thousand years and an equally epic variation and complexity. It can be traced from the Atharvaveda, where unmāda was first discussed, to the early canonical texts of Caraka, Susruta, and Vāgbhaṭa, where diagnostics and treatment modalities were more formally discussed. Contemporaneous with these texts, it is discussed in the Mahābhārata and Buddhist medical treatises. From these sources it was dispersed to a variety of Tantras and āyurvedic texts beginning in the late first millennium and finally drifted into dharmasāstra and astrological literature. It will come as no surprise, then, that, given this deep history, the idea of negative, invasive spirits has not ceased in India. Rather, under the influence of thousands of years of new religious movements and sociopolitical formations, as well as updated scientific notions, it has been reconfigured and reintroduced in recent medical and paramedical texts, the latter represented by Tantras composed in the past few centuries. One of the chief characteristics of this history has been a flow of ideas on bhūtavidyā and bhūtonmāda from educated elites who control texts into the popular imaginaire and, conversely, from popular, non-Brahman, culture into texts. The juxtaposition of and exchange between classical and folk is, in fact, nowhere more evident than in the area of bhūtavidyā. It appears from the material presented copiously in the preceding pages that the banishment of the subaltern from much of classical discourse is a thin and tenuous construction or, perhaps more generously, an artifact of orthodox conditioning within the minds of both classical Indian and modern Western-trained scholars.
One of the dynamic aspects of bhūtavidyā, the dialectic between continuity and change, is the identity of possessing entities. The vedic literature recognizes an array of such beings, most of them male, including the gandharva, nāga, siddha, sādhya, vipra, yakṣa, rākṣasa, piśāca, and yātudhāna. Only rarely are female possessors mentioned in the Vedas, notably the apsaras, the wife of the gandharva. Many of these appear to be early instances of Sanskritization, imported into the pantheon of vedic deities as minor but essential players in a sumptuous cosmogony. Contemporaneous with the latter phases of the vedic period, the second half of the first millennium B.C.E. or perhaps a few centuries later, an alternative genus of female possessors appears to have been imported from popular culture into the MBh and the early canonical āyurvedic texts.259 These were the bālagrahas (child-snatchers) who attacked and possessed pregnant women and children. They were much more malevolent and far less playful and Sanskritic than the male grahas who preceded them in the literary record. Most grahas found in the bhūtavidyā sections of the āyurvedic texts, the Bower manuscript, the ĪŚP, the Madanamahārṇava, and, to a lesser extent, the ethnographies of Varanasi and the Maldives are male and described by their personalities. It is as if they have life histories that can be deciphered through their names and literary contexts. By contrast, the grahas of the MBh, the pediatric sections of the āyurvedic texts, and, to a great extent, those represented in popular culture, are female, usually demonesses known only by their names, with little discernible life history, such as Mukhamaṇḍikā and Lohitāyanī (MBh).260 The more general terms found in modern Indian culture, such as the bhavānī and the cuḍail, signify categories of demoness, rather than historically situated ones. This representation in popular culture is, however, consistent with the earliest vedic literature and with a number of intermediate genres in which minor goddesses are given much less definition than male gods or demigods.
Also evident in the modernization of the texts is an updating that has taken a turn perhaps unexpected by Caraka. It is clear, as Meulenbeld has pointed out, that Caraka and Susruta attempted to forge an empirical medical system, in part by expunging from their texts material from the AV and its supporting literature, such as the Kauśikasūtra, that was viewed as unscientific or otherwise opposed to the system that they were attempting to construct. However, with the rise of Tantra and the application of astrology to nonritual, personal situations, a good deal of material from these increasingly defined bodies of knowledge was adopted into āyurvedic literature and the discourse of āyurvedic practice, as the medical textuality of the seventeenth to twentieth centuries testifies. The tantric and astrological literature differs considerably from that associated with the AV and could not have been derived from it in any significant way. Rather, the infusion of textual elements of Tantra and astrology into medical texts, as well as the reverse, an infusion of medical textuality, including extensive passages on bhūtavidyā, into texts on astrology, Tantra, and karmavipāka, indicates a blurring of boundaries between these genres in the past five hundred years. Thus, Āyurveda has, in a limited and updated manner, come full circle. Caraka and Susruta might not approve, but the project of scientizing and sanitizing, as well as the introduction of Tantra and astrology into Indian popular culture have afforded them access into the expansive domain of Āyurveda. The evidence from Kerala, with its history of intertextuality and the overlapping concerns of vaidyas, mantravādins, and jyotiṣīs, appears to resist the notion that the rise of Western educational models, the ascendance of allopathic medicine, a national āyurvedic curriculum, and an increasing reliance on text rather than apprenticeship as authoritative,261 could end the diagnosis and treatment of exogenous mental illness any time soon.
NOTES
1. The composition was cumulative, dating from the second or third century B.C.E. to the fourth or fifth century of the Common Era. For a summary of the evidence see Wujastyk 1998:39–41. For a more precise dating of the core of the text to 50–150 C.E. and exhaustive evidence for this date, see Meulenbeld HIML IA:105–115 and references.
2. Caraka-sūtrasthāna (CaSu) 11.54, Caraka-vimānasthāna (CaVi) 8.87; also Aṣṭāṇgasaṃgraha-sūtrasthāna (ASaṃSū) 12.3; and Hārītasaṃhitā 3.2.3 for variants daivapathāśraya and yuktipathāsraya. On the word sattvāvajaya, see Murthy and Singh 1987, Shankar and Unnikrishnan 2004:168, and note 221 below.
3. None of my informants knew the textual source of this verse, and I am unable to trace it. However, for a number of similar passages in Purāṇas and Smṛtis see Thite 1982:17–18. One vaidya told me that it is from a karmavipākaprāyascitta, or section on “expiation of ripened karma,” found in various Tantras and Dharmasastra texts. For more on karmavipāka, see the discussion below on the Madanamahārṇava.
4. Daiva, lit. “divine,” from deva, “deity,” has the extended (and usual) meaning “fate.” See CaSā 1.116: “Action that a person performs in a previous body is known as daiva. In time this also becomes the cause of diseases”; also CaSā 2.44: “That which is done in a previous birth is called daiva, while whatever action [karma] is seen in the present birth is called pauruṣa. That which is seen to be unbalanced is the cause of the onset [of disease], while that same thing brought into balance is the cause of its cessation.” Thus, daiva is contrasted with pauruṣa (human), from puruṣa (man or person); cf. CaVi 3.33–34.
5. Scant mention is made of sattvāvajaya, following the āyurvedic texts themselves.
6. Engler 2003:438 translates this as “science of demonic seizures.” This is a misunderstanding of the term.
7. See Kleinman and Seeman (1998:237–252), who show, through comparative analysis, that religious healing is a form of moral practice. They write: “Calculations of locally defined utility, construction of self and cosmology, and the micro-politics of local social life are all central to healing as moral practice” (p. 251). This assertion is abundantly supported by the āyurvedic and folk texts and practices presented in this chapter.
8. For example, CaCi 9.33:93–94. Zysk (1985:63) translates this as “accidental” and distinguishes it from madness caused by external agents such as gods, sages, deceased ancestors, and a variety of demons. My reading is that the latter are the primary agents of āgantuka unmāda. Zysk states the general theme as follows:
Causes of diseases are not attributed to physiological functions, but rather to external beings or forces of a demonic nature who enter the body of their victim and produce sickness. The removal of such malevolent entities usually involved an elaborate ritual, often drawing on aspects of the dominant local religion and nearly always necessitating spiritually potent and efficacious words, actions and devices.… The Vedic Indian’s attitude toward disease, therefore, was dominated by the belief that evil spirits, demons and other malevolent forces invade the body and caused their victims to exhibit a state of dis-ease. (p. 8)
See Zysk 1991:15 for another similar assertion.
9. Dash and Kashyap 1980:1. These authors, following Ṭoḍarānanda (hence the intellectual lineage of Susruta), divide Āyurveda into ādhyātmika duḥkha (śārīraka and mānasika), ādhibhautika duḥkha, and ādhidaivika duḥkha, employing familiar Indic hermeneutical categories.
10. For convenient summaries see, e.g., Keith 1925:231–242; Oldenberg 1988:131–134. Bhattacharyya 2000 supplies much more detail in his diffuse and irregularly documented study of bhūtavidya; see esp. ch. 2 on “Vedic Demonology.” I cannot here enter into all the detail Bhattacharyya discusses, because the nature of the present project is to discuss only those bhūtas that are involved in possession, rather than all manner of vedic bhūtas.
11. Cf. also Chāndogya Upaniṣad (ChU) 7.1.4, 7.2.1, 7.7.1. The best accounts of bhūtavidya are Filliozat 1937 and Sukla 1985. Filliozat’s remarkable work provides a succinct introduction to the topic (p. 24–28) and tracks bhūtas through a stunning array of texts, languages, and cultures. Sukla contains more data than anyone else, except Bhattacharya 2000, but he organizes it much more effectively. I do not here address postcolonial interpretations or what Langford calls the “narrative of decline,” in which Āyurveda has been subjected to a secularization process. As part of this narrative, bhūtavidya has been regarded by many indigenous scholars and physicians dedicated to discovering modern science in ancient knowledge as superstition or psychotherapy, and bhūtas as microorganisms such as fungi, bacteria, and viruses; see Langford 2002:85ff. and notes.
12. Cf. Śāṇkhāyana Gṛhyasūtra (GS) 4.9.3, Asvalāyana GS 3.4.2; cf. also Pāraskara GS 2.12.2.
13. Dhaky 1984:240. His interpretations are often far-fetched, but he may be right to take bhūtas back to their roots in the elements, the bhūtas or mahabhūtas. He posits bhūtas as “elementals” and bhūtanāyakas as “captains of the elementals.” Cf. MBh 9.44.74–108; Matsyapurāṇa 53.1–46, 182.64–66, for physical features and other details of gaṇas, gaṇanāyakas, and gaṇeśvaras. The elements—earth, water, light, air, sky—have bhūtas and gaṇas (clans or tribes of the divine retinue) as their representatives, their symbols of noumenal and phenomenal reality.
14. yás ta ūr viháraty antar dámpatī śáye | yóniṃ yó antár āréṭhi tām itó nāsayāmasi || (He who opens your thighs, who lies between husband and wife, who licks inside the womb; him do we cast away from here). On 10.162.4c, see Sāyaṇa’s commentary, yonim antaḥ praviśya āreḷhi niṣiktaṃ reto jihvayā āsvādayati, bhakṣayatīty arthaḥ. See Sharma 1972:99; also Zysk’s perspicacious rendering, 1985:51–53.
15. Mayaram 2001:215. See also, notably, Taussig 1987.
16. For example, the case of Indra, the king of the gods, who becomes disfigured and degraded because of his untoward sexual advances on respectable women, or Śiva, who destroys Dakṣa’s sacrifice because of his exclusion.
17. Cf. AVŚ 3.11.1, 20.96.6. On yákṣma, see Bahulkar 1994:129f.; Filliozat 1964:43f., 96; Zysk 1985:12–17.
18. On the pravargya, see van Buitenen 1968; Houben 2000.
19. Āpastamba ŚS 15.19.4; Bhāradvāja ŚS 11.19.12; cf. Baudhāyana GS 3.7.27,28.
20. Āyurveda is often considered an upāṇga or upaveda of the AV; cf. e.g., Sukla 1985, in Sanskrit, intro. p. 10; cf. e.g., Aṣṭāṇgahṛdaya Saṃhitā (AH) 5.84; Aṣṭāṇgasaṃgraha (ASam) 7.118cd: daivavyāpāśrayaṃ tadvad atharvoktaṃ ca pūjitam. However, Wujastyk notes, “such medical material as is recoverable from the Vedic literature is remarkable more for its differences from classical āyurveda than for its similarities.… Of course there are some points of contact, but the overall sense is that, culturally speaking, āyurveda comes from somewhere else” (1998:17). Wujastyk’s translation of this chapter of the AH (ibid.:294–301) is the clearest exposition of its kind yet published. See also Meulenbeld: “In my opinion, Caraka’s list of sages, many of whom are known from Vedic literature, may have been inserted in order to stress the connection between āyurveda and the Vedic tradition, the orthodoxy of its teachings, and its association with the brāhmaṇa” (HIML IA:10).
21. See also the commentary by Sāyaṇa, cited in Whitney 1905:1:17.
22. Cf. Kausikasūtra 28.9 for a linked remedial rite against possession by evil spirits. The commentary on the KausSū explains that this rite is a mahāsānti called gāndharvī; cf. Bahulkar 1994:159f.; Whitney 1905:I:211.
23. See Zysk 1985:52 and, especially, Bahulkar 1994:126f. for the ritual and pharmacological prescriptions of the KausSū and the commentary of Dārila, as well as a summary of scholarly opinions on the hymn. The hymn itself, AVŚ 2.25, is addressed to a pṛśniparṇ
, a “spotted leafed” plant, which was used to help redress the condition.
24. agna … sa gráhyāḥ pśān vi cṛta prajānán. Cf. Bloomfield 1894:cxix ff., 1897:521 ff.
25. This is a much recognized and discussed aspect of healing traditions in ancient India. See, e.g., Bahulkar 1994 and Zysk 1985 for extensive discussions of this.
26. Beyer (1973:100) describes a Tibetan ritual in which the yogin makes a wax effigy into which he subjugates an enemy—a king or a demon are mentioned—by visualizing and bringing it into the effigy, which he then torments over flames and tramples.
27. Shukla 1980:135ff. and infra.
29. Ibid.:63, Sanskrit words added.
30. Cf. AH 6.6.55cd: “One should examine the patient for supernatural possession. The signs which were earlier described as applying to madness appear exaggerated” (trans. in Wujastyk 1998:301).
31. Sanderson 1985:213n89.
34. Ibid.:211–212n69, abhiniveśam is translated as “obsession.” Sanderson notes a further symbolic or, from the Indian point of view, adhyātmika reading in Jayaratha’s commentary.
35. Bahulkar 1994:160; his translation of amati- is “unconsciousness,” but this does not quite fit the context.
36. Weiss 1977. This remains unpublished, though the author is now preparing it for publication. For a comprehensive bibliography of works on unmāda, see Meulenbeld, HIML IB:65n93.
37. Cf. Weiss: “The object of the present study is to gain an understanding of the psychiatric conceptual formulations of the early Indian medical authors in the light of Western nosology, and our interest in therapeutics is limited to the extent to which it serves to elucidate the diagnostic categories” (1977:3).
38. Meulenbeld 1997:186–198.
39. One of the primary alleged causes of abhiṣaṅgaja disease is abhicāra (cf. Türstig 1985); that is, a bhūta that “sticks to” (abhiṣaṅga) a person is regarded as abhicāraja, sent by a hired sorcerer to possess the person. Cf. Dwyer 1998 and Chapter 4, note 35, and above p. 517f.
40. Lewis 1966:308 (reprint 1986:24).
41. Sanskrit lexicographers also acknowledge bhūtāveśa; cf. Trikāṇḍaśeṣa 3.230 and Hemacandra’s Anekārthasaṃgraha 4.161. The Rasaratnasamuccaya (5.2, 6.7), a thirteenth-century medical text prescribes a bhūtāveśapraśānti (procedure for pacifying possession by bhūtas). For further attestations of bhūtāveśa and piśācāveśa (and variants), see Kalhaṇa’s Rājataraṅgiṇī (c. 1150) 8.114.6 and the late dramas Vasantatilaka (c. 1800) 25.19 and Amṛtodaya (1710) 3.7, 3.18. Synonymous with āveśa is upahata (struck, afflicted); thus, also attested in the lexicons are bhūtopahata (afflicted by bhūtas); bhūtopahatacitta (the mind possessed by bhūtas); and, of course, the antidote, assuming the former: bhūtapratiṣedha (warding off bhūtas).
42. Other texts, many of them now lost, existed in both periods. See Meulenbeld HIML IA:689ff. (and notes), for a discussion of ancient authorities cited by Caraka, Susruta, Vāgbhaṭa, and their commentators, and HIML IIA:117–55 (and notes) for discussion of a number of other texts from the seventh to the eleventh centuries.
43. Meulenbeld (HIML IA:42) calls them “character types,” which is more accurate than “mental faculty,” “psyche,” etc., used by other translators. For discussions of the category sattva, see the many sources cited in Meulenbeld HIML IIB:82n112.
44. Cf. CaSā 4.36–39, SuSā 4.88–98, Kāsyapa 28.8–36.
45. For additional details, see Caraka Saṃhitā, ed. Sharma and Dash:3:416–420.
46. Cf. also Śārṅgdhara Saṃhitā (composed around the beginning of the fourteenth century) 1.7.38. Meulenbeld states that this expansion represents an “incorporation of material from popular medical lore” (1997:189).
47. Thanks to Robert Svoboda, a trained āyurvedic doctor, for this information. On the relationship between bhūtas and apasmāra described in different āyurvedic texts, see Meulenbeld 1997:212–215.
48. āveśo grahaduḥkādyair apasmāro yathāvidhiḥ bhūpātakampaprasvedalālāphenodgmādayaḥ (4.31); for references in other texts on poetics and dramaturgy see Haas 1912:119.
49. Cited in Sabdakalpadruma.
50. This is discussed in the Cidambara Māhāytma, a text describing the greatness of the Naṭarāja (Śiva) temple in Chidambaram, Tamilnadu; cf. David Smith 1996:222ff.
51. I present my own translations because I find that I differ with Weiss 1997 in a significant number of places.
52. In a rather different vein, see AVŚ 5.4 and commentaries, in which hrūḍu, the secret name of the graha, is uttered to appease the spirit in charge of giving consumption (yákṣma) or fever (takmán); cf. Zysk 1985:53.
53. One word for “comet” is upagraha; cf. MBh 3.216.1.
54. For a good brief synopsis of the evolution of the application of the word graha to the planets, see Yano 2003:380ff.
55. Cf AVŚ 19.9.7,10. See Caland 1926. Verse 10 reads: śáṃ no gráhāś cāndramas
śáṃ ādityś ca rāhúnā | śáṃ no mṛtyúr dhūmaketúḥ śáṃ rudrs tigmatejasaḥ || (Weal for us be the planets belonging to the moon, and weal the sun with Rāhu; weal for us smoke-bannered death, weal the Rudras of keen brightness) (trans. Whitney 1905:2:914).
56. Bṛhatsaṃhitā 5.11; BṛS 5.13 reads uparāga (darkening) for eclipse; cf. also Bhāgavata Purāna (BhP) 5.24.3.
57. I am not able to find the words āveśa or praveśa in these contexts to indicate the sun or moon becoming possessed by Rāhu or Ketu (or vice versa). Rather, these two grahas assail (abhidhāvati, BhP 8.9.26, Rāhu only) or devour (grasati, MBh 1.17.8, again only Rāhu) their quarry, rather than “enter” or “possess” them. For references and discussion see, e.g., Bedekar 1967; O’Flaherty 1975:273–280. Although the word ketú is found as early as AVŚ 19.9.8–10, where the word appears to mean “comet,” its reconfiguration as the descending node of the moon occurs only in the early second millennium C.E.; cf. Gail 1980.
58. This is consistent with the use of √gṛh in BĀU 3.3.1, 3.7.1 described above.
59. For example, Bṛhatsaṃhitā 43.37. Cf. Bṛhatparāśara 84.17–20; Kane HDh II:749 ff.; Gaur 1979; Karmaṭhaguruḥ 74ff., 110ff. In the Purāṇas, see, e.g., Matsya Purāna 17.56, 24.46, 93.2ff., instructions for a grahahoma. On the growing importance of astrology in India during the first millennium C.E., see Inden 1992.
60. Cf. Kātyāyana Śrautasūtra 9.14.14, Āśvalāyana Śrautasūtra 8.1.10.
61. Palsule 1967:11. On avatāra, see note 141 below.
63. Lutgendorf 1991:250, 322. See also Eck 1985:38.
65. On its Tamil synonym aruḷ, see Chapter 4, p. 128.
66. Cf. ṚV 7.103.4 (anugṛbhṇāti), where frogs kindly receive one another; 2.28.6 (anugṛbhāya), in which the supplicant implores Varuṇa to accept or extend his grace to him. Cf. Kāṭhaka Saṃhitā 25.8, AB 7.18.
67. For dating, see Olivelle 1992:8ff.
68. manyamānau ca kaumāraṃ puṣpitaṃ tadanugraham (KSS 1.2.76).
69. Devasenapathy 1983:10–11. Upadhyaya and Upadhyaya 1984:7 (cf. Honko 1998:251) speak of nigraha and anugraha in Tulunad, meaning the protective and punitive activities, respectively, of different spirits. Honko notes that the Siri epic was performed at possession rituals that were similar to bhūta rituals “with the purpose of supporting the mental health of participants with a history of illness” (p. 261).
70. On this, see Zysk 1993.
71. Eventually Āyurveda was used to confirm Sāṃkhya. The Sāṃkhyasaptativṛtti, an early commentary on the Sāṃkhyasaptati of perhaps the sixth century, uses examples from Āyurveda to support its arguments. See Larson and Bhattacharya 1987:193ff.
72. For these lists, and accompanying discussions, see ibid.:59~61, 161, 177, 206, 284. Larson’s speculations linking the eight realms of the SK to the eightfold prakṛti (p. 61) are thought provoking, but entirely lacking in textual support.
73. For the use of some of these categories in the Mālinīvijayottaratantra, see Vasudeva (2004:326).
74. Cakrapāṇidatta glosses abhidharṣaṇam as āveśaḥ and niyamavratādi as niyamety ādinā aihikaṃ karma devādyāveśakāraṇaṃ brute (by niyama, etc., worldly action caused by possession [āvesa] by gods and others is indicated). The term niyamavratādi may be a reference to Yogasūtras 2.30–32. In these sūtras the five yamas ([external] restraints) (YS 2.30)—non-injury, truthfulness, abstaining from theft, sexual continence, non-covetousness—are called mahāvrata (the great vow) (YS 2.31). The niyamas ([internal] restraints) listed (YS 1.32) are cleanliness, contentment, austerity, studiousness, and devotion to Īśvara. Āyurveda and Yoga have long claimed a connection, but hitherto it has been difficult to actually find textual connections between the two, especially for the early periods. However, a sixteenth-century text called Ayurvedasūtra, which closely links the two, was published in 1994. This late, and rather labored, sutra collection contains within it the entire text of the YS, interspersed with material drawn from āyurvedic texts. Whereas it is fairly certain that Caraka predated YS by several centuries, it may be assumed that many elements of the YS were well known long before the final compilation of the sūtra text. Thus, there is no reason that the yamas and niyamas, as well as other elements of the eightfold (aṣṭāṅga) path, which were quickly incorporated into yogaśāstra and may be found in Buddhist texts, could not have been known to Agnivesa or other compilers of the Caraka Saṃhitā.
75. This verse is adopted as a basic definition of possession in many later texts; cf, e.g., Mādhavanidānam 20.17.
76. On the evolution of the pitṛ in Indic thought, especially in the Harivaṃśa, see Saindon 1995.
77. One wonders about the models on which some of these categories are based. Mozart, it is reported, would suddenly burst out laughing in the middle of a conversation, jump up and down, turn somersaults, and impulsively leap over tables and chairs.
78. Often the texts do not distinguish between waxing and waning fortnights. We must assume that both fortnights are assumed unless otherwise specified.
79. Ḍalhaṇa: bhinnamaryādaṃ tyaktocitakramam abhakṣyabhakṣaṇe agamyāgamane; abandonment of propriety, such as eating what should not be eaten or going where one should not go.
80. Ḍalhaṇa: sṛkkiṇyau oṣṭhaprāntau.
81. Ḍalhaṇa: uddhasto vikṛtadarśanaḥ (deformed or disfigured). Verses 8, 9, 13, 15, and 16 contain the word juṣṭaḥ (pleased, welcomed, visited). This is a rather milder and more positive manner of speaking than Caraka’s equivalent verses: unmattaṃ vidyāt. Verses 10 and 11 read, for “possession,” paripīḍitaḥ; verse 12 reads abhibhūtaḥ; and verse 14 gṛhītaḥ.
82. Ḍalhaṇa: adriḥ parvataḥ, dvirado (two-tusked) hastī, nago (non-moving) vṛkṣaḥ.
83. After Ḍalhaṇa, who cites the variant reading vardhakena rather than vardhakena: vardhakena chedakena hiṃsārthinā kenacid graheṇa juṣṭo gṛhīta iti vyākhyānayanti.
84. This apparently general description could also be an indication of oracular possession, which we have seen from the account of Pātañcala Kāpya and his daughter in the BĀU, was well known by then.
85. Ḍalhaṇa: the eight attributes (guṇāḥ) are the eight siddhis: aṇimā (minification), laghimā (lightness), mahimā (physical expansion), garimā (heaviness), prāptiḥ (instant acquisition), prākāmyam (irresistable will), īśitvam (lordliness), vaśitvam (power of influence). Ḍalhaṇa also cites CaŚā 1.140, which has a different list: āveśaś cetaso jñānam arthānāṃ chandataḥ [or ca svataḥ] kriyā | dṛṣṭiḥ śrotraṃ smṛtiḥ kāntir iṣṭataś cāpy adarśanam-ity aṣṭavidham aiśvaryam āhuḥ | (Possession [āveśa] of [another’s] mind, knowledge of objects according to one’s will, and [the powers of] action, vision, hearing, memory, beauty, and invisibility, all gained in accordance with one’s desire, these, they say, are the eight kinds of majesty). Ḍalhaṇa glosses vyastaḥ as dvitricaturaḥ and samastāḥ as sarve, which is to say fewer of the eight siddhis are manifested according to the power of the possessing agent. He concludes that fewer of these attributes are found in asuras, etc., and all in divinities (tatra vyastā guṇā asurādigrahāṇāṃ samastā devagrahāṇām).
86. Ḍalhaṇa confirms that saṃviśanti refers to sexual intercourse (te devayonayo manuṣyaiḥ saha na saṃviśanti). Thus, medical practitioners of the day must have encountered such claims, based, no doubt, on references in the epics and elsewhere to divinities and semidivine beings cohabiting with humans, e.g., Purūravas and Urvaśī. Ḍalhaṇa glosses āviśanti with anupraviśanti. Cf. Weiss 1977:54 for discussion of this verse.
87. “Infinite numbers,” koṭisahasrāyutapadmasaṃkhyāḥ: koṭi, ten million; sahasra, one thousand; ayuta, ten thousand; padma, according to Ḍalhaṇa, fourteen times the multiples of the previous numbers.
88. Ḍalhaṇa asks, rhetorically, how a being who shares at least part of the essence of deities can relish blood, fat, and flesh, etc. He answers that Nirṛti is the grandfather of the demons (nirṛtī rakśasāṃ pitāṃahaḥ), who, through their association with the deities, etc., acquired part of their natures.
89. Aruṇadatta glosses: sadyaḥ adhunātanaḥ pūrvajanmakṛto vā hetuḥ (action performed at the present time or during a previous birth).
90. See this chapter, n. 236, for more on the continued use of this term.
91. The editor of the text cites “Candra” (Candranandana’s Padārthacandrika): śmaśānacatvaracaityādiṣu (burning ground, crossroad, or [Jain or Buddhist] funeral mound).
92. parasūtakasaṅkaraḥ, cf. Candra: strīprasavasaṅkāryam; also Meulenbeld HIML IB:563n90, on this difficult term.
93. The critical text reads kāṣṭhāśvaṃ tathā (who mounts a wooden horse). This makes no contextual sense and at best could be an idiom (found nowhere else, so far as I am aware) for delusional sensory activity. In the absence of certainty with regard to that, however, the translation I offer is of a minority reading: kāṣṭḥāśmarāśim, for which, compare Caraka 6.9.20(8): aśmakāṣṭhādhirohaṇaratiṃ (who enjoys climbing on assorted heaps of garbage and walking on rags).
94. tṛṇachidram, lit. “one [who digs] holes in the grass”; Weiss: “who burrows in the grass” (1977: 291). Unfortunately no commentary addresses this compound. I prefer to take it metaphorically: chidra in its known meaning of “imperfection”; tṛṇa as “straw,” “grass,” “insignificant,” “minute,” “worthless.”
95. AṣṭāṅgaSam 7.2 replaces kūṣmāṇḍa with kārkhoda, but includes the kūṣmāṇḍa elsewhere. See Chapter 11, n. 97; this is surely a variant of khakhōrda. For more complete information on the kūṣmāṇḍa see Meulenbeld HIML IB:563n101.
96. The notes to the text provide several alternative readings: maukiraṇa, uttaruṇā, ramautkiraṇa. None of these help in identification of this enigmatic graha. For other occurrences of this word see Meulenbeld HIML IIB:564n103.
97. Cf. ibid.:564n104, for a discussion of this word. Cf. KSS 73.290 for vetālānupraviṣṭa, and 73.292 for vetālādhiṣṭhita.
100. Ibid.:183. It may be of interest to note here that what is now regarded as obsessive-compulsive disorder (OCD) was interpreted in Europe as demonic possession and treated with exorcism until the nineteenth century, when clinicians decided that this condition was a disease rather than the devil’s work. Freud recommended psychoanalytic guidance to free such victims from their obsession, which he believed was the result of being torn between guilty fear and defiant rage born of draconian toilet training. One treatment to which OCD has been surprisingly responsive is prefrontal lobotomy (PFL). Since the heyday of this treatment, genetic markers for OCD have been discovered in laboratory animals, and drugs such as Prozac and Zoloft, which are selective serotonin uptake inhibitors, have become cornerstones of treatment, though Pflis still often recommended in extreme cases.
102. Ibid. It is important to recognize that Weiss is not the first to attempt such an analysis. His work represents significant advances over that of Sinh Jee, who wrote in 1896: “The demoniacal diseases of the Hindus are but other words for hysteria, epilepsy, dancing mania, and other disorders of the nervous system” (cited by Weiss, 1977:53–54).
103. Though see Suśruta, Śārīrasthāna 1.18 for a similar, if less-detailed, passage. See also Manu 12.40–51 for a different model of sattvas.
104. Wujastyk (1998) gives an accessible account of the contents of these texts.
105. See also Weiss 1980:105, for a discussion of sattva. He says that Ca 4.4 uses it in the sense of “personality,” but that Ca 4.3 identifies it with manas.
106. Suddham here is equivalent to sāttvikam, characterized by sattvaguṇa.
107. Weiss mentions the “insane character” (1977:34) in the Mattavilāsaprahasana, but fails to mention that the character is possessed.
108. Meulenbeld states: “Weiss tried to correlate [mental disorders according to Āyurveda] with the syndromes of Western psychiatry. In my opinion, the results of his efforts are not convincing” (1997:197). However, I also agree with Meulenbeld’s more nuanced sentiment on Weiss’s work: “I would misrepresent the basic assumptions and views of M. Weiss if I were to attribute to him a simple scheme of one-to-one equivalents of Indian and Western diagnostic entities. He explicitly rejects this in theory, but, nevertheless, he often suggests it” (ibid.:192).
109. Murphy 1976:1027; cf. Weiss 1977:5.
110. Clifford 1984:xiii, 12.
111. This was the same decade that Obeyesekere published his initial studies and that Kakar (1978), Goldman (1978), Masson (1974, 1976), and others initially applied Freudian psychodynamics to Indian cultural and literary history. These studies stand out today as period pieces. Kripal 1995, discussed in Chapter 2, is much more substantive and sophisticated, though, as I pointed out there, not beyond criticism.
112. On Candranandana and his works, see Meulenbeld, HIML IA:665f. and notes; IIA:120ff. and notes.
113. See Fenner, who asserts: “All Tibetan writers have accepted the tradition that the rGyud bzhi is a translation of a Sanskrit medical book titled Amṛtahṛdayāṣṭāṅgaguhyopadeśa Tantra” (1996:458). This text needs to be found and checked, as there is much in Fenner’s article that is questionable. It is not mentioned at all by Meulenbeld in HIML. For a good discussion contextualizing the Gyu-zhi, see Obermiller 1989.
114. For an outline of the entire text, see Clifford 1984:237–241.
117. Ibid.:151–152. Also: “‘Ghosts’ and ‘evil spirits’ are not regarded simply as the projections of an unhealthy mind. For although they arose from projections, they are said to be real in a relative sense. That is, they have to be dealt with” (p. 159). Along the same lines, she states: “Two predominant themes emerge in the Dharma view of ‘demons.’ One, from the level of absolute truth, they do not exist. Like everything else, they have no self-nature. Two, from the level of relative truth, they have conditioned existence. One should protect oneself from their negative influence” (p. 161).
118. Ibid.:214. Heinroth’s most famous work was Störungen des Seelenlebens (Disorders of the Soul, 1818). Heinroth was the first to use the term “psychosomatic” to describe the connection between the psyche and the body.
120. Ibid. On the relationship between emotional and humoral imbalance, see Meulenbeld 1997:208–209.
121. Indeed, Clifford repeatedly defends Buddhist positionings. I look forward to a more perspicacious approach to cross-cultural, particularly Buddhist, psychology. Clifford’s effort may be compared to Vogt 1999, discussed in Chapter 4, in that the point of reference for both is abhidharma and that neither inspires confidence in their psychoanalytic expertise. In this sense, Weiss’s study is more reliable. As mentioned earlier (see Chapter 2), a more productive cross-cultural methodological exercise is Littlewood 1996.
135. Nevertheless, her study would have been greatly enhanced if she had utilized Weiss’s work, which was completed several years before the publication of her book, or if she had used the published and translated āyurvedic texts as a check on her textual history.
136. For example, Lutz 1988; Lynch 1990. Recall Lutz’s argument that emotion is an “ideological practice.”
137. On this, and much more about the sources and variation of emotion, see the comprehensive study of Nussbaum, 2001:151ff., passim.
138. This opposes Freud, who argued, at least in Totem and Taboo (1946), for the universal applicability of his system.
139. Very few studies have been conducted on the origins of any of these classifications. See, however, Mukherjee (1981) for speculations on an early group of five: pitṛ, gandharva, deva, asura, and rakṣas.
140. In addition to those mentioned below, some Purāṇas describe a certain class of beings as simply “possessors” (āvesaka); cf. VayuP 2.8.40, ŚivaP 2.2.33.22a, SkandaP 1.2.26.42c.
141. A brief text that I cannot deal with extensively here (because I did not discover it until the present manuscript was sent off to press) is the Āviṣṭaceṣṭāvidhiparivartaḥ (ĀCVP), “Chapter on Instructions Regarding Behavior of the Possessed,” appearing in the Mahāyānasūtrasaṃgraha, Part 2, pp. 255–259. This text is taken from the Mañjusrimulakalpa (MMK), the exact dating of which is uncertain, but appears to have roots as early as the fourth century C.E., but was probably not fully compiled until the tenth century. This fragment, like the rest of the MMK, is presented as a dialogue between Sākyamuni and Mañjusri Kumārabhūta. The question Mañjusrī poses to Sākyamuni concerns the behavior of people whose bodies have been possessed by other beings, including Buddhist teachers (ārya), divine beings, mendicants, perfected spiritual practitioners, gandharvas, yakṣas, rakṣasas, piśācas, great serpent beings (mahoraga), and so on, whether human or nonhuman, what kinds of physical actions and symptoms of these beings are revealed by these individuals, as well as the different courses of action to be taken in such cases.
Sākyamuni then delivers a discourse in 59 verses that covers two topics. The first deals with the symptoms of positive possession, describing the symptoms of auspiciousness in a person thus possessed by a deva. The second topic is a linguistic circumnavigation of India, attributing the peculiarities of speech in different parts of the subcontinent to a primal ancestor that could be identified as a yakṣa, mahoraga, kinnara, piśāca, garuḍa, and so on, along with a caricature of the speech of these regions, in phonological detail. Thus, this text should be of interest to scholars who study the dialectology of first millennium C.E. South Asia. Also of interest is the fact that the word most commonly employed in the ACVP for possession is avatāra (“descent,” which we have discussed above in reference to bhakti texts), though the words āveśa and āviṣṭa also appear in the text. It appears then, that the ACVP employs demonology as a sociolinguistic trope, rendering it qualitatively different from its use in the āyurvedic and tantric literature.
142. For the full story of the discovery of the manuscripts and an accounting of the contents, see Wujastyk 1998:195ff.
143. Hoernle 1909:2:223, 227; plate 49, part 6, leaf i, Reverse. Compare this to a sixth-century manuscript fragment from Bāmiyān; cf. Ojihara 1984:308.
144. Perhaps this is identical to Vinatā, the śakunigrahaḥ, from the MBh (3.219.26) noted earlier.
145. Cf. MBh 3.269.2, in which Pūtana is an enemy of Rama. But this is probably the same as the Pūtanā-graha (despite the short “a” at the end of the word) from MBh 3.219.26–27.
146. Cf. AVŚ 8.6.21, chyaka; this hymn is meant for the protection of pregnant women and therefore precedes the notion of bālagraha in the MBh and the medical literature. This hymn contains a large number of difficult words; cf. Whitney 1905 II:493ff. Hoernle suggests “unnatural change in appearance” (1909:2:227n5) after Caraka 5.7, Susruta 1.31: “morbid appearance” (ibid.:172n383). On chāyā-graha, see Jamison 1991:142ff.
147. Hoernle regards this problematic word as most likely an orthographic error for dustāraka (difficult to cross over), hence “evil eye” (1909:2:227n6). However, this word appears in the ACVP unambiguously as the name of a possessor.
148. It is not possible that in this context maruta indicates either the vedic deities known as the Maruts, or Hanumān, even as a possessing entity, one of whose names more than a millennium later was Mārutī; cf. Chapter 11, on possession of Hanumān.
149. Cf. Manu 12.71cd, amedhyakuṇapāśī ca kṣatriyaḥ kaṭapūtanaḥ (a kaṭapūtana is a kṣatriya [spirit] that eats impure objects and corpses). Kullūka glosses amedhyakuṇapāśī as purīṣaśavabhojī (an excrement and corpse eater). Manu 12.71–72 contains equally horrific fates for members of other varṇas who neglect their ordained duties, but for some reason kṣatriyas alone were singled out here.
150. This term is still widely used in Kerala; cf. Caldwell 1999:213.
151. There is a class of tantric texts, now mostly lost, called bhūtatantras. One of them, the Totula, is quoted by Kṣemarāja in his commentary on Netra Tantra 19.70.
152. This division (or variants of it) is often found in tantric texts. See, e.g., NT 19.30c–31: balikāmas tathā cānye bhoktukāmās tathāpare || ratikāmā hantukāmā vātajāḥ pittajāḥ pare | śleṣmajāḥ saṃnipātotthā bhūtā vividharūpakāḥ || (Various species of bhūta, fond of sacrifice, of eating, of sex, and of killing, are produced from disorders of wind, bile, phlegm, and all three together [saṃnipātajāḥ], respectively). The commentary on NT glosses bhoktukāmāḥ as māṃsaraktābhilāṣiṇaḥ (desirous of meat and blood).
153. Verse 1 reads heḍhra. I am not able to locate this word elsewhere in Sanskrit or Prakrit. I am grateful to Robert Zydenbos for the following: “It could be a Sanskritised Dravidian word. Kannada has a word heḍḍa, ‘a dull, stupid man, a fool’ [Kittel 1968–71:4:1755], with derived words like heḍḍatana / heḍḍu ‘stupidity, foolishness,’ and heḍḍāṭa, ‘stupid, foolish behaviour’” (e-mail communication, February 16, 2003). It appears, then, that we can be fairly certain that heḍhra was Sanskritized from this or a related South Dravidian word. See Prasnamārga, an astrological text in Sanskrit from Kerala (dealt with below), for the word hedva (15.62c), with probably the same meaning. The description could be that of spasticity or another kind of muscular or motor disorder, or, just as likely, severe mental retardation.
154. sutkāravān (makes the sound sūt). This could also refer to whistling.
155. Dwyer reports: “As many Hindus often comment, ‘Women have a weak heart; so spirits severely attack them’” (aurato ke kamzor dil hotā hai; islīe bhūt-pret us par jyādā hamlā karte hai)” (2003:45).
156. Kane (HDh I.2:800 ff.) dates it to 1360–1390.
157. The other important work is the Vīrasiṃhāvaloka; see below, n. 248.
158. For a more extended explanation of the principles of this literature, see Nambiyar’s introduction to the Sanskrit text of the Madanamahārṇava, pp. 15ff.
159. Nambiyar, in his introduction to the MM, writes: “Most of the causes are more or less connected with immorality and violation of the social customs and manners” (p. 25). For an account of karmavipāka in more canonical dharmasāstra texts, see Rocher 1980.
160. On the problems of violence and killing in the ritual arena, see Knipe 2004: “goats and humans have identical structures and therefore goats are the best offerings to the gods. No goat is ever ‘killed,’ that is, annihilated. Rather, it is transformed by ritual speech and practice” (p. 438). See also Houben 1999; Tull 1996.
161. Cf. Taittirīya Āraṇyaka II.7. This is performed before most śrauta rituals, rites of passage (saṃskāra), and other complicated brahmanical ceremonies.
162. Cf. Kane HDh IV:134ff.
163. Ashok Aklujkar has suggested to me that the reading apastamba could be wrong, that it could be avaṣṭhambha, which can have the meaning “arrogant” or “resolute.” However, the text does not record a pāṭhabheda here, nor do the orthographic possibilities support such a reading.
164. Probably he was a Ṛgvedin, as he quotes the entire sūkta, ṚV 6.24 in his comments on the karṇagraha. This sort of block quotation is unusual in Sanskritic commentarial tradition.
165. The discussion on the navagraha, probably a “new” or “recent spirit” rather than one of the nine planets, refers explicitly to a sexually transmitted disease. This spirit afflicts a man who has had sex with the wife of a friend, an ascetic, a guru, of a master. He is afflicted with a burning penis (meḍhradāhī), among other things.
166. This is the view of Kātyāyana, cited by the MM. Viśveśvara Bhaṭṭa comments that the rice and ghee can be offered separately.
167. Cf. Śāṇkhāyana ŚSū 9.9.5 (cf. Āśvālāyana ŚSū 8.6.13 for a slightly different application). That Viśveśvara Bhaṭṭa would know the ritual applications for this also suggests that he is a Ṛgvedin.
168. The Vidhānamālā (Garland of Ritual), a text probably dating to the nineteenth century, deals with this topic in much more abbreviated and bloodless fashion. Cf., e.g., VM, p. 26, for an account of the karṇagrahapīḍāśamana-vidhānam (ritual that allays affliction by a karṇagraha), p. 27, for the skandagrahapīḍāśamana-vidhānam (ritual that allays affliction by a skandagraha), and p. 28, for the skandāpasmārapīḍāśamana-vidhānam (ritual that allays affliction by a skadāpasmāragraha), and other childhood and planetary grahas.
169. Thanks to Ashok Aklujkar for this suggestion.
170. Shukla 1980:170f. For a lively but fragmentary account of a local North Indian demonology, see Kakar 1982:20-31.
171. Shukla 1980:170f. Also see Parry 1994:230f. for discussion of some of the same material. Note Parry’s statement, which we have countered often here, “that spirit possession is really an oblique aggressive strategy by which the weak and the exploited get back at those who oppress them” (p. 231). An example of a more recent “orientalist” missionary who dealt with this is McClintock (1990), dealt with in Chapter 2. His detailed account of bhūtavidyā draws largely from others’ ethnographies, rather than from his own work in Pakistan in the 1980s. Thus, his account does not situate the material in a specific local tradition or geographic area. He lists five categories of bhūta among Muslims in India (and presumably Pakistan): (1) jinn (a class of beings between angels and men); (2) pari (fairy of entrancing beauty); (3) shayatin or shaitan (devils, demon hosts, fallen angels); (4) ifrit (a powerful and malignant genius, the ghosts of the wicked are sometimes so termed) (5) marid (an evil genius of the most powerful kind) (pp. 39–40). McClintock’s Christian evangelism biases his description and views of other people’s beliefs and practices; it is thus comparatively unhelpful. Although his outlook is more generous than that of Ziegenbalg in the early eighteenth century, he still does not understand academic ethnography. His main source is Crooke 1894. It is valuable that he brought together Crooke’s data in a single organized list; yet to him it serves no purpose other than to vouchsafe the indigenous people’s need for a liberating nonsuperstitious Christianity.
173. This word is uncertain, possibly from būṛnā. Shukla’s transliteration is random, at best.
174. Another problematic word, possibly marua.
175. The word ḍīh refers to a mound on which its shrine is situated. I am grateful to Philip Lutgendorf for assisting me with these Hindi terms and their significance. For another extensive catalogue of local bhūtas and a discussion of their relations with disease, see Prabhu 1977.
177. Maloney suggests, rather, that this is from “Buddha,” but I question this.
178. I have doubts about this etymology.
179. For these descriptions, see ibid.:242–246.
181. Such numbers occur in Tantras as well. For example, see NT 19.62cd-63: “I made thirty trillion (triṃśatkoṭī sahasrāṇām) fearsome Vināyakas with the brightness of blazing fires from my thumb. If someone is impeded by them he will be overcome.” Cf. McHugh 2000.
182. The idea of hordes, gaṇas, acting for good and ill is a persistent and powerful idea, dating from the earliest vedic texts to the adālat of Bālājī described in Chapter 4.
183. For the cakrapūja and tantric sexual ritual, see Bharati 1965:228ff.; Marglin 1985:227f., White 2003.
184. Rāmāyaṇa 3.36.20ab: krīḍāratividhijñānāṃ samājotsavaśālinām; cf. Pollock 1991:79, 163.
187. In the NP, however, this ritual of apparent fertility appears, rather oddly, in the fall, perhaps as a final opportunity for revelry before the Kashmiri winter sets in.
188. See the translations earlier of Ca 6.9.20.8, Su 6.60.15, AH 6.4.30–34b, ĪŚP 42.22ab.
189. See the Kāśīkhaṇḍa of the Skanda Purāṇa (uttarārdha, ch. 54) and trans. Tagare (1997:28–35). See also the Pisācamocanamāhātmyam, published in 1910, which contains an introduction in Hindi describing the Pisācmocan tīrtha and its importance, the SkP chapter, the Śrīkapardīśa Stotra, and instructions for the ancestral rites to be performed at the temple tank (piśācamocane tripiṇdīśrāddhavidhiḥ). See Parry 1994:242ff. for an ethnography of an exorcism at Pisācmocan.
190. Daniel Cohen reports a phenomenon known as pretpañcayat, modeled after the local village governing body called pañcāyat (cf. Chapter 4, on the venue for exorcism at Bālājī modeled on another judicial paradigm, the adālat). The pretpañcayat features competition between exorcists, usually in family disputes, brother against brother. Cohen observes that the primary locus of exorcism in Varanasi is the exorcist, not any temple or specific place, including Pisācmocan. What this means in practice is that exorcisms are generally performed in the home of the client or the “office” of the exorcist, not usually in a public arena like a temple. I am grateful to Cohen for sharing with me his research and insights on the Pisācmocan temple.
191. This term is used by Dārila, in his commentary on Kausikasutra 25.22; cf. Bahulkar 1994:90.
192. Wujastyk 1998:212–235; cf. also Kāśyapa Saṃhitā, pp. 351–374.
193. For a fairly comprehensive discussion of bālagrahas in the major āyurvedic texts and the Agnipurāṇa, see Kumar 1994:284–312. See especially p. 287, table 42, which compares the names of the bālagrahas in different texts (though without explanation). Kumar does not theorize about balagrahas, except to enter into highly suspicious speculations on the diseases wrought by them. For example, affliction by Revatī (“Lady Opulence,” in Wujastyk’s felicitous translation) results in diarrhea, dehydration, anemia, and vitamin B deficiency; Mukhamaṇḍikā afflicts with “Indian Childhood Cirrhosis,” etc. (p. 304).
194. Also bhūtābhiṣaṅga (intimate contact with a bhūta) (Su 6.39.60; Ca 6.3.116; Bhāvaprakāsa 8.1); and bhutopasarga (touched by a bhūta) (BhP 4.29.23). The Śabdakalpadruma cites under āveśaḥ (1:194): bhūtādinā rogaḥ; bhūtasaṃcāraḥ, grahādayaḥ (Rājanighaṇṭu); āsaktiḥ, abhiniveśaḥ (Raghuvaṃśa 5.19).
195. tasyās tu dharma eva nivṛttikāraṇam uktam iti | KS kalpasthānam 6(?).3.7 (Kāśyapa Saṃhitā, p. 352).
196. To take one further example from the later āyurvedic tradition as a case in point, see Aḍhamalla’s Dīpikā on the varieties of bālagraha presented in the thirteenth-century Śārṇgdhara Saṃhitā. Aḍhamalla states that “seizers [graha] beginning with skandas possess children” (skandagrahādayo bāleṣv āviśanti). Thus, by the time of Śārṅgdhara, followed by Aḍhamalla in the fifteenth century, graha was defined in terms of āveśa (rather than the more expected praveśa). Śārṅgadhara lists twelve varieties of bālagraha: tathā bālagrahāḥ khyātā dvādaśaiva munīśvaraiḥ | skandagrāho viśākhaḥ syāt svagrahaś ca pitṛgrahaḥ | naigameyagrahas tadvac chakuniḥ śītapūtanā | mukhamaṇḍitikā tadvat pūtanā cāndapūtanā || revatī caiva saṃkhyātā tathā syāc chuṣkarevatī | (1.7.189–191b).
197. On Khaḍgarāvaṇa and the lost Khaḍgarāvaṇabālacikitsā, see Goudriaan 1977. On Khaḍgarāvaṇa as a form of Śiva, also called Caṇḍeśvara, Rudra, and Caṇḍāsidhārāpati (Lord of the Edge of the Fierce Sword), see Bühnemann 1999:325. On the importance of Rāvaṇa in this context, see Zysk (1985:67) for discussion of the Rāvaṇaproktabālācikitsāsūtra, a pediatrics text; also Filliozat (1937:159–178), on the Buddhist transformation of Rāvaṇa into a peaceful king with healing powers. Filliozat’s volume is an expansion of his earlier article (1935). The Kumāratantra draws heavily from the MBh, AV, Caraka, and Suśruta, all of which Filliozat discusses in detail. The Nārāyaṇīya mentioned here is undoubtedly the Tantrasārasaṃgraha, known in Kerala as the Viṣanārāyanīya. For this text, see below.
198. The text prescribes the manner in which the mantras are to be recited. The terms tārapūrvaṃ namo ’ntam and praṇavādinamo ’ntam (beginning with o
and ending with namaḥ) or svāhāntaḥ (ending with svāha) often indicate the recitation of a full mantra or series of mantras. This is irrespective of whether the mantra actually ends with namaḥ or svāhā. Tāra is equivalent to praṇava while ṭhaṭha indicates svāhā.
200. On Rajasthan (and rural UP), see the video Kusum (see Chapter 4, n. 17). The healer, who was not an āyurvedic physician, gave the father of the possessed girl (Kusum) a handful of packets of an undisclosed substance with instructions to burn them regularly in front of the girl.
201. I am grateful to Vaidyamadham Cheriya Narayanan Nambudiri for this information.
202. On aṣṭavaidya physicians, see Zimmermann 1987:212–213, and below.
203. This sort of “hot” substance is a tantric innovation; it is not commonly offered to Śiva, generally considered a “cool” or ascetic deity.
204. For this reason neem twigs have long been a common form of toothbrush in India and are still widely used in the countryside.
205. In modern-day India, sprouts of various kinds are germinated on a small raised plot of ground inside the house as part of the ritual of the nine-day festival to the goddess (Durgāpūjā) in the month of Asvina (September–October). It is likely that this is a very old practice.
206. Cf. Meulenbeld HIML IB:237n6. Meulenbeld does not suggest a botanical name or identification for this plant.
207. For closer examinations of the KausSu on this, see Bahulkar 1994:109–110, Bloomfield 1897:263–264, and Zysk 1985:29–31.
208. Chapters 11–16 deal with bālagrahas and unmāda. This text is in print and can thus be compared to the account in the ĪŚP also see Meulenbeld HIML IIA:456–457, IIB:468–469.
209. Mother of the dark graha Rāhu, daughter of Dakṣa or Kaśyapa and wife of Kaśyapa or Vipracitti; cf. MBh 1.59.12,30.
210. This bija is exclaimed at the funeral pyre when the skull cracks and the jīva exits the body.
211. Many of the words here (e.g., mañja, kaḍhḍha, and ṭhaṭha) are incomprehensible and would be accepted for their mantric value alone were it not for the fact that this mantra reads rather differently, and more comprehensibly, in the text of the Nārāyaṇīya, viz. the Tantrasārasaṃgraha (pp. 158–60), demonstrating that the ĪŚP had either corrupted it or was itself in possession of a corrupt version. The mantra in the Nārāyaṇīya reads: oṃ namaḥ sarvamātṝṇāṃ hṛdayaṃ moṭaya bhañja paṭa sphoṭaya sphura gṛhṇa ākaṭu troṭaya | evaṃ siddhiṃ jñāpayati | hara nirdoṣaṃ kuru bālakaṃ bālo vā sarvagrāhīṇām upakramyatu | (Om, salutations to the heart of all the Little Mothers. Crush! Break! Rip! Flash! Spring forth! Possess! ākaṭu [You whose pungent odor comes forth(?)]. Break apart! One should announce this clearly. O Hara, make this child faultless or let the child approach all the possessing demonesses!) (The commentator advises doubling the impv. 2 sg. forms, e.g., moṭaya moṭaya, bhañja bhañja, consistent with what we have seen in the South Indian manuscripts examined in Chapter 11.) The text continues: oṃ cāmuṇḍe namo divye hraṃ hrīṃ apasarantu duṣṭagrahāḥ | hrūṃ tathā gacchantu guhyakāḥ | anyatra sthāne oṃ rudro jñāpayati ṭha ṭha | One of the striking features of these mantras is the extensive use of retroflex consonants, e.g., kaṭa, ṭhaṭha, troṭaya. This could be due to a brahmanical assignation of these mantras to lower-class, non-Sanskritic strata of society, in this case an assumption or recognition of Dravidian or tribal origins, as these languages have much more retroflexion than Sanskrit. For somewhat similar speculations regarding the secret speech of Himalayan tantras, see Davidson 2002:267ff.
212. Cf. the discussions in Chapter 4 of the work of Srinivas, Mumford, Pakaslahti, Dwyer, and Vogt, which provide very different models of healing processes associated with possession.
213. For a review of scholarship on the term yukti see Engler 2003:426, 428. Engler describes it as “empirically based inference,” drawing principally on CaSū 11.21–25 and arguing against its modern interpretation as scientific reasoning.
214. See F. Smith 1987 for an idea of how the vedic sacrificial rituals varied within traditions defined by affiliation with a single Śrautasūtra, such as Apastamba or Baudhāyana.
215. The eight powers (guṇāḥ) are probably the eight archetypal siddhis; see note 85.
216. SuUt 60.21cd: ye tv āviśantīti vadanti mohāt te bhūtavidyāviṣayād apohyāḥ|| Though Caraka says that grahas are innumerable (Ca 6.9.21.8, above, p. 490; also Su 6.60.22, above p. 492), he does not attribute the possession to lesser spirits in a hierarchy of fiends.
217. See Stephens for accounts of sex with demons as an “aspect of the more general European fixation on physical interaction with demons” (2003:14). Agreements were contractual, including the sale of one’s soul to the devil. This idea does not appear to have resonance in India, where such sexuality is more prominently formalized and medicalized. Nor did this lead to crises of belief among the Hindu clergy, as appears to have been the case among Christian clergy in Europe.
218. This is not so far-fetched as it may seem at first glance. See Slater 2003 on the work of the psychologist David Barlow of the Center for Anxiety and Related Disorders at Boston University. Slater writes:
Barlow’s method for treating anxiety disorders is surprisingly simple, although its philosophical and clinical implications are anything but. He aims to reduce anxiety not by teaching customary relaxation techniques involving calming mantras or soothing imagery, but by doing just the opposite: forcing the patient to repeatedly face his most dreaded situation, so that, eventually, he becomes accustomed to the sensation of terror. Barlow claims he can rid some people of their symptoms in as little as five to eight days. His treatment promises to be psychotherapy’s ultimate fast track, but while many clinicians praise its well-documented results, others take a dimmer view of what one clinician calls “torture, plain and simple.” (p. 34)
There can be little doubt that the extreme forms of psychological treatment recommended by Caraka et al. were as controversial in their day as Barlow’s are today, testifying, at the very least, to the composite nature of the āyurvedic compendia.
219. CaCi 6.9. 31cd: vismṛter hetor nayanti prakṛtiṃ manaḥ; cf. also 6.9.79–98.
220. On daivavyapāsrayacikitsā, see, in addition to the above, AHCi 1.171, 5.83–84; ASaṃSū 9.76; cp. Hārīta 3.3.2–3. Rasaratnasamuccaya 22.60–69; Chattopadhyaya 1979:314–320.
221. It is also possible that this term referred to internal practices, such as yoga, which do not require priestly or other formal mediation. Neither Caraka nor Susruta mentions yogic or Buddhist practices, to which sattvāvajaya might refer. Such practices might have been expunged from these texts in an attempt to enhance their scientific credibility. Their empiricism would have admitted observable or inferred phenomena, which to them included at least some ritual, use of gemstones, etc. By the same token, possession was admitted as a valid category, legitimated by repetition of characteristic behavior. For the same reason, sattvāvajaya, which must have been an inherited category with which they were uncomfortable, was relegated to the scantiest offhand mention. On the word sattvāvajaya, see references in note 2.
222. Oracular possession also may be characterized as a vāta disorder. Knipe observes that most goddess ritualists in coastal Andhra Pradesh are “by nature excitable, sometimes volatile and raucous, their monologues punctuated with belches, yawns, hiccups and other marks of incipient possession” (2004:440). All of these are characteristics of vātaprakopa (affliction due to vāta imbalance).
223. dṛṣṭaṃ spṛṣṭam athāghrātaṃ tad dhi sarvagrahāpaham; CaCi 6.9.62cd.
224. I am grateful to Robert Svoboda for this information (e-mail communication, March 4, 2003).
225. See the comments of Raman 1991:1:383. He states, on PrMā 12.62, that the proportion of dung, urine, milk, and yogurt should be 1:2:16:5. Raman does not note the proportion of ghee, but presumably the mixture should consist predominantly of ghee, the quantity of which might remain unspecified. Raman does not state his source. Vaidyamadham says that the ghee should measure four times the quantity of yogurt.
226. Cf. CaCi 6.9.50–52, for dog’s bile (śunaḥ pittam), beak and feather, probably incinerated, from inauspicious birds including the owl, vulture, and crow. These were likely recommended for their native properties that could contribute to an atmosphere inimical to bhūtas.
227. Recall that the mixing of animal with vegetal substances goes back to vedic times. Note in particular that surā, a kind of multigrain wine, produced for the Sautrāmaṇī sacrifice, is mixed with milk and the hair of a lion, wolf, and tiger (cf. Kane HDh II.2:1224–26). The Śatapatha Br. 12.7.3.19–21 (Eggeling 1900:229–230) demonstrates that the mythic idea is to include potent elements of wildness (raudra), because this is for Rudra. A similar notion of integrating different aspects of nature is presented here, the purpose of which is to present a multilevel front against offending bhūtas or grahas.
228. For recipes from the Tibetan Gyu-zhi, see Clifford 1984:88.
229. For a summary of the text, see Meulenbeld HIML IIA:86–90 and notes.
230. For dāruharidrā, elsewhere called dārunisā, see Meulenbeld HIML IIB:305n900, given as Berberis aristata DC. Like the barberry in Europe and the Oregon grape in North America, it is more a bush than an actual tree and possesses at least some of the medicinal properties of turmeric (haridra). The reason for preparing this under puṣya nakṣatra is not stated, but possibly the name puṣya itself, “prosperity,” is sufficient.
231. Vacā (Acorus calamus Linn.) is a widely used root that is dried and chewed to alleviate the discomforts of a sore throat. As mentioned earlier, it is also recommended for the enhancement of mental clarity. It is “hot” in the sense that it induces the release of heat in the body.
232. According to two of my āyurvedic informants, this is included because the cat has the capacity to recognize a bhūta, and the dried and ground up excrement captures that quality. Whether this kind of associative thinking was also that of the authors of the classical texts is uncertain.
233. The text says kesa (hair). However, according to my āyurvedic informants, this means jaṭāmāṃsī, a cooling and sweet root that is easily acquired and is often used in havan (tantrically inspired fire ritual). The word jaṭā means “dreadlock,” specifically those of sādhus, grown in the image of Śiva. Thus, jaṭāmāṃsī may be viewed as a kind of keśa.
234. Priya Vrat Sharma, the editor and translator of the published text of the Cakradatta, translates pakvaindrī, as “ripe fruits of indravāruṇī mixed with cow-urine and ghee,” following the commentary (pakvaindrīphalamūlajam sājyam).
235. In fact, mayurapiccha does contain trace elements of copper and gold, which was apparently known by classical practitioners of tantric medicine. Baidyanath, the well-known Indian pharmaceutical firm, manufactures an antispasmodic under the trade name Mayuracandrikā Bhasma, in which the central ingredient is mayurapiccha. The indications on the container state: “Antiemetic. Relieves hiccough, asthma, and vomiting.” All of these are vāta conditions. Similarly, sarpanirmoka (snake skin) is recommended today as an ingredient in a dhūpa meant for alleviating piles—another vāta affliction.
236. Some vaidyas have altered their views, however, under the influence of Western scientific and psychological theory. Dr. Raghavan Thirumulpad of Chalakudy (Thrissur District), the most prolific modern author of āyurvedic texts in Malayalam, explained to me that bhūtonmāda is always the result of one’s mistaken or non-dharmic actions, prajñāparādha, a word taken from the āyurvedic texts (cf. above AH 6.4.4), which I have often heard vaidyas (and many others, including a variety of vedāntins) use as a descriptor for moral transgression. In his view, bhutonmāda is comparatively rare today, because of changes in contemporary belief systems. This, he says, is an artifact of outdated cultural notions. This view expresses the voice of modern, Western-influenced Sanskritic culture. On his diagnostic approach to bhutonmāda, he said: udāharaṇamātraṃ tu budhimatāṃ, yadoktānusāram evāśrayo maṇḍabudhīnām (An intelligent physician is free to design a clinical approach to each problem, irrespective of the texts, but a slow-witted one strictly adheres to the letter of the texts). This has apparently been the practical approach of Āyurveda for millennia.
237. Shadows of this might have once been found in Buddhism, where Āryaśūra, in his Jātakamālā, refers to experts in bhūtavidyā as Siddhavidyās; cf. White 1996:58.
238. For an account of the possession at the Chottanikkara “Bhagavathy” temple near Tripunithura in the eastern suburbs of Cochin/Ernakulam, see F. Smith 2004.
239. The institution founded by P. S. Varier, the Arya Vaidya Sala, maintains its home base in Kottakkal, but has dispensaries and pharmaceutical outlets throughout India. According to one of my informants familiar with the Arya Vaidya Sala educational system, this institution has trained more than four hundred non-Hindu vaidyas, most of them Muslims. I have also encountered a number of non-Brahman and Christian vaidyas trained at the aṣṭavaidya medical colleges. The education in the latter institutions is still conducted by Nambudiris, some of whom know the entire AH by heart. I have met a few non-Brahman Hindu pandits in the Varier tradition who have also memorized the entire AH. I mention this to counter the common assertion that only Brahmans study Sanskrit and only Hindus study Āyurveda. In both Kerala traditions, however, I found the non-Hindu vaidyas considerably less skilled because they did not have access to the Sanskrit texts.
240. A measure of his prestige is that he also bears the title sālāvaidyam. This title indicates that he has the unique eligibility (adhikāra) to sit in a vedic sacrificial arena (yāgasāla) during the performance of a soma sacrifice. In Kerala a doctor is required in a sacrifice, and only a sālāvaidyam is qualified. Vaidyamadham is a follower of the Dakṣiṇāmūrti temple at Shukapuram in Mallapuram District. Historically, he says only members of his family have served as sālāvaidya. Frits Staal does not mention this position in Agni, his massive 1983 study of a vedic agnicayana performed in central Kerala in 1975. Thus, it is doubtful that this tradition is or has been in general currency throughout Kerala. In order to confirm his eligibility for the position of sālāvaidyam, at one time he had to spend forty-one days at this temple “doing dhyāna and bhajan,” meditating and singing hymns to the deity. In addition, he says, yajñikas or sacrificial patrons (āhitāgnis) assemble at this Dakṣiṇāmūrti temple to practice for their sacrifices, as Dakṣiṇāmūrti is “the deity for right thinking.”
241. The botanical identity of this herb is disputed, as several plants have been identified with this name. One possibility is Convolvulus pluricaulis. Another is Canscora decussata, reportedly used for śaṅkhapuṣpi in Bengal and parts of north India. A third is Clitoria ternatea, reportedly used for śaṅkhapuṣpi in Kerala. In any case, śaṅkhapuṣpi is said to increase sattva (brightness, mental clarity, positive qualities) and was prescribed for use in brāhmighṛta and other tonics for the nervous system. (I am grateful to Robert Svoboda for this information.) For similar information, see AH Uttarasthana 1.43, 6.24–25, 7.24, 39.44–51. It has also been called viṣṇukrānta, probably denoting that parts of the plant were used in Hindu worship. See www.thehimalayadrugco.com/herbfinder/h_evolvulus.htm; for the results of research on it, see http://ccras.org/publication/clinical&experimental_1.htm.
242. This is the family of P. Nilakanthan Nambudiri, with whom I also spoke at length. Their burgeoning practice is limited to the treatment of mental disorders. A large majority of their clients are Muslim girls, as Mallapuram District, listed in the 1991 census of India at 68 percent Muslim, now probably has an even higher proportion (this information is missing in the 2001 census). The family of Nilakanthan has a residential facility and empowers their medicines with mantras and puja, all of which are tightly guarded family secrets. For a more complete description of the Poomkutil treatment center, see Unnikrishnan 1998.
243. For more on this bhūta, see Tarabout 2003.
245. For all of this, see the rather hurried and unreliable edition and translation by Raman (1991). It is to a great extent this systematic elegance (or elegant systematics) that makes astrology (jyotiṣa) so attractive in India to this day.
246. See Brahmānand Tripāṭhī’s note on this in his Hindi commentary on the AH titled Nirmalā: āyurved kā jyotiṣaśāstra se ghaniṣṭh sambandh hai (Ayurveda’s relationship with the sastra of astrology is of the closest kind).
247. Trans. Keith 1914:118.
248. The connection between astrology, tantric ritual, and Āyurveda is found elsewhere in India; it is not limited to Kerala. Another text that is much more expansive on this than the PrMā is the Vīrasiṃhāvaloka, composed in North India, possibly in Gwalior, dated 1383 (see Meulenbeld HIML IIA:229–230 and notes). This text, never the subject of serious academic study, deserves far more attention than I can give it here. It purports to be a jyotiśśāstrakarmavipākāyurvedaprayoga, a term that appears in the colophon to each of its seventy-one chapters. It is a handbook discussing the astrology (jyotiḥ) and karmic causes of disease (karmavipāka, see the above discussion on the Madanamahārṇava), and at length describes medicinal treatments for more than fifty diseases. Curiously, the astrological significations are absent from the account of unmāda, though they are present in the account of apasmāra. The bulk of the section on unmāda consists of recipes for medicated ghees, powders (cūrṇa), and fumigants (dhūpa). It is possible to argue, drawing from Langford’s view of Āyurveda as a cultural commodity in the West today (2002:263ff.), that the four of these taken together—astrology, medicine, ritual, and karmavipāka—have constituted a form of local knowledge and correlative local commodity for more than a thousand years. For more on the situation at present, see Friedman 1986; Perinbanayagam 1982; Pugh 1983.
249. Clifford 1984:157–158. Kakar reports the following on Tibetans in exile in India: “If the patient is possessed by a bhūta, his eyes move downward toward the left; in the case of possession by a dakini—a female spirit often addressed as ‘Mother,’ who is said to bestow esoteric knowledge and power on its devotees—the eyes move upward toward the right” (1982:110).
250. Cf. Meulenbeld HIML IIA:348–52. This text is widely cited in Maharashtra as an authority on strīroga, (diseases of women), especially regarding questions of childbirth.
251. This is from the first few pages of the YR, pp. 12–14 of the Varanasi edition, verses 1–3, 11–20, 22.
252. The same vaidyas prescribe pearl (muktā) and silver (rajata) bhasmas for nija unmāda.
253. The nine gems are: Sun–ruby (māṇikya), Moon–pearl (mauktika, muktā), Mars–red coral (pravāla, vidruma), Mercury–emerald (gārutmata, garudodgāra), Jupiter–yellow sapphire (puṣparāga = topaz), Venus–diamond (vajra, hīra, or often simply ratna), Saturn–blue sapphire (indranīla), Rāhu–hessonite or onyx (gomeda), Ketu–cat’s eye (vaiḍūrya). For variations on these gems and how to prepare them, see the sixteenth-century encyclopedia Āyurvedasaukhya 4.649–701. For information on gems in ancient India, see the references cited in Meulenbeld HIML IIB:742–743nn.389–403. Navaratna bhasma is, needless to say, very expensive. It is, however, available.
254. Cf. Meulenbeld HIML IIA:581–787 (Part 10), works on rasaśāstra and ratnaśāstra.
255. An edition of this was published in the 1990s in Lhasa. See Byams-pa ’Phrinlas and Wang Lei n.d.:32, thangka 63 and description, pp. 402–405, and thangkas 66–68, pp. 33–34, descriptions, pp. 430–443. This extraordinary text deserves much closer work than that volume gives it.
256. One of these is reproduced schematically in Clifford 1984:158; two are reproduced in Tsarong et al. 2001:116.
258. Tsarong et al. 2001:118.
259. While virtually all āyurvedic compendia of the past few centuries have an obligatory section on bhūtavidyā, this was not always the case in the first millennium. For example, the Siddhasāra of Ravigupta (from Kashmir, c. 650 C.E.) does not. See Siddhasāra:2:307–313, ch. 20, on unmāda, Āveśa is not mentioned. Ch. 29, called kumāratantra, is somewhat more relevant in its concern with pharmacopeia. The Siddhasāra was translated into Tibetan, Uighur, and Khotanese (at least). This text includes Ronald Emmerick’s edition of the Sanskrit and Tibetan versions.
260. However, see White 2003:35ff., and above Ch. 6 note 52.
261. See Reddy 2000 and Langford 2002 for illuminating theoretical accounts of the modernization of Āyurveda.