Parpola - Vedic time human sacrifice

VIII. HUMAN SACRIFICE IN INDIA

IN VEDIC TIMES AND BEFORE

Asko Parpola

Human sacrifice in the Veda has been the object of scholarly study for two hundred years now. It was first discussed in the pioneer ing article of 1805 by Henry Thomas Colebrooke, who knew horse sacrifice (aśvamedha) and human sacrifice (puruṣamedha) at first hand from the Vājasaneyi-Saṁhitā and the Satapatha-Brāhmaṇa.¹

For an overall account of Vedic religion see H. Oldenberg, Die Religion des Veda (Berlin, 19172); A.B. Keith, The Religion and Philosophy of the Veda and Upanishads, 2 vols (Cambridge, Mass., 1925); J. Gonda, Die Religionen Indiens, I: Veda und ālterer Hinduismus (Stuttgart, 1960, 19782); J.C. Heesterman, Vedism and Brahmanism’, in M. Eliade (ed.), The Encyclopedia of Religion, 15 vols (New York, 1987) 15.217-42. Sac rificial rites occupied a prominent position in Vedic religion. The earliest text collec tion, the R̥gvedal-Saṁhitā), attests to the worship of gods with recited and chanted hymns and offerings of an invigorating drink called soma; see T. Oberlies, Die Reli gion des R̥gveda, 2 vols (Vienna, 1998-99). The next oldest text collection, the Athar vaveda(-Saṁhitā), attests to a rather different ritual background, with emphasis on royal rites (with bloody sacrifices), rites of ‘white’ and ‘black’ magic as well as domestic ceremonies; see M. Bloomfield, The Atharva-Veda and the Gopatha-Brāhmaṇa (Strassburg, 1899). In subsequent literature (the later Saṁhitās, the Brāhmaṇas and the śrautasūtras) we meet a very complex srauta ritual, with hundreds of rites divided into different categories according to the sacrificial substance (vegetable, animal or soma offerings or combinations of these) and their duration (from one day up to a thousand years), and with a varying number of specialized priests (as many as sev enteen); see A. Weber, “Zur Kenntniss des vedischen Opferrituals (1-11)’. Indische Stue dier 10 (1868) 321-96 and 13 (1873) 217-92; A. Hillebrandt, Ritual-Litteratur, bedische Opfer und Zauber (Strassburg, 1897); F. Staal (ed.), Agni: The Vedic ritual of the Fure Altar, 2 vols (Berkeley, 1983). The relatively simple domestic (grhya) rites are discussed at greater length only in the youngest category of Vedic texts, the Gr̥hyasūtras; see J. Gonda, Vedic ritual (non-solemn rites) (Leiden, 1980).

158

According to him, the Veda teaches their performance ‘as emblematic (that is, symbolic) ceremonies, not as real sacrifices’ (p. 61), and hence ‘human sacrifices were not authorized by the Véda itself’ (p. 61). This is evident from the fact that in the purusamedha,

‘a hundred and eighty-five men of various specified tribes, characters, and professions are bound to eleven posts; and after the hymn concerning the allegorical immolation of Náráyana has been recited, these human victims are liberated unhurt; and oblations of butter are made on the sacrificial fire.’

159

The issue has since been debated on a much broader textual basis than was available to Colebrooke. In 1864 Albrecht Weber published an impressive study on human sacrifice in India in Vedic times, republished in an enlarged version in 1868.³ Weber offers an exhaus tive and accurate description of all places in Vedic literature known to him at that time that contain material related to human sacrifice. In 1876, Rājendralāla Mitra published a well-reasoned paper on the subject, in which he defended the view that human sacrifice has been a reality in India, in both early and later times. New textual material of high interest was added in 1926-28 when Willem Caland published extracts from the previously unknown Vedic text Vādhūla Sūtra. It is obvious that the whole discussion and its results are hardly known to many contemporary scholars writing on this subject. In a widely used textbook on Hinduism published for the first time in 1996, for example, we find altogether three statements on this topic that do not show much progress beyond Colebrooke:

There was also a human sacrifice (puruṣamedha) modelled on the horse sacrifice, though the human victims were set free after their consecration ‘Indeed the human sacrifice, the sacrifice of the ‘great beast’ (mahā paśu), is regarded in the Veda as the highest sacrifice, even though human sacrifices may never have actually taken place.’

We do possess texts which refer to a human sacrifice in the Indian traditions, but such a practice may never have actually occurred, existing only as an ideal or possibility.

• G. Flood, An Introduction to Hinduism (Cambridge, 1996) 41, 184, 218, respec tively.

160

The last statement is factually wrong. Early on, William Crooke, in addition to providing numerous textual references to human sacrifice, especially to the goddess Durgā and her multiforms as well as to demoniac village godlings in the Indian narrative literature, found copious evidence of actual human sacrifices available in ethno graphic records and official documents.

One important reason for statements like the ones quoted above is probably the criticism leveled by such an authority as Hermann Oldenberg in his very influential handbook of Vedic religion in 1894 and its revised edition in 1917. Oldenberg states that to his knowledge there is no sure evidence for the existence in Vedic India of a genuine human sacrifice. However, Oldenberg had his own narrow definition for this term and other reservations. ⁴

A short summary of the practice of human sacrifice in the Veda was offered in 1960 by one of the best twentieth-century experts of Vedic and Hindu religion, Jan Gonda, who underlined its being a debated issue. And in any case, as he noted, human sacrifice is no longer topical in the śrauta ritual. This being the only treatment of the subject in his handbook, one gets the impression that Gonda did not consider it a matter of great importance.⁵

161

Lack of space does not allow me to cover many points in this chapter. Most of the debate has centered on isolated text passages, and concrete textual evidence might indeed be of great interest. I therefore discuss first a few important textual references and their interpretation, hoping to establish beyond reasonable doubt that Vedic texts do indeed attest to real human sacrifices performed within the memory preserved by the authors, and that by the time of the Brāhmaṇa texts, the actual practice of bloody offering had already begun to diminish. Then I underline the direct association of human sacrifice with some pivotal concepts of Brahmanical religion - cosmic man and his sacrifice. I conclude with a few words on the wider context of human sacrifice in the Veda and my own views in this regard. The alleged archaeological evidence of Vedic fire altars is not taken up here, as it is ably discussed and criticized by Hans Bakker in the next chapter of this volume.

1. The Śunaḥśepa legend

In 1852 Horace Hayman Wilson published an article ‘On the Sacri fice of Human Beings as an Element of the Ancient Religion of India.’⁶ In this paper Wilson translated the Vedic variant of the Śunaḥśepa legend as preserved in the then unedited Aitareya Brāhmaṇa, which he sagaciously placed around 700-600 BC, in between the R̥gveda and the later epic Rāmāyana. For the benefit of the readers of this book who may not be acquainted with this text, some excerpts from Wilson’s translation follow:⁷

Hariśchandra [Hariscandra) the son of Vedhas, was a prince of the race of Ikshwáku [Ikṣvāku]: he had a hundred wives, but no son. … [Sagel Nárada [Nārada) advised Harischandra to pray to Varuṇa for a son, promising to present him as an offering to that divinity. So be it,’ said the prince; and repairing to Varuṇa he said: ‘Let a son be born unto me, and with him, I will sacrifice to you.’ - ‘So be it,’ said Varuṇa, and a son was born to the king, who was named Rohita. ‘A son has been born to you,’ said Varuṇa, ‘sacrifice with him to me.’-‘An animal, replied the king, ‘is fit for sacrifice only after ten days from birth. When the term of purification has passed, I will sacrifice to you.’ - ‘Very well,’ said Varuṇa.

162

The ten days expired, and Varuṇa said, Now sacrifice with him to me.’ … [Four more times the king asks for, and is granted, a postponement. The youth grew, and was invested with arms; and Varuṇa said, ’now sacrifice to me with him.’ The king replied, ‘Be it so.’ But he called his son, and said, ‘My child, Varuṇa gave you to me, and I have also promised to sacrifice with you to him.’ - ‘By no means,’ said the youth; and taking his bow, he set off to the forest, where he wandered for a twelvemonth.

Upon Rohita’s disappearance, Varuṇa inflicted the descendant of Ikshwáku with dropsy; which when Rohita heard he set off to return home. On the way he was met by Indra in the shape of a Brahman (who urged him to wander on. This happened four more times, so that eventually] Rohita returned for the sixth year to the forests. Whilst wandering thus in the woods he encountered the Rishi (Rși, Sage] Ajigartta (Ajīgarta], the son of Suyavasa, who was distressed through want of food. He had three sons, Śunahpuchcha [Śunahpucchal, Śunahsephas (Śunaḥśepa], and Śunalángula [Sunolāngüla). Rohita said to him, ‘Rishi, I will give thee a hundred cows for one of these thy sons, that by him I may redeem myself.’ But the Rishi, taking hold of the eldest, said, ‘Not this one;’ No, nor this one,’ said the mother, securing the youngest; but they both agreed to sell the middle son Śunahsephas, and Rohita, having paid the hundred cows, took the youth and departed from the woods. He proceeded to his father and said, ‘Rejoice, father, for with this youth shall I redeem myself.’ So Hariśchandra had recourse to the royal Varuṇa, and said, ‘With this youth will I sacrifice to you.’ And Varuṇa replied, Be it so - a Brahman is better than a Kshatriya [Kṣatriya];’ and thence directed the king to perform the sacrificial ceremony termed the Rajasúya (Rājasūya, the royal consecration); and he, on the day of initiation, appointed Śunahśephas to be the human victim.

At that sacrifice of Hariśchandra, Viswamitra (Visvāmitra) was the Hotri [Hot] or reciter of the Rich (Rc); Jamadagni, the Adhwaryu (Adhvaryu], or repeater of the Yajus …; but they had no one who was competent to perform the office of binding the victim, when consecrated, to the stake, whereupon Ajigartta said, If you give me another hundred cows I will perform the duty;’ and they gave him the cows, and he bound the victim. But for the victim thus consecrated and bound, sanctified by the divinities of sacrifice, and thrice circumambulated by the priests bearing burning brands of sacred grass, no immolater could be found, when Ajigartta again offered himself, saying, ‘Give me another hundred cows and I will immolate him;’ accordingly they gave him the cows, and he went forth to sharpen his knife. In this interval Śunahśephas reflected,

163

‘These [people] will put me to death as if I were not a man but an animal; my only hope is the aid of some of the gods, to whom I will have recourse.’ [Śunaḥśepa now prays to various gods with Rgvedic verses quoted here by their initial words. While he was repeating the three concluding stanzas], his bonds fell off, and he was set free; and the king, the father of Rohita, was cured of his complaint … [The legend ends with Ajigarta’s claim to have his son back, the rejection of Ajīgarta by Śunahśepa, and Śunaḥśepa’s eventual adoption by Viśvāmitra, who curses the eldest 50 of his 100 sons who refused to acknowledge Śunaḥśepa’s primogeniture, so that these rejected sons became the ancestors of various despised tribes.]

The legend had been known from a later version in the first book of the Rāmāyana, but that version ’leaves it doubtful whether an actual sacrifice of the victim, or one only typical [= symbolical], is intended, while there is no question of its purport as it is found in the Aitareya Brāhmaṇa’ and ‘it may be received as authority to a qualified extent for … the sacrifice, on particular occasions, of human victims.’ In 1859, Max Müller published a new translation of the legend and edited its text with the variant readings of the Śānkhāyana-Śrautasūtra.⁸ In his opinion, the legend ‘shows that, at that early time, the Brāhmans were familiar with the idea of human sacrifices, and that … Brāhmans were ready to sell their sons for that purpose.’⁹

Wilson compared Hariscandra’s vow with Abraham’s readiness to offer up his son,16 noting that the purport of the divine command could hardly have been wholly unfamiliar to him, as a similar sort of sacrifice occurs in later Jewish history in the vow of Jephtha (Judges 11). The Jews borrowed ’the offering of children to Moloch … from their idolatrous neighbours.’[^17] Müller retorted: ‘it does not necessar ily follow from this legend that the Rishis, the authors of the Vedic hymns, offered human sacrifices’ for ’no one would conclude from the willingness of Abraham to sacrifice his own son in obedience to a supposed command from Jehovah, that the Jews had been in the habit of offering their sons as victims.’¹⁰

The older version of the Aitareya Brāhmaṇa explicitly quotes the seven hymns attributed to Śunaḥśepa in the R̥gveda (1,24-30); the later version of the Ramayana refers to ‘sacred verses’. However, to Friedrich Rosen, who edited the first book of the R̥gveda in 1838, the Śunaḥśepa hymns, ’except in one or two doubtful passages, bore no relation to the legend of the Ramayana, and offered no indication of a human victim deprecating death.’¹¹ A similar opinion was expressed by Rudolf Roth, who also published on the Śunaḥśepa legend of the Aitareya-Brāhmaṇa. Roth concluded that the features central to the legend came into being only after the R̥gveda, and that they are of a didactic nature. The story is ethical and directed against the gruesomeness of human sacrifice ¹² According to A.B. Keith too, ’the whole story reveals the slaying as proposed as something utterly monstruous … it is enough to show that human sacrifice was for the Brāhmaṇa period a horror beyond words.’¹³ Rājendralāla Mitra, however, does find indications of human sacrifices in the R̥gveda, and regards these as pointing to a real and not a merely symbolical sacrifice.¹⁴ According to Julius Eggeling, the Rgvedic verses 1,24,11-13 and 5,2,7 contain the earliest reference to the story of Śunaḥśepa. The verse in the fifth book, addressed to Agni, says:

‘Even Śunaḥśepa, who had been bound for the sake of a thousand (cows), didst thou let loose from the stake, for he had already been prepared (namely, for the sacrifice)! '

This legend, so far from bearing witness to the existence of human sacrifices as a generally recognized practice at the time when it originated, would rather seem to mark this particular case as an exceptional one.+++(5)+++¹⁵

165

Weber thought that the royal consecration (rājasūya), where Śunahśepa was supposed to be sacrificed, originally contained such a sacrifice, but the ritualized recital of the legend during the rājasūya is its only reminiscence ¹⁶ Interesting proposals about the meaning and function of the legend at the royal consecration have been proposed by Jan Heesterman, and we shall return to them presently.¹⁷ Unfortunately, space forbids following the discussion of the Śunaḥśepa legend further.¹⁸ Yet the principal arguments for and against the legend’s value as evidence for human sacrifice should be apparent from the early studies quoted above.

2. From human and bloody sacrifice to vegetable offerings

Wilson found vicarous sacrifice evidenced by the Śunahśepa story, in which ‘one human victim is substituted for another, whilst in the parallel cases of antiquity the substitutes were animals. ¹⁹

Although he was sceptical about persistence from the R̥gveda, Max Müller saw no reason to doubt the previous existence of human sac rifice suggested by the Śunaḥśepa legend. In his opinion the Aitareya Brāhmaṇa offers a striking confirmation of this.

It is said there (Ait.-br. 6.8) that the gods took man for their victim.

“As he was taken, medha … went out of him. It entered the horse. Therefore the horse became the sacrificial animal. Then the gods took the horse, but as it was taken, the medha went out of him.
It entered the ox. Therefore the ox became the sacrificial animal. The same happened with the ox.
Afterwards the sheep, then the goat, and at last the earth became the victim.
From the earth rice was produced and rice was offered in the form of purodāśa, in lieu of the sacrificial animal. The other beings, which had formerly been offered and then been dismissed, are supposed to have become changed into animals unfit for sacrifice; man into a savage, the horse into a Bos Gaurus, the ox into a Gayal ox, the sheep into a camel (ushtra), the goat into a Sarabha. All these animals are amedhya or unclean, and should not be eaten.”

Müller comments:

The drift of this story is most likely that in former times all these victims had been offered. We know it for certain in the case of horses and oxen, though afterwards these sacrifices were discontinued. As to sheep and goats, they were considered proper victims to a still later time. When vegetable offerings took the place of bloody victims, it was clearly the wish of the author of our passage to show that, for certain sacrifices, these rice-cakes were as efficient as the flesh of animals. He carries out his argument still further, and tries to show that in the rice the beard corresponds to the hair of the animal: the husk to the skin; the phalikaranas to blood; the meal to the flesh; the straw to the bones.[^28]

Jan Houben has recently discussed in detail the embarassment about violence in sacrifice to which the texts of the Middle and Late Vedic period attest, finding it to agree with the general trend towards non violence and vegetarianism that can be observed from these times onward in Indian culture.²⁰

This ahimsā tendency was underlined already by Auguste Barth in 1879:

Originally [Vedic sacrifices) were … feasts … in token of which the participants, priests and yajamana, consume each a small portion of the different offerings. ²¹

The necessity to eat the idā portion of each victim is implied in the reason given in Satapatha-Brāhmaṇa 13,6,2,12-13 for the release of the human victims at the puruṣamedha:

“Now, the victims had had the fire carried round them, but they were not yet slaughtered, - Then a [bodiless) voice said to him, “Purusa, do not consummate (these human victims): if thou wert to consummate them, man (puruṣa) would eat man.”
Accordingly, as soon as fire had been carried round them, he set them free, and offered oblations to the same divinities, and thereby gratified those divinities, and, thus gratified, they gratified him with all objects of desire. ‘²²

167

Repugnance against cannibalism is evidenced likewise by the following legend of the Vādhūla-Sūtra on the piling of the fire altar:

‘Previously they used to immolate a man as a victim to Prajāpati. Karṇājāya piled his fire altar with this (sacrifice). But as the gods were retreating, Dhārtakratava Jātukarni did not want to eat the ida portion of this (human victim).
Then they made the horse the victim. But as the gods were retreating even further, Rahahkṣita Jātakarni did not want to eat its idā portion. …
Finally, they made a hornless goat the victim, saying: “The hornless all-coloured goat is Prajāpati’s sacrificial victim”.[^32]

However, this sort of historical development is not necessarily the only explanation for the formation of the set of these five victims, since ‘cows, horses, men, sheep and goats’ are mentioned as sacrificial animals (paśú) belonging to Rudra as early as Atharvaveda 11,2,9.

3. The five animals and the fire altar

In the ritual piling of the fire altar (agnicayana), all these five victims are bound to the stake, appeased, have fire carried around them, and are slain.[^33]

According to the Satapatha-Brāhmaṇa (6, 2, 2, 18), the slaughter takes place on the first night of the year. The man who is bound to the stake with the longest rope is sacrificed first for Viśvakarman, then the horse for Varuṇa, the bull for Indra, the ram for Tvaṣṭr̥, and the he-goat for Agni.

The Kātyāyana-śrautasūtra (16, 1, 14) adds that the man is to be slain in a screened shed. The head of each of the five victims is cut off, and the trunks are thrown into water at a place where mud is later fetched for baking the bricks of which the fire altar is built. Mixed with the mud, the bodies become the five layers of the altar. The body of the goat, however, is prepared and partly eaten according to the normal rules of an animal sacrifice. According to another opinion, this is done with all five bodies. The heads are skinned and the brain removed - or neither skin nor brain is removed - and the skull is smeared with melted butter and deposited for later use.

168

The heads will be placed (the human head in the middle and the others around it) in a newly made fire-pot (ukhā), and the heads and the pot are consecrated and laid down as six bricks in the lowest layer of the fire altar.

The Satapatha-Brāhmaṇa (6, 2, 1, 37) criticizes those who procure the heads of the five victims without sacrificing them, for such sacrificers will become mortal carcasses, like Aṣādhi Sauśromateya, who died quickly after such heads had been put into his fire altar.²³ The same text also does not approve of golden or earthen replicas (38-39), but recommends the slaughter of the five victims, as was first done by Prajāpati and then by others up to Syāparṇa Sāyakāyana. The text admits, however, that Syāparṇa Sāyakāyana was the last to do so and that ’nowadays only these two (animals) are slaughtered, (a he-goat) for Prajāpati, and (a he-goat) for Vāyu’ (²⁴).

In the sequel (6, 2, 2, 1 ff.) the Satapatha-Brāhmaṇa ascribes the sacrifice of a he-goat to Prajāpati to the Carakas, a rivalling school of the Black Yajurveda. Indeed, according to the Kaṭha-Saṁhitā (19,8; 20,8) and the Taittirīya-Saṁhitā (5,1, 8, 1 ff.; 5, 2,9,1 ff.), the human head is obtained by buying it for 21 beans, the other (unspecified) sacrificial animals are set free after the fire has been carried around them, and only a he-goat for Prajāpati is sacrificed; a he-goat for Vāyu is also mentioned. These texts mention specifically a human head and the heads of a horse and a bull, the Taittiriya-Saṁhitā also the head of a snake, among the animal heads to be placed in the ukhā pot. The mantras of both the White and Black Yajurveda, however, are practically identical and mention the five victims as man or two-footer, horse, bull, sheep and goat,²

J. McDaniel, ‘Interviews with a Tantric Kāli priest: Feeding skulls in the Town of Sacrifice’, in D.G. White (ed.), Tantra in Practice (Princeton, 2000) 72-80, has interviewed a Bengali Tantric priest, who feeds the skulls of his ancestral Kali temple, ‘Under the altar (vedi) of this temple there are 108 skulls buried. Some altars have 1,008 skulls. Skulls awaken the Goddess, and make her present here. Male gods have stones (Silas) or lingas [of Siva], but goddesses have skulls’ and ‘The skulls in this temple mostly come from people who died in epidemics, especially cholera epi demics. Large numbers of people used to die, and there was no effective system of cremation at that time. Corpses would lie on the roadside or in the forests’ (p. 77). I should like to point out that the numbers of the skulls (108 or 1,008) represent sacred numbers that ultimately seem to come from the number of bricks in the Vedic fire altar (10,800): they represent the 360 x 30 ‘moments of the year.

169

The Vādhūla-Sūtra, which belongs to the Taittiriya school, states:

Previously they used to immolate these five animal victims: a man for Prajāpati, for Agni; a horse for Agni kṣatravat;²⁵ a bull for Agni kṣatrabhrt; a sheep for Agni braltmaubat; and a goat for Agni pustimat. After a firebrand has been taken around them and the hornless goat for Prajāpati, these five are set free and the sacrifice concluded with the hornless goat for Prajāpati. - He should make these five animals out of rice and barley. The Vādhūla-Sūtra further tells that Agni had revealed to the king of the Kurus the (secret) knowledge concerning the heads of the sacrificial victims (paśusīrṣavidyā). The learned Brahmins of the Kurus and Pancālas wanted to have it, but the king did not reveal it to any body until Suddhojas Māndavya sent his student Mārisābhagi to the king. The king asked Māriṣābhagi’s group to perform the sacrifice on his behalf according to that knowledge. The sacrificer whose priests do not have this knowledge will die 37

Even Oldenberg accepts that human beings had been killed during the construction of the fire altar not too long before the Brāhmaṇa texts were composed, though he denies that this killing had a sacrifi cial character. In his opinion, it was a charm purporting to give firm ness to the construction, an example of a custom widely attested in India and elsewhere.²⁶ Indeed, according to the commentator Karka, the bodies make the bricks firm; Weber further compares this with Roman, German and Slavonic traditions of embodying a human or animal victim in a wall in order to make it firm, and thinks that a Proto-Indo-European origin is possible.

170

But the Satapatha-Brāhmaṇa insists on the real heads of the victims, because only if the heads of living animals are united with their bodies (represented by the five layers of the altar) do they become alive again.²⁴

4. The Fire Altar and Cosmic Man

The 10,800 bricks of the fire altar represent the 30 x 360 moments of the year, and the body of the exhausted Creator God Prajāpati is conceived of as the year:²⁷

In building the great fireplace one restores and reintegrates Prajāp ati, whose dismemberment had been the creation of the universe, and makes him whole and complete. At the same time and by means of the same ritual acts the sacrificer, who is identified with Prajāpati (cf. [Satapatha-Brāhmaṇa] 7.²⁸.[^1].⁹), constructs himself a new sacral personality and secures the continuance of his existence (amstam).’²⁹

In 1883, Monier Williams wrote:

The most preposterous of all the ideas connected with the sacrificial act was that of making it the first act of creation. In the Purusha hymn of the Rig-veda (X. 90) the gods are represented as cutting up and sacrificing Purusha, the primeval Male, and then forming the whole Universe from his head and limbs… The Tāṇdya-Brāhmaṇa makes the lord of creatures offer himself up as a sacrifice. Even Sacrifice (Yajña) itself was sometimes personified as a god …Indeed it is evident that human sacrifice was once part of the Brāhmanical system.³⁰

It was early pointed out that the sacrifice of the primeval Puruṣa is likely to go back to Proto-Indo-European times, as it has a parallel in the Nordic myth of the Giant Ymir, out of whose body the gods built the heaven and earth, sea and mountains, and so on.³¹

171

Other Indo-European peoples, too, have preserved comparable ideas, including the Slavonic (with Adam in Old Russian literature), the Romans (with Romulus), and the Iranians (with Gayömart in the Bundahi&n).³²

Sylvain Lévi, in his pioneering study of the sacrifice in the light of the Brāhmaṇa texts, briefly discussed human sacrifice. He observed that the only authentic sacrifice would be suicide, which has been known and practised in India at all times, probably also in Vedic times. In the sacrificial system of the Brāhmaṇa texts, how ever, it is represented by its closest counterpart, human sacrifice, in which man redeems himself by sacrificing man. The legend of the Śunaḥśepa is an important monument of this cruel practice.³²

In his article ‘Self-Sacrifice in Vedic Ritual’, Jan Heesterman writes:

The vedic ritual texts abound in statements equating the sacrificer with the sacrificial victim and, generally, with substances offered in the fire … Generally, when fasting in preparation for sacrifice, the sacrificer becomes himself the oblational substance. Self-sacrifice, then, is a commonplace notion in the ritualistic discussions of the Brāhmaṇa texts. The sacrificer’s prototype is Prajāpati, the Lord of Creatures, who is both sacrificer and victim. Holding both ends together in his person, Prajāpati is himself the sacrifice, as the texts never tire of stating…. [T]he Prajāpati-sacrifice identification harks back to the celebrated Puruṣa hymn (R̥gveda 10.[^90]). By immolating the puruṣa, the primordial being, the gods break up the unchecked expansiveness of his vitality and turn it into the articulated order of life and universe. Life and order must be won out of their oppo sites, sacrificial death and destruction. Fittingly, this paradox is expressed in the enigmatically involuted phrases that conclude the hymn: “With sacrifice the gods sacrificed sacrifice, these were the first ordinances.” The riddle is the more critical since the purusa is not just a mythic figure. The word simply means ‘man’. The enig matic phrase is the riddle of man’s life and death.³³

172

Heesterman has rightly questioned Friedrich Weller’s conclusion that the Śunaḥśepa legend originally had nothing to do with royal rites.³⁴ In his opinion, the rājasūya is originally not an investiture of the king, but a yearly recurring festival by which the regeneration of the powers of fertility and the renewal of the universe are effected. The festival centres round the king, whose rebirth is equivalent with the universal regeneration and renewal. Now the chief interest of the legend lies in the birth of a son, conceived of as a real rebirth of the father […] Hariscandra and Śunaḥśepa are ritually reborn through the sacrifice, through the brahman power extant in the rc stanzas [i.e. verses of the R̥gveda) recited by Śunaḥśepa. Against this background it becomes clear why a brahmin should act as the sacrificial victim: he represents the brahman sacrifice out of which the sacrificer is reborn. […] In the last resort it is, however, not the brahmin, who is immolated, but the sacrificer himself, who is in the case of Hariscan dra doubled by his son and the brahmin Śunaḥśepa. He himself must pass through death in order to be reborn. Out of himself he is reborn, ātmā hi jajña ātmanah, as Nārada teaches […] The idea of the rebirth of the sacrificer out of himself is not in opposition to the idea of rebirth out of brahman sacrifice. The equation of the sacrificer with the sacrifice is well established; Prajāpati, the first sacrificer, is at the same time the sacrificial victim while he is also interchange able with brahman.”

5. The wider context: Connections with Sakta Tantrism, the Near East and the Eurasian steppes

The theme of the severed head and its restoration, involving resur rection by means of a cultic drink (possessed by the victim), is piv otal in the Pre-Vedic, Asuric’ religion sketched by Kasten Rönnow in 1929,50 The theme is important also in the ‘preclassical phase of development that, according to Heesterman, preceded the ‘classical Vedic ritual of the Brāhmaṇa and Sūtra texts.³⁵

173

During the past forty years Heesterman has built up an impressive model of an ‘agonistic’ and ‘cyclical system, in which violence, sexuality, feasting and changes of purity and impurity between two competing parties played central roles.³⁶ Developing studies of J.W. Hauer, Samaren dranath Biswas³⁷ and others, Heesterman has shown that sodalities of warring young men called vrátya were important agents in this ritual, and that their rites, the vrátyastomas, and other rites that share with the vrátyastomas some unusual components (such as ritual cop ulation)55 are fossilized remnants of the archaic preclassical ritual.³⁸ In 1966, Paul Horsch was able to connect the vrātyas with the proto epic gāthā and sloka verses of the Middle Vedic texts - and with the Śunaḥśepa legend in which such verses figure prominently.[^57] At one point, however, that is, when the Brāhmaṇa texts were codified, the Vedic ritual was fundamentally changed, and the violent and sexual elements were mostly reduced to symbols. Heesterman has argued that this internal development of the Vedic ritual was the result of social and economic changes.

But there were outside influences, too. The human sacrifice of the Veda and the change in the ritual can and must be studied in a wider perspective as well. I conclude by mentioning some of my own endeavours in this regard. I am convinced that the ‘preclassi cal’ Vedic ritual prevailed in northern India before the arrival of the main wave of Rgvedic Aryans around 1350 BC, and continued to do so outside the Vedic realm and for some time even inside the Vedic area.³⁹ The concept of ‘cosmic man’ (puruṣa) and his sacrifice becomes suddenly important in the youngest hymns of the R̥gveda, soon after the Rgvedic Aryans had settled in India and been in con tact with its previous occupants, largely an earlier wave of Aryans

whose traditions seem to be continued in the Atharvaveda. Previ ous research has strictly denied any genetic connection between the human sacrifices of the Vedic and Sākta Tantric traditions, and this is understandable for chronological reasons alone. Unfortu nately, the earliest texts on sākta cults are rather late (from the fifth or sixth century AD onwards, starting with the Buddhist adapta tions of Hindu Tantras), and bridging the gap admittedly requires some act of faith, but there is an inclination to see an Atharvavedic background to many of the ‘magical practices of Sakta Tantrism.⁴⁰ As I do not want to leave my claim for a direct connection between the Vedic and Sākta Tantric human sacrifice entirely in the air, I would like to mention some of the principal arguments, also in order to point to different sources of origin. I have discussed these topics at length elsewhere, and limit myself here to little more than an enumeration.

There is evidence of human sacrifice in the iconography of the Harappan alias Indus Civilization that flourished c. 2500-1900 BC. On a seal from Mohenjo-daro, a severed human head (of a warrior, to judge from its ‘double-bun’ headdress) is placed on an altar in front of a sacred fig tree inhabited by an anthropomorphic deity, attended by a kneeling worshipper whose hands are raised high in adoration or prayer.⁴¹ One Harappan seal also depicts a ‘sacred marriage between a bison bull and a prostrate human priestess, reminding one of the ‘sacred marriage of the Vedic horse sacrifice and the Lamaist thangkas depicting Yama’s buffalo mating with a female corpse ⁴² There is clear Near Eastern influence in other Harappan art motifs,

175

such as the ‘contest’ theme.⁴³ The Harappans are therefore likely to have had ‘sacred marriage’ rituals in which the male partner was put to death, as in the royal Inanna-Dumuzi cult of Mesopotamia. Willibald Kirfel in fact suggested in 1951 that this element of the Vedic horse sacrifice and the parallel human sacrifice might have been inherited by the Vedic Aryans from the agriculturalists of the Indus Valley. 66

The water buffalo plays a dominant role in the Early and Mature Harappan religion, which included a buffalo sacrifice. The water buffalo bull, in Sanskrit mahiṣa, may therefore in earlier Indian ritu als have occupied the place taken by the horse after it was imported by the Aryans as their ritually most important animal. The female partner in the ‘sacred marriage of the horse sacrifice, the chief queen, has the title mahiṣl, which also denotes ‘water buffalo cow’, Hundreds of buffaloes were offered to Indra before battle to increase his warring strength according to the Rgueda (5,29,8; 6,17,11; 8,12,8, 8,77,10), but there is only a single reference (Maitrāyani Saṁhitā 3,14,10 = Vājasaneyi-Saṁhitā 24,28) to a water buffalo sacrifice (to Varuṇa) in the extensive Middle Vedic texts; this suggests a con scious suppression of buffalo sacrifice in Vedic religion.⁴⁴ In later South Asian religions, buffalo sacrifice is almost exclusively part of the goddess Durgā cult.⁴⁵ Hundreds of buffaloes used to be decapi tated on the vijayadasami, the ’tenth day (dasamī) of victory (vijaya)’ that concludes Durgā’s navaratri festival ’lasting nine (nava) nights (rātri) and celebrates her victory over the Buffalo demon. Tradi tionally, war expeditions were started on the vijayadasamī day, and in Karnataka a human victim used to be offered to the Goddess on this day by the warring bands until British times.⁴⁶

176

A lion-escorted martial goddess imported from the Near East is depicted on the seals of the ‘Bactria and Margiana Archaeo logical Complex’ (= BMAC) of the Bronze Age (c. 2500-1500 BC) in Afghanistan and Turkmenistan; she apparently kept her Sumerian name Nana(ya) for two millennia, as her counterpart worshipped in Afghanistan in Kuṣāṇa times was so called, and is worshipped in Afghanistan even nowadays as ‘Bibi Nanni’.⁴⁷ This BMAC culture interacted with the Indus Civilization, and may be a principal source of the ‘Gangetic Copper Hoards’.⁴⁸ A BMAC type cylinder seal from the Harappan site Kalibangan bears an Indus inscription and a tiger-escorted goddess in the midst of two warriors spearing each other. Vedic texts know Vāc (‘Voice, Speech’) as a goddess of war identified with the lioness and con nected with the Vrātyas.⁴⁹ The Vedic prātyastomas were performed before and after raiding expeditions, and closely resemble the later Hindu navarātri festivals of Goddess Durgā, which involve sexual licence and feasting with the meat of many different sacri ficial animals,74 The Vedic lists of ‘unclean’ animal victims (to be released) agree with Purāṇic lists of victims pleasing the goddess; in both cases, a human victim as the most appreciated offering heads the list.[^75]

Excavations at the BMAC site Dashly-3 in northern Afghanistan brought to light a palace with the layout of a Tantric mandala and a temple-fortress surrounded by a moat and three concentric circu lar walls. In Afghanistan and Pakistan, the Rgvedic Aryans encountered an inimical people, called Dāsa, whose chief Sambara had in the mountains ‘autumnal forts’ (sāradī pur), possibly venues of the autumnal navarātri festival, like the fortress called Sār(a)dī in Kashmir. Many of the goddess’s names (Durgā, Kottavī, Tripura sundari) designate her as the guardian of the fortress (durga, kotta, tripura).[^76] Sambara has survived as the name of a fierce divinity of

177

Saiva origin in Tantric Buddhism, and, like Dása, appears to be Irano-Aryan in etymology.⁵⁰ The ethnicon Dāsa is cognate with Khotanese and Wakhi words meaning ‘man, hero’, which along with the linguistic peculiarities of Nūristānī and Magadhi, suggests that the Dāsas were distant ancestors of the later Scythians or Sakas, who introduced headhunting and the skull cult to Afghanistan (Nuristan). As the acculturated rulers of the BMAC culture, the Dāsas then took these gruesome cults to the Indus and Ganges Val leys by the end of the third millennium BC.⁵¹

Human sacrifice and skull cults have survived to the present day especially in eastern India, in the sākta Tantric worship of Durgā, Kāli and related goddesses. This Sākta tradition appears to be a direct continuation of the cultic practices involving human sacrifice that prevailed in South Asia before the coming of the Rgvedic Aryans. In the millennia older testimonia preserved to us in the Vedic texts, on the other hand, this pre-Vedic form of human sacri fice has a somewhat modified form, as it was first incorporated into a different religious tradition in an early phase of syncretism, only to be eradicated from the Vedic religion during a later phase of this process.