EIGHT RELIGION

In any ancient society, religion is notoriously tricky to reconstruct without having access to scriptures. Think of the tantalizing European Palaeolithic cave paintings of animals and human hunters at Lascaux and Altamira, where no writing exists. What sense could scholars make of the convoluted myths in the Egyptian Book of the Dead, simply by looking at tomb paintings without reading their hieroglyphic texts? Or of the imagery from the ancient Mesopotamian myths, such as the epic of Gilgamesh, without reading Sumerian cuneiform tablets?

In the Indian subcontinent, the Vedic period does not suffer from this handicap, because of its extensive literature, that is, the Vedas, the Upanishads and other early Hindu scriptures, composed orally between about 1500 and 500 BC and in due course written down in Sanskrit. On the other hand, the Vedic period signally lacks any archaeological remains; the only significant ones from the period are some city walls at Rajagriha in Bihar dating from the early to mid-first millennium BC. The Indus civilization, by contrast, certainly lacks any scriptures, since its script (which may or may not contain religious thought) is undeciphered. Yet it offers extensive architectural remains. But which of the Indus remains can be shown to be religious? Was the Great Bath at Mohenjo-daro essentially a religious building, rather than an elaborate swimming bath? Were the city’s numerous, small, stone objects – roughly cylindrical in basic form, in some cases shaped into a tip – phalluses comparable with the much later Shiva lingam of Hinduism, rather than, perhaps, game pieces? Marshall was inclined to answer both of these questions in the affirmative during the 1920s. Today, some scholars agree with him, while others are more inclined to caution, especially about the purpose of the stone objects.

Even Marshall, however, openly confessed there were no recognizable Indus temples, to his considerable surprise and disappointment given the obvious constructional skills of the Indus builders. In Mohenjo-daro, he writes:

no building that can definitely be stated to have been a temple has yet been found either at Mohenjo-daro or Harappa. This does not by any means prove that no temples were built. Many buildings have been found that are clearly not ordinary dwelling places or administrative buildings . . . though at present we cannot determine their uses with any degree of certainty, and the objects found in them, unfortunately, prove nothing. Indeed, for all we know, the temples of Mohenjo-daro may, for conservative reasons, have been of wood and perished altogether. Whatever their form and material, one thing is certain, viz., that they did not in anyway resemble the temples of Sumer or Babylonia. No trace of a ziggurat with its associated temple has been unearthed at either Mohenjo-daro or Harappa . . . Up to date not a single building has been found whose plan in any way resembles that of the Babylonian temple with its temple-tower, its large open court for worshippers well supplied with water, and its especial shrine for the god or goddess at the end of the court. This alone would, in my opinion, suffice to show that the religions of the Sumerians and the Indus Valley peoples were dissimilar.1

Despite almost a century of subsequent excavation, no self-evidently religious structures have been found at any Indus site to contradict Marshall’s account. The only exception might be the sacrificial hearths dubbed ‘fire altars’, discovered at several Indus sites, including Kalibangan, Lothal, Nageswar, Rakhigarhi and Vagad. At Kalibangan these consist of clay-lined pits containing ash, charcoal, the remains of a clay stele and terracotta cakes; at Lothal, a terracotta ladle with smoke marks was found near such a hearth. However, not every expert is convinced they are sacrificial hearths: at Nageswar, ‘the “fire altar” is in all likelihood a regular Indus funnel-shaped updraft kiln’, according to Possehl.2 The most convincing examples – from Kalibangan – recall the Hindu ritual of libation of the five products of the cow (milk, sour milk, clarified butter, urine and dung) in the presence of a fire as offerings to a clay lingam in the worship of Shiva. There is a parallel, too, with the Vedic fire ritual, which involves the libation of milk at sunrise and sunset. In this, the heated milk is considered to be the sun or the sun’s seed poured into the womb: ‘Surya [the sun] and Agni [the fire] were in the same receptacle [yoni, “womb”]. Thereupon Surya rose upwards. He lost his seed. Agni received it . . . he transferred it to the cow. It (became) this milk.’³ But while there is no doubt that sacrifice at fire altars was integral to the Vedic religion, there is no proof that the excavated Indus hearths – if that is what they really are – constitute fire altars in the Vedic sense. ‘The similarities have been overemphasised and the shared elements of fire and animal sacrifice are too common, being found in many religions, to be a culturally diagnostic link’, notes McIntosh.⁴

In the absence of scriptures and temples, speculation about the Indus religion must rely on imagery: in figurines and sculptures such as the ‘priest-king’, and on pottery and seals, many of which show mythical scenes. It begins with the very early ceramic female figurines from Mehrgarh. What was their purpose? Was it religious, or were the figurines merely toys? The excavators found them frequently in rubbish deposits, where they seemed to have been haphazardly discarded, encouraging the view that they were toys. But more careful study by Catherine Jarrige has revealed that the rubbish deposits in question were often within the household area, which suggests the possibility that the figurines had a cultic significance. Some of them also contain small holes running through the figurine, probably created by small twigs when the clay was soft, which add to the impression of cultic significance (as in the device of sticking pins in a voodoo doll). Moreover, the fertility attributes of the figurines, such as their bulbous breasts, broad hips, prominent hairstyles and even a child in the arms, are combined ‘so systematically that they must conform to certain rules with definite symbolic connotations’, comments Jarrige. ‘With good reason, therefore, we can discount the suggestion sometimes made that these figurines are in fact children’s toys.’⁵

‘Fire altars’ at Kalibangan. These clay-lined pits containing ash, charcoal and terracotta objects may have been sacrificial hearths.

But it is the seals that naturally invite the most attention and prompt the most ingenuity in interpretation. Not only do they depict composite fantasy animals – some with three heads – as well as real ones, they also depict interactions between animals, humans and what can only be deities. Let us take a closer look at three particularly promising seals. The first shows a godlike human figure surrounded by animals, excavated by Mackay and quickly dubbed ‘a prototype of the historic Shiva’ by Marshall.6 The second concerns a mysterious object that generally accompanies the unicorn motif. The third is a highly complex scene of worship, apparently of a goddess in a peepal tree, which may involve human sacrifice.

Female ceramic figurines from Mehrgarh. The earliest figurines date from 7000 BC.

Marshall’s ‘proto-Shiva’, which may well be the most celebrated of all Indus seals, is a male figure seated upright in a clearly yogic posture, with his legs folded under his body and his feet pointing downwards (the posture known as mulabandhasana) on what appears to be a throne. He has three faces, one looking straight ahead, the other two looking sideways, to the left and right. His head is crowned with a spectacular three-pointed headdress including two curved buffalo horns. On his arms there are bangles. He appears to be ithyphallic. Arranged about him are four vividly depicted wild beasts, two on either side: a rhinoceros with a water buffalo and an elephant with a tiger. Above his headdress are characters from the Indus script.

Why did Marshall feel licensed to connect this Indus figure from the third millennium BC with the chronologically much later figure of the Hindu god Shiva, known only from the first millennium BC, who is never mentioned by name in the intervening Rigveda? Four attributes suggested this interpretation. First, the three faces of the figure fitted with the fact that Shiva is frequently three-faced (trimukha), a trinity, although Marshall qualified this argument by writing:

I do not mean by this that the philosophic idea of a triad associated with the doctrine of the absolute had taken shape at this early period, but simply that the cult of this particular god – call him Shiva or by whatever name we like – had been amalgamated with [two] other cults, and that the fact was signified by giving him three faces instead of one.7

Second, the yogic posture fitted with Shiva’s reputation as Mahayogi, the prince of yogis. Third, Shiva is also known as Pashupati, the lord of the animals – symbolized here by the rhinoceros, water buffalo, elephant and tiger. Fourth, a headdress with horns was commonly used to denote a deity, and also worn by priests and kings, in ancient Mesopotamia; and more speculatively, this particular three-pointed headdress resembles in shape the famous trident (trisula) carried by Shiva. Finally, though this is not an attribute, the connection fitted Marshall’s belief that the roots of Hinduism – in this instance, the god Shiva – lay not only in the Vedic culture of the Aryans but also in the pre-Vedic Indus civilization.

Indus seal impression from Mohenjo-daro, with drawing, showing a three- or four-faced figure wearing a horned headdress and seated in a yogic posture, surrounded by four wild beasts. John Marshall, who discovered the seal, dubbed the figure ‘proto-Shiva’; but solid evidence for this Hindu attribution is lacking.

None of these arguments is entirely convincing; indeed, ‘the evidence for any kind of continuity between this prehistoric god and Shiva is rather weak’, according to the historian A. L. Basham.8 Many scholars have come up with significant criticisms. For example, it makes more sense to imagine the deity as having four faces (one of them looking backwards, and therefore hidden), with each face matching one of the four animals. This would associate him with the Hindu god Brahma, successor of the Vedic creator-god Prajapati, rather than with Shiva. His erect phallus is not clear; this detail may instead be part of a knot in his waistband (as Marshall himself recognized). His yogic posture, which occurs in various other Indus seals, seems to have a non-Indian antecedent in the proto-Elamite art of neighbouring Iran dating from the early third millennium BC. The four animals that surround him are wild animals, with the possible exception of the water buffalo; yet the Hindu god Shiva did not protect wild animals. And since the horns he wears are buffalo horns, he should probably be associated with the Hindu buffalo demon, Mahishasura (mahisha is the Sanskrit for ‘buffalo’), rather than with Shiva, who is usually associated with the bull (Nandi), though it is true that Mahishasura is sometimes identified with Shiva. How very much one wishes that the accompanying characters, which may well contain the deity’s name, could be reliably read.

Consider now the second example of possible religious imagery, the ‘unicorn’. When an animal motif, rather than a deity, is the focus of an Indus seal, it is sometimes depicted on its own, as is usual with the elephant, rhinoceros and tiger – perhaps because these are wild animals. But often it is shown with a specific object in front of the animal. In some cases, such as the short-horned and humpless bulls and the water buffalo, the object is almost always a feeding trough, which would have been filled with offerings of grain or water; similar types of shallow basin with downturned rims have been found in excavations. However, the unicorn motif is never depicted with the feeding trough, only with an unfamiliar object that has not been recovered from excavations. Most scholars believe this object had a ritual purpose, and it is difficult to disagree, despite much uncertainty about the nature of the ritual.

Kenoyer describes the object as follows:

The ritual offering stand is made up of three parts, a tapering shaft or column which stands on the ground and pierces a hemispherical bowl-shaped container that is sometimes held on the shaft by a small pin. Projecting above the bowl, the shaft supports a square- or dome-shaped object. This top component is usually cross hatched with a grid or zigzag lines, and the bowl portion is variously depicted with cross hatching or horizontal lines. The edges of the bowl often have tiny dots or radiating lines along the bottom edge and sometimes even along the top edge.9

To Marshall, the object was an incense burner. The incense was placed in the upper container, which may have revolved around the shaft, while the bowl contained a fire, with its flames indicated by the dots or radiating lines along its top edge. Iravatham Mahadevan, however, identifies the object quite differently. To him, it is a filter for preparing a sacred, intoxicating drink, soma, that played a crucial role in later Vedic rituals. (The botanical identity of the plant used to make soma has been much debated but is now widely accepted as belonging to the genus Ephedra, the source of the drug ephedrine, a banned substance among modern sportspersons.) In this second interpretation, the upper container was a filter through which the liquid soma dripped into the lower bowl, with its overflow indicated by the dots and radiating lines on the bottom edge.

Our third example of religion in the seals includes an animal motif, a puzzling object, a deity and a worshipper, as well as other figures. This fascinating ‘narrative’ seal from Mohenjo-daro has been the subject of much speculation and is far from being fully understood. The deity stands inside a sacred peepal tree (Ficus religiosa), a fig tree divided into two main branches with their characteristic heart-shaped leaves – the same species as the Bo tree (or Bodhi tree) under which the Buddha achieved enlightenment and which today commonly provides shade over the tombs of Muslim saints in Pakistan. The figure wears a long braid at the back, numerous bangles on its arms and a horned headdress, so it most likely depicts a goddess, since tree deities in India (yakshis) are usually female. In front of the deity the suppliant, also wearing a horned headdress, kneels, and behind the suppliant stands a giant, human-faced markhor goat with a beady eye and two long curly horns, above which float several characters from the script. Beneath this scene, along the bottom of the seal, stands a procession of seven human attendants of the deity, all with long braids at the back, bangles and single-plumed headdresses, who are sometimes referred to as seven priestesses, although there is no definite evidence that they are female.

But the most intriguing detail is yet to come. ‘What the small object is near the feet of the suppliant is uncertain, as the seal is slightly damaged at this point; possibly it is an offering to the deity on a small altar, or possibly an incense table’, comments Marshall.10 This seems reasonable, and as far as one may go in identifying the object. However, Parpola and some other scholars, including Kenoyer, boldly go further. To them, the object is probably a severed human head; indeed Parpola, after studying two curious protrusions from the object, maintains that it is a man’s head with a double bun at the back typical of a male hairstyle seen elsewhere in Indus art and sculpture (as on page 129). If so, this ‘fig-deity’ seal would provide the only known example of human sacrifice in the Indus civilization – unless we propose the same interpretation for some seated male figurines with their legs together and hands clasped at the knees. ‘This type of figurine has been interpreted as a worshipper,’ writes Kenoyer, ‘but perhaps it represents a sacrificial victim with hands and feet tied together.’¹¹

In looking, however briefly, at Indus religion, it has been hard to resist drawing comparisons with Hinduism – whether with the god Shiva, Vedic rituals or tree deities. Another interesting example is the swastika motif. This ancient symbol, the name of which signifies ‘well-being’ in Sanskrit, pre-dates the Indus civilization and occurs frequently in Indus seals and inscriptions – both in its left-facing and right-facing (Nazi) variant, sometimes even side by side on the same tablet. Its Indus meaning is unknown: possibly the two variants stood for different cults or philosophical schools. Certainly in later Hinduism and Buddhism the variants represented the opposing forces of the universe. Today in the Indian subcontinent, the swastika is very widely used both in Hindu religious rituals and secular life as a decorative symbol supposed to bring wealth and good fortune, for example on public buildings, domestic dwellings and the human body. Further intriguing examples of such continuities between the Indus civilization and present-day Hindu South Asia will appear in the penultimate chapter on the origins of Hinduism.

‘Fig-deity’ seal made of steatite, from Mohenjo-daro: one of the most intriguing Indus seal stones. This impression, with drawing, shows the deity on the right, a kneeling worshipper, a goat and seven human attendants. Some scholars identify it as a scene of human sacrifice.

Having said this, we must remain mindful of the lack of solid evidence for Indus religion. No Indus temples, priests or rituals can be identified for sure. Moreover, there is little consistency in the conceivably ‘religious’ structures at different sites, by contrast with the consistency in other kinds of structure such as brick sizes, flood platforms and ‘citadels’, as well as the consistency in ornaments such as beads and bangles. To draw yet another comparison with later Indian religion, enormous diversity exists within Hindu beliefs, practices and rituals, not to mention the tribal religions that have influenced, and been influenced by, Hinduism. For example, human sacrifice has been practised in India, among certain Hindu devotees and certain tribes, at various times, though never as a dominant tradition. If Indus religion really is one of the roots of Hinduism, then it, too, may well have been remarkably diverse.

On the present limited evidence, it is surely possible there may have been less religion in the Indus civilization than the later Indian civilization might suggest, or than some scholars – beginning with Marshall – might wish. In the absence of a decipherment of the script, religious explanations have conveniently filled some yawning gaps in scholarly understanding. But is this wise? To my mind, the current Indus situation is uncomfortably reminiscent of the situation in ancient Mayan studies before the decipherment of the Mayan script in the 1980s and 1990s. According to the leading Mayanist of the 1970s, Eric Thompson, the ancient Maya rulers of Central America were a theocracy with a deeply spiritual outlook. Their ideal was ‘moderation in all things’, their motto ‘live and let live’ and their character had ‘an emphasis on discipline, cooperation, patience, and consideration for others’.12 Theirs was a civilization unlike any other, said Thompson, who looked to the Maya as a source of spiritual values in a modern world that placed far more importance on material prosperity. Only thanks to the Mayan decipherment did Mayanists come to know that Thompson had been utterly wrong. The real Maya relished internecine war and the extended torture of captives; and both the Mayan rulers and their gods liked to take hallucinogens and inebriating enemas using special syringes. I am far from suggesting that the same volte-face will one day occur in Indus studies, if and when the script is deciphered. For now, though, it may be wise to assume that the Indus civilization was not a civilization unlike any other civilization – not least in the circumstances of its perplexing decline and disappearance.

Male head from Mohenjo-daro, probably broken from a seated sculpture, showing combed hair tied into a double bun at the back.