Indus Valley Religion

views updated

INDUS VALLEY RELIGION

INDUS VALLEY RELIGION is the goddess-centered religious system of the urban civilization that emerged in the Indus Valley of western India around 2500 bce and declined into a series of successor posturban village cultures after 1750 bce. The antecedents of this religion lie in the village cultures of Baluchistan and Afghanistan, which were part of a larger regional cultural system in western Asia that also included the village cultures of southern Turkmenistan and the Elamite culture of southwestern Iran. Common religious patterns within this larger region continued into the early stages of urbanization in Elam, Turkmenistan, and the Indus Valley, after which the unification of the local regions and subsequent historical changes led to separation: Elam was drawn into the orbit of Sumerian and Akkadian culture; Turkmenistan was settled by new groups from the northern steppes; and Indus settlement shifted eastward into the Ganges-Yamuna Valley in the North and Gujarat and the Deccan Plateau in the South as the original cities in the Indus Valley were abandoned. After the entry of Aryan tribes into northern India around 1500 bce, the continuity of Indus Valley religion is found mainly in the Dravidian cultures of South India, although various elements were also preserved in the village cultures of North India and in the synthesis of Aryan and non-Aryan cultures that marked late Vedic and post-Vedic developments in the Ganges-Yamuna Valley.

The Western Asian Setting

The evolution of the Neolithic cultures of western Asia that preceded the Indus civilization cannot yet be reconstructed in detail, but a pattern is emerging from current evidence that sheds new light on the basic features of the Indus Valley religious system. Archaeological research in southern Turkmenistan has revealed a continued sequence of village cultures north of the Kopet Dagh Mountains from at least 6000 bce onward, culminating in a regional urban culture at Namazga and Altin around 2500 bce. Research on the proto-Elamite and proto-Dravidian languages points to a common proto-Elamo-Dravidian ancestry among a pastoral people moving southward from Central Asia into Iran sometime between 8000 and 6000 bce, combining the herding of goats, sheep, and cattle with the cultivation of wheat and barley, and gradually separating into two branches: a proto-Dravidian branch that settled eastward in Afghanistan and Baluchistan, and a proto-Elamite branch that continued westward across southern Iran to the Zagros Mountains. The broadly based set of common cultural features established throughout western Asia in this early period and reinforced by later interregional contacts is reflected in similar patterns of proto-urban development and urbanization between 3500 and 2500 bce in the various localized regions.

Because the Indus civilization's script has not been deciphered, the proto-Dravidian identification of the Indus language remains uncertain. There is broad scholarly consensus, however, that a form of proto-Dravidian was the dominant language of the Indus urban culture, and this is substantiated by parallels between cultural and religious features of the Indus civilization and later Dravidian village culture. These parallels, in conjunction with the pre-urban cultural affinities with Elam and Turkmenistan, provide a framework for interpreting the evidence from village and urban sites in the Indus Valley region and constructing a hypothetical picture of Indus Valley religion.

The single most significant religious feature in all of the western Asian village cultures is the importance of female powers or goddesses, as evidenced by stylized clay and terracotta female figurines in a variety of types that appear—often in conjunction with figurines of bulls or rams—from the early levels of village culture on into the urban periods in Turkmenistan, Elam, and the Indus Valley. Whether they represent specific goddesses or powers is impossible to determine without more information, and the villages are mute. Evidence of coherent mythologies only appears in the richer range of artifacts at the urban level, and by then, in all of the urban regions, clearly defined goddesses had become part of complex urban cultic systems that reflect at least in part differences in regional urbanization. Enough affinities remain, however, to provide clues to the Indus Valley system.

Turkmenistan, where extensive excavation has been carried out, provides a valuable point of reference for parallel developments in Afghanistan, Baluchistan, and the Indus Valley. Turkmenistan village sites show four millennia of clay and terracotta goddess figurines and figures of male animals, most often rams and bulls. The goddesses appear in a variety of styles and are often marked with painted stripes, dot-centered circles, or clusters of pocked depressions; most have concentrated attention on the breasts, thighs, and buttocks, and they often have either no arms or vestigial stumps. One distinctive type, the so-called foot profile style, shows the truncated legs and torso of a female in a semireclining posture.

Evidence from early sites indicates that special areas were set aside as shrine rooms for a likely domestic cult. Enclosed village shrines appear by the fifth or fourth millennium along with a new type of Namazga III "foot profile" figurines with elaborate hairdos. The Bronze Age Namazga IV culture, concurrent with developments in early Elam and the pre-Indus village cultures of Baluchistan and Afghanistan, has evidence of more elaborate shrines and a range of figurine types. Finally, around 2500 bce, a full urban culture appeared in the Namazga V period that was contemporary for several centuries with the early phases of the Indus cities.

Urbanization in Turkmenistan brought not only greater complexity but also dramatic new religious forms. A massive brick platform, with three stepped tiers reaching forty feet in its final height, was built on the edge of the Namazga V site of Altin. This was certainly a center for public rituals, and implies a class of professional priests or priestesses. A richly endowed burial of a woman holding two female figurines in her hands has been tentatively named a "priestess's grave." Namazga V figurines are in a new and highly abstract style: flat fiddle-shaped cutouts with no legs, stylized triangular arm extentions, pinched maskedlike faces usually without a mouth, conical breasts, and a stippled pubic triangle with a vaginal line. There is evidence of a more standardized iconography and a clearer identification of individual figurines, including the use of different hair styles and engraved markings that resemble signs found in the Elamite and Indus writing systems.

Goddess worship in Turkmenistan clearly survived the transition from Neolithic to Bronze Age culture and subsequent urbanization. The styles of representation changed, the identities and meanings of individual goddesses may have varied, and the form of cultic practice certainly differed dramatically at the urban level from that in early villages, but the goddesses and their powers remained the central focus of religious life throughout the millennia. Much the same pattern can be seen also in the Elamite culture of southwestern Iran.

The foundation of Elamite culture was laid by proto-Elamite-speaking settlers who brought wheat and barley cultivation and the herding of sheep, goats, and cattle into the southern Zagros Mountain region of Iran sometime after 7000 bce. By 4000 bce, cultivation had been carried into the lowlands of Khuzistan at the western base of the Zagros, providing an agricultural base for urbanization. Sumer, across the Tigris in southern Mesopotamia, achieved urbanization around 3500 bce. By 3200 bce, Khuzistan and the Zagros highlands had been united in the rival urban civilization of Elam, with a highland capital at Anshan near later Persepolis and a lowland capital at Susa. Within the next two centuries, a proto-Elamite script had been developed and Elam had extended its influence eastward along a trade network that passed through Tepe Yahya in southern Iran as far as the Nal village culture of southern Baluchistan.

The expansion of Elamite urban culture was limited to the early third millennium, and its eastern trading centers had been abandoned several centuries before the first Indus cities emerged. The similarity between the later Indus script and the proto-Elamite script provides circumstantial evidence for the transfer of writing during this period, because the proto-Elamite script had been replaced by cuneiform by the middle of the third millennium; there is, however, no evidence of direct Elamite influence on Indus urbanization. Yet if Elam cannot be assigned a significant causal role in the creation of Indus civilization, it nonetheless provides an important model for understanding Indus Valley religion because of the many evident parallels between the two traditions.

Pre-urban cultural levels at Persepolis and Susa reveal a familiar pattern of female figurines and goddess worship, and painted pottery at these sites reveals a related concern for serpents as objects of religious veneration—a combination found also in the Dravidian villages of South India, and further evidence for an earlier common culture. Terracotta female figurines from Susa and other sites in the early third millennium show that goddess worship survived the transition from agricultural villages to urbanization in Elam as in Turkmenistan. The religious data from urban proto-Elamite sites such as Susa, however, are much richer than that of Turkmenistan and reveal not only the importance of goddesses in urban religious life but also an elaborate system of myths, symbols, and cultic practices.

Cylinder seals in a distinctive proto-Elamite style provide the most valuable evidence for the symbolism of this period. Many of the motifs in the proto-Elamite seals can be traced back to painted designs on earlier village pottery, but the more elaborate seal designs reflect a new urban sophistication: complex mythic or ritual scenes; symbolic designs involving mountains, trees, and animals (bulls and rams most often, but also lions and other felines); and an androgynous bovine in a variety of humanlike poses characteristically found in the figurines of goddesses. This latter figure is most likely the animal form or surrogate of the main Elamite fertility goddess, a moon goddess who was born from the Primeval Bull and was both the protector and soul of cattle—roles certainly consistent with the symbolism of the seals. None of the other figures can be identified with any certainty, and the meaning of individual symbols and scenes remains obscure in the absence of explanatory myths. In general, however, the symbolism reflects a developed fertility religion with its roots in the village past—probably a mountain past—but with new dimensions and new meanings in the urban culture: a religion in which the village goddesses have become patron deities of the city as well.

Cylinder seals, supplemented by other data, allow at least a partial reconstruction of proto-Elamite cultic practices. One seal shows a goddess or priestess being drawn in procession in a chariot flanked by moon symbols and horned cattle; another shows a tree in procession in a similar chariot, also with horned cattle and moon symbols; another shows an image or shrine on a palanquin flanked by attendants carrying moon symbols and what are either snakes or snake symbols in their hands. The goddess being honored in all of these scenes is almost certainly the moon goddess, whose connection with trees, serpents, and horned animals is indicated on a seal from the late third millennium that shows priests wearing belts or girdles of snakes around their waists and a device on their heads that combines the symbolism of crescent-shaped horns and trees.

Ritual processions and pilgrimages to sacred sites were apparently important features of Elamite religion. Elamite reliefs from around 2000 bce depict long lines of worshipers in procession, confirming the evidence from earlier proto-Elamite seals. Sumerian texts from the same period describe similar practices associated with the moon goddess Inanna, whose characteristics closely match those of the main Elamite goddess with whom she was later assimilated. A hymn to Inanna vividly describes a parade of priestesses, musicians with harps, drums, and tambourines, and a priest who sprinkles blood on the goddess's throne, and notes that men in the procession "adorn their left side with women's clothing" while women "adorn their right side with men's clothing" (Wolkstein and Kramer, 1983, p. 99).

The centers of Elamite worship were the shrines and temples erected for the goddess, usually in high places and with an associated sacred grove. In Susa, temples to the major deities were located on an elevated sacred area on the western edge of the city between the river and the royal establishment. The main ritual activity in the temples was animal sacrifice, and raised altars with drains attest to the emphasis on blood in these sacrificial rites. The major sacrificial festival to the goddess at Susa was descriptively called a "day of the flowing sacrifice" in tribute to the quantity of blood offered on this occasion.

The Elamo-Dravidian Pattern

The significance of these features of Elamite religion for comparative purposes is the light they shed on those aspects of Indus Valley religion for which there is no available Indus evidence. Indus seals and votive figurines, for example, suggest that Indus religion was based on some form of animal sacrifice centered around goddesses, but there is no direct evidence of ritual practice. Indus stamp seals, however, depict goddesses, trees, tigers, and horned animals such as rams, bulls, and water buffaloes in various combinations in mythic or cultic scenes. In other symbolic settings snakes appear as sacred animals. One scene portrays a line of androgynously appareled worshipers parading before a buffalo-horned goddess in a tree. This is the same basic set of symbols—goddesses, trees, lions/tigers, horned animals, snakes, and androgynous figures—found on Elamite cylinder seals, and indicates a significant body of shared religious concepts that reflect the common proto-Elamo-Dravidian ancestry and presuppose a common ritual practice. The relevant ritual in Elamite religion was blood sacrifice to the goddess, as it was also, along with many of the same symbols, in later Dravidian village religion. It is thus highly likely that Indus Valley religion followed the same Elamo-Dravidian pattern.

This is not a case of wholesale borrowing of Elamite religion or of basing Indus urban culture on external models. There are well-documented influences from Turkmenistan for at least a millennium prior to urbanization in the Indus Valley, and there must have been some degree of contact with Elam during the late stages of Indus urbanization to account for the similar scripts. Indus urbanization may have been stimulated by these contacts, just as trade and interregional contacts stimulated urbanization throughout western Asia. The similarities between Indus urban culture and other western Asian cultures, however, were general family resemblances, like those between the Elamite and Dravidian language systems. Indus urban culture was both unique and uniquely Indian, as much a product of the regional setting as of the common western Asian heritage, with characteristic features that were deeply rooted in the pre-urban cultures of Afghanistan and Baluchistan.

The groundwork for Indus urbanization was laid by a series of village cultures in the highlands west of the Indus Valley, the earliest of which dates from around 6000 bce. Archaeological research since the 1950s has revealed several early aceramic settlement sites with subsequent pottery development and domestication of local plants and animals, proving that the Indus region contributed to the Neolithic revolution and was not just a recipient of imported culture. By the early fourth millennium local village cultures had been established in northern and central Baluchistan, and by the mid- to late fourth millennium these cultures had been linked by trade with the Namazga III culture of southern Turkmenistan.

By around 3000 bce, when Elam was extending its influence eastward, the Nal culture had emerged in southern Baluchistan and the Nal-related Amri culture had expanded into the southern Indus Valley. By early in the third millennium another related culture known as Kot Dijian had expanded northward along the Indus from the region of later Mohenjo-Daro as far as the later sites of Harappa and Kalibangan. These new developments laid the foundation for urbanization.

Goddess worship was an integral part of Indus village culture, as can be seen from the example of Mehrgarh, the oldest known continuous settlement site within India proper. Discovered in the 1970s at the eastern end of the Bolan Pass, Mehrgarh spans the range from aceramic settlement around 6000 bce to the brink of urbanization around 2600 bce. Goddesses in the form of female figurines appear at every cultural level, and their evolution is intertwined with the development of the Indus region.

The earliest figurines from Mehrgarh date from the sixth and fifth millennia, a period when pottery was being developed, cultivation was expanding, and local animals—especially humpbacked cattle (zebus)—were being domesticated to replace the earlier reliance on hunting. The style of these first Indian village goddess figurines was the "foot profile" style found also in Turkmenistan, and this style continued essentially unchanged down to around 3000 bce. At that point, reflecting Mehrgarh's greater involvement in regional trade, there was a convergence toward the pinched-faced, goggle-eyed "Zhob mother goddesses" from the Zhob culture in northern Baluchistan. During the village's final century, from 2700 to 2600 bce, nude goggle-eyed female figurines were being produced commercially by the thousands at Mehrgarh and are for the first time in a standing position like Zhob figurines. Nude male figures, also standing, with shoulder-length hair and Zhob-like goggle eyes, suggest for the first time the possibility of a divine couple.

The end of Mehrgarh's final century marks the beginning of urbanization. As in Turkmenistan, the new iconography of female figurines and the appearance of male figurines coincided with the building of monumental ritual platforms. At Mehrgarh, the platform was a massive structure of brick faced with plaster with a colonnade of square mud-brick columns in front. This was contemporary with others in the larger contiguous area, one of which, near Quetta, had drains in the center and a stone-built hollow containing a jawless human skull, perhaps evidence of a "building sacrifice" in the platform's construction. Near this platform were found figurines of females and cattle with painted stripes.

Other platform structures of note belonged to the Kulli culture, which gradually replaced the Nal culture in southern Baluchistan after about 3000 bce and was still flourishing when the Indus civilization emerged on its eastern boundaries about five centuries later. Near the fertile stretches of the Porali River elaborate ceremonial centers were built on a new vast scale: At one typical site are two stone-built platforms about thirty feet high with ramps to the top and, nearby, a complex of over forty buildings. The Kulli ceremonial centers are set apart from the nearby agricultural villages with which they share common artifacts such as pottery and figurines. As the latter include both goddess figurines and striped cattle, it is likely that the cattle were votive offerings to the goddess. The whole combination, with platforms, wells, and drains, clearly suggests a ritual pattern involving sacrifice and ablution. The goddess figurines show a combination of originally Elamite postures with other styles (Zhob, Mehrgarh) developed regionally in western India. Although the Indus Valley civilization synthesizes elements from all these cultural styles, Kulli figurines and ceremonial centers provide its most direct prototypes.

Harappa and Mohenjo-Daro

The Indus civilization has been widely noted for its rapid development and continued stability over a seven-hundred-year period from around 2500 to 1750 bce. Over an estimated 500,000 square miles, the same basic cultural features recur from the cities to the several hundred towns and villages so far discovered. Such uniformity is striking, because unlike the concentrated settlement patterns in Mesopotamia, Indus sites were often well over fifty miles apart. It was the long-established base of village agriculture—wheat and barley cultivation along with cattle herding—that by the mid-third millennium provided the base for urbanization. The new cities, however, also broke with traditional village cultural patterns, and imposed new developments upon them. The two largest, Harappa and Mohenjo-Daro, both seem to go back to the civilization's beginnings. Their founding was no doubt decisive in setting the new political, economic, and religious styles. To the north, Harappa was built over an earlier Kot Dijian farming village, while to the south, Mohenjo-Daro's new urban culture dominated and soon replaced the neighboring village of Kot Diji and the local culture of Amri farther south. Similarly, at the smaller Indus city of Kalibangan farther east, the imposition of the new urban culture included construction of an Indus style ritual platform on the mound of an earlier fortified Kot Dijian agricultural settlement.

Mohenjo-Daro and Harappa were both built on a similar plan, one that smaller sites replicate. The cities were divided into two basic components: a lower city, about three miles in circumference with rectangular grid streets, and an upper area on each city's western edge formed by a brick-walled platform on an artificial mound that leveled some twenty feet above the surrounding plain. The massive exterior walls of the two major cities, over forty feet thick at the base, served to protect against flooding: Mohenjo-Daro from the Indus River, Harappa from the Ravi. Mohenjo-Daro was the larger of the two, and in most matters preserves the best evidence, as most of the Harappa mound was destroyed either by erosion or by its dismantling in 1856 by British engineers to provide ballast for a railroad. It was not until the early 1920s that the antiquity of Indus sites was recognized.

The lower city's residential and commercial character is evident at Mohenjo-Daro. The exterior baked-brick walls lining the main street were for the most part without adornment or direct street access. Residences range from barracklike dwellings to multistoried complexes, two of which have been dubbed a palace or hostel, but the typical residence was of a still-common South Asian type: small rooms around a central courtyard. Interspersed among residences were various shops and ateliers, and a large area with threshing floors has been found at Harappa. The lower city shows no clear evidence of dominant religious structures. In continuation of village patterns, there probably was a domestic cult centered in the home, perhaps connected with the terracotta female figurines and the elaborate drainage system that suggests a concern for hygiene and purity. But for the culture's larger religious patterns one must turn elsewhere, and first of all to the raised platform mound on the western edge of the urban complex.

The standardization of the urban plan suggests that dominant political and religious sanctions lay behind the civilization's conservatism. It is noteworthy that the two major cities show none of the gradual growth that occurred in Mesopotamian cities, but were built from the very beginning with their dominant platform mounds. One may assume from this that the civilization's basic values were set and preserved by those who established these structures, and that their functions were connected with the architectural eminences they created.

These platform structures—often misleadingly called citadels—did not have a primarily defensive purpose. Though the heavily walled mounds at some of the more decentralized locations like Harappa and Kalibangan may have been used defensively, and the Mohenjo-Daro platform had watchtowers fortified with pellets, it is noteworthy that at a time when Mesopotamian rulers had for several centuries raised large armies to extend their power, the Indus cities leave no traces of arrows, spears, or swords.

Rather than citadels, the monumental platforms are thus no doubt continuations of the structures found at pre-urban village sites, but on a far grander scale, with a surface area large enough for several big buildings. With such structures, they differ from unoccupied platforms elsewhere (most notably Altin in Turkmenistan). But they bear a resemblance to the "acropolis" at Elamite Susa, also on the west of the city, and the purpose is clearly similar: to give prominence to the institutions and activities set apart and above. It is even possible, with the Kulli culture as an intermediary, that Elam provided the model for the Indus platforms, as for its script. But in specific features the Indus platform reflects an independently emergent tradition with its own cultural dy-namics.

Most of the structures on the Mohenjo-Daro mound have been variously identified. A large columned building was probably an assembly hall; another has been dubbed a college. Definite is a granary; with grain as the primary measure of wealth and medium of exchange, the control of grain distribution was tied to civic authority. Yet most distinctive was a structure called the Great Bath.

The Great Bath itself was both literally and no doubt also symbolically the center of this complex. A large rectangular bitumen-lined tank in a colonnaded courtyard, it had steps leading down into its water from both ends. Clearly the steps were for bathing, and possibly for crossing from one end to the other. Moreover, the trouble taken to build such an elevated bath probably reflects an intensified concern for purification already evident in the lower city. In later Indian notions, higher waters are purer. Quite likely the whole complex—with wells, bathing rooms, and bath—served for the performance of purification rites supervised or enacted by priests. And because it was situated adjacent to the granary, such concerns probably also tied in conceptually with an interest in agricultural fertility.

In all this one is faced with a combination of concerns similar to those that underlay the practices connected with the platform mounds of pre-Indus villages. There is a new assertion of political, economic, and religious authority in the building of such massive structures in the heart of the riverine plain. The platforms themselves, however, must have been more than assertions of power by a new urban elite; they must also have been intended as sources of power: not because they were dominant physically, but because they provided a stage for rituals that would bring the ascendant cultural forces into harmony with the divinely empowered order of nature so evident in the nearby rivers, herds, and fields of grain.

One representation of divinity in Indus sites has been met: the terracotta female figurine, a surviving type from pre-urban village cultures, with the closest analogues being from Kulli and Zhob. Similar figurines also reappear in classical periods of Hinduism and serve as models for the yakṣī s on early Buddhist stupas. The basic Indus type has bare breasts, tapered or full-length legs, a girdle, heavy pendant necklaces, and an elaborate hairdo. Whether one or more goddesses is represented is uncertain, but parallel evidence from Elam suggests an iconographic differentiation. A few male figurines have also been found. While human figurines are predominantly female, animal figurines are invariably male. Most common are various bovines: zebu, short-horned bulls, and water buffalo. Most likely different kinds of potency were represented: that of the female in the form of the anthropomorphic goddess, that of the male in animal form, perhaps linked with symbols of civic power.

Indus Valley Seals

For further insights into the religious conceptions of the Indus Valley civilization, however, one must turn to a new iconography that has no precedent in the pre-urban village cultures. This comes from the controversial evidence of the Indus Valley seals. Here again, however, one must reckon with prior developments in Sumer and Elam, which produced cylinder seals earlier than the flat steatite (soapstone) stamp seals—measuring about ¾–1¼ inches per side—of Indus sites. Although Elamite seals are linked typologically to Sumer, their subjects are distinctively Elamite, and in certain cases present images with Indus counterparts, indicating the likelihood of iconic cross-fertilization. Thus, two cylinder seals from Susa seem to draw on familiar Indus motifs: one a series of bulls eating from a manger, and the other a composite bull-antelope with long wavy horns facing a stylized pipal tree.

It is not, however, only the older urban civilizations that shed light on the Indus seal iconography, but also the likely continuities from Indus urban culture to the Dravidian village culture of South India. For not only is there the likelihood of linguistic continuities, but there is also archaeological evidence of cultural continuities from the Indus civilization, through Gujarat, to the Dravidian culture of the Deccan Plateau. Moreover, the urban models that the Indus cities provided during the Indus period not only reshaped village life in Indus times, but transmitted patterns that long outlived the Indus cities.

The Indus Valley stamp seals, found by the hundreds, confirm impressions gained from the terracotta figurines, most notably the tendency to accentuate female power in human form and male power through animal forms. But the situation with the seals is also more complex, as there are also humanized males and human-animal and even human-animal-plant composites of apparently both genders. Of single animals, many are drawn from nature: short-horned bulls, zebu, buffaloes, rhinoceroses, tigers, elephants, antelopes, crocodiles. Composite human-animal forms reach such complexity as one with tiger hind quarters, ram forepart, bull horns, elephant trunk, and human face. Others show animal heads radiating from a central trunk.

The most frequently depicted seal animal is a "unicorn" bull or ox of generalized bovine traits with a single erect horn that faces an apparently sacred object, perhaps a brazier or incense burner. The unicorn's horn is sometimes shown as a thin curved shaft crossed by lines that taper toward the tip, suggesting an affinity between animal and plant forms. Some of the naturally drawn animals also sometimes face simpler brazier or manger type objects. It is thus likely that "real" animals were linked with ritual symbols as well, and mythologically marked no less than the more clearly "mythic" figures like the composites, multicephalics, and unicorns.

It is the seals with humanoid figures, however, that take one beyond the general sense of mythic and ritual markers to evidence for a cult with a complex of sacrificial symbols. Females appear in various such scenes, but the most important are a series of scenes that portray a recurring ensemble involving goddesses, trees, tigers, and water buffaloes. Three seals show these interrelationships most dramatically. In one, a slender goddess with a crescent-shaped headdress kneels on a branch of a neem (margosa) tree, her arm outstretched toward a tiger below that turns its neck around to face her. The goddess's position replicates a worshiping pose in other seals and suggests that she beckons the tiger with her outstretched arm. In the second seal, the goddess, now descended from the neem tree, stands behind the tiger about to seize it from the back. Yet both goddess and tiger are strikingly transformed. The goddess has assumed multiple traits of the water buffalo: Along with filling out the stylized crescent horns, her legs and feet have become flanks and hoofs and her ears pointed and flapped. Meanwhile, the tiger has sprouted horns that replicate the serrated V branch and leaf pattern of the neem, which now stands behind the goddess.

These two seals seem to suggest that the goddess has her primary affinity with the buffalo, and an opposition to the tiger. But a third seal shows a fusion of the goddess and the tiger, joined together so that the goddess retains a standing human form as the forepart of a tiger's body that extends back from her hips and rear. Here, where the goddess's affinity is with the tiger, she has wavy ram's horns rather than buffalo-like horns. There is thus the suggestion that while the goddess has affinities with both the tiger and the buffalo, the two animals themselves remain in an oppositional and unfused tension.

This sense of a tiger-buffalo opposition is further reinforced in the so-called proto-Śiva seal from Mohenjo-Daro. But before discussing this seal, one should observe the primary seal evidence for a connection between the transformational themes that link the goddess to these and other animals and to a sacrificial cult. Clearest in this regard is the so-called ritual seal from Harappa depicting a goddess in the U-shaped twin branches of a stylized pipal tree, which rises from a circular platform base or altar. The goddess has crescent buffalo-like horns and a pony tail like the goddess who descends toward the tiger from the neem. But her horns have a third central peak, and she does not have the hind legs or hoofs of the buffalo. Standing at rest, her bangled arms loose at her sides, she observes a horned and pony-tailed figure much like herself kneeling at the base of the pipal in the same "suppliant" posture that the goddess in the neem adopts toward the tiger. The kneeling suppliant has led before the goddess in the pipal a composite animal with buffalo hind quarters, ram horns and forepart, and a large masklike human face. Seven figures in thigh-length tunics, single backward-curving horns, pigtails, and bangled arms stand or possibly file before this scene, which clearly depicts the essentials of a sacrifice. The horned goddess is the recipient of an offering, its composite nature no doubt representing something of the range of victims she receives: the ram, buffalo, and the human face. Whether all are real offerings, or the human face solely symbolic, cannot be ascertained.

The kneeling figure making the offering in the "ritual seal" has an intermediate status between the horned goddess whose dress he affects and the composite human-animal he offers. The precise nature of his sacrificial role eludes researchers. But it is striking that the combination of elements that the seals configure remain coherent in the setting of still-current South Indian village rituals. In that Dravidian context, the neem, a female tree, is itself a form of the goddess. It is linked to her fierce side, and more specifically to the forms she takes to "cool" and thus overcome violent forces like smallpox, fevers, and various demons. The pipal, on the other hand, is the male tree the goddess marries, so the two trees will be planted to actually intertwine. It is highly suggestive that while the seal goddess on the neem branch descends to overcome the fierce and wild tiger, the goddess in the pipal stands tranquilly in a position to receive as a sacrifice the composite of human and domesticated animals that would seem to reflect the range of her regular cult.

Since the discovery of the Indus civilization, the one seal most central to a succession of different interpretations of the religion has been the so-called proto-Śiva seal from Mohenjo-Daro. This designation, however, now appears to have been based on a combination of misattributions: most notably the "three heads" that actually outline the dewlap of a buffalo face, and the "trident" headdress that actually consists of buffalo horns enclosing a central fan-shaped and stylized tree or sheaf of grain reminiscent of the tree-and-horns headpiece worn by Elamite and Sumerian priests. Moreover, above this "mitre" is what looks like a stylized pipal tree. The main figure is thus a humanlike water buffalo with a buffalo head and horns, axially centered on representations of a plant and/or tree. He sits with his knees out to the side and his feet drawn in below an erect phallus. The posture has usually been identified as yogic, though it is also reminiscent of the posture of the androgynous bovine seated in the pose of the goddess on an Elamite seal. Possibly the series of V-shaped stripes that end at his waist—sometimes regarded as necklaces—represent tiger stripes, making him a figure in whom the tiger-buffalo tension finds a resolution in a yogic or more likely regally dominant self-discipline.

The central buffalo figure in the "proto-Śiva" seal seems to be the male counterpart to the goddess, who herself combines both tiger and buffalo attributes in various transformational modes. But in this male figure, what remains tense and dynamic in the goddess seems to find poise, dominance, and resolution. This is especially suggestive in view of the four animals that surround him, for the tiger and buffalo appear among these along with the elephant and rhinoceros. By analogy with later Indian iconographies, the four together are likely to have had a directional symbolism: the elephant linked with the east, the rhino with the west, the tiger with the north, and the buffalo with the south. Most strikingly, while the elephant and rhinoceros appear indifferent to each other, both facing east, the tiger and buffalo, which most directly flank the buffalo-man's horns, face each other in a state that has the look of combative arousal. Furthermore, a stick figure appears over the back of the charging tiger: possibly a form of the goddess herself.

Of the four "wild" animals on the seal, two—the elephant and water buffalo—are susceptible to some degree of domestication even though they retain their "wild" traits. There is some evidence that elephants were captured and trained for heavy forest work during the Indus period, and it is likely that domestication of the water buffalo for agricultural use in the river valley was one of the major achievements of the Indus civilization, complementing the village cultures' earlier domestication of the zebu in the highlands. It is significant that the water buffalo played the main symbolic role in Indus urban culture instead of the zebu, despite the latter's longstanding economic importance. This suggests that the buffalo had a critical role in the riverine agriculture on which the Indus system was based, and that it symbolized the control of both nature and culture that made urban civilization possible.

The central water buffalo figure on the proto-Śiva seal seems to have the same general symbolic meaning of power-under-control as does the bull in Mesopotamian symbolism, and like the latter it probably also represents the king or ruling authority. Because the figure is male, it may be assumed that the Indus rulers also were male, as the few examples of protrait sculpture at Mohenjo-Daro suggest. There is little doubt, however, that the Indus people considered the goddess to be the real power and the ruler only her surrogate, empowered by her and thus responsible to her and for her. This is certainly consistent with the symbolism of the seal, where the central figure seems to bring into a regulated and authoritative image the various forces that the goddess oversees: agriculture, animal sacrifice, and the dangerous forces associated with the truly wild regions beyond the domain of civilization.

What kind of authority—priestly, political, economic—this figure represented still remains uncertain. The urban background of the ritual complex on the platform mound leaves all these possibilities open, and the distinctively Indian features of Indus symbolism make it risky to explain the Indus civilization on the basis of other urban cultures, even the closely related proto-Elamite culture. There is little doubt, however, that Indus Valley religion played a major role in establishing and maintaining that authority, and there is even less doubt that sacrifices to the goddess were the primary form of cultic practice. All of the external evidence—earlier village cultures, contemporary and related western Asian urban cultures, and later Dravidian culture—points to this conclusion, and the Indus evidence seems to confirm it.

Concluding Observations

Taking the evidence as a whole, it is possible to construct a model of Indus Valley religion that explains its major known features and its place in the Indus civilization. The central element was certainly worship of the goddess at both the domestic and public levels, with corresponding levels of cultic practice. At the domestic level, votive sacrifices involving figurines were the likely form of worship, with a related emphasis on bathing and ritual purity. At the public level, represented by the raised platform mounds, worship must have involved more powerful blood sacrifices.

Mohenjo-Daro seems to have been the major cultic center for the system as a whole and the site for the most important sacrifices. Indus symbolism and later Dravidian practice point toward water buffalo sacrifices as the most important cultic rituals. The buffalo is the husband of the goddess in Dravidian cult sacrifices, and on Indus seals he appears as both the goddess's surrogate and the symbol of centralized rule; it is likely that Indus cultic practice involved these elements, at least on major ceremonial occasions, but there is no direct evidence for how this might have been conceptu-alized.

The interpretation of Indus Valley religion cannot proceed beyond such speculation at the present time. Much has been learned about the Indus civilization since its discovery in the 1920s, and the pattern of Indus Valley religion is beginning to emerge from the growing body of data, but there are still many gaps to fill. The major task, moreover, has hardly begun: to trace the contributions of the Indus system to later Indian religious developments and to understand the place of the Indus system in the larger pattern of religious history.

Bibliography

The most useful single source on village cultures and urbanization in Turkmenistan and western India is S. P. Gupta's Archaeology of Soviet Central Asia and the Indian Borderlands, vol. 2 (Delhi, 1979), which presents a judicious overview of the archaeological evidence and current interpretive theories. An assessment of the archaeological data from Turkmenistan sites is provided by the Soviet prehistorians V. M. Masson and V. I. Sarianidi in Central Asia: Turkmenia before the Achaemenids (London, 1972), edited and translated by Ruth Tringham and published initially in English as volume 79 of the "Ancient Peoples and Places" series. The best and most comprehensive survey of cultural development in India from earliest times through the establishment of Indus urban culture is Walter A. Fairservis, Jr's. The Roots of Ancient India, 2d rev. ed. (Chicago, 1975), which examines the relevant evidence from all of the regional village cultures, discusses the factors that led to urbanization, and describes the basic features of the Indus civilization. The most accessible summary of the evidence from Mehrgarh is Jean-François Jarrige and Richard H. Meadow's "The Antecedents of Civilization in the Indus Valley," Scientific American 243 (August 1980): 122–133.

A variety of data on proto-Elamite religion is available in A Survey of Persian Art, edited by Arthur U. Pope and Phyllis Ackerman (1938–1939; reprint, London, 1964–1965), especially in the sections by Phyllis Ackerman on cult figurines and early seals (vol. 1, chaps. 11 and 14, and accompanying plates in vol. 7). A summary of Elamite history, culture, and religion is provided in Walther Hinz's The Lost World of Elam: Recreation of a Vanished Civilization (New York, 1973). Diane Wolkstein and Samuel Noah Kramer's Inanna: Queen of Heaven and Earth (New York, 1983) supplements what is known of Elamite goddesses with a comprehensive portrait of the closely related Sumerian goddess Inanna. The argument for a common origin of the proto-Elamite and proto-Dravidian languages is presented in David W. McAlpin's Proto-Elamo-Dravidian: The Evidence and Its Implications (Philadelphia, 1981). The significance of the Elamite trading center at Tepe Yahya for understanding interregional contacts in West Asia is discussed in Carl C. Lamberg-Karlovsky and Martha Lamberg-Karlovsky's "An Early City in Iran," Scientific American 224 (June 1971), and in Carl C. Lamberg-Karlovsky's "Trade Mechanisms in Indus-Mesopotamian Interrelationships," Journal of the American Oriental Society 92 (1972).

Data from the early excavations at Mohenjo-Daro are contained in John Marshall's Mohenjo-Daro and the Indus Civilization, 3 vols. (London, 1931), and E. J. Mackay's Further Excavations at Mohenjo-daro, 2 vols. (New Delhi, 1938). A valuable synthesis and interpretation of Indus evidence is presented in Mortimer Wheeler's The Indus Civilization, 3d ed. (Cambridge, U.K., 1968), and in his Civilizations of the Indus Valley and Beyond (London, 1966). A survey of Indus data up to the mid-1970s is provided in Walter A. Fairservis, Jr's. The Roots of Ancient India (Chicago, 1975), supplemented by his Allahdino I: Seals and Inscribed Material (New York, 1976). Many of the most important contributions to the ongoing study of the Indus civilization are found in two volumes edited by Gregory L. Possehl: Ancient Cities of the Indus (New Delhi, 1979) and Harappan Civilization: A Contemporary Perspective (New Delhi, 1982). Possehl's own initial work on Indus culture in Gujarat is presented in his Indus Civilization in Saurashtra (Delhi, 1980).

Understanding Indus Valley religion has been a concern of investigators since Marshall and Mackay offered their first tentative interpretations, and the issue receives significant attention in the cited works by Wheeler and Fairservis. Indus seals in particular have been studied for clues to Indus religious concepts, with the so-called proto-Śiva seal receiving the greatest interest. A major new interpretation of the symbolism on this seal is found in Alf Hiltebeitel's "The Indus Valley 'Proto-Śiva,' Reexamined through Reflections on the Goddess, the Buffalo, and the Symbolism of vahanas," Anthropos 73 (1978): 767–797. Earlier interpretations of this seal and of other supposedly related Indus data are challenged in Doris Srinivasan's "Unhinging Śiva from the Indus Civilization," Journal of the Royal Asiatic Society (1984 pt. 1): 77–99. New insights into the meaning of Indus seals, especially those dealing with goddess-and-tiger motifs, are presented in Pupul Jayakar's The Earthen Drum: An Introduction to the Ritual Arts of Rural India (New Delhi, 1980), a beautifully illustrated book that interprets Indus religious data by drawing comparisons with the art and religious practices of later Indian folk traditions.