Archaeologists and historians studying religious structures frequently tend to classify temples by the initial dynastic period of their construction, and the literature abounds with phrases like the ‘Chola temple’ or ‘Satavahana stupa’ [i] However, in the academic quest for order in data, we underestimate how frequently monuments are in constant flux. Religious structures in particular cannot be fixed in time, although they might be so in space. By pinning these structures within specific temporal and dynastic periods, we often ignore the fact that religious structures are living entities. We forget that these are complex entities that have complex life histories extending long after that of their initial construction—they were constantly added on to and altered, often spanning the rule of multiple dynasties. By tracing the life-histories of religious structures archaeologists and historians can access an ever-changing pattern of cultural memory formation and religious practice.
At Banavasi (Karnataka, India) where I worked for several years [ii], my team and I studied several Buddhist stupas, hemispherical structures constructed to enclose Buddhist relics. Site 71 is an extremely overgrown and eroded circular brick mound located about a mile north of the village of Banavasi (Figure 1) [iii]. Based on the form and size of the bricks used in the structure, the stupa was constructed around the second-third centuries CE. Ceramics and terracotta roof tiles found on the structure also date it to an early period, at least prior to the 7th century CE [iv]. It thus falls within a period when Buddhism was widespread in southern India and Banavasi itself was likely an important religious and economic center. The limited historical research on these monuments has so far focused on their form and temporal context and once the structures have been neatly categorized by these criteria their later histories have been largely ignored.
 |
| Figure 1: Site 71, eroded stupa. Photo by author. |
It is likely that the core period of the stupa’s use and worship as a Buddhist structure was limited to an early period and declined starting from the fourth-fifth centuries as Buddhist worship in south India was largely replaced by a resurgent Hindu tradition. In Karnataka, Shaivite Hinduism, which focused on the primacy of the God Shiva, emerged as predominant. As Buddhism gradually became less popular, stupas across the region were abandoned and fell into ruin. And yet, even as Hindu temples increasingly became the focus of social and religious life, fragments of “material memory” remained. At site 71 (and at other stupa locations in and near Banavasi) the mound has a looter’s hole on the top. From colonial travellers accounts from the 17th and 18th centuries, we know that the ‘topes’ were often mined for reliquaries by the rather straightforward, though archaeologically unsound, method of digging a hole in the top into the relic chamber. While the looter’s holes in the Banavasi stupas cannot be dated, it is an interesting remnant of a memory or belief that there might be ‘treasure’ in the centre of these structures.
There is also clear evidence of the later use of site 71. In fact, at present the structure is considered a Hindu shrine although there is some memory among the present inhabitants of surrounding villages of its early history as a Buddhist structure. The hemisphere has been flattened on top, and brick fragments mined from the structure have been used to construct a makeshift shrine consisting of a platform surrounded on three sides by low, roughly-built walls (Figure 2). The shrine itself contains an extremely eroded figure of the elephant-headed god, Ganesha, as well as a fragmentary sapta-matrika panel that represents seven mother goddesses who are a part of the Hindu pantheon (Figure 3). These items have clearly been appropriated from one or more Hindu temples and date to a period after the 16th century. This fits with evidence of a second episode of roof construction on the stupa, where the terracotta tiles are of forms that can be dated to between the 16th and 19th centuries CE.
 |
| Figure 2: Shrine on top of stupa. Photo by author. |
 |
| Figure 3: Shrine elements. Photo by author. |
Even more recently, within the last couple of years, a set of cement reinforced steps lead up to the shrine. When we talked to people living and worshiping at the shrine there was no recognition that it was originally a site of Buddhist worship, instead the mound itself has been absorbed into a modern mythos that weaves tales of ancient mounds or 'guddas' that were the palaces of ancient (and unnamed) kings). At most of the stupas that survive in the area, there is evidence of later use and worship, including the construction not just of shrines but of simple stone alignments of unclear purpose.
Sites like these offer an interesting contrast to other stupas that have been completely forgotten and destroyed. For instance, at site 207 we initially noticed a low circular mound, barely more than an undulation on the ground. Since there were no structural fragments (like brick or tiles) visible on the surface it was difficult to identify it as a stupa. On a visit a couple of months later, the farmer who owned that field had decided to level the ground for cultivation and was using a large mechanical backhoe to dig up the mound. With this excavation, the true nature of the structure was revealed and the distinctive bricks and terracotta tiles that emerged clearly identified it as a stupa (Figure 4).
 |
| Figure 4: Site 207, destroyed stupa. Photo by author. |
Yet another example of the complex life histories of religious structures comes from a consideration of folk religious practices that often occur outside the traditional ritual spaces of the temple. Throughout South India, folk beliefs populate the landscape with a variety of divine and semi-divine beings, as well as spirits (bhutas) and other inimical forces. In many cases, these small sacred sites do not have built shrines. Instead, they could consist of rounded stones or earthen pots worshiped as forms of the mother goddess (Chowdamma); or places identified as residences of spirits or natural symbols (termite mounds, snake holes). In other cases, these shrines can include miscellaneous architectural or sculptural fragments appropriated from larger structures. These ephemeral forms of construction are a crucial part of the wider religious landscape and as important in lived practice as the larger stupas and Hindu temples. Such small village shrines are simply made of easily available materials and require little labor. Due to their very impermanence the materials they are made of require maintenance and they are continuously cleaned, added to, worshiped. These small shrines are a more organic feature of the village landscape- a rounded stone tucked away under a banyan tree, appropriating the hole of the village cobra, or a broken sculpture under a palm leaf shed. I cannot imagine that such places would leave easily identifiable traces for the archaeologist. And yet, they must have been a part of village life for generations.
However, the boundaries between these local traditions and more institutionalized Hinduism, where worship was sited within stone temples and mediated through priests, are extremely fluid. Traditionally, if flaws or cracks developed in the central lingam (typically a phallus-shaped symbol of the Hindu god Shiva, worshiped as a generative force) within a temple it was no longer considered worthy of worship. And yet, as sacred items they had to be disposed of carefully and were, by being submerged in the nearby river. Periodically throughout the year these items re-emerged during the dry season when the water level falls drastically. Over some time, these discarded items become the focus of smaller folk shrines, with small walls enclosing them (Figure 5). In many cases worship at these shrines are the province of local families and do not require the intercession of the priest who is attached to the larger temple. However, as the shrine becomes more permanent, the priest re-enters the picture and begins to make more formal ritual offerings on behalf of the people.
 |
| Figure 5: Linga on dried river bed. Photo by author. |
A more careful exploration of the life histories of small and large structures thus adds greatly to our understanding of the complexity of cultural memory in the communities we study. By foregoing some of our desire to classify the material indicators of history we can begin to explore something of the messiness of human action, past and present!
Acknowledgements
This blog post derives from research that will be published in an article that is under review: ‘The Multivalence of Landscapes: Archaeology and heritage’. In Himanshu Prabha Ray (ed.), Preserving Plurality: Heritage in South and Southeast Asia. Routledge.
Endnotes
[i] ‘Chola’ and ‘Satavahana’ refer to pre-modern dynasties known to have ruled in south Asia. The Satavahanas controlled the central section of the Indian subcontinent from the 1st c. BCE to the 2nd century CE. The Cholas ruled large areas of southern India between the 9th and 13th centuries CE.
[ii] Uthara Suvrathan, “Spoiled for Choice?: The sacred landscapes of ancient and early medieval Banavasi”, South Asian Studies, Vol. 30.2 (2014); “Regional Centres and Local Elite: Studying peripheral cores in peninsular India”, Indian History (The Annual Journal of the Archive India Institute), Vol. 1 (2014).
[iii] During my research we recorded and studied over 600 sites, large and small, dating from the third century BCE to the present day. Each site was assigned a unique identification number.
[iv] Evidence from similar structures elsewhere in the subcontinent, as well as inferences drawn from the low quantities of roof-tiles found at 71 indicate that only certain sections of the structure were roofed.
