‘From all quarters of the Indian world’: the temple at Rameshvaram, Hindu kings, and Dutch merchants

Abstract

On Rameshvaram island in the south-east corner of India lies one of Hinduism's most important temples—the Rāmanāthasvāmi, one of the four dhams (‘holy abodes’) and the site of two Śiva-liṅgas said to have been consecrated by Rāma himself. A temple has existed here since at least the eleventh century, although most of the present temple dates to the seventeenth and eighteenth centuries when the island was protected by the Setupati rulers of nearby Ramnad. In several of the long corridors and halls for which this temple is famous are brightly painted life-sized standing images of over 100 male figures attached to columns. Though such images are characteristic of many south Indian temples from the seventeenth and eighteenth centuries, there are far more at Rameshvaram than at any other south Indian temple. This article examines the number, location, and significance of these numerous standing images within this temple. By exploring the significance of the temple as a long-standing site for the royal performance of devotion, this article seeks to address whether the great number and identity of the life-sized donor images can be explained by both Purāṇic ideas of kingship and seventeenth- and eighteenth-century Dutch observations of the pan-Indian status of the temple.

Introduction

On the island of Rameshvaram in the far south-east corner of India is one of Hinduism's most important sites of pilgrimage. A temple has existed on the site since at least the eleventh century CE, although most of the present temple dates to the seventeenth and eighteenth centuries when the island was under the protection of the Setupati rulers of nearby Ramnad. Many early European visitors came to Rameshvaram island in the nineteenth century, remarking upon the monumental scale of the very long corridors built in the eighteenth century around the sacred heart of the temple. In several of these long corridors and halls are brightly painted, life-sized standing male images with their hands placed palms together in greeting (añjalimudrā) attached to the columns—a sculptural tradition characteristic of many south Indian temples built in this period. There are around 100 such images throughout the temple—more than in any other south Indian temple—that are located at important axes and routes through this monumental site of pilgrimage, appearing to greet passing devotees as they enter. Previous scholarship has only briefly mentioned the existence of these images, suggesting that they represent the Setupatis, their ministers, and family members. This article aims to re-examine the number, distribution, and potential identities of the numerous standing male images within this great temple. By exploring the significance of the temple as a long-standing site for the royal performance of devotion, this article seeks to address whether the great number and the identity of the life-sized male images in this important temple can be explained by both Purāṇic ideas of kingship and seventeenth- and eighteenth-century Dutch observations of the pan-Indian status of the temple. Following an outline of the temple, its numerous images, and its significance to many south Indian rulers over the centuries, we consider what early modern Dutch sources suggest about the status of Rameshvaram across India and traditional ideas about India's constellation of kings. Finally, bringing all these sections together, we propose a preliminary conclusion for the identity of the numerous statues of devotees in Rameshvaram.

Worshipping Śiva at Rameshvaram

The Tamil region of south India is dense with sites of pilgrimage but only some of these, including Rameshvaram, have attained pan-Indian significance. Together with Badrinath, Dwarka, and Puri, it is one of the four directional dhams—the southernmost of the major pilgrimage sites that mark the sacred boundaries of India (see Figure 1). But Rameshvaram is best known for its significance in the Rāmāyaṇa as the site from which Rāma, Lakṣmaṇa, Hanumān, and their army built a bridge (Sētu) to cross to Lanka, defeat Rāvaṇa, and rescue Rāma's wife, Sītā. The Rāmanāthasvāmi (also Rāmaliṅgasvāmi or Rāmaliṅgēśvara) temple dedicated to Śiva in Rameshvaram town on the north-east side of the island is built around the two liṅgas said to have been consecrated by Rāma himself: the Rāmēśvara liṅga initially made of sand and the Viśvanātha liṅga brought by Hanumān from the Himalayas. From at least the twelfth century, the former has been considered to be one of Śiva's 12 liṅgas of light (jyotirliṅgas) that map Śiva's presence across the country.¹ Within the temple are a series of 22 wells or tanks (tīrthas) whose water is auspicious and that are visited in sequence by pilgrims prior to their encounter with Śiva and his consort.² Across the island are further tīrthas, such as the Agni tīrtha on the seashore directly in front of the temple, and other sites identified with events in the Rāmāyaṇa that are all visited by pilgrims.³

Figure 1. Map of south India. Source: Crispin Branfoot.

In common with many important pilgrimage sites in southern India, the temple at Rameshvaram has been built, expanded, modified, and renovated over an extensive period. The temple was not built all at once on a clearly defined layout and phases of expansion were followed by infilling. The temple visited today has three concentric walled enclosures (prākāras), the outermost two each with long corridors around the darker, enclosed inner area of the first prākāra containing the most important shrines to Śiva in his various forms, his consort, and other deities (see Figure 2). The outermost wall of the third prākāra measures circa 263 metres by 200 metres with pyramidal gateways (gopuras) on the north, south, and west sides. The earliest surviving archaeological evidence for the antiquity of this long-established pilgrimage destination is the group of five small stone shrines on the west side of the third prākāra that date to the tenth to eleventh centuries, in addition to a number of fragmentary inscriptions with Pandya regnal dates of the eleventh to thirteenth centuries.⁴ The outermost enclosure wall and the three gateways were built in the fifteenth century, defining the greatest extent of the temple before 1900.⁵ A further gopura on the east side of the second prākāra wall built in the 1650s—though the superstructure was not completed until 1904—is aligned with the ritual and spatial centre of the whole temple: the Rāmanātha shrine (see Figure 2: A). A smaller adjacent gopura is aligned with the shrine to Rāmanātha's consort, Pārvatavarttiṉi Ammaṉ (see Figure 2: B). From an architectural perspective, the most celebrated parts of this temple are the monumental corridors of the third prākāra constructed between 1740 and the 1770s at the height of Setupati power.⁶ On the west side of the vast outer enclosure, the Cokkattam corridor, which is circa 100 metres long, was built between the west gopura and a small entrance through the second prākāra wall.⁷ This corridor intersects with a similar north–south corridor at its halfway point that continues all around the temple, with the longest east–west corridors on the north and south sides. It is along these corridors and at important axes or entrances that the great mass of figural column sculptures is located.

Figure 2. Plan of Rāmanāthasvāmi temple, Rameshvaram. Source: Adapted from James Fergusson, History of Indian and Eastern Architecture, (rev. and ed. with additions) James Burgess and R. Phené Spiers (London, 1910), p. 381.

Upon entering the temple at Rameshvaram from either the east or the west into the long stone corridors with their high ceilings and wide aisles, pilgrims walk past row upon row of life-sized sculpted male figures (see Figure 3). Often raised above eye level, peering between recently assembled make-shift shops selling temple souvenirs or above signs with details of the current temple festival, these likenesses stare straight ahead with their hands placed together in greeting. Some are brightly painted—garishly to some eyes—as if to suggest their continual presence; a few are periodically honoured with garlands as if to recognise their singular status. Within this one temple, there are just over 100 of these life-sized sculpted male images, some of which are placed alongside further large architectural sculptures of mythical animals and deities (see Figure 2; red dots indicate their location, with largest groups numbered in blue). Previous visitors and scholars have made some reference to their presence but have not examined the number, distribution, or identities of the numerous life-sized sculpted male figures lining many of the corridors of the temple.

Figure 3. Images on the east side of the third prākāra corridor. Source: Crispin Branfoot.

In January 1796, Colin Mackenzie—the future Surveyor-General of India—visited the temple and, though not able to enter the innermost prākāras, commented upon the statues of ‘chiefs (Pulitaver)’ with their ministers and attendants in the ‘vestibule or building on the east front of the pagoda [temple]’ and ‘a number of statues, of another chief and his followers, placed on a raised stone terrace, on either side of the covered passage leading to the inner gate’ on the western side.⁸ In 1842, the naval officer W. Christopher not only produced the earliest known plan of the temple, but also described the ‘statues of men the size of life stand in front of some of the primary rank of columns, all in the same attitude, with the head erect and the open hands joined palm and palm over the breast’. He was impressed by this ‘effort in statuary of a superior kind’ given the huge monoliths from which the sculpted columns were created and, though not enumerating or identifying any of the images, he remarked that ‘from what we heard, a statue in Ramiseram is desired by the ambitious of Ramnaad equally as much as a monument in a niche of St Paul ['s Cathedral] is by our countrymen’.⁹ During his visit to Rameshvaram in 1868, the photographer Edmund David Lyon was informed that the 12 images in the western Cokkattam corridor that he photographed were of the Rajas of Ramnad on the right (south) and their secretaries opposite.¹⁰ These earlier visits were followed by the most detailed examination of the temple, its inscriptions, and its rituals, including these standing male images, in early 1883—around a century after they had been created—by James Burgess (1832–1916), then head of the Archaeological Survey of Western and Southern India, and his talented Tamil brahmin assistant, S. M. Natesa Sastri (1859–1906).¹¹ One of their chief informants was ‘Satāvadhānam Muttusvami Ayyangar, an able Paṇḍit and poet’, from whom they obtained the identifications for some of the numerous statues that have been reiterated since, as outlined below.¹² Yet few scholars have gone further than to state that the images at Rameshvaram—that is, those that are most accessible in the east and west corridors—represent the Setupatis.¹³ Some may be images of the Setupati rulers of the nearby capital Ramanathapuram but, given the enormous number, can they all be identified as rulers or ministers of the dynasty? Or can the great sacred status of the temple across the south and all of India account for the many images?

Rameshvaram, the Setupatis, and the performance of kingship

The island upon which Rameshvaram is situated lies at the narrowest point in the Palk Straits between India and Sri Lanka. The distance from Dhanushkodi, the site from which some consider Rāma's bridge to have been built, on the long sandy spit of land that stretches east to south-east and across the shallow stretch of sea to Talaimannar on the island of Mannar just off the coast of Sri Lanka is only around 20 miles. For pilgrims, traders, and armies, this is the route between the mainland and Sri Lanka (former Ceylon). Furthermore, seaborne trade around the south coast between the Indian Ocean and the Bay of Bengal would also pass through the straits. The islands off the south-east coast, stretching west from Rameshvaram, have also been important for their pearl fisheries.

Given its religious significance, many of the major rulers and dynasties of south India before 1600 left records recording their devotion to Śiva in the temple, and also their conquest of Rameshvaram or their performance of royal ceremonies at the site. Indeed, for a ‘world-conquering monarch’ (cakravārtin), the Sētu was sometimes cited in Gupta-period literature and inscriptions of the fifth century CE and later as the southern limit of the whole world of Bhāratavarṣa.¹⁴ In around 930, the expansionist Chola King Parāntaka (r. 907–55) performed the tulābhāra ceremony three times, weighing himself against gold for distribution to priests and brahmins. As a declaration of his claim to be an independent sovereign lord with imperial aspirations, Parāntaka performed this ritual not only at ‘beautiful Rāmatīrtha, where the ablest monkey flocks built the bridge’, but also at Kanyakumari at the southernmost tip of India and at Srirangam at the centre of the Tamil country.¹⁵ The tulābhāra or tulāpuruṣa ritual—the weighing of a person against gold or other precious items such as pearls that were then given away as gifts—was one of the most important of the mahādānas (‘great gifts’) of Purāṇic kingship in medieval India, especially in the Deccan and farther south.¹⁶ The ritual may have been performed as a declaration by kings of their claims to independent imperial status at the outset of expansionist conquests or immediately afterwards following victory in battle. For example, the Chola King Arunmolivarman (or Rājarāja, r. 985–1014) is stated to have performed the tulābhāra ceremony directly before the start of his conquest of the four quarters (digvijaya) in the eleventh-century Tiruvalangadu copperplates.¹⁷

Shortly after Parāntaka's ritual performance, the Rastrakuta King Kṛṣṇa III (r. 939–66) claimed in copperplates dated circa 959 to have erected a pillar of victory (jayastambha) at Rameshvaram following his brief conquest of Chola territory and subordination of the Cheras, Pandyas, and Sinhalas across the far south.¹⁸ In subsequent centuries, the Pandyas and Hoysalas, in addition to the kings of both Jaffna and Polonnaruwa in Sri Lanka, demonstrated claims over Rameshvaram, whether by making gifts of land or lamps to the temple or by performing royal and imperial rituals.¹⁹ As part of his consolidation of power across Sri Lanka and the establishment of alliances with southern Indian kings, the Sinhala Buddhist ruler Niśśaṅkamalla (r. 1187–96) of Polonnaruwa performed the tulābhāra ritual, set up several jayastambhas and built a temple at Rameshvaram dedicated to Niśśaṅkesvara (Śiva).²⁰ On several occasions between the 1220s and 1340s, Hoysala kings from the southern Deccan marked their territorial claims over the far south by erecting pillars of victory at the Sētu.²¹

Following the incorporation of the Tamil country into the Vijayanagara empire in the 1370s, Rameshvaram was again the site of royal performance. Virūpākṣa, son of Harihara II (r. 1377–1404) and governor of the Tamil country, performed the tulābhāra ceremony before Rāmanātha according to two sets of similar copperplates dated 1387 and 1390.²² In this period, Rameshvaram received the patronage of the Hindu Arya Chakravarti kings of Jaffna in northern Sri Lanka—the location of several important Śaiva temples including the Tirukketīśvara (Śiva) near Mannar and others on Jaffna Island. In 1414, Pararājaśekhara of Jaffna led the renovation of the temple at Rameshvaram with stone brought from Trincomalee (Trikonamalai) on the north-east coast.²³ Rameshvaram was thus an important site for the performance of kingship by many southern kings from at least the tenth century, especially at moments of the expansion or consolidation of authority in new frontier territories.

In the early seventeenth century, the temple at Rameshvaram took on greater political significance with the rise of a local dynasty of kings called the ‘Guardians of the Sētu’ or the Setupatis, whose emergence to power in the region is connected with wider political changes across southern India. First, a new level of Vijayanagara imperial engagement with the Tamil country developed in the early sixteenth century, when Kṛṣṇarāya and Acyutarāya visited many of the great Vaiṣṇava and Śaiva temples from Tirupati and Srikalahasti through to Tiruvannamalai, Chidambaram, Srirangam, and Rameshvaram, making lavish donations to the deities at each place.²⁴ During the same period, the Nayakas of Gingee (Senji), Tanjavur, and Madurai were established as regional governors of the northern, central, and southern parts of the Tamil country. At a time when their subservience to their nominal Vijayanagara overlord was weak in the early seventeenth century, the regionally powerful Madurai Nayakas established the Setupatis as chiefs of Ramnad, the south-eastern coastal area surrounding Rameshvaram. As warriors of the local Maṟavar caste, their duties included the protection of the pilgrimage route to the Rāmanāthasvāmi temple. The first of these chiefs, Śaḍaika Tēvar, was installed by Madurai's Muttu Kṛṣṇappa Nāyaka in around 1605. Assuming the regal name Uḍaiyān Sētupati, Śaḍaika founded the Setupati Dynasty reigning over Ramnad, which, by the eighteenth century, had grown practically independent from Madurai.²⁵

There is little historical evidence that, prior to Śaḍaika's appointment, his family already held an exalted status or otherwise commanded much power. Yet, the Setupatis claimed high antiquity as custodians of the Sētu, including Rameshvaram. Indeed, some of Ramnad's foundation stories declare that, on his way to Lanka, Rāma himself installed the very first member of the Setupati line: either the oarsman Guha who had rowed Rāma across the Ganga River or a local relative of Rāma's commander of the Ganga. Upon his return, Rāma would have mandated the Setupatis to guard the newly established Rāmanāthasvāmi temple and other sanctuaries in the Sētu.

On later occasions before the seventeenth century, according to various traditions, these early Setupatis protected, defeated, or reigned over the Pandya or Chola kings—or the regions identified with these dynasties—as suggested by their Tamil titles, such as ‘establisher of the Pandya realm’ (pāṇṭimaṇṭala tāpaṉācāriyaṉ) and of the ‘Chola realm’ (cōḷamaṇṭala). As some stories have it, the Setupatis even employed the Pandyas as their ministers and generals for 12 generations while they ruled over much of the Tamil-speaking lands for 608 years. At one point, however, this old Setupati line supposedly lost its position, as its last ruler, Jayatuṅga Tēvar, was imprisoned to die by an envious Pandya king or because the dynasty was deceitfully overpowered by the Vijayanagara empire. The aforementioned Śaḍaika Tēvar, appointed by Madurai in the seventeenth century, was said to be related to these earlier, ‘mythical’ Setupatis, with chronicles variously stating that he descended from them or was himself the formerly jailed Jayatuṅga.²⁶ In sum, the later, historical Setupati house fashioned itself as a continuation of an ancient line of legitimate and powerful rulers going back to the age of the Rāmāyaṇa.

From the early seventeenth century, the Setupatis began issuing inscriptions of their own and, throughout the seventeenth century, established a greater degree of regional autonomy, as their Madurai overlords’ power declined and the Nayakas’ attention turned northward.²⁷ The changing political fortunes of the Setupatis from the 1670s are evident not only in the shift of their capital from the early coronation site at Pokalur to the elaborately fortified site at Ramanathapuram 10 miles east, but also their performance of royal mahādānas, the increasing patronage of the great temple at Rameshvaram, and the celebration of Navarātri (‘Nine Nights’).²⁸ Navarātri, or the Mahānavami festival, was the most important public royal ritual in the Vijayanagara empire and its successor states—a devotional festival that was focussed on the reigning king and the annual revitalisation of his kingship and his realm within a broader cosmological hierarchy.²⁹ The Setupatis’ sovereignty was based upon their devotional relationship with their lineage's local tutelary goddess Rājarājeśvari, from whom they now received the ‘sceptre of authority’ (cēṅkōl), with Rāma and with Śiva in the form of the liṅga installed within the Rāmanāthasvāmi temple on Rameshvaram island, in addition to their status as ‘protectors of the Sētu’ (Sētupati).³⁰ Emerging from beneath the shadow of subservience to the Madurai Nayakas, the expansion and rebuilding of the temple at Rameshvaram in the seventeenth and eighteenth centuries, together with the inclusion of numerous life-sized male images lining the long corridors, may be seen as one way in which the Setupatis’ new royal status was declared.

Image, donor, portrait

Sculpted images of donors and devotees have a long history in South Asian religious art. Some have been understood to be representations of specific individuals, identified in an adjacent inscription and differentiated by their body language or attire. In south India, small-scale reliefs of standing or seated donor figures appear alongside identifying inscriptions on some ninth- and tenth-century temples and a few examples of free-standing copper-alloy or stone images of specific individuals, usually less than half life-size, are known from later centuries.³¹ But, from the early sixteenth century in south India, a tradition developed of placing in temples large, often near life-sized, images of male figures, depicted standing with their hands in añjalimudrā—a gesture of greeting or devotion. From the later sixteenth century, such images became widespread in many temples, especially in the far south of the Tamil region—one element in the increasing prevalence of portraiture across South Asia in this period.³²

The degree to which such images may be regarded as portraits of specific named individuals varies, for, in common with portraiture in general, these south Indian images engage with both the likeness of the individual and the typical, conventional, or ideal. Inscriptions that identify the subject adjacent to such stone images from the late sixteenth through to the nineteenth century are rare. For many such images, it is the details of dress, additional jewellery, ornament, or other items as well as their posture that may suggest their identity as not just a ruler, but an identifiable one. While further scholarship continues to identify images and better understand both patterns of patronage and the dating of images, paintings, and buildings, the function that these portraits served within the construction of sovereignty needs to be emphasised.³³ The essentially social and political purpose of these images of rulers from noble lineages qualifies describing them as ‘state portraits’ for which the primary purpose is not the portrayal of an individual as such, but the evocation through his image of those abstract principles for which he stands.³⁴ They embody the ideals of south Indian kings as the gods’ paramount devotees, in a prominent display of devotion important to the maintenance of dharma and social well-being in a dynamic political era. As Caleb Simmons has eloquently demonstrated in his Devotional Sovereignty—a study of kingship and religion at the courts of Mysore under Tipu Sultan (r. 1782–99) and Krishnaraja Wodeyar III (r. 1799–1868)—devotional portraiture (stone and metal images, as well as paintings) served alongside genealogies in both literature and epigraphy as a primary means of king fashioning in early modern South Asia.³⁵

In south Indian temples, portrait images such as these are located alone or in pairs primarily in corridors, in the gateways of gopuras, and in festival maṇḍapas (detached columned hall). The common theme is that these are all processional routes or places where deities are normally absent but at certain times present, principally during the regular festivals that increased in both number and scale from the thirteenth century. The devotional portraits greet deities when they are moving or temporarily enthroned during festivals; they are to be seen taking darśana of the deity. But this is a three-cornered relationship between deity, king, and devotee. Not only do the god and king greet each other, but the priests, devotees, or worshippers see both the king and the deity greeting each other when the two are assembled for a festival. The king, a frequently inaccessible figure in his palace, is given permanent presence in the temple in a life-sized representation and in locations there that are widely accessible and visible. At festival periods, the ruler's public display of devotion to the deities of the temple was seen by worshippers. Individual kings often appear with one or more of their queens on a smaller scale attached to the same stone column.

Though such likenesses more often appear alone or in pairs on either side of a corridor or within a festival maṇḍapa, in some temples, a long row of such figures may be found, sometimes facing one another. The multiplicity of such figures may create genealogical galleries of rulers, the present ruler being joined by his historic predecessors—a visual counterpart to the genealogies in royal inscriptions that traced a ruler's descent from the Sun or Moon, the Surya- or Candra-vamśa, respectively, and the gods themselves.³⁶ The best-known of such portrait galleries is the series of Madurai Nayakas, including Tirumala (r. 1623–59) and his predecessors, in the ‘New Hall’ (puṭumaṇṭapam) built by him in Madurai in the 1630s but other genealogical groups of sculpted images can be identified from the 1560s–80s and later (see Figure 4). In sixteenth- to eighteenth-century South Asia—a period when political relationships were fluid and dynastic succession never assured—both pictorial and sculpted portraiture was an additional means to assert current legitimacy and age-old authority; it constituted a performance of devotional sovereignty.

Figure 4. Portrait gallery, Nellaiyappar temple, Tirunelveli. Source: Crispin Branfoot.

Before considering the potential identities for some of the images at Rameshvaram, an outline of their overall number throughout the temple would be helpful. There are currently 108 across the whole temple—a number of rich numerological significance within South Asian religions. Tempting though it may be to ascribe significance to this, the images identified are not evenly distributed nor do they necessarily date to the same period; they are not indicative of a single wholly inclusive conception. The 108th, for example, which is a standing image of Bāskara Sētupati (r. 1889–1903) (see Figure 5), was only added to a new extension on the east side of the temple in around 1970 prior to the kumbhabhiṣekha (reconsecration) in 1975. Even if some images of Setupatis were reportedly damaged in the nineteenth century, as discussed below, the varied locations suggest alternative, less coherent interpretations. No other south Indian temple has so many life-sized male portrait images as at Rameshvaram. The closest contender is the similarly dated seventeenth- to eighteenth-century Ātmanāthasvāmi temple at Avudaiyarkoyil (Tirupperunturai), 80 kilometres north of Ramanathapuram. The identities of the images at Avudaiyarkoyil are similarly difficult to determine but include the eighteenth-century Tondaiman rulers of nearby Arantangi as well as some of the Setupatis from Ramnad.³⁷

Figure 5. Image of Bāskara Sētupati (r. 1889–1903) added circa 1970, Rāmanāthasvāmi temple, Rameshvaram. Source: Crispin Branfoot.

Though the great number is extraordinary, the pattern of their distribution is shared with many other temples in the far south, primarily being located along corridors, in the gateways of gopuras, and in festival maṇḍapas. At Rameshvaram, only nine are located within the innermost first and second prākāras. Two very large two-metre-high brightly painted figures, each bare-chested to the waist, wearing a long patterned veṣṭi around their legs and with tall conical cloth hats (kuḷḷāyi) on their heads, stand either side of the large image of Śiva's bull-mount Nandi before the entrance to the innermost enclosure leading to the Rāmanātha shrine. This style of tall cloth hat, sometimes with detailed embroidered patterns, was adopted by Vijayanagara kings and courtiers from the late fifteenth century and was worn with some variations in form (straight or pointing forward) by both their successors and by the many Nayakas into the first two decades of the seventeenth century. In this period, a tighter-fitting cloth hat that fell to one side was worn in its place by many Madurai Nayakas including the many figures identified as Tirumala Nāyaka (r. 1623–59).³⁸ These two figures in Rameshvaram were identified by Burgess in 1883 as the first two Madurai Nayakas, Viśvanātha (r. circa 1530–63) and Kṛṣṇappa (r. circa 1563–72); he reports that an inscription on the nearby entrance to the first prākāra identifying the two had been destroyed shortly before his visit.³⁹ This identification is supported by their formal similarity to images identified as of the same rulers in Madurai's ‘New Hall’ (puṭumaṇṭapam) though the pair in Rameshvaram are taller.⁴⁰ A further two portraits before the Ammaṉ shrine located alongside 14 images of smaller goddesses, including the Aṣṭalakṣmis (Eight Goddesses), in the Śukravāra (Friday) Maṇḍapa are identified as Setupatis on the basis of their formal appearance.⁴¹ Burgess was informed that these are the early eighteenth-century Hiraṇyagarbhayājī Ravikula Vijaya Raghunātha Sētupati (r. 1710–25) and his brother Kadamba Tēvar.⁴² Within the second prākāra corridor in the east gopura, there are three images in the south-east corner (see Figure 6) and a further two at the north end just before the Yāgaśālā (sacrificial hall). Two of these were understood to be Tirumalai Sētupati (r. circa 1645–73) and his son Raghunātha, the patrons of the second prākāra corridor in which they are situated.⁴³

Figure 6. Three images, south-east corner of second prākāra. Source: R. K. K. Rajarajan.

The remaining 98 standing male images in añjalimudrā attached to columns are all in the outermost areas of the temple: a group of 12 in the Cokkattam corridor near the west gopura and the remaining 86 located on the east side of the temple around the main entrances. As indicated on the plan (see Figure 2), these 86 images are located in five places: outside the eastern prākāra wall, there is a group of 22 in the kalyāṇa (wedding) maṇḍapa (see Figure 2: C) and a further group of 10 in the adjacent corridor on axis with the Rāmanātha shrine. The remaining 54 are clustered on columns along the north–south corridor of the third prākāra, around the entrance leading towards the Amman shrine (20) and Rāmanātha shrine (28), and six more facing each other across the aisle further north along the same corridor towards the south-facing Sabhāpati (Naṭarāja) shrine. As discussed below, the latter four groups, with a total of 64 approximately life-sized standing male images (10+20+28+6), are the most problematic to explain simply with reference to the patronage of the local Setupati Dynasty. But first, we shall consider each group in turn.

The Cokkattam corridor leads from the western gopura directly to a door into the second prākāra, though this is usually closed. In the first section before the axial junction are 12 such standing male images either side of the aisle (see Figure 7). Beyond these, both before and after this corridor meets the perpendicular one beyond, are further large sculpted images of various deities, folk figures, ascetics, and yāḷis (mythical rearing lion-like animals). All but one of the 12 near life-sized male figures stand alone and with their hands placed together in greeting. They wear turbans on their heads, as is more common from the late seventeenth century; are bare-chested in standard Hindu manner, with a veṣṭi around their legs; and wear elaborate jewellery and punch-daggers (katār) at their hips, all suggesting their elite status.

Figure 7. Cokkattam corridor between west gopura and second prākāra. Source: Nicholas and Co., circa 1884. © British Library Board, Photo 1003/(2347).

The most prominent and elaborately sculpted figure at the centre of the south side of the aisle, and the only one with two smaller wives alongside him, is identified as Muttu Rāmaliṅga Sētupati in Burgess's account (see Figure 8).⁴⁴ He reigned from 1763 to 1795, though temporarily deposed from power from 1772 to 1781, and it was in the 1770s–80s that the third prākāra corridor was completed. Like many of the eighteenth-century figures at Rameshvaram, he wears a turban rather than the rounded cloth cap falling to one side favoured by the Madurai Nayakas from the early seventeenth century. Such sartorial change is indicative of the increasing Persianisation or Mughalisation of court dress across southern India in the eighteenth century, evident in the wall paintings in the Rāmaliṅga Vilāsam in Ramnad palace dated to circa 1715–25 and in contemporary Dutch accounts.⁴⁵

Figure 8. Muttu Rāmaliṅga Sētupati (r. 1763–95), south side of Cokkattam corridor. Source: Crispin Branfoot.

All but one of the 12 images have the three horizontal lines of ash (vibhūti) denoting a Śaiva devotee on their foreheads, appropriate for this great temple to Śiva; however, the one facing Muttu Rāmaliṅga has the u-shaped Vaiṣṇava nāmam on his forehead. Burgess was informed in 1883 that the statues in this corridor are of Muttu Rāmaliṅga and of his two mantrīs (ministers): Muttiruḷappa Piḷḷai, son of Saundara Pāṇḍiyan Pillai; and Kṛṣṇa Aiyaṅkār, with members of his family—a more detailed identification than any previous scholar had suggested.⁴⁶ Given his name, the latter may be identified as the figure with the Vaiṣṇava forehead mark. Though a Śaiva temple, a shrine for Paḷḷikoṇḍa Perumāḷ (Viṣṇu) is present in the north-west corner of the first prākāra of the Parvatavarttiṉi Ammaṉ temple.

While many of these images are relatively generic in appearance, sartorial details and the sculptors’ depiction of more life-like postures, with the weight shifted from one leg to the other, suggest some intention to depict the presence of a series of historic individuals. Unlike the majority of others within this temple, these stone sculptures are not plastered and painted to the same degree. Though it is not possible to identify the remaining nine images with named historical Setupati predecessors, the seeming omission of several of the very short-lived rulers who reigned for less than a year in the many disputes over dynastic succession between 1605 and the start of Muttu Rāmaliṅga Sētupati's reign in the 1760s suggests that this images of this corridor were one element in the Muttu Rāmaliṅga Sētupati's demonstration of his dynastic genealogical claims.⁴⁷

Burgess and Natesa Sastri were able to obtain identifications for some of the nine donor images of Setupatis within the first and second prākāras in the 1880s, in addition to those in the western Cokkattam corridor. But this still leaves the identities of the greatest concentrations of these figures, over 80 in number, on the east side of the temple uncertain. These include the many situated along the main entrance corridors leading toward the two east gopuras and the inner enclosures of the Rāmanātha and Ammaṉ shrines. Visibility may have been a factor in their distribution, for only the third prākāra was accessible to all castes (including Europeans); before the twentieth century, untouchables were not even allowed that far within.⁴⁸ They are all attached to columns looking down upon pilgrims entering the temple from the east side: a short distance away is the sea where pilgrims bathe. The images may be placed above the heads of pilgrims, but they are on a similar level as the deities when they emerge from their dark, central shrines and are carried in procession on palanquins or other vāhanas (vehicles, often in the form of a mythical animal) upon the shoulders of groups of men, not only all around the long corridors, but also to the streets outside the temple.⁴⁹

With the exception of one group, none of these life-sized images has clear identifying adjacent inscriptions or painted labels, but the majority may be dated to around the 1740s–80s when the third prākāra corridor was under construction. The majority are very similar to one another, standing bare-chested wearing a knee-length or longer cloth veṣṭi with hands pressed together before the chest, and with little to distinguish one from another. Some have one or more smaller subsidiary male figures alongside. Elsewhere in the temples of the Tamil region, smaller adjacent women are more commonly placed alongside portrait images. An unusual feature of this temple is the degree to which the interior stonework of columns, sculptures, platforms, and ceilings is both plastered and brightly painted. In part, this is to obscure the rough appearance of the sandstone used to build the third prākāra corridors, for harder granite is routinely used for temples built on the mainland in the similar period where it is more readily available. The selective use of granite at Rameshvaram for the series of 12 images of the Setupatis in the western corridor suggests a deliberate decision to make these images, among all of those in the third prākāra corridors, of finer quality and individuated though now unpainted. However the great number of standing male images in the temple are explained, the minimal sculptural distinction and formal sameness among many (especially those on the east side of the temple) enhances the appearance of collective cohesion—that they should be perceived as a group and not primarily as individual likenesses.⁵⁰

Inscriptions that identify the subjects of stone Tamil Temple portrait images are rare between the sixteenth and nineteenth centuries. The only clearly visible labels alongside the many portrait images in this great temple are those in the kalyāṇa (wedding) maṇḍapa (see Figure 2: C) built beyond the outermost enclosure wall on the east side and axially aligned with the Ammaṉ shrine.⁵¹ Described by Colin Mackenzie in 1796 and just visible in Henry Salt's 1804 drawing of the east side of Rameshvaram, one of the earliest European depictions of the temple, this maṇḍapa was probably completed shortly before in the late eighteenth century.⁵² Inside are 22 images with painted labels above each that create a visual genealogy of the Setupatis from Śaḍaika Tēvar alias Uḍaiyān Sētupati (r. circa 1605–22) through to Ṣaṇmuka Rājeśvara born in 1909 and installed as Setupati in 1929 (see Figure 9). The inclusion of five Setupatis who ruled in the nineteenth and early twentieth centuries long after the maṇḍapa was completed in the late eighteenth century suggests subsequent modification: the identifying painted labels—rather than incised inscriptions—have been added later and the images of the most recent Setupatis may then have been remodelled in contemporary dress. For example, the image of Bāskara Sētupati (r. 1889–1903) in this maṇḍapa wears a long knee-length jacket covering his upper body, unlike the majority of images of earlier rulers—or indeed the image of this same Setupati added in the early 1970s—who are bare-chested and wear a veṣṭi. After a century of disputes over the management of the temple, the Setupatis were assured a hereditary position on the five-member executive committee of the temple in 1911. Hence, this remodelled maṇḍapa with the inclusion of a full and labelled genealogy of the Setupatis would seem to be a visual reassertion of their hereditary rights to temple honours in the 1920s–30s after over a century of conflict.

Figure 9. Three Setupati images on south side of kalyāṇa maṇḍapa. Source: Crispin Branfoot.

Temples are built, added to, modified, and even completely rebuilt so care is needed in evaluating the historic architectural fabric dating to before 1800 on the assumption that little has subsequently changed. This is not the place to develop a detailed building history of this vast temple, but any evaluation needs to take into account the significant renovations that took place from the late 1890s through to the 1920s. The only published detailed surveyed plan of the temple was produced in 1883 before these transformations.⁵³ As mentioned above, the superstructure of the east gopura—ruined, or at least incomplete in Salt's 1804 view—was only completed between 1897 and 1904. But the most significant modifications to the nineteenth-century temple in the first decades of the twentieth century were concentrated on the shrines and corridors of the first and second prākāras (see Figure 10). As indicated in the photo produced in 1905 for the Archaeological Survey at the time of the renovation programme, the sacred heart of the temple was previously open to the sky while today it is completely roofed over like the many other Śaiva temples built or substantially renovated in this period.⁵⁴ A significant aspect of this change was to replace the weathered, friable locally sourced sandstone of the earlier temple with finer-quality, harder granite columns that were brought ready-carved from inland quarries near Ambasamudram.⁵⁵ Halfway along the south side of the second prākāra corridor, the incomplete renovation is demonstrated at the juncture between the new and higher granite columns and the shorter, rougher old sandstone columns that in 2011 had yet to be replaced.

Figure 10. Main shrine to Rāmaliṅgēśvara under renovation in 1905. Source: ASI © British Library Board, Photo 1008/8 (2019).

Why does this early twentieth-century renovation project matter? Because these substantial—and at the time deeply controversial—renovations may have resulted in the removal of images of the Setupati donors created in the eighteenth century or the modification of their appearance to suggest an alternative identity. Even though the most substantial changes to the temple took place in the inner two prākāras, which, as discussed above, have the fewest number of portrait images, there are documented occasions on which the Setupati images in this temple were deliberately targeted by their rivals for control of the temple. In a long-running dispute with its origins in the disruption to temple affairs of the 1770s and later played out in the colonial law courts between the 1830s and the 1880s, the Setupatis’ control of the administration of the Rameshvaram temple and of its rich endowments was challenged by a series of priests. The conversion of several Setupati images into ascetics by adding beards, together with the disappearance of some copperplate Setupati inscriptions and the forging of ones that claimed to be ancient with no mention of the Setupatis patronage or rights, were all used to successfully argue that the Setupatis did not have ancestral privileges. It was claimed instead that a great sage (muni) named Rāmanātha had built maṇḍapas and renovated the temple in the early 1600s and not the Setupatis, at precisely the time at which the Setupati lineage had been established by the Madurai Nayakas.⁵⁶ A Setupati portrait in the second prākāra was reported to have been broken in the late 1870s and the pieces thrown away.⁵⁷

Until the early 1890s, it was the priests and the paṇḍāram (a senior priest in control of temple resources) rather than the Setupatis who were in control of the temple at Rameshvaram. After a brief period of greater authority in the early years of Bāskara Sētupati's reign (b. 1868, r. 1889–1903), a wealthy family from the Nakarattar (Nattukkottai Chettiar) caste from nearby Devakkottai led by Al. Ar. Ramasamy Chettiar was able to take control of the management of the temple in 1900 and initiated the controversial renovation. The details need not detain us here, but it is worth noting that Bāskara himself was concerned in 1901 that, if the renovations went ahead, more of the Setupati images in the temple would be moved, altered, or removed. He was worried that the temple at Rameshvaram may meet a similar fate to that at Avudaiyarkoyil, where recent renovations under Chettiar patronage had led to statues of the Setupatis being demolished and replaced with those of the Chettiar and the paṇḍāram who had led the renovation being put in their place.⁵⁸ These events demonstrate the continued role of portrait images of past donors and inscriptions on the walls and copperplates in the possession of the temple in demonstrating temple rights and privileges.

With this cautionary note that the present number of standing male donor images in the temple at Rameshvaram may not necessarily correspond to the number erected in the eighteenth century, how the remaining 60 or more images—especially those on the east side of the third prākāra corridor (the groups of 20, 28, 6, and 10; see Figure 2)—may be identified remains to be addressed. One possible explanation for this large number is that the statues represent the long sequence of the early ‘mythical’ Setupatis, who preceded the clearly historical line established in around 1605. In this scenario, the unidentified images portray over 60 Setupatis, stretching from Rāma's original appointee up to—and perhaps including—Ramnad's contemporary kings. Although their historical existence is unknown, depicting these many alleged ancient ancestors in rows along the corridors of the Rāmanāthasvāmi temple could have served to convey the image of an age-old, more or less continuous and divinely sanctioned role as guardians of Rameshvaram and the Sētu.

But, while this may be one solution to identifying these statues, there seems to be no other evidence as yet to support it, particularly with regard to the number of over 60 kings. Depending on whether these images do or do not include the clearly historical Setupatis—about 15 men in the 1780s—we are left with between circa 49 and 64 unidentified rulers. There are no written sources that can somehow be linked to these specific numbers. Inscriptions, chronicles, or other texts listing all the names of the Setupatis’ mythic ancestors or otherwise detailing their ancient genealogy have not been identified. As mentioned above, all that is available for this early period are data such as a few named individuals, a dozen anonymous generations, and an undated span of six centuries. Those figures can hardly account for this particular number of statues. However, it is of course still possible that the mythical Setupatis are actually portrayed here, in which case the host of rulers may symbolise more generally the notion of an unquantified multitude or completeness, further promoting the claimed antiquity of the dynasty and uninterrupted protection of the temple.

Another possibility, rooted in the political structures of the south-east corner of the Tamil country, is to identify these figures with the pāḷaiyakkārars (‘poligars’).⁵⁹ In the foundation histories of the Madurai Nayakas, Viśvanātha (r. circa 1530–63) was understood to have divided his country into 72 regions (pāḷaiyams), each ruled by local chief.⁶⁰ Each pāḷaiyakkārar was designated to defend one of the same number of bastions around the fortifications of Madurai, thus emphasising the hierarchical protective role of this group. According to some traditions, within this group of 72, the rulers of Pudukkottai, Shivagangai, and Ramnad were considered to be of higher status than the others.⁶¹ Therefore, by depicting the pāḷaiyakkārars in the Rāmanāthasvāmi temple, the Setupatis may have wished to demonstrate their superiority over these chiefs. Once again, this figure is close to the maximum number of 64 unidentified figures discussed above—closer still if three or more are excluded as being members of the elevated inner circle of pāḷaiyakkārars.

But, given the pan-Indian status of this temple that transcends its regional significance to the Setupatis alone, it might be more logical that these are images of other Indian kings or their representatives rather than local chiefs. Besides, various versions of a Tamil text on the Setupatis’ Maṟavar caste, called Maṛavar jāti kaifīyat or closely related titles, state that, while some pāḷaiyakkārars were indeed subordinated to the Setupatis to various degrees (having to stand or prostrate themselves before them), others actually held an equal status (being invited to sit with them).⁶² So one may wonder why the Setupatis would depict this diverse group as a collective in this temple. Perhaps examining seventeenth- and eighteenth-century Dutch sources on the Rāmanāthasvāmi temple—rarely considered by many scholars for the interpretation of temples—may offer further clues to the puzzle of the massed images.

Rameshvaram and the Dutch

By the time the sculptures in the third prākāra of the temple were created, the Dutch East India Company, also known under its Dutch acronym VOC (Verenigde Oostindische Compagnie), had been the main European power in the waters around Rameshvaram for about a century.⁶³ The Dutch set up their first trading station in Nayaka-ruled Madurai in 1645, at the port of Kayalpatnam. In 1658, the VOC concluded its first treaty with the Setupatis, after which it established trading stations in Ramnad's ports at Adirampatnam in 1674 (functioning for a few years) and at Kilakkarai in 1690 (lasting until 1825).

The Company settled down on Ramnad's shores not merely for mercantile reasons. This kingdom was also of strategic relevance because it controlled the Pamban Channel—one of only two places where vessels of a certain size could pass the string of sandbanks, reefs, and islets stretching between India and Sri Lanka (see Figure 11).⁶⁴ Called Adam's Bridge by European visitors, to Hindu devotees this string was of course the sacred Sētu, and the Pamban Channel lay right between Rameshvaram island and India's mainland. This passage was thus located at the junction of two important routes: one of high religious significance as it led to a major pilgrim destination and the other of great commercial value for being a vital link in the Indian Ocean trade.⁶⁵ The twofold nature of the channel made it a delicate spot and a potential source of conflict, as the relationship between Ramnad and the Dutch demonstrates. Even though treaties between the two parties declared that only the VOC was permitted to use the Pamban Channel, in practice, the Setupati court regularly let other traders, Asian and European, sail through it. This resulted in frequent disputes between Ramnad and the VOC, and even a few military confrontations, the latter in particular providing us with valuable Dutch observations of Rameshvaram island and its Rāmanāthasvāmi temple.

Figure 11. Detail of a Dutch map of south India and Sri Lanka showing VOC settlements in Madurai, Ramnad, and north-west Sri Lanka, probably an eighteenth-century copperplate print. Source: Nationaal Archief, The Hague, collection of foreign maps: supplement (access no. 4.VELH), no. 114. Courtesy: Nationaal Archief.

Early modern Dutch writings—such as VOC records, travel accounts, and treatises by Christian ministers—generally contain very little information about the Rāmanāthasvāmi temple itself. If anything, most texts briefly mention the pan-Indian significance of the temple or its fabled riches. For instance, in 1663, VOC Governor Rijckloff van Goens of Ceylon (Sri Lanka) wrote that this ‘very old pagoda’ was visited by ‘people from the whole of Hindustan, Orixa [Orissa] and Bengale’. The seventeenth-century pastor Philippus Baldaeus spoke of ‘a beautiful pagoda, very rich with cattle’, adding that ‘people say that infinite treasures are kept here’.⁶⁶

By contrast, numerous VOC documents deal with the Pamban Channel together with the adjacent Rameshvaram island as a whole, calling the latter ‘Pambe’, ‘Pembenaer’, and the like, or—somewhat inaccurately—variations on ‘Ramanacoil’, a corruption of Rāmanāta kōyil (Tamil for ‘Rāmanāta temple’). Because of the strategic value of the site, these records usually concern the VOC's continuous fear that other traders would use the channel and the Company's desire to take possession of the passage and the island. In fact, this became something of an obsession for the Dutch. In 1671, for instance, the VOC suggested to Madurai's governor in the Tirunelveli province, Kumāra Svāmi Mudaliyār, that they should attack Ramnad together, after which the Company would gain control of Rameshvaram island. This was strongly rejected, with the Tirunelveli governor stating that even listening to this proposal was a great sin. As a second example, two Dutch letters from 1674 discussing another possible occupation of Rameshvaram were written in a secret code, hiding their content in case they would be intercepted by competitors. This was a most exceptional measure, of which very few other cases remain in the VOC archives. Demonstrating the crucial importance of the island to the Dutch, the letters even include a special symbol for ‘Ramanacoil’, as the key to the code reveals (see Figure 12).⁶⁷

Figure 12. Detail of a VOC letter from Colombo to Amsterdam in secret code, with key, January 1674. Source: Nationaal Archief, The Hague, archives of the Verenigde Oostindische Compagnie (VOC, Dutch East India Company, access no. 1.04.02), no. 1292, f. 544. Courtesy: Nationaal Archief.

Yet another instance of the fixation of the VOC on controlling this location was an expedition in 1690 to construct a permanent watch post on one of the Sētu's sandbanks. Calling it ‘Kinkhoest’ (whooping cough), the Dutch intended to place poles and plants on the sandbank and raise its level. While this proved too dangerous because of unpredictable sea currents, during the same exploration, the VOC stationed a few guards on an islet close to Rameshvaram island, from where the Rāmanāthasvāmi temple and the nearby Pamban fort were visible. As the report from the expedition makes clear, the sacred status of the area was well known to the Dutch. They wrote that the ‘river’ separating the islet from Rameshvaram island (probably the small channel east of Dhanushkodi) was visited by ‘heathens’ to bathe in, thereby ‘honouring the brave and strong Rama … for killing the ten-headed Rawanna’.⁶⁸

Despite this awareness, later in 1690, their decades-old frustration finally got the better of the Dutch and in August they attacked Rameshvaram island.⁶⁹ The direct causes of this campaign were the forced removal by Ramnad of the VOC guards from the abovementioned islet and an incident in the Pamban Channel in which VOC vessels had halted due to wind currents. As the Dutch claimed, Ramnad troops then fired at them and killed one or two men. The subsequent invasion of Rameshvaram island by the Company proceeded smoothly. The VOC easily took the small Pamban fort next to the Pamban Channel, while most of Ramnad's forces fled to the Rāmanāthasvāmi temple, described by the Dutch as ‘a heavy building of suitable blue stones and high walls’. Reluctant to assault the temple itself and foreseeing problems with the many pilgrims arriving ‘daily from all quarters of the Indian [Indische] world’, the VOC quickly negotiated a new treaty with the Setupati, Kiḻavan Tēvar (r. 1673–1710), and withdrew from the island in September.⁷⁰ All in all, the attack had been successful: one of the clauses of the treaty stated that Ramnad now permitted the Dutch to erect a small dwelling next to the Pamban fort on the island and place a few men there to monitor the traffic using the Pamban Channel (see Figure 13).⁷¹

Figure 13. Details of a Dutch map of south India and Sri Lanka (north on the right) showing Ramnad, the Sētu, and north-west Sri Lanka, and inset of Rameshvaram island indicating the Rāmanāthasvāmi temple and the Pamban fort, from a manuscript by J. C. Toorzee, circa 1690–1700. Source: Nationaal Archief, The Hague, collection of foreign maps (access no. 4.VEL), no. 923. Courtesy: Nationaal Archief.

Still, the following decades continued to see much discord between Ramnad and the VOC, often caused by disagreements about who was entitled to pass through the Pamban Channel.⁷² In May 1746, these tensions escalated into another Dutch attempt to conquer Rameshvaram island, bolstered by the Company's fear that Indian powers invading the region in this period—such as Arcot and the Marathas—might capture the allegedly wealthy Rāmanāthasvāmi temple and next cross the Sētu to attack Dutch-controlled Sri Lanka.⁷³

But this time, the campaign of the VOC turned out to be a failure. Although the Dutch quickly occupied the Pamban fort, they were confronted by several insurmountable complications. The VOC was incapable of dealing with Ramnad's unexpected, often nocturnal, guerrilla-like raids on the fort from the surrounding bushes, after which the assaulters hid in the various temples on the island. Further, many Company soldiers were smitten with hot and cold fever, tightness of the chest, and dropsy—a regularly fatal condition that they labelled ‘the Pambe disease’ and was perhaps malaria. Disciplinary and logistic problems arose as well, lowering the morale of the forces even more. Meanwhile, as the Dutch had blocked entry onto the island, pilgrims were stuck on the mainland across the Pamban Channel, numbering 4,000 by mid-July. Eventually, in January 1747, the VOC had no choice but to withdraw the few troops that were left, without gaining anything from the expedition.⁷⁴

But there was an additional factor compelling the Dutch to vacate Rameshvaram island, which provides an alternative clue to the many unidentified statues in the Rāmanāthasvāmi temple. For, as the VOC found out during its two invasions, Ramnad's Setupatis were apparently not the only kings holding a mandate to guard this sanctuary. During the campaign of 1690, while the Dutch contemplated the wisdom of attacking the temple, the Indian merchant Babba Prabhu (‘Babba Porboe’ in VOC documents) informed them about the special relationship between the temple and a large number of Indian rulers. Babba Prabhu's explanation was included in a report of the expedition in 1746, from which a section is translated below:

He [Babba Prabhu] told us [the Dutch on Rameshvaram island in 1690] that nowadays the entire sanctuary [Rāmanāthasvāmi temple] fell under the number of 56 kings and monarchs [koningen en vorsten], who each—the one more, the other less—were used from the olden times onward to keep together a number of people in that pagoda, who were sent there as protection men [schuts mannen] for them and represented them over there; that even now at this hour there were still such persons who had been sent on behalf of the old kings of Carnatika [Vijayanagara] or their descendants. Regardless of the fact that that empire was overpowered so long ago, first by the king of Golconda and later by the Mogol [Mughal], however, this devotion [divosie] was still observed and followed.

Over all these protection men, the Theuver [Tēvar, the Setupati] was the chief protection lord [opperste schuts heer], without being allowed in the least to meddle in the government of the pagoda, as it belonged here to this island, while each [protection man] belonging to the pagoda did not have to recognise anyone but the king by whom he was sent, which kings in their countries have endowed the pagoda with—the one these, the other those—lands and revenues, through which their representatives are fed two, three and more times, and all offerings [are] settled, and the Braminees [Brahmins], depending on this, [are] maintained. But the militia of the Theuver is never being fed and only twice a day provided from the other side [of the Pamban Channel] with nelij [nelly, rice in the husk], betel [-leaf], tobacco, vegetables, oil and so on, wherefore one might believe that these [the Setupati's militia] could be starved out in a short time.

The duty of the mentioned protection men was that they should stay at the pagoda during their entire life and in case of a siege or war had to defend [it], without being able to yield or flee, but being obliged to give their life for that [temple], in which case the king has to avenge their death and has to wage war against the one who has shot dead and taken the life of his people for the pagoda, whoever that is, wherefore the protection men are like sworn and sacrificed [verswooren en opg'offert] to the pagoda. Everything being like that, it was a very dangerous and delicate matter for the Honourable Company [VOC] to proceed and enter that pagoda.⁷⁵

Added to this passage, first recorded in 1690, was a remark that this situation had not changed by 1746. Additionally, after the Dutch had taken Rameshvaram island in 1690, merchant Babba Prabhu sent a letter on behalf of the VOC to ‘all Tannataars [thānadārs, chiefs] or overseers of the Ramanacoijl pagoda and inhabitants of the village around it, as well as envoys of all heathen kings, foreigners and saints arrived there to do service [worship]’. While this letter was an attempt to justify the Company's occupation, it also suggests that the story of the connections of the temple with multiple kingdoms was well known among local parties:

[Y]our Honourables know that the Ramanacoijl pagoda and Cassie [Kashi, Benares or Varanasi] are highly estimated places of all heathen kings and [are] common [gemeen] to all peoples of 56 countries but not to one [country], therefore the Honourable Company will take care of the same.⁷⁶

Judging from Babba Prabhu's explanations, it appears that a whole range of Indian kings—56 of them to be exact—were responsible for the protection of the Rāmanāthasvāmi temple and endowed it regularly. Among those rulers, Ramnad's Setupatis allegedly had a somewhat special position but not an exclusive mandate, unlike what was claimed in their temple inscriptions and foundation stories. They were not supposed to interfere in the management of the temple or exercise control over representatives of other kings. This arrangement apparently existed in the late seventeenth and eighteenth centuries but was said to date from the period of the Vijayanagara empire (circa 1340s to 1640s), so it was perhaps several centuries older—in which case the Setupatis were probably only added to the group of protecting rulers in the early seventeenth century.

The merchant mentioned here as the source of this information was most likely the celebrated Babba Prabhu—a Sārasvat brahmin with Konkani origins operating from the Malabar (Kerala) coast. He was one of the most prominent south Indian trading partners of the VOC from the 1660s to the 1690s, and also conducted activities in the Madurai region. In the years surrounding the first Dutch attack on Rameshvaram island, he maintained a warm personal relationship with the VOC commissioner, Hendrik Adriaan van Rheede, who played a central role in this campaign. Besides, Babba Prabhu occasionally served as an official envoy of the Dutch, was granted special privileges by them, and had provided them with local knowledge before.⁷⁷ Considering his background, his close ties with the VOC and his aforementioned letter to Rameshvaram's people, it seems his description of the special status of the Rāmanāthasvāmi temple was largely reliable. As one Dutch document states, there was no doubt that ‘merchant Babba will offer his old diligence, assistance and help regarding this for the benefit of the Honourable Company’.⁷⁸

Babba Prabhu's story proved accurate indeed. Almost as soon as the VOC had set foot on Rameshvaram island—during both the 1690 and 1746 expeditions—it started receiving letters from various south Indian kings, urging the Dutch to withdraw immediately because of the sacred status of the spot. In 1690, within two weeks after the first conquest, Madurai's King Muttu Vīrappa Nāyaka III (r. 1682–91) and his governor Venkaṭādri Nāyaka in Tirunelveli wrote to Babba Prabhu in his capacity as a VOC representative. They warned him that an army had been raised to assist Ramnad in regaining Rameshvaram island. They further threatened to destroy all Company settlements in Madurai territory so that ‘the name of the Hollanders was heard no more’ in the kingdom. Sometime later, a prominent courtier of the Tanjavur kingdom, Bābojī Paṇḍidar, informed Babba Prabhu that the conquest by the VOC was shocking since Rameshvaram was a sanctified place, unlike any other, and was located ‘in the middle of the world’. If the Dutch harmed this ‘holy washing site’, the Tanjavur king could not possibly live in peace with them.⁷⁹

In 1746, about a week after the second invasion of the island, the ruler of Shivagangai wrote to the Dutch, saying that they could occupy Ramnad all they wanted but should leave Rameshvaram island. Failing to do so would lead to ‘numerous enemies from the region Casie [Kashi, Varanasi] down to Caab Commorijn [Cape Comorin, Kanyakumari]’, as pilgrims from these lands were not going to accept the presence of the VOC at Rameshvaram.⁸⁰ Soon after, the Dutch received a letter from King Pratāpasimha Bhonsle of Tanjavur (r. 1739–63) that recalls Babba Prabhu's explanation of 1690:

[T]he security and safety on Pambe [Rameshvaram island] is a point to which the religion [geloof] obligates most kings and monarchs of these Indian regions, with all their servants. Pambe, my lord, is a miraculous place, widely famous for the innumerable wonders that have been performed there since many centuries by our God Ramanade.⁸¹

Tanjavur's king further stated that, because of its attack, the VOC had made ‘uncountable enemies’, just like Shivagangai's ruler had declared. About a month later, King Pratāpasimha wrote to the Company again, repeating that the sacredness of Rameshvaram was ‘known in the world’ and many people had ‘an interest [belang] in it’. He warned the Dutch that their actions had left him no choice but to ally with Shivagangai and Ramnad itself, after these kingdoms—because of the occupation of Rameshvaram—had already been forced by him to end a war. And this coalition was only the beginning because (as the Dutch translation of the letter goes) ‘through the similarity of religion, all heathen kings of these eastern sea regions, as well as their servants, have come with us, standing ready to defend together the safety of Pambe against all attackers’.⁸²

After another month, the Dutch received news that the mother of ‘Nimaliger, a king of the Moratten’ (perhaps the Maratha Chief Jānojī Nimbāḷkar of Phaltan in western India) had gone on a pilgrimage to Rameshvaram, but was now stuck in mainland Ramnad because of the invasion by the VOC. Her son had allegedly assembled 60,000 troops to expel the Dutch if they would not retreat soon and ‘certainly all the other heathen kings would put on the harness against the honourable Company’. Rumour had it that the Maratha army had advanced as close as a four-day journey from the island and consulted with Ramnad, Shivagangai, Tanjavur, and Madurai about further steps.⁸³ It was also said that Tanjavur's king had received letters from rulers all over India asking him to attack the VOC at the earliest or else wait for the arrival of their armies. These rulers may have included the kings of Travancore and other Malabar states, given the repeated efforts by the Dutch to convince them that the Rāmanāthasvāmi temple would not be harmed.⁸⁴

One more indication in the VOC archives of the close connections between Rameshvaram and various Indian kings is found in the report of the conclusion of the treaty between Ramnad and the Dutch after their invasion in 1690. Taking place under a tamarind tree near the capital Ramanathapuram, the signing of the agreement was attended by ambassadors of several Indian kingdoms. Following an extensive discussion about the permission given by Ramnad for the Company to station a few men on the island, the Setupati, Kiḻavan Tēvar, requested the VOC envoys to read aloud all clauses in the Dutch version of the treaty. An interpreter then had to explain everything to the Setupati, who compared this with the Tamil version of the treaty. Next, Kiḻavan Tēvar asked the various Indian ambassadors whether they agreed with all of the clauses. Upon their approval, he signed the two versions of the treaty, showed his signatures to the Indian ambassadors, and had the VOC envoys sign as well.⁸⁵ Clearly, the Setupati needed or desired the consent of other rulers before he would allow any Dutch presence on Rameshvaram island.

This close involvement of several kingdoms rather than a single one when a temple came under attack does not seem to have been a common phenomenon. In February 1649, during a conflict with Madurai, VOC forces landed on the shore of Madurai at the town of Tiruchendur—the site of one of the most significant temples dedicated to Murukaṉ (or Subrahmaṇya) in the Tamil country and an important pilgrimage destination. Not only did the Dutch occupy, burn, and heavily damage this temple and let their flag fly from it, but they even took some of the deity images from the temple with them to Sri Lanka, reportedly including the stone image of the main god ‘Perimal’ (Perumāḷ), as the Dutch called him. The VOC later returned the images to the temple, where, in early 1653, elaborate reconsecration ceremonies were conducted so that worship could resume.⁸⁶

This campaign had an enormous impact on the temple in Tiruchendur, damaging its building, desecrating its images, and disrupting its activities for several years. In contrast, during the two Dutch invasions of Rameshvaram island, no VOC troops ever entered the Rāmanāthasvāmi temple and its access for pilgrims was interrupted for only a few months. Yet no VOC documents suggest that, after the expedition to Tiruchendur, any Indian kingdom apart from Madurai itself urged the VOC to withdraw from the temple or to return the statues, let alone threatened to attack the Dutch. It therefore seems that Rameshvaram really enjoyed some pan-Indian status among kings, as demonstrated by the exceptional engagement of several Indian rulers when they felt that the Rāmanāthasvāmi temple was put in jeopardy by the VOC.

Given the close connections between Rameshvaram and a range of kings, the question arises as to whether these multiple rulers may have been connected to the unidentified statues in the Rāmanāthasvāmi temple. Do the circa 60 images along the eastern corridors perhaps represent the 56 kings mentioned by merchant Babba Prabhu as the protectors of the temple from all over India?

The 56 kings of Bhāratavarṣa

As we have seen, VOC records describing the repercussions of the Company's two invasions of Rameshvaram make it clear that several south Indian rulers indeed felt a responsibility to safeguard the sanctuaries of the island. But Babba Prabhu's statement that there was a formal and permanent protective relationship between the Rāmanāthasvāmi temple and 56 kings does not appear to be supported by other sources, either Indian or European. At the very least, no earlier research on Rameshvaram points to anything resembling this arrangement.⁸⁷ For instance, various studies have considered the aforementioned eighteenth- to nineteenth-century dispute between the Setupati Dynasty and the paṇḍārams (senior priests) over the management of the Rāmanāthasvāmi temple.⁸⁸ There is nothing to suggest, however, that this conflict somehow stemmed from the alleged limited role of the Setupatis in the affairs of the temple, as related by Babba Prabhu.

It is thus worthwhile exploring what Babba Prabhu's explanation—recorded but perhaps not entirely understood by the Dutch—could actually have meant. To begin with, the number 56 may not have literally referred to 56 specific rulers alive at that time. This number had long been used throughout India in various settings and was perhaps considered significant for being the product of seven times eight. It seems that the number seven often denoted the highest desirable quantity or a sense of completeness. One may think of notions of the seven constituents of the body, the seven limbs or elements of the kingdom, the king's seven duties, and the seven worthy gifts, to mention some examples among many others in contemporary Indian texts.⁸⁹ In Purāṇic cosmology, seven seas and seven continents (dvīpa) surround Mount Meru toward the eight directions of space, and so the multiplication of the two indicates the expansive spread of the world. Furthermore, it has been proposed that the number eight, resulting from adding one to seven, implied a number beyond completeness: infinity.⁹⁰ Somewhat speculatively and paradoxically, multiplying these two numbers may thus have indicated a form of endless totality—encompassing all, but always leaving room for more, as it were.

At any rate, the number 56 has been playing a role in various contexts. In the religious sphere, it is indeed often the product of a multiplication, again suggesting a notion of wholeness. In Kashi (Benares, Varanasi), Purāṇic cosmology is mapped onto the topography of the sacred city and pilgrimage destination (tīrtha) with seven concentric circles of protective Gaṇeśas or Vināyakas arranged in the eight directions of space, all centred near the Viśvanātha temple at the heart of the city.⁹¹ In Puṣṭimārga temples for Kṛṣṇa in the Vallabha sampradāya (tradition), for example, there are eight services per day over the seven days of the week and there are seven svarūpas or principal images to which these eight daily services are made. In temples at Vrindavan, Mathura, and elsewhere in the Vraja region (or Braj, south of Delhi)—but also at other places throughout India—Kṛṣṇa receives 56 food items (chappan bhog) all at once during the Kṛṣṇa Janmāṣṭamī festival that marks the birthday of the deity. One interpretation of this event points to Kṛṣṇa's eight daily dishes multiplied by the seven days he lifted the Govardhan mountain to protect his people from Indra's massive rains and had to forgo his meals. Another explanation for this offering refers to the 14 worlds of the cosmos multiplied by the four basic food substances (chewed, non-chewed, and licked foods, and beverages).⁹² In Puri (Orissa), a lunch of 56 courses is even served every day to the deity of the Jagannātha Temple, also a manifestation of Kṛṣṇa.⁹³

In a more political and geographical context, the idea of 56 kings is already evident in the Mahābhārata, for this number of kingdoms of Bhāratavarṣa participated in the Kurukshetra war; this number of kings is also cited in some later Purāṇic sources. Another early reference is found in the Candragarbhasūtra, dating from the sixth century or earlier, which mentions 56 countries connected to manifestations of the Buddha. Late medieval and early modern works listing 56 kingdoms include the Sanskrit Ṣaṭpañcāśaddeśavibhāga, part of the Śaktisaṅgama Tantra (circa seventeenth century), aiming to guide pilgrims to holy sites around India.⁹⁴ Several south Indian texts from that period refer to the number 56 as well. For instance, the late fourteenth-century Kannada Padmarāja Purāṇa relates how the Hoysala King Nārasiṁha Ballāḷa receives envoys with gifts from 56 tributary countries all over India.⁹⁵ In the later Rāmarāja Bakhair (post-1565), available in various Marathi and Kannada versions, three sultans (including those of Ahmadnagar and Golkonda) also sit amidst representatives of 56 kingdoms.⁹⁶ Other early modern texts, in Telugu and Tamil, claim that the emperors of Vijayanagara reign over ‘the 56 kingdoms or provinces’ and their rulers, and that the ‘Dravida’ land (here probably the Tamil-speaking region) is one of India's 56 original realms.⁹⁷ In sum, there was a long-standing and widespread notion of 56 ‘traditional countries’, apparently signifying a totality or at least a multitude of kingdoms and dynasties that spanned the whole of India's space and time, commonly stretching from Kashmir to Kerala, and including ancient states long vanished by the early modern period, such as Magadha, Anga, or Kamboja.⁹⁸

Considering all this, merchant Babba Prabhu's mentioning of 56 rulers must have denoted the traditional number of kingdoms that together constituted India. Thereby, he made it clear to the Dutch that many kings, from all over India, maintained ties with Rameshvaram. As said, it is uncertain whether these links were as institutionalised as the Dutch rendering of Babba Prabhu's explanation suggests, given the apparent lack of other sources referring to such connections. What is certain, however, is that, over the centuries, Rameshvaram saw many visitors from all over India. These included not only pilgrims and devotees, but also kings harbouring more political aspirations, as outlined at the outset.

To return to the question posed above: Do the circa 60 images along the eastern corridors perhaps represent the 56 kings mentioned by merchant Babba Prabhu as the protectors of the temple from all over India? If a correlation between the ‘56 kings’ and the proliferation of standing, life-sized male images within the temple corridors at Rameshvaram is sought, then the 48 on the east side of the third prākāra corridor around the axial entrances leading inward towards the Rāmanātha and Pārvatavarttiṉi Ammaṉ shrines may be considered possible candidates. The addition of the six nearer the Sabhāpati shrine to the north results in 54 in this long corridor, which is tantalisingly close to the traditional mythical ‘56 kings’. The recollection that renovations—sometimes stemming from contestation over temple rights and privileges—are known to have resulted in the deliberate replacement or modification of life-sized portrait images at both Avudaiyarkoyil and here at Rameshvaram itself in the later nineteenth century may explain any potentially ‘missing’ images. This still leaves the ten in the entrance corridor on an axis with the Rāmanātha shrine to the north of the kalyāṇa maṇḍapa, placed—in a similar manner to the Setupati lineage in the western Cokkattam corridor—alongside sculpted and painted column images of ‘folk’ subjects including men and women carrying one another on their shoulders, often identified as members of the Kuravar community. If these may also be identified as Setupatis on the basis of this formal similarity, then pilgrims to the temple entering from either the west or the east through the corridors built in the middle of the eighteenth century would first pass images of the Setupatis—‘the protectors of the Sētu’—and then a protective group of the ‘56 kings’ standing in eternal devotion before passing through a monumental gopura toward the heart of one of Hinduism's most sacred shrines. By constructing these monumental corridors with row upon row of these images in the eighteenth century, the Setupatis wished all pilgrims to see not only their eternal protective presence with the deities of the temple, but also their position among the great classical kings of India.

Conclusion

So, what may we conclude from this exploration of the temple at Rameshvaram? Famous as a pre-eminent site of Hindu pilgrimage given its significance within the Rāmāyaṇa, the temple and the Sētu may also be understood as an important site for the performance from at least the tenth century of royal rituals such as the erection of a jayastambha (victory column) or the tulābhāra ritual, by rulers seeking to expand, augment, or define their authority. Though the present temple has been closely associated with the Setupati Dynasty, this wider political base for the patronage of the temple may then offer an insight into understanding of the proliferation of life-sized images of kings and devotees added to many temple halls and corridors in the seventeenth and eighteenth centuries. Previous studies have mentioned these images and simply stated that they are of the Setupatis, their family members, or ministers, but not sought to enumerate them all, nor be precise about their locations, scale, or purpose.

This article has not only mapped the location of the 108 life-sized images currently seen in the temple, but also suggested more precise identities than previously considered. These include images that can indeed be considered to be Setupatis, in small groups or larger historical genealogies, even without the additional evidence of an adjacent inscription or painted label. Two groups of images may persuasively be identified as genealogies of Setupatis: the 12 images in the western Cokkattam corridor (though two may be their ministers), one of the main entrances to the temple for many pilgrims, and a second group within the kalyāṇa maṇḍapa used for the annual ‘wedding’ festival on the east side of the temple. Both series of standing images date to the later eighteenth century, though, as argued above, the inclusion of early twentieth-century Setupatis and identifying labels in the kalyāṇa maṇḍapa suggest later modifications. But the great number of remaining images attached to columns along some of the long corridors within the temple suggest that they too are Setupatis or other royal donors, given the similar size, posture, depiction, and dress. Yet, as argued here, though such extensive visual dynastic genealogies are known in other south Indian temples, the great number of these life-sized images preclude any convincing identification with the Setupatis’ claims to a mythic genealogy stretching back to Rāma himself.

Another perspective on the identities of the circa 60 life-sized images on the east side of the temple that—together with the long corridors—are such a striking feature of this temple is suggested by contemporary Dutch sources stemming from the VOC's occupation of Rameshvaram island in 1690 and 1746. For the Dutch were informed that the temple was said to be protected by not only the Setupatis of nearby Ramnad, but also ‘56 kings’ of all India. Rather than a reference to contemporary kings, this is a reference to long-standing Purāṇic tradition that can be traced to the Mahābhārata that is cited in contemporary Deccani and other south Indian sources. But there is no mention in any of these Indian sources of the traditional ‘56 kings’ having a protective role over the temple at Rameshvaram—a connection evident from the Dutch archives. It is this observation that may offer clues to the presence of the rows of male images standing in devotion along the temple corridors.

By exploring the long-standing significance of Rameshvaram and the Sētu-tīrtha as locations for the performance of kingship by many medieval rulers and not only the Setupatis in the seventeenth and eighteenth centuries, this article has suggested that a significant number of the remaining anonymous images may be considered to be not simply ‘donors’, but specifically other kings of India or their representatives. Inscriptions on stone and copperplates, architectural sculpture in the temple, the VOC archives, and Purāṇic texts all combine to reveal further insights into the religious, political, and cultural significance of one of the largest and most sacred temples in India.