Cosmopolitanism in Early Historic and Medieval Sri Lanka

Whilst the Early Historic and Medieval periods of Sri Lanka are often presented as a golden age of Sinhalese and Buddhist achievement, there is also substantial evidence for a multiplicity of communities residing within the island during that time. This is unsurprising, given the island’s location on trade routes spanning the Indian Ocean, linking its communities with East Africa, the Middle East, South-east Asia and China. The physical evidence of this trade is clear from excavations within the island's ancient capital Anuradhapura and its major port, Mantai, with their assemblages of fine Chinese ceramics, glazed and bitumen-coated vessels from Mesopotamia, and semi-precious stones from Afghanistan (Carswell et al. 2013; Coningham 2006). The spread of Buddhism into the island also forged strong regional links with communities in northern India but, at the same time contributed to increasing differentiation between communities in the south of the Subcontinent. Indeed, tensions were also evident within and between the island’s Buddhist sects, which led to conflict on a number of occasions. Archaeological research in the hinterland around Anuradhapura has also identified a variety of competing, yet complementary ritual foci, indicating that whilst Buddhism may have been regarded as the official religion of the island, other traditions and belief systems were strong and pervasive (Coningham & Gunawardhana 2013). These divisions may have often been superficial or have had ‘fuzzy’ boundaries in the past, yet they have been used to legitimate colonial and post-colonial religious and political inequalities. Despite these uncertainties, Sri Lanka's place at the heart of international trade routes, linking East and West is undisputable, creating an island of diverse communities and traditions, and prospering in the process.


Introduction
This chapter will examine the applicability of the concept of cosmopolitanism within Sri Lanka during the Early Historic and the Medieval period across a timespan of between c.500BC and AD1200 utilising archaeological evidence augmented by epigraphic and textual sources. It will focus on North Central Sri Lanka and the city of Anuradhapura, but will draw wider references, comparisons and analogies where appropriate. Before proceeding, it is necessary to define the possible meanings of cosmopolitanism within an archaeological context. Cosmopolitanism is a relatively new and underexplored phenomenon to have been introduced to the discipline of archaeology and, whilst some volumes have recently explored issues of identity and cosmopolitanism (such as Meskell 2009), those relating to cosmopolitanism have been more concerned with how this may have been represented in the present (largely through cultural heritage), as opposed to exploring the nature and character of ancient cosmopolitanism. Philosophically, cosmopolitanism may be taken to refer to the concept that all humans belong to a single community with shared moral codes and philosophies, and that such a concept should be nurtured (Meskell 2009). However, to a wider public community, cosmopolitanism has often been used to reflect multiculturalism, sophistication and a general worldliness. In an archaeological context, the conceptualisation of the former definition is inherently problematic and challenging; however, the latter set of definitions is more achievable, but to varying degrees as will become apparent.
How archaeologists may commence the process of defining and identifying cosmopolitanism within archaeological communities is, in itself, a challenge although one may simply acknowledge or recognise the presence of multiple communities within the past. On a more ambitious level, archaeologists may investigate the relationships between such communities more deeply and the influences they may have had on each other. In such a way, the concept of cosmopolitanism may assist the development of more nuanced understandings of the complex and multi-faceted identities of individuals and communities in the past. For instance, these individuals may have had allegiances to multiple communities, may have spoken numerable languages and may have participated in numerous religious, ritual and belief systems. However positive an ambition, the inclusive and integrating nature of cosmopolitanism makes it difficult to define and even more difficult to identify within the ephemeral material remains with which archaeologists have to contend. There is also the danger that over-deconstruction can leave archaeology bereft of meaning or tools for interpretation, leaving it subject to accusations of 'pseudo-science'.
In all cases, the focus of this discussion must also acknowledge the underlying and underpinning concepts of identity; concepts which have had a chequered history within the discipline of archaeology. Early twentieth century archaeologists, such as Gustav Kossinna (1858Kossinna ( -1931, linked material archaeological remains with cultures, and variations within these cultural groups were attributed to ethnic diversity, and each clearly defined cultural province correlated with major ethnic peoples (Trigger 1989: 165). Pioneering British archaeologists, such as Gordon Childe (1925) and Stuart Piggott (1950), continued this tradition by identifying and mapping cultural provinces across time and space in both Europe and colonial South Asia, and maintained Kossinna's assumption that cultural groups correlated with ethnic and linguistic groups. As such, in Western Europe and by imperial proxy South Asia, archaeologists utilised concepts of diffusion and migration to explain cultural and linguistic variations, most succinctly highlighted by the debate over the development and spread of Indo-European languages in South Asia and its underpinning ubiquitous Aryan invasion theory (Chakrabati 1999;Coningham and Lewer 2000;Erodsy 1995;Leach 1995). As will be discussed, Sri Lanka may be perceived as representing a microcosm of this latter Aryan question. Indeed, underpinning Tamil and Sinhalese ethnic identities have been created and curated on the basis of relatively modern distributions of Indo-European and proto-Dravidian linguistic communities, combined with references to oral and literary traditions relating to the Vijayan colonization of the island (Coningham and Lewer 2000). Despite this long scholarly tradition, there has been a more recent rigorous examination of concepts of ethnicity within archaeology, leading some scholars to reject the concept that ethnic identity was ever concrete or could be traceable to a definable point. Indeed, Jones (1997: 13) has suggested that "ethnic identity is based on shifting, situational, subjective identifications of self and others, which are rooted in ongoing daily practice and historical experience, but also subject to transformation and discontinuity". Indeed, Archaeologists have also focused on issues of identity within the archaeological record, challenging preconceptions relating to age, gender ethnicity and religion, and recognising that "identity…is not a static thing, but a continual process…Identities are constructed through interaction between people and the process by which we acquire and maintain our identities requires choice and agency" (Diaz-Andreu & Lucy 2005: 1-2).
Crucial within this quote is the recognition that identity is not singular, but a plural concept. Individuals may hold many different identities simultaneously, and this is something that becomes increasingly evident when examining the complex Sri Lankan past. This is equally true of the challenge of trying to discern religious identities from archaeological remains, individual objects or artefactual corpora. As discussed elsewhere, many monuments and motifs were commonly shared by a number of major religious traditions (Coningham 2001), making it difficult and inadvisable to offer firm affiliations. With regard to Sri Lanka, this issue is certainly challenging as to the transition from a predominantly 'Hindu' belief system to a 'Buddhist' one. Whilst recognisable deities, such as Ganesh, Vishnu or Kubera, continued to be venerated, they were reconstituted within a cosmography which placed the Buddha centrally. The survival of old beliefs and the introduction of new initiatives and traditions occurred throughout the island's archaeological sequence and range from the introduction of the Buddha image in the first half of the first millennium AD to the spread of the 'Tabbova-Maradanmaduva Culture' terracottas at the beginning of the second millennium AD (Coningham et al. 2012). In order to investigate the nature of cosmopolitanism present in ancient Sri Lanka, or even whether it is an appropriate concept to utilise at all in archaeological investigations, this chapter will examine a series of case studies. These range from the role of religious pilgrimage, in particular Buddhist, to and from the island; local, sub-continental and international trade networks within the island and the impact this has had on its inhabitants; patronage within the island and Sri Lankan patronage elsewhere in South Asia; and the religious and economic landscapes within Anuradhapura and its surrounding hinterland. At all times, this study will focus primarily on archaeological data, but will introduce textual and epigraphic evidence where appropriate. Indeed, it will begin by examining these sources and critically discussing how modern ethnic constructs in Sri Lanka have been intrinsically linked to the island's ancient past.
Textual Narratives and the linking of archaeology to ethnicity Traditionally, the pre-colonial history of Sri Lanka has been constructed from a variety of textual sources, in particular the Dipavamsa, Mahavamsa and Culavamsa. Geiger argued that the Dipavamsa's contents relied upon an earlier chronicle known as the Atthakatha-Mahavamsa (Geiger 1912: x). Whilst the Dipavamsa is viewed as a first attempt at collating Pali verses, the Mahavamsa is viewed as a younger, more elaborate, treatment of the same material. Geiger even goes as far to suggest that the Mahavamsa represents "a conscious and intentional rearrangement of the Dipavamsa" (ibid.: x-xi). Whilst its authorship is unknown, the Dipavamsa is believed to have been compiled in the fourth century AD, whilst the Mahavamsa has been argued to have been written by various monks of the Mahavihara and compiled into a single document by the Buddhist monk Mahanama in the late fifth or early sixth century AD (ibid.: xi). It narrates the history of the island from its initial colonisation by Prince Vijaya through to the reign of King Mahasena (r. AD 275-301) (Coningham and Lewer 1999a: 707;Strathern 2009). The Culavamsa was a continuation of this narrative, detailing the history of the island up to the eighteenth century (Geiger 1929). Initially believed to be oral legends (Wickramasinghe 2006: 89), the rediscovery of palm leaf manuscripts by George Turnour at a monastery at Mullgiri-galla near Tangalle, led to the serious reconsideration of their contents as historical, leading Sir James Emerson Tennent, Colonial Secretary of Ceylon between AD1845 and 1850, to state that this "long lost chronicle… [had] thus vindicated the claim of Ceylon to the possession of an authentic and unrivalled record of its national history" (1859: 315).
This led to an increase in the numbers of Western studies of the island's history (Devendra 1959: 24), paralleled by the significant research undertaken by members of the Sri Lankan Sangha whose translations of Pali works into Sinhalese and correspondence with European academics facilitated the development of European-based Oriental scholarship (Guruge 1984: xiv, xvii). Unique across South Asia, the Chronicles provided a historical framework for the island from before the Mauryan Empire through to British rule, and with colonial endorsement, the Chronicles became the privileged source of evidence for scholars studying Sri Lanka's past and this focus has produced what has been termed by Seneviratne the "Mahavamsa view" of the island's past (1997: 6). His comments reflect the fact that ever since the rediscovery of the Chronicles, the disciplines of Sri Lankan history and archaeology have been largely overshadowed by the Mahavamsa's narrative (Valentine Daniel 1997: 49). As will be discussed below, whilst Seneviratne undoubtedly referred to 'popular' culture and history, he has also suggested that archaeological evidence from excavations within Anuradhapura have been utilised to bolster academic narratives derived from the Chronicles (Seneviratne 2005).
The narrative itself details the arrival of Prince Vijaya, the exiled heir to a kingdom in northern India, with his 700 followers on the uninhabited island of Lanka in the middle of the first millennium BC. On arrival he slays the demonic Yakkhas who reside on the island, whilst at the same time having two children by the Yakkhini, Kuvanna. Descended from a lion, Vijaya referred to his followers as Sinhala, or 'people of the lion'. However, having borne his children, Prince Vijaya spurns Kuvanna in favour of an Indian princess, and the former and their children retreated to the jungle, forming the Pulinda people (Coningham and Lewer 2000: 707). After the conversion of the Sinhalese to Buddhism in the third century BC as a result of Asoka's proselytising documented in the Mahavamsa, this chronicle makes its first reference to differentiated communities mentioning Damilas, a term often associated with Tamils, though this is contested (Monius 2001). With the exception of those Tamil-speakers  brought across as indentured labour for the colonial tea plantations, the Tamil communities  of present day Sri Lanka have often been directly linked with the invading South Indian  Pandyans and Colas during the later phases of the Sinhalese rule from Anuradhapura (ibid.). As such, the Mahavamsa establishes within its narrative, three distinct communities, which have often been perceived to have been at odds with one another, often with an underlying question of who were the rightful autochthons, rather than recognising this narrative as a frame for a multi-cultural island with a shared history.
This link of past to present has often been translated into the notion of the Sinhalese as rightful "heirs" to the island (Strathern 2009: 3), whilst Tamils were portrayed as latecomers or outsiders. Conversely, other scholars sought to attribute a much deeper antiquity to the Tamil communities of the island (Ragupathy 1987: 180). The reasoning behind this polarisation are complex, but partially originate from British colonial interpretations of Sri Lankan history. As well as endorsing the Mahavamsa as history, Tennent equated the Pulinda with the modern communities of hunter-gatherers or Veddas, often described as the aboriginal inhabitants of the island; the Sinhalese as the civilised creators of the architectural and engineering masterpieces of the northern plains or the Rajarata; and finally, the Tamils as the "debased" destroyers of that civilisation (1859: 340; Coningham and Lewer 2000: 708). These views became mainstream historical and political opinion, although others did suggest that sites such as Mantai were part of a separate early Tamil trading civilisation (Bertolacci 1817: 13), or that an early Dravidian population were already present on the island at the time of the Vijayan colonisation (Parker 1909: 709). However, these views never garnered widespread acceptance as central to British colonial interpretation was the Aryan invasion; the concept that Indo-European speaking-people had invaded South Asia from the north and west around the first millennium BC, bringing with them a cultural package that included writing, iron, horse-riding and advanced social institutions (Chakrabarti 1999: 3). This event was portrayed as bringing civilisation and progress to the previously backward and stagnating populations of the Subcontinent. The Vijayan colonisation from northern India was therefore an extension of this concept and also neatly explained the presence of Indo-European speakers off the southern tip of the South Asia separated from the rest of their language family by millions of Dravidian speakers. Within South Asia, the Indo-Aryan invasion was portrayed as part of a long pageant of historical precedents which helped to legitimate British control of the region as part of a long line of incursions by Aryan, Greek, Persian and Turkic armies (Coningham and Lewer 2000: 711). The civil servant and historian, H.W. Codrington, pursued these legitimacies in his Short History of Ceylon, when he reminded readers that the British invasion of Kandy and exile of the last king, Sri Vikrama Rajasinha (r. AD 1798-1815) was to deliver "the Kandyans from their oppressors and the subversion of the Malabar dominion' (Codrington 1939: 172), referring to the fact that Rajasinha was a South Indian Tamil by birth (1939: 172).
Episodes and events of oppression were also portrayed within the Chronicles and they frequently referenced the destruction of Buddhist heritage by South Indian aggressors. For instance, during the reign of Mahinda V (r. AD 982-1029)  These descriptions were also used during the anti-colonial Buddhist revival by leaders of that movement, such as Angarika Dharamapala (AD1864-1933), who identified both modern Europeans and ancient Tamils as "barbaric vandals" of Sinhalese culture (Coningham and Lewer 2000: 709), fitting a framework promoting Sinhalese and Buddhist concerns whilst noting European interference. However, colonial archaeologists also laid the blame for the desecration and destruction of monuments in antiquity at the hands of Tamils, utilising the same narratives from the Chronicles (Strickland 2011). Early archaeological interpretations drew from such descriptions and H.C.P. Bell, the first Archaeological Commissioner for Ceylon between 1890 and 1912, described the damaged condition of the stone Buddhist railing at the Jetavana monastery of Anuradhapura (1904a: 7): "The indescribable confusion in which the fragments were found heaped one upon another, and the almost entire wreck of the railing, leave little room for doubt that this unique relic of Ceylon Buddhist architecture must have perished under the ruthless destruction of those invaders from South India at whose door lies the mutilation and ruin of the best works of the sculptor's art in Anuradhapura".
Such interpretations were not rare as illustrated by the interpretation of the discovery of a Buddha sculpture with a fractured nose in Jaffna. This led Sir Paul Pieris to record that earlier scholars, such as the Government Agent for Jaffna Sir William Twynam, had suggested that Buddhist sculptures found in the North "have been similarly mutilatedan undoubted sign, he thinks, of Dravidian invasion" (Pieris 1925: 41). Such viewpoints were not restricted to the infancy of archaeological enquiry, but continued through the twentieth century. For example, excavations at the Abhayagiri monastic complex in Anuradhapura in the 1980s revealed Buddha statues lying flat with their heads removed and this phenomena was cited as evidence of the Cola destruction as narrated in the Culavamsa (Wikramagamage et al. 1983: 48).
The latter findings were recovered from excavations conducted as part of Sri Lanka major heritage program, the UNESCO Cultural Triangle, established by President J.R. Jayewardene in 1980. Tasked with excavating, conserving and presenting the ancient cities and Buddhist monuments of Sri Lanka, Anuradhapura, Polonnaruva and Sigiriya were inscribed on the UNESCO World Heritage list in 1982, followed by Kandy in 1988 and Dambulla in 1991. The program was expanded under Jayewardene's successor, President Premadasa, to include conservation of the Mirisavatiya stupa in Anuradhapura (Coningham and Lewer 1999: 865). Although colonial Galle was inscribed in 1988, this singular focus on Buddhist sites was recognised by Tambiah who stated that whilst there should be no barriers to the sponsorship of the restoration of Buddhist monuments "It would also behove a Sri Lankan government to recognise at the same time that there are monuments, archaeological remains, and literary and cultural treasures that are neither Sinhalese nor Buddhist as these labels are understood today" (1986: 126). Indeed, the early results and methods of the Cultural Triangle were also queried by its own Director-General and Director of the Cultural Triangle's excavations at Anuradhapura's Jetavana monastery, Professor Sudharshan Seneviratne, who observed that artefacts including "Beautiful beads, sculpture, statues, ceramics were displayed at site museums along with the historical narration of the Middle Historic texts" but that "analytical studies were sporadic and interpretative studies were mainly commissioned to strengthen the Buddhist history of Anuradhapura and to authenticate the Mahavamsa narration" (Seneviratne 2005). One of the unintended consequences of the increasing alignment of the state-sponsored promotion of Buddhist heritage with the Mahavamsa's narrative was to focus the attention of the separatist Liberation Tigers of Tamil Eelam (LTTE) on the symbolic importance of such monuments with an attack on the Bodhi tree at Anuradhapura in 1985 and bombing of the Temple of the Tooth at Kandy in 1998 (Coningham and Lewer 2000).
However, the meaning and character of the ancient heritage of Sri Lanka was far more complex, diverse and fluid than these recent constructed identities and representations suggest. For example, although Sinhalese monarchs were guardians of Buddhism within the island, close marriage ties with non-Buddhist South Indian dynasties were formed, culminating in the accession of the Nayakkar dynasty to the Kandyan throne in the eighteenth century (De Silva 2005: 197). The current Temple of the Tooth in Kandy was therefore partly constructed by a Nayakkar, Sri Vikrama Rajasimha II (AD 1798(AD -1815, a Tamil/Telegu-speaker from a Hindu dynasty of South Asian origin. The attack on the Temple of Tooth by the LTTE in 1998 therefore not only resulted in damage to a monument constructed by a South Indian dynastic king from the Nayakkar dynasty, but also damaged adjacent shrines to Pattini and Vishnu, Lewer 2000: 709, Coningham andHardman 2004: 158), which were important to both Buddhist and Hindu communities. This patronage and protection afforded by 'non-Buddhists', to use more recent characterisations of communities in Sri Lanka, is further reinforced by a Tamil inscription on a stone slab beside the Tooth Relic Temple in Polonnaruva. Known as the Adatage, this structure was built under the patronage of Vijayabahu I (r. AD 1055-1110) and the epigraph instructs guards from South India, Velaikkaras, to protect the Buddha's Tooth Relic within (EZ2: 242-255).Part of a long tradition of 'Sinhala' states employing South Indian Mercenaries and guards, the Velaikkaras are stated to be adherents of the Mahatantra, and this further highlights the diversity and complexity of identity, religiosity and the construction of royal legitimacy within Medieval Sri Lanka. In the light of these complexities, this chapter will now explore evidence for cosmopolitanism through a series of case studies, further challenging a number of the static and monolithic identities ascribed to archaeological heritage through the simple application of textual narrative to archaeology.

Indian Ocean Trade
Although investigated for over one hundred years, intensive archaeological analysis of the Citadel of Anuradhapura only commenced in the 1990s with the programme of fieldwork instigated by Dr Siran Deraniyagala. One of the excavations associated with this programme was the Sri Lankan-British excavation at trench Salgha Watta (ASW2), which began in 1989 and continued into the 1990s. The trench measured 10 by 10 metres and was excavated to a depth of almost 10 metres (Coningham 1999c). Whilst trench ASW2 identified and dated an architectural sequence with over a thousand years of occupation, rebuilding and eventual abandonment at the site, it was also designed to develop a typological artefact sequence for the island and, in so doing, also allowed the identification of local and international trade networks within that sequence. As noted in 1996, many early studies of the island's archaeology have attributed its apparent peripheral position within Early Historic South Asia sequences to its seemingly marginal geographical location, off its extreme southern tip (Coningham et al. 1996). As a result, the island was assumed to have adopted a number of innovations, such as writing and urbanisation later than North India (Deraniyagala 1972). On a broader scale, this interpretation reflected another deep rooted colonial concept, that contact with the Roman world was the catalyst for the beginnings of Indian Ocean trade. Indeed, in line with his tradition of linking South Asian archaeology with established western chronologies, Mortimer Wheeler presented the early, pre-Roman levels of his excavations at the port site Arikamedu in South India as populated by "simple fisher-folk" living "a leisurely and enterprising fashion just above subsistence level" (1946: 174-5). An adherent of the theory of culture change through invasion, diffusion or trade, Wheeler believed that Roman traders provided the stimulus for the settlement to develop into an international trading hub (ibid.) supported by the presence of finds of Roman Arretine Ware and other Roman and Hellenistic goods. Whilst he focused on the archaeological sequence of a single South Indian trade entrepot, the ideas and concepts that he developed were transposed to Sri Lanka. Begley's re-excavations at Arikamedu (Begley 1996) and Coningham's at trench ASW2 at Anuradhapura (2006) demonstrated the weakness of such models and the latter confirmed the presence of well-developed trade networks across South Asia and across to the Malabar Coast centuries before contact with the Roman or Hellenistic world (Coningham 2002;Coningham 2006). This ancient trading pattern is also proving far more complex and the simple model of pioneering Roman traders is being replaced by one with a highly diverse character with regard to its potential trading communities. For example, the Italian excavations at Khor Rohri on the Omani coast have provided evidence of South Asian wares (Avanzini 2008) and the presence of sherds with Early Brahmi inscriptions, the lingua franca of early trade, at the sites of Myos Hormos and Berenike in the Red Sea suggest that South Asian traders may have been residing in Egypt as early as the first century AD (Tomber 2007). The excavations at trench ASW2 also provide evidence of the repertoire of diverse traded materials, indicating links across the Early Historic and Medieval Indian Ocean. These objects arrived early in the sequence and demonstrated links as far as Afghanistan and Gujarat with finds of lapis lazuli and carnelian in the first millennium BC but later expanded to include glazed ceramics from the Persian World and from South-east and Eastern Asia with delicate monochrome lustre ware bowls and Changsha stone wares in the first millennium AD. Whilst some of the objects, such as the Egyptian glass kohl sticks, were already well known artefact categories in copper alloy within the sequence, other objects represented the introduction of new behaviours and tastes. For example, during Period F (AD 300-600), 'torpedo' jars were imported from Sasanian and Early Islamic regions. Lined with bitumen to make them watertight, they were used to transport liquids. Gas chromatography-mass spectrometry (GC-MS) and stable isotope analysis (carbon and deuterium) of torpedo jars from trench ASW2 identified that the bitumen was derived from Susa (Stern et al. 2008) and whilst it was not possible to determine what liquids were transported within the Anuradhapura torpedo jars, it is likely that one of the commodities was wine (Tomber 2007: 976). Representing newly imported ingredients and tastes within the island, torpedo jars have also been found at Mantai (Carswell et al. 2013), Sigiriya (Stern et al. 2008: 426) and Tissamaharama (Tomber 2007: 980) and undoubtedly represent a broadening of the consumption habits of the elite.
A rather more utilitarian development was also recognised within the sequence of trench ASW2, in the form of sherds with inscriptions. Pre-dating Asokan contact with the island, the presence of Early Brahmi script in levels dating to the c.400 BC was striking but also raised questions as to its presence within the island (Allchin 2006: 456). This was also discussed in a paper in 1996 in which the presence of a North Indian Prakrit, the direct ancestor of Sinhalese, was postulated as well as the concept that it may have been adopted as a trade language and that the communities utilising it within the Citadel of Anuradhapura may have been bilingual and steadily replaced their own language in favour of Prakrita process which resulted in Sinhala (Coningham et al. 1996). The striking nature of its presence within the city has been further stressed by the fact that despite five years settlement survey within the hinterland, only a single inscribed sherd was identified outside the wallsrestricting the cosmopolitan nature of its use to the capital and differentiating this settlement from its rural hinterland . The presence of early trade links at Anuradhapura and Mantai also demonstrate that distant links existed long before scholars had suggested. Indeed, the excavations across the ramparts at Anuradhapura (Coningham & Cheetham 1999) also demonstrated that the urban nature of the site was defined long before contact with Asoka in the third century BC and the supposed 'Mauryanisation' of the island, whilst the earliest levels of the site (c.800BC) also detail an extensive intra-island network of trade and exchange.
However, although this evidence of trade goods within Anuradhapura may have informed us of the extensive trade networks flowing through the site, it still remains to be seen how this trade was organised. Were international merchants and traders residing within the city itself or was the port of Mantai established as a trading entrepot with goods shipped locally to Anuradhapura by local traders (Carswell et al. 2013)? Were Sri Lankan traders setting forth to procure goods from around the world to bring home for local markets? The recent discovery of the Godavaya shipwreck off the southern coast of Sri Lanka promises to shed more light on the identities of the sailors and their cargos but initial reports include finds of quern stones, glass ingots and iron dating to between the first century BC and first century AD (Carlson & Trethewey 2013). As is often the case in archaeology, it is difficult to be absolutely certain and the answers are likely to be a combination of all of the above. Certainly, there do not seem to have been ascribed rules or dictates surrounding the purchasing or selling of goods within the city, nor restricting the presence or absence of foreign merchants. However, at the same time there is little to no mention of trading within ancient textual sources -unsurprising given their religious and ritual nature. The later travel itineraries of the Chinese Pilgrim Faxian noted the presence of "The houses of Sa-pho (Sabaean) merchants" within the city (Beal 1869: 154) although scholars differ in their identification of the ethnicities involved. Finally, note must be made of the discovery of a Nestorian Cross within the Citadel of Anuradhapura as it was interpreted by early archaeologists as marking the presence of a church (Hocart 1924). This discovery has been recently reassessed by Prabo Mihindukulasuriya, who coupled the find with records by the historian Cosmas and the presence of a Nestorian bulla from Mantai to suggest that Anuradhapura hosted a thriving Christian community within the ancient Sri Lankan capital (2011). Alhough the presence of these isolated artefacts may not necessarily equate to the permanent presence of such a community are contested, this is not the only example of the introduction of external influences. For example, Fynes has argued that Roman and Egyptian traders brought the concept of the goddess Isis to South Asia, where she developed over time to become the goddess Pattini (1993). While such a theory is largely conjectural, are most models with a reliance on western diffusion of ideas and concepts, his work does highlight the potential for a broader study of the introduction of external influences and practices.

Pilgrimage
The earliest contact between Sri Lanka and northern India recorded in the Mahavamsa's narrative was the Vijayan colonization. The second notable contact was between Devanampiya Tissa's (r. 250-210BC) kingdom and the Mauryan Empire, which had emerged in the mid-fourth century BC from the internecine warfare between the various mahajanapadas of South Asia (Allchin 1995: 187). After his own conversion to Buddhism, it is recorded that the Mauryan Emperor Asoka (272-235 BC) sent missionaries to various neighboring states to spread the Dharma {Mahavamsa 7}. Asoka's own son Mahinda was sent to Sri Lanka and, after converting Devanampiya Tissa and his entourage, Mahinda was supported by the King in spreading Buddhism to the rest of the population throughout the island {Mahavamsa 14.59-64}. Later, Asoka's daughter Sangamitta brought a branch of the Bodhi tree from Bodhgaya, under which the Buddha gained enlightenment, to Sri Lanka {Mahavamsa 18} forming a centerpiece which is still venerated today at the Sri Maha Bodhi in Anuradhapura. Further relics were brought to Sri Lanka, such as the Buddha's alms bowl, whilst his collarbone was enshrined within the stupa of Thuparama in Anuradhapura {Mahavamsa 17}. This textual evidence records the movement of relics across South Asia as diplomatic and religious gifts and exchanges, operating within established political networks as well as demonstrating how regions not historically visited by the Buddha during his lifetime could still benefit from proximity to his relics. Furthermore, archaeological evidence supports this contact and exchange during the Early Historic period with Deraniyagala suggesting that the presence of Northern Black Polished Ware at Anuradhapura constituted "physical evidence of links between the core of the Mauryan culture sphere and Sri Lanka" (Deraniyagala 1986: 47), further augmented by the presence of sherds of Northern Black Polished Ware in Trench ASW2 (Coningham 2006). The exact nature of this link is less certain, however, as the ware pre-dates the rule of Asoka and may represent down-the-line trade rather than courtly exchange. One may also note the close parallel between the recorded arrival of the Bodhi Tree and the advent of Tree and Swastika coinage in the third century BC.
Once established, the monasteries of Anuradhapura gained an international reputation and links with Buddhist institutions throughout Asia continued to expand. In the late fourth and early fifth century AD, the Chinese Pilgrim Faxian journeyed to Sri Lanka during his travels around South Asia visiting sites associated with the life of the Buddha and major Buddhist centres. In addition to describing the city of Anuradhapura and the religious ceremonies that occurred at the Citadel, he also described the wealth of the monasteries of the Sacred City.
Here, Faxian reported that over 10,000 monks and nuns resided, with the Abhayagiri vihara housing 5000 monks and the Mahavihara 3000 monks (Legge 1886: 102, 107). The treasury of Abhayagiri was said to contain jewels and gems of incalculable value (Beal 1869), garnered from the patronage the Sangha received. The international networks of the Sangha continued into the medieval period but, after the purported damage to the Sangha caused by Cola incursions in the eleventh century AD, it was reported that no ordination had been conducted in Sri Lanka for many decades. To restore the Sangha, Vijayabahu I was aided by King Anuruddha of Ramanna, whose realm is thought to coincide with modern Burma/Myanmar, to bring monks to Sri Lanka to fulfill these duties (Panditha 1955: 133). After the decline of Polonnaruva, a major mission of monks from Chiangmai and Pegu came to Sri Lanka in the 1420s AD to receive further training, worship the Tooth Relic and receive higher ordination (Ilangasinha 1992: 173-177). This Indian Ocean-wide Buddhist worldview may have been encapsulated symbolically in more concrete materials and not only in textual sources. The layout of the twelfth century Sacred Quadrangle at Polonnaruva has been identified by Frasch as a cosmogram of international Buddhist contact (Frasch 2010). He has interpreted the Tooth Relic Temple as having been surrounded by four religious structures reflecting differing architectural traditions of major Buddhist regions at the time, including areas of modern Thailand, Sri Lanka, Myanmar and India, "standing like devotees in a semi-circle around the Tooth Relic" (Frasch 2010: 655).
However, this is not to suggest that Sri Lanka acted solely as a receptacle for external cultural impulses as there is also evidence for communities within Sri Lanka influencing their Indian Ocean partners. For example, Pliny noted that a Sinhalese monarch sent an embassy to Rome in the second century AD (Gokhale 2004: 136-137) and an inscription on the Asokan period stone railing at the Mahabodhi Temple in North India was translated by Sir Alexander Cunningham as the "Gift of Bodhi-rakshita of Tambaparna (Ceylon)" (Cunningham 1892: 16). References to Sri Lanka have also been identified at other sites in India. At Nagarjunakonda in Andhra Pradesh, an inscription recorded the Sihala-vihara and the dedication of a shrine to the fraternities of Tambapanni (Ramachandran 1953: 5). Tambapanni is thought to reference Sri Lanka, and this third century AD inscription also noted the presence of Buddhist nuns at Nagarjunakonda (Walters 2000: 111). Further evidence of a Sri Lankan monastic presence at the site was suggested from the discovery of a decorated moonstone at one of the apsidal temples. Unlike other moonstones at Nagarjunakonda, which were plain, this example resembled the ornately carved moonstones of Anuradhapura, decorated with elephants, lions, deer, horse, bull and buffalo and was regarded a find that "may easily be expected as Buddhists from Ceylon had settled at Nagarjunakonda" (Ramachandran 1938: 13).
Later cultural and intellectual contact is confirmed by an inscription dating to AD792, found at a monastic site in the Ratubaka plateaux of Java, commemorating the founding of a branch of the Abhayagiri vihara of Sri Lanka in this locale (Gunawardana 1979: 17, Sundberg 2004. Sundberg has argued that the pendopo architectural unit at Ratubaka has similarities with the layout of padhanaghara parivena, or double-platform monasteries, which have been identified on the western outskirts at Anuradhapura and at Ritigala. These sites have been linked to the fraternity of monks known as the Pamsukulika or 'those clothed in rags from dustheaps' (Coningham 1995: 235), who rose to prominence from the eighth century AD onwards. The architecture of a padhanaghara parivena is usually characterised by two quadrangular units connected by a stone bridge. These platforms are surrounded by an enclosing wall, occasionally a moat, as well as cisterns and ponds. Padhanaghara parivena do not possess typical Buddhist structures or iconography such as stupas, but are often associated with meditational pathways (ibid.). Sundberg suggested that the Javanese pendopo shares these features such as a lack of ornamentation, cardinally-oriented doubleplatforms, artificial rock-cut ponds and the presence of a compound wall (Sundberg 2004: 100). Although it is not clear whether the Pamsukulika associated with the padhanaghara parivena of Sri Lanka were present in Java, or vice versa, it is clear that there was communication between these regions across the Indian Ocean and shared architectural concepts.
Furthermore, inscriptions within Sri Lanka, such as two twelfth century AD inscriptions from Polonnaruva record the construction of a Temple in South India (EZ2: 148-152) and the construction of alms-houses abroad (EZ2: 165-178). Artefactual evidence is also suggestive of international Buddhist contacts and it has been argued that a tenth century AD bronze Buddha figure found in Thailand originated in Anuradhapura (Von Schroeder 1990: 206). Such Asia wide links are corroborated in the Chronicles and, as stated above, it is recorded that Vijayabahu I (AD 1055-1110) sent envoys to South-east Asia to aid him in purifying and re-establishing the Sangha in Sri Lanka (Culavamsa 60.4-6;Sirisena 1978: 59-60). Textual sources also illustrate the influence of Sri Lankan monarchs overseas not just in religious matters. For example, Parakramabahu I (AD1153-1186), is recorded as instigating military campaigns against South-east Asian polities as well as sending an army to South India to assist a Pandyan ruler against the Colas in AD1169 (Strathern and Biedermann, this volume;Sirisena 1978: 59-60).
Finally, it is worth noting that not all pilgrimages were to 'Buddhist' monuments or 'Buddhist' relics as it has been suggested that the site of Sigiriya, creation of Kassapa I (r. AD473-491), was constructed symbolically to recreate the city of Alakamanda, the celestial home of Kubera, god of wealth (Paranavitana 1950). Historically a site which attracted large numbers of visitors and pilgrims, inscription 28 of the Sigiriya graffiti recorded that "The resplendent rock named Sighigiri captivates the minds of those who have seen (it) as if (the mountain) Mundalind, which was adorned by the King of Sages, had descended to the earth" (Paranavitana 1959: 17). Mundalind has been equated with Mount Meru, and thus the large outcrop at Sigiriya was identified as celestial Meru. Continuing this symbolism, Paranavitana suggested that the lake at Sigiriya represented the celestial lake Anottatta, the white-washed boulders before the outcrop as the snow-clad Himalayas, and the royal palace was that of Kubera on the summit of Meru (Paranavitana 1950: 137). Whilst it has been argued that Sigiriya was in fact not an urban centre, but rather a vast Buddhist monastic complex (de Silva 2002), if viewed as the creation of Kassapa, the graffiti and cosmological symbolism of Sigiriya created what is argued to be the clearest example of a cosmic city in early Sri Lanka (Duncan 1990: 54). This symbolism suggests the physical modeling of South Asian-wide concepts, such as Mount Meru at the centre of a cosmic-universe, already present previously in Anuradhapura (Wickremeratne 1987) and continued into the medieval and post-medieval periods at the urban forms of Polonnaruva and Kandy (Duncan 1990). The Sigiriya graffiti also demonstrates that individuals from various communities travelled from all over Sri Lanka to visit the site (Paranavitana 1959), and not necessarily for religious reasons. However, if Raja de Silva's view that Sigiriya represented a large Mahayana-Theravada monastery, another intriguing possible explanation for Sigiriya's design may be postulated. Rather than Mount Meru, he argues that the paintings of female figures were representations of Tara, rather than ladies of the royal court, or apsaras surrounding the summit of Meru (de Silva 2002). There is also the possibility that these two interpretations were both held side by side, again highlighting the diversity in symbolic meaning that could be attached to physical remains in early Medieval Sri Lanka. This evidence from Polonnaruva and Sigiriya suggests that communities and concepts from the wider region were incorporated into the fabric of Sri Lankan society, and whilst evidence has been provided of the adoption of worldly ideas and materials within Sri Lanka, it is equally appropriate to note that the island's presence and influence was felt across the Indian Ocean.

Patronage
As the above section has intimated, there are numerous textual sources and artefactual evidence associated with Sri Lanka's monumental centres that demonstrate the breadth and depth of the networks of pilgrimage and trade throughout the Indian Ocean region. However, until recently the extra-urban networks around these hubs have been somewhat neglected. Recently completed fieldwork in the hinterland of Anuradhapura ) has begun to redress this imbalance, providing an archaeological dataset that has been combined with geoarchaeological and textual studies in order to understand the development of the city in relation to its wider landscape context. One of the key findings of this fieldwork was the central role of Buddhist monasteries in the administration of landscape Coningham et al. 2013c), a state of affairs that has previously been suggested from archaeological landscape surveys in Sri Lanka (Bandaranayake 1994;Gunawardhana 2009) and India (Shaw 2007;Hawkes 2009). The two major site categories identified during six years of field survey in Anuradhapura's hinterland were Buddhist monasteries and small-sized ceramic scatters. These sites were differentiated by deep occupation sequences at monasteries, compared with shallow ephemeral traces at ceramic scatter sites. In addition, artefacts such as coins, precious and semi-precious stones, fine ware ceramics, as well as monumental architecture and writing were restricted to monastic sites. With the general absence of high order secular settlements with such indicators in the hinterland, monasteries appear to have acted both religious and secular administrators with jurisdiction over large temporalities in the hinterland (Coningham et al. 2013c). This pattern reached its climax in the Early Medieval hinterland when the most dominant form of Buddhist patronage was through immunity grants, preserved in inscriptions on stone pillars, rather than the direct construction or maintenance of religious structures. Accounting for almost half of all donations during this period, and found across Anuradhapura's hinterland , these immunities alienated vast tracts of land and transferred authority from the Crown and local officials to the Sangha (Gunawardana 1979: 97). On the one hand, this created an integrated landscape administered by monastic institutions. However, simultaneously, the growing diversity and division between and within Buddhist and other religious sects ensured that such homogeneity was not as fully relaised as it could be. Further inscriptions, found throughout Sri Lanka from the Early Historic Period onwards, corroborate the links between monasteries of the hinterland and urban monasteries of the Sacred City, recording the affiliation of monasteries to the Mahavihara, Abhayagiri vihara and Jetavana vihara (Coningham 2011: 941).
The fact that three distinct monasteries are referred to also indicates that the Buddhist Sangha at Anuradhapura was not a monolithic organisation. The Mahavihara was founded in the reign of Devanampiyatissa with the arrival of Mahinda's mission and incorporated the Bodhi tree and the Ruwanwelisaya stupa {Mahavamsa 15}. The first major recorded schism occurred during the reign of Vattagamani (r. 89-77BC) and led to the founding of Abhayagiri vihara {Mahavamsa 33.83}, often cited as a centre of Mahayanist learning and doctrines. Mahasena (r. AD 275-301) not only founded the Jetavana vihara at Anuradhapura, but also under the influence of a "lawless bhikkhu" {Mahavamsa 37.4}, withdrew support from the Mahavihara. This led to the abandonment of the Mahavihara for nine years with the monks leaving for Malaya and Rohana. Construction materials were then taken from the complex and brought to Abhayagiri, which became wealthy under state support {Mahavamsa 37.7-8}. Later under the rule of Mahasena's son, Sirimeghavanna (r. AD 301-328) the Mahavihara was reconciled to the royal lineage {Culavamsa 37.53-66}, but all three major fraternities continued to receive state patronage.
Different sects of monasteries have been identified architecturally within the Sacred City of Anuradhapura. In addition to lena, natural rock-shelters with Early Brahmi inscriptions engraved along drip ledges, which represent the earliest extant category of monastic establishment known at present in Sri Lanka (Coningham 1995: 228-229;Dias 2001: 12), Bandaranayake (1974) identified three categories of monastic complex. The first was the organic or centric monastery, dating from the first century AD onwards. Termed organic due to associations with locales with pre-existing traditions (Bandaranayake 1974: 33), they are also termed centric due to a layout focused around a colossal stupa, and in this regard, the Mahavihara, Jetavana, Abhayagiri, Vessagiriya in Anuradhapura, as well as Mihintale fit this model (Bandaranayake 1974: 48). The second was the padhanaghara parivena, also known as double-platform monasteries, mentioned in the previous section and thought to be associated with the Pamsukulika fraternity. The best preserved examples of this form are a group known as the Western Monasteries at Anuradhapura, as well as at the site of Ritigala. To reiterate, the typical layout of the padhanaghara parivena consisted of two quadrangular units connected by a stone bridge, usually surrounded by an enclosing wall, with ponds and cisterns occasionally located within the compound (Bandaranayake 1974: 127;Wijesuriya 1998). In general, these sites were built from plain ashlar blocks and did not possess typical Buddhist structures or iconography, with the only decorated features being the urinal slabs that seemingly depict images of 'orthodox' ornate viharas. Such iconography can be viewed as a visible, physical representation and reaction against the wealth and grandiose nature of the ornate monasteries of Anuradhapura (McAlpine and Robson 1983: 29). Decorated urinal slabs, meditational pathways and architectural style suggest that the padhanaghara parivena represented a contestatory discourse from the other monastic categories of Anuradhapura (Coningham 2001: 87). Thirdly, the pabbata vihara is a monastic form thought to date from between AD700 to 1200 (Bandaranayake 1974: 81). Believed to be royal foundations, they exhibit evidence of a pre-planned scheme with a core monument zone of stupa, image house, Bodhi tree shrine and chapter house surrounded by individual residential structures within a major moated enclosure (Bandaranayake 1974, Wijesuriya 1998. They have also been tentatively identified as having Mahayana affinities due to their resemblance to prescriptions outlined in the Mahayana architectural treatise the Manjusri Vastuvidyasastra (Jayasuriya et al. 1995). This Mahayana influence has been illustrated at these sites through the recovery of copper plaques and plates inscribed with text from the pabbata vihara of Vijayarama (Bell 1893: 4-10), as well as from stupas at Mihintale and Jetavana vihara. This suggests evidence of the Mahayana practice of Dharma-dhatu, or venerating the word of the Buddha (Perera 2005: 298), at monasteries with Mahayana leanings (Dias 2001: 108). Furthermore, bronze figures depicting Mahayana deities have been excavated in pabbata viharas within the Sacred City of Anuradhapura (Prematilleke and Silva 1968).
Such Mahayana traditions have also been recorded in textual sources from the foundation of the Abhayagiri vihara dating to the first century BC onwards. For instance, the first recorded Bodhisattva image is attributed to the reign of King Mahasena (r. AD275-301) and was produced on his request {Mahavamsa 37.102}. Other Mahayana practices, such as Dharmadhatu mentioned above, are attributed in the Chronicles to the sixth century AD {Culavamsa 41.37} Silva 1968: 62, Dias 2001: 95), and a group of Sanskrit inscriptions provide evidence of Mahayana traditions referencing concepts such as Trikaya (Perera 2001: 260-261). The cosmopolitan nature of Buddhism is also demonstrated in the Culavamsa when the Queen of Udaya I (AD797-801) donated a monastery to a "Damila bhikkhu community" {Culavamsa 49.24}. Although the term Damila is contested (Coningham and Lewer 1999), the fact that a distinction was drawn suggests a group with possible differing practices. Whilst identified in textual sources and in the architecture of monumental sites, such a pattern of architectural monastic variation was also apparent in the Early Medieval (AD600-1200) hinterland of Anuradhapura (Coningham et al. 2103a). A pabbata vihara site was identified at Parthigala (Z001) in the vicinity of the Nachchaduwawewa, however, it was later found to have been located 4.8 kilometres away to a padhanaghara parivena site at Marathamadama (C112). Although both appear to have belonged to different Buddhist traditions, both also appear to have been occupied contemporaneously. Furthermore, 'orthodox' monasteries were widespread within the landscape, represented by the Focal monastic type, identified during recent archaeological fieldwork in the hinterland, and were contemporary with these other monastic complexes and sects (Bailiff et al. 2012;Coningham et al. 2103a;Coningham et al. 2013c).
At an individual object/artefact level of analysis, scholars have also been able to identify significant areas of differentiation within chronologically contemporary schools of sculpture and image-making. Perhaps representing different fraternities, such as those with Mahayana traditions, Arjuna Thantilage's (2010a) analysis of the chemical composition of bronze sculptures suggests a distinct divergence. In a study of six sculptures, he identified two clear groups through lead isotope and trace element scatter plots. The two groups of sculpture were indicative of technological differences, interpreted as representing two separate schools of image production co-existent during the Anuradhapura period. Architecturally, there are also differences in stupa construction across the island. Whilst perhaps reflecting differences in patronage or access to building materials, the brick and stone constructions of Anuradhapura and Polonnaruva, particularly the examples of "Buddhist giganticism" (Coningham 2011: 937) are in stark contrast to the coral and limestone stupas in the Jaffna peninsula, such as those at the monastic complexes of Kantarodai (Ragupathy 1987). That such diversity existed elsewhere within the island is illustrated by the presence of the freestanding crystalline limestone Avalokitesvara statues, measuring 10 metres tall and dating to the seventh century AD, close to a free-standing 14.5 metre high Buddha statue at Maligawila. Again, their co-existence demonstrates the variability and diversity of Buddhism, as reflected though worship, patronage and architecture, within Sri Lanka; although even more diversity is illustrated from amongst the ancient Buddhist communities of the Maldives (Bell 1940).
It is also important to note that the population of Rajarata was not particularly homogenous and Early Brahmi inscriptions dating to between the third century BC and first century AD, document a broad spectrum of patrons of early Buddhism, highlighting the diverse nature of society during this period (Paranavitana 1970;Davis et al. 2013). In an analysis of the 458 inscriptions dating to the Early Historic Period (340BC-AD200) in the 50 kilometre radius from the city of Anuradhapura, donations mentioning monarchs only accounted for 20.22% of the corpus. Parumakas, identified as representing local chiefs, were the most prevalent at 25.22%, whilst those where no definitive rank could be assigned accounted for 24.35%. The role of those thought to represent monks was also important in the early patronage of Buddhism providing 18.91% of donations; with other ranks such as Gamikas (6.09%), Gapatis (3.70%) and Brahmans (1.52%) contributing, as well as those where the donor was unknown (5%) . When anaylsed island-wide, the prevalence of donations by monarchs drops to only 6.4% . This is in contrast to what would have been expected from the Mahavamsa's narrative, whereby one might expect an elite-driven process of conversion, under the patronage of Devanampiya Tissa, leading to the majority of donations being royal in origin. The disparity between the donations made in the Chronicles and those in the epigraphic record may be due to the rise of the Mahavihara, and it has been suggested that the Chronicles "may represent a contrived ecclesiastical tradition legitimising the contemporaneous status quo by awarding a central position to the successful kings of Anuradhapura and ignoring the contributions of the failed kings" (Coningham 1995: 231), as well as other segments of society. Indeed, it has previously been noted in the Early Brahmi epigraphic corpus of Sri Lanka that only ten kings mentioned in the Chronicles have been found in the epigraphic record (Sirisoma 1990: 30-31) and Paranavitana reported failing to identify a single inscription relating to a donation by Devanampiya Tissa (Paranavitana 1970). Moreover, the corpus of inscriptions reveals genealogies of previously unknown royal lineages, and lineages which have been either ignored by or edited out from the Mahavamsa's narrative (Coningham 1995: 231), and, many diverse communities led to the establishment of Buddhism in Sri Lanka.
Whilst this evidence suggests that many different communities and segments of society were patrons of Buddhism, there is also evidence that the Sangha was not the sole recipient. Indeed, pre-Buddhist beliefs are attested to in the Chronicles with references to the presence of Yakkhas and demons are recorded in the Vijayan narrative {Mahavamsa 6, 7}, and in the laying out of Anuradhapura by King Pandukabhaya in the fourth century BC {Mahavamsa 10.84}. In this description, additional religious groups, such as "ascetics", "heretical sects" and "Brahmans" were alluded to but were located outside the city {Mahavamsa 10.95-102}. Many of these orders are recorded as having received state patronage with Pandukabhaya building "a monastery for wandering mendicant monks, and a dwelling for the ajivakas, and a residence for the Brahmans" {Mahavamsa 10.101-102}. Brahmans were recorded as undertaking important religious roles prior to the arrival of Buddhism and it is noted that during the reign of one of Sri Lanka's first monarchs, King Panduvasudeva, that the wisdom of "Brahmans skilled in sacred texts" {Mahavamsa 9.1-2} was called upon for important matters. This importance continued after the arrival and adoption of Buddhism in Sri Lanka and the 22 Early Brahmi inscriptions that mention Brahmans are also a valuable resource in understanding religious plurality. Though these inscriptions only mention that these donors to the Buddhist Sangha were Brahmans, the presence of such inscriptions, contemporary to Buddhism, suggests that Brahmans continued as an important group within Sri Lankan society. Whilst there is a possibility that the title may have continued to be applied despite changes in religious affiliation, there is a greater possibility that Brahmanism continued alongside early Buddhism in Early Historic Sri Lanka (Paranavitana 1970).
Traditionally, the historical developments of the later phases of the occupation of Anuradhapura and the shift of the capital to Polonnaruva have been presented as a period associated with the emergence of a multi-vocality of religions, with an increased South Indian influence. Indrapala has suggested that, in tandem with the widespread appearance of tenth century Tamil inscriptions dated in the regal years of Cola rulers, there was also an increase in Saiva temples (2005: 208). Furthermore, in the Chronicles there are some records of Anuradhapura's kings supporting non-Buddhist institutions, although other reigns are recorded only with the destruction of such institutions as with Mahasena's demolition of the temples of brahmanical gods {Mahavamsa 37.41}. Later, the chronicles recorded that Mahinda II (r. AD777-797) "restored many decayed temples of the gods here and there and had costly images of the gods fashioned" {Culavamsa 48.143-144}, an act which was repeated by Parakkamabahu I (r. AD1153-1186), who constructed 24 temples to the gods {Culavamsa 79.81} and Sena II (r. AD853-887) who was recorded as supporting Brahman rituals {Culavamsa 55.65}. The transference of the capital to Polonnaruva has also been portrayed as a religious shift with a more pluralistic and eclectic patronage at state-level incorporating Buddhist, Brahmanical and Saivite practices (Indrapala 2005: 251). In support, archaeological investigations at Polonnaruva have identified Saiva and Vaisnava shrines with bronze Nataraja, Siva and Parvati images (Paranavitana 1955: 79, 82) and a twelfth century inscription of Nissanka Malla (r. AD1187-1196) at Dambulla recorded the construction of a Hindu temple as well as the restoration and construction of Buddhist temples (EZ1: 121-135). In Anuradhapura itself, structures north of Abhayagiri dating to the later phases of the occupation of Anuradhapura, were identified as 'Hindu ruins' on the basis of their architectural layout and the recovery of several lingams (Bell 1904a;1904b), though this identification has been contested (Strickland 2011: 230). As noted above, the building in the north-east of the Citadel where an inscribed cross was recovered, may have been the location of a community of Nestorian Christians and a church (Hocart 1924: 52), suggesting religious communities that were not Buddhist within the urban core.
Beyond the Citadel walls, artefactual evidence from the Anuradhapura hinterland survey also illustrates additional 'non-Buddhist' religious and ritual practices about which the Mahavamsa's narrative appears silent. These comprise a total of 489 terracotta artefacts from eight sites, although the majority were excavated at the site of Nikawewa (D339) (Coningham et al. 2103b). Dating to between AD900 and 1300 on the basis of radiocarbon determinations from Nikawewa and Rajangane (Coningham et al. 2012: 10;Wijayapala 1997: 34), these artefacts include human and animal figurines as well as anthropomorphic phalli. It is apparent that they had been deliberately broken, and based on ethnographic analogies, it is suggested that these terracotta caches of broken figures may represent a practice not far from that described in studies of the Gammaduva ceremony. First recorded by Knox in 1681, Gammaduva ceremonies are invoked to ward off disease and to ensure good luck as well as personal and agricultural fertility. In addition to dances and food offerings, terracotta objects representing plants, animals, and people are created. The terracotta objects are broken into pieces after being placed on an altar and such practices provide a possible explanation for the archaeological assemblages of terracotta artefacts, which contain fragmentary terracotta animal and human heads, bodies, and limbs (Coningham et al. 2012: 12).
Deposited in caches across the hinterland and the Dry Zone of Sri Lanka. Known from more than 20 sites, they display a distinct uniformity of design and were clearly restricted to non-Buddhist and non-urban sites. Traditionally dismissed as 'folk' art, they were found in a monumental structure at Nikawewa and we have reinterpreted them as representing a powerful shared corpus of practice forming a formalised ritual structure parallel to Buddhism (Coningham et al. 2012: 12), due to the high degree of uniformity within the design, manufacture and usage of the terracotta figurines. The presence of such communities is absent from textual sources, and highlights the ability of archaeology to recognise the presence of groups and individuals who operate outside of official state or elite circles. Indeed, it is likely that further groups, invisible within historical sources, may have resided within both the urban and non-urban landscapes of Sri Lanka. The Nikawewa terracotta corpus, alongside the various Buddhist monastic fraternities identified in the urban core and hinterland, as well as the evidence for Hindu and Christian communities of the Citadel and Sacred City, suggest that Early Medieval Anuradhapura was highly diverse in nature, capable of incorporating multiple contemporary religious and ritual networks (Coningham et al. 2013c).

Urbanism
Nestled within this diverse monastic landscape was the secular Citadel, measuring roughly one kilometre square, and defined by a ditch and rampart, and has been subject to over a century of excavations, the most recent of which have been briefly introduced earlier in this chapter (Coningham 1999a(Coningham , 2006. Early Historic treatises, such as the Arthasastra, contain details of how urban forms should be planned and it states that a city should be quadrangular, surrounded by three moats and a rampart {Arthasastra 2.3.4-6}, and be internally demarcated by cardinally-orientated roads and gateways {Arthasastra 2.4.1-2}. Within the city, the Arthasastra advises that the inhabitants should be divided according to varna and occupations with heretics and Candalas (outcastes) banished outside the city walls {Arthasastra 2.4.7-23}. Furthermore, as outlined in the previous section, the description in the Mahavamsa of the laying-out of Anuradhapura in the fourth century BCE by Pandukabhaya records that separate areas of the city were allocated for differing social groups, and that the city was divided into four quarters {Mahavamsa 10.88-89}. Such textual descriptions have been transplanted onto the archaeological and architectural evidence, suggesting that different communities lived divided according to caste and profession within the city (Hocart 1928).
Architecturally, it would appear that Anuradhapura matched these textual descriptions as the moat, rampart and cardinally orientated streets and structures of Anuradhapura appear similar to the Arthasastra's precepts, particularly the description of Pandukabhaya's city. It has been argued that Anuradhapura's layout "was no casual cluster of buildings but a cosmography that reflected the universe" (Wickremeratne 1987: 45) and Hocart suggested that during the Early Historic period "the doctrine of the four quarters… had a considerable influence on the planning of cities" (Hocart 1928: 156), noting Anuradhapura's roads ran according to the cardinal directions inside roughly quadrangular ramparts (1928: 151). He also inferred that the royal palace would have been east-facing, located in the eastern part of the city (Hocart 1930: 86), as advised by the Arthasastra {Arthasastra 2.4.6-7}. As stated above, other Sri Lankan urban forms correspond to cosmological symbolism, with their plans recreating the universe in microcosm (e.g. Coningham 2000;Duncan 1990). However, archaeological evidence from Anuradhapura suggests that such plans were ideals, rather than actualities.
Working on the premise that ancient social groups may be identifiable across a site through artefactual variability, Young (1999, 2007), analysed craft waste and faunal remains from different areas across the Citadel of Anuradhapura. They found that rather having distinct areas associated with specific crafts, as would be expected if different professions were provided with differing localities within the city, the co-location of metal and semi-precious stone working was found throughout the Citadel. This trend was mirrored in the faunal record where, referencing the Laws of Manu, forbidden and permitted species of animals were found in the same locations throughout the city Young 1999, 2007). Whilst the antiquity of caste is disputed, with some suggesting that caste rigidity is a more recent phenomenon (e.g. Dirks 1997Dirks , 2001Sharma 1999), this analysis illustrates that social divisions based on material differentiation were not present in Early Historic Anuradhapura, and that if such social systems, such as caste, were present and can be identified then more refined archaeological methodologies are required (Coningham and Young 1999: 92). This is not to suggest that Anuradhapura was not cosmopolitan, containing various differentiated communities. As outlined above, Anuradhapura hosted a thriving trade network, including several different groups. However, the similarity of artefacts across Anuradhapura suggests that whilst there is abundant evidence elsewhere for plurality, there were also shared practices and ideals throughout Anuradhapura's urban cityscape, illustrating cosmopolitanism both through differentiation and unifying concepts and lifestyle choices. These shared concepts not only included the homogenous artefactal record of Anuradhapura, but also long-lived patterns of settlement in the Citadel, Sacred City and through low-density urbanism in the hinterland, which will be explored below.

Future perspectives for cosmopolitanism in the archaeology of Sri Lanka
This chapter has presented the evidence for diversity and cosmopolitanism apparent within the archaeological record of Sri Lanka during the Early Historic period through to the Medieval in terms of trade, urbanism, pilgrimage and patronage. We have utilised the four case studies of trade, pilgrimage, patronage and urbanism to demonstrate the breadth and variety of the ancient and historic communities within the island and it is clear that the evidence base which earlier scholars utilised to bind modern ethnicities to ancient identity was extremely limited in comparison with the information available today. Their early championing of the Mahavamsa's narrative has led to an under-development of more research-oriented archaeology and the development and spread of critical self-awareness amongst archaeologists and historians. Indeed, as Sudharshan Seneviratne reflected on his time as Director General of the Central Cultural Fund: "analytical studies were sporadic and interpretative studies were mainly commissioned to strengthen the Buddhist history of Anuradhapura and to authenticate the Mahavamsa narration" (2005). Our own approach within this chapter has been to select a combination of extremely well-known examples, such as Anuradhapura's Nestorian Cross and the Sri Lankan Sangha communities across the Indian Ocean, as well as drawing on the results of new research-oriented fieldwork within the hinterland of Anuradhapura survey and excavation program.
The latter has been extremely significant as it has demonstrated the existence of a diverse patchwork of intricate networks of religious and secular cooperation and communication. As noted above, six years of field research Coningham et al. 2011; has identified over 750 archaeological sites, ranging from small scatters of eroded ceramics covering a few square metres, to abandoned monastic complexes several hectares in size. In between these extremes lie rock-cut caves, stone bridges over the Malvatu Oya, metal-working sites and a corpus of sites containing terracotta figurines. The two most numerous categories of sites were ceramic scatters, interpreted as short-term villages engaged in slash-and-burn (chena) agriculture, and monastic sites ranging from rock-cut shelters to large complexes. This latter category of sites was much more permanent, and highly visible within the landscape, and led to an initial model of a 'Theocratic Landscape' in which it was postulated that the monasteries functioned as centres of economic and political control in lieu of towns, whilst villages shifted around them over time ). Since then, we have presented a more complex model of 'Buddhist Temporalities' and low density urbanism, that reflects a growing awareness of multiple heterarchies in the hinterland of Anuradhapura. The Buddhist Temporality model reflects the complex relationship between monastic institutions and secular authority within the island in the modern era, yet which is also evident within the archaeological record. Indeed, it is clear that "the city's surrounding landscape of villages and rural communities was not centrally regulated by the state through higher-order settlements and royal officials but through a network of viharas, closely linked to the great monasteries of the city rather than the throne" (Coningham 2011: 940). There is little evidence of higher order settlements, save for two short-lived attempts to secularise authority outside the city, and the vast majority of secular occupation was in small, ephemeral settlements engaged in chena agriculture, a model of settlement present prior to Buddhist monasteries in the protohistoric, through to the Early Medieval period (Coningham et al. 2103c).
From the arrival of Buddhism, monasteries appear to have been the central hubs in the landscape. Whilst non-Buddhist rituals are also apparent archaeologically from the Early Medieval onwards with the appearance of terracotta artefacts, monasteries appear to have continued as organisational elements within the hinterland, surrounded by dispersed settlement. However, this was a highly contested landscape and one which was simultaneously closely integrated into the secular and monastic core, yet at the same time divorced from the same networks and linkages enjoyed by the urban elite. Tensions exist between the pressing need to be integrated to larger networks of exchange, patronage and religious merit that can only be derived from the royal centre and major monastic institutions of Anuradhapura, whilst simultaneously non-Buddhist rituals continued to be practised in the landscape, and hinterland monasteries were very much focused on administering their immediate surrounding lands. The terracotta corpus evident in the hinterland during the Early Medieval Period, as with the diversity within Buddhist monasticism, does not always fit within the Mahavamsa's narrative. Such diversity continued into the Late Medieval period accompanying the expansion and extension of networks of international trade, attested by the fifteenth century tri-lingual inscription of Admiral Zheng He in Galle. Recorded in Tamil, Chinese and Persian script it records the veneration of an incarnation of Vishnu, the Buddha, and the shrine of a Muslim saint, it provides ample evidence for the continued diverse and dynamic linguistic and religious framework of the island that had emerged and developed from the Early Historic Period onwards (Strathern and Biedermann This volume [Introduction]).
However, the continuity of settlement in the hinterland, combined with continuity of a homogenous artefactual record across the Citadel suggests that there were also distinct agreed and shared urban concepts of how a city should be laid out within society, even within the pluralistic context of religion and communities. Furthermore, such a model of landscape settlement has strong parallels with dynamics of civilisations studied archaeologically elsewhere. Such a model of central places and dispersed settlement has earned the term, 'Low Density Urbanism' and Anuradhapura shares many characteristics with communities living in other densely forested, tropical landscapes, such as Mesoamerica and South-east Asia. Whilst initial links were drawn between these areas on the basis of shared environmental and economic constraints (Coe 1957(Coe , 1961Bronson 1978), more recently scholars have identified shared social and ritual adaptations. Most notably is the presence of unifying public spacesthe stupa platforms of Anuradhapura and the public plazas of the Maya (Inomata 2006). Likewise, all areas were dependent on the large-scale management of water resourcesabundant at some times of the year and scarce at others and the long-term adaptation of the landscape to achieve agrarian sustainability (Fletcher 2009). Critical to this was the recognition of small-scale settlements in the landscape ) and approaching our view of urbanism as extending outwards from city walls, incorporating dispersed settlement patterns as a network of urban or semi-urban locales, interdependent with one another (Fletcher 2012: 288), and that low density urbanism "creates highly repetitious spatial patterns over vast regions…and affects the natural landscape in a variety of ways through economic extensification of the stable crop" (ibid.: 314). Such a system existed within the hinterland of Anuradhapura from the Early Historic period up until the final abandonment of the city in the eleventh and twelfth centuries AD, altohuhg it is not yet clear whether such a system was adopted around later cities such as Polonnaruva, Kotte and Kandy As the urban and monastic core grew and developed at Anuradhapura, the non-urban hinterland remained fairly stable in terms of its organisation, traditions and communities (Coningham et al. 2013c: 468-470), with evidence of various communities sharing unifying concepts.
This critical reappraisal of previous archaeological evidence in Sri Lanka, combined with new research in Anuradhapura and its hinterland has illustrated a large degree of diversity and connectivity within the Early Historic and Early Medieval periods. However, previous archaeological evidence of such connectivity has not been fully appreciated or set within a more nuanced discussion of identity in Sri Lanka. Whilst Polonnaruva, the successor capital to Anuradhapura has long been archaeologically attested as a centre of cosmopolitanism, similar evidence from Anuradhapura has been largely undervalued. The remains of Polonnaruva, occupied between AD 1017-1293, have revealed Buddhist monasteries and Hindu Temples with evidence of bronze sculptures of Hindu deities. Such evidence has been utilised by some scholars to furnish a narrative of religious plurality and harmony with both Buddhist and Hindu institutions patronized by Sinhalese monarchs (Indrapala 2005). Moreover, excavations within the Alahana Parivena in Polonnaruva (Prematilleke 1982a(Prematilleke , 1982b(Prematilleke , 1985(Prematilleke , 1988 uncovered quantities of pottery with applique designs, including swastika, srivvasta and vajra or trisula. Whilst bronze figurines representing deities, such as Siva and Pavarti, excavated at Polonnaruva have been put forward as evidence of the presence of Hinduism in the city (Paranavitana 1955: 79, Ratnayake 2010Thantilage 2010b), such objects have not gained as much attention in Anuradhapura, even though such evidence, was not restricted to Polonnaruva. For example, figures of Siva, Pavarti, Kevalamurti and Nrtyamurti, and potentially Ardhanarisvara, were recovered from a pillar foundation at Jetavana during Cultural Triangle excavations in the 1980s (Ratnayake 2010). Furthermore, three appliqué ceramic sherds with symbols similar to those from the Alahana Parivena were discovered in the later sequence of trench ASW2 at Anuradhapura (Coningham 1999c), as well as in the vicinity of Jetavana (Bell 1904c: Plate XII). Notwithstanding the evidence of a Nestorian community, the so-called Hindu ruins, in combination with sherds of applique pottery and bronze Hindu sculptures in Anuradhapura, the later evidence at Anuradhapura is striking in its similarity to the evidence usually presented for plurality of belief in Polonnaruva, yet Anuradhapura is generally not advanced as a similar example of religious plurality. As stated above, this is more likely the result of the scholarly promotion of Anuradhapura as a Buddhist capital and this bias may be due to historical research priorities and questions within the archaeology of Sri Lanka, many of which were outlined earlier in relation to dominant textual narratives.
Furthermore, it is often problematic to assign certain artefactual forms or symbols to particular groups. Sharing of artefactual categories, especially symbols such as the swastika and vajra/trisura, highlights the innate difficulties in South Asian archaeology, where without privileged knowledge, the affinity of whether such symbols belong to a Buddhist, Hindu or Jain context, or a combination, remains unknown. Indeed, it can lead to questions of when does an artefact become a Buddhist, Hindu or Jain object? The fluid identity of artefacts and monuments has recently been noted at a modern shrine to Ayanayake, lord of the jungle, in the Anuradhapura hinterland. The shrine was constructed using re-used pillars from nearby monasteries, but had also re-used an eroded Buddha head, which was painted with the trunk of Ayanayake, as the focus of the shrine (Coningham and Gunawardhana 2012: 293). Whilst objects may be used in more than one context with changed meanings, similar architectural motifs may be utilised by several traditions. For instance, stupas and viharas are not architectural forms restricted to Buddhism, sharing traits with Jain and Hindu architecture (Mitra 1971: 52;Ray 2004). Even though it is occasionally difficult to ascertain the religious affiliation of an object or structure, especially if these have the potential to change affiliation over time, such evidence does suggest cosmopolitanism through the sharing of ideas and concepts between differentiated communities. This chapter has presented the archaeological evidence for varying degrees of cosmopolitanism within Early Historic and Early Medieval Sri Lanka, in terms of trade, pilgrimage, patronage and urbanism. One aspect which is abundantly clear is that the archaeological evidence presents a far greater degree of international exchange and local variation than is evident within the textual sources. Sri Lanka's unique position at a crucial apex of Indian Ocean trade routes, as well as its rich natural resources, allowed it to fully engage in exchange networks. Rather than being the passive recipient of materials and goods from Roman ships, Sri Lankan traders were dispersed across the region, whilst Anuradhapura (and quite possibly other cities) was home to an international cohort of merchants. With them, they brought new languages, religions and conceptsas well as customs such as wine consumption, brought across the ocean in 'torpedo jars'. Such international connections were not only economically derived, but also prevalent in the Buddhist world. From the initial spread of Buddhism to the island in the third century BC to Faxian's travels in the fourth and fifth centuries AD and the reinvigoration of Buddhist monks from Myanmar, Sri Lanka has been part of a global network of communities and individuals. Indeed, Sri Lankan monasteries and shrines were established across India and South-east Asia. Within the island, there was a complex relationship between religious and secular institutions, in both an urban and non-urban context. The major monastic institutions of Anuradhapura had a fractious history, both soliciting support and condemnation from kings over time. It is also clear that Sri Lanka was not a uniquely Buddhist island, but had strong Hindu influences as well as more localised traditions as evidenced by the emergent terracotta cults in the hinterland of Anuradhapura. Indeed, the monastic institutions of Anuradhapura seemed to have played a much more dominant role in the colonisation, management and development of the wider landscape surrounding the city. Monasteries remained the only viable long-term structures within the hinterland, as attempts to secularise the management of the landscape faded. The epigraphic record reveals how land was donated by kings and other elites to these monasteries, to the degree of completely alienating it from state control. Sri Lankan scholars have long recognised this with Leslie Gunawardana stating in 1979 that "considerable powers were transferred to the monastic administration by withholding the authority of government officials to intervene in their affairs " (1979: 190) and, rather more starkly, by Dias that "lands and villages beyond the control of the central authority were given to the monasteries to bring some control over them…This way the monastic institutions became the landed intermediary between the central political authority and the people" (2001: 115). Bringing these four strands together, we have presented evidence of a highly cosmopolitan society with broad international links and outlooks; a society which sat within an ever expanding network of religious and economic exchange and patronage. Yet, at the same time, there were also distinctly pervasive traditionssubsistence, ceramic forms, craft manufacturing and patterns of low density urbanism. This contradictory stance is not overly surprising, and reflects the many facets of community identity, from the local to global. To what degree different communities had access to these opportunities probably acted on a sliding scale over time, but it is clear that Sri Lanka was not a peripheral part of the Indian subcontinent but at the very heart of the social, economic and religious development of the region.
Most of the evidence presented within this chapter has been derived from over twenty years of archaeological research at Anuradhapura (Coningham 1999a(Coningham , 2006, which have identified a high degree of plurality and cosmopolitanism, despite the city being portrayed as predominantly Buddhist and Sinhalese in nature. It is only through the adoption of nuanced archaeological methodologies, supported by scientific analysis that these new models are emerging. In order to extend and refine current models of Sri Lankan history derived from textual sources, further work is needed to analyse further artefactual categories, included applique ceramic forms in the later sequence at Anuradhapura and at Polonnaruva, and more generally at sites such as Polonnaruva, widely recognised as more plural in its development. The transmission of the capital from Anuradhapura to Polonnaruva has been projected as an abrupt and singular event, yet recent research has shown the abandonment of Anuradhapura and its hinterland to be a slow process that happened over several centuries (Gilliland et al. 2013). We would argue that around Polonnaruva there is likely to be evidence of much earlier settlements and communities that would shed more light on the nature of Early Medieval Sri Lanka. One further challenge that needs addressing is the archaeological visibility of women within early Sri Lanka's history and archaeology. Indeed, Faxian noted the presence of nuns within the Sacred City of Anuradhapura (Beal 1869) and inscriptions record donations by female devotees throughout Sri Lanka (Paranavitana 1970). Furthermore, communities of Buddhist nuns from Sri Lanka are recorded in an inscription at Nagarjunakonda in India (Ramachandran 1953).In spite of this there has been little discussion of them within both archaeological interpretations and Buddhist discourse. Gender is notoriously difficult to identify within the archaeological record, especially when dealing with the ephemeral remains of a subtropical landscape, however, if we want to truly understand the cosmopolitan nature of Early Historic and Early Medieval Sri Lanka, such challenges have to be acknowledged and then addressed.