The History Of Arakan (Arakan Kingdoms)

Ancient geography and recent archaeology: Dhanyawadi, Vesali and Mrauk-u.

Bob Hudson Archaeology Department, University of Sydney, Australia. “The Forgotten Kingdom of Arakan” History Workshop. Chulalongkorn University, Bangkok, November 2005.

Recent archaeological excavations and surveys at the old Arakanese sites of Dhanyawadi, Vesali and Mrauk-u raise new issues about each. It appears that Dhanyawadi is not the eccentric shape portrayed in early archaeological plans, but an oval site with some notable similarities to the walled Pyu sites of Upper Myanmar. Vesali shares one of these characteristics, an inwardcurving brick gate. A radiocarbon date, the first for Vesali, intriguingly places another city gate in the 13th century AD. A review of the huge earth banks that surround Mrauk-u suggests that the popular notion that these were defensive may be a romanticised interpretation of what was essentially a water management system.

Location. The early polities of Arakan were located in the valleys of the Kaladan and Lemro Rivers. While some traditional accounts locate early settlements, “royal capitals”, as far north as the Kyaukpandaung plateau (Tun Shwe Khine 1992: 20-21) the available evidence points to the alluvial lowlands. Satellite imagery (Figure 1) shows how restricted the area available for irrigated rice agriculture was. The old settlements occupy a strip of land that is only between 15 and 35 kilometres wide, and perhaps 60 kilometres from north to south. To the north, west and east are hills, and to the south the combined deltas of the two rivers meet the sea. Dhanyawadi. There are traditional claims in Arakan of royal capitals dating back to 3000 BC (Tun Shwe Khine 1992: 20). However the historical record begins with the c. AD 729 Anacandra inscription which describes how the founding king of the Candra Dynasty, Dvancandra (c. 370-425 AD), “built a city adorned by surrounding walls and a moat” (Johnston 1944; Gutman 1976: 63, Vol 1). This is Dhanyawadi, whose Gupta-period sculptures point to the 5th century AD (Gutman 2001: 29). It is the home of the Mahamuni shrine, an important pilgrimage site for Buddhists (Forchhammer 1892; Tun Shwe Khine 1994). The shrine is pretty much in the geographical centre of an oval outer wall which encloses an area of 5.6 square kilometres. Southwest of the shrine is a relatively square enclosed area, with another square series of walls inside it (Figure 2). These two sets of inner walls are generally interpreted as a palace. Apart from the walls themselves, and a couple of small brick structures, there are few brick foundations evident, suggesting that if this area enclosed an elite centre, then the inhabitants must have lived mainly in wooden structures built directly on the ground. Excavations on the eastern side show the walls curving inward to form a corridor, providing a narrow entranceway to the complex. These walls, like the outer city walls, are several metres thick, faced with brick, and filled with rubble (Kyaw Zan 2004). One thing that immediately strikes the observer on seeing the curved brick gate is the similarity with curved brick corridor gates that have been excavated at Halin, Beikthano and Sriksetra (Aung Myint 1998). We seem to have no written information from ancient times to tell us just why the gates were built in this shape. Was their function defence

against enemies, administrative (perhaps for the collection of taxes as people went through- some of the Upper Burma gates had niches that could have housed guards or officials) or cultural, to ensure that only members of the community that owned the walled city could enter? Until now, this kind of entranceway had appeared unique to First Millennium AD Upper Myanmar, but it now seems that the ancient architects must have exchanged a few ideas across the Arakan Yoma. Several important features came to light during field survey in 2005. The author, U Nyein Lwin, of the Archaeology Department in Mrauk-u and U Maung Maung Than, a staff member of the Mahamuni museum who was raised in the local area, undertook a program of “ground-truthing”, directly checking features that had previously been mapped or detected from aerial photos or satellite imagery. A key discovery was that the huge earth banks to the southeast of the Mahamuni, which have appeared on maps as part of the outer city wall, form quite a separate feature. They very likely became incorporated into the archaeological plan due to a misinterpretation of aerial photographs (Thin Kyi 1970) and were cheerfully accepted as giving the city an inexplicably eccentric outline by subsequent scholars, including the author (Gutman & Hudson 2004: 162). However inspection on the ground shows that there are brick remains in a field between the earth banks which form a continuous line with brick walls that run under the earthworks (Figure 2). The earth bank, sometimes known as the “gold and silver road”, has more than one folk tale attached to it. In one story, it was a twin road to Mrauk-u. In another, it was an artificial lake built by rival royals to hold boat races. Its walls are now breached, and crops are grown on its floor. Other finds from the ground survey include a curved brick gate on the outer east wall and a stone quarry, characterised by the remains of drill holes in the grey sandstone, at Kyauktalon, beyond the west wall. The early sculptures of Dhanyawadi and Vesali largely employ red sandstone, so the Kyauktalon quarry cannot be claimed as a source for these artworks. Outside the southern part of the outer wall we located a cluster of brick and/or stone platforms, typically about 8 metres square. They appear as low mounds on the ground. Many are preserved as field corners, presumably too hard to plough and too dense to make it worth the effort of removing the brick or stone. Perhaps they are religious monuments or graves. Careful excavation of one or two of them may provide valuable new information. Vesali. Art history and numismatic studies place Vesali between perhaps the 6th and 10th centuries AD (Nyunt Han 1984; Gutman 2001: 41). It is enclosed by a brick wall, with an area of 6.2 square kilometres. Excavations in the 1980s revealed several brick buildings. Regular finds of stone and bronze artifacts were noted then (Nyunt Han 1984) and since (Shwe Zan 1995). An inner walled area, known as the “palace site”, is obscured by the present village of Wethali, although brick remains are widely seen in the village pathways and roads. Recent excavations have unearthed a curved brick gateway on the northern side of the outer wall, which can be seen where the road to Dhanyawadi crosses the wall (Figure 3, VSL 8). This curved gate appears to have been overbuilt by later structures (Kyi Khin 2004), suggesting long-term use of the site. In the northwest corner, a different kind of gate was excavated, a gap in the wall with a large timber post set at each side (VSL 6). One of these posts has been radiocarbon dated to the period between AD 1260-1400 at 95.4% probability (sample OZH970, 670±40 BP, Australian Nuclear Science and Technology Organisation 2005). We should not rush to judgement on the basis of a single radiocarbon date, but at face value, the result suggests we should at least

not discount some kind of construction activity in a period that had previously been considered to be well after the time the city was occupied. Vesali has been called in the Arakanese chronicles the “city of stone stairs” (Gutman 1976: 21; Nyunt Han 1984). Local people point out a section of the bank of the Rann Chaung about 500 metres from VSL 6 where they say stonework has been seen, but none is visible today. Mrauk-u. The Mrauk-u period went from the 15th to 18th centuries AD, and seems to have been preceded by settlement activity along the Lemro River to the east in several centres including Sambawak/Pyinsa, Parein, Hkrit and Launggret (Harvey 1925: 137-149, 370371; Thin Kyi 1970; the Lemro sites were recently re-surveyed by Berliet 2004: 234239). A characteristic view of Mrauk-u is that the earth banks that surround particularly the eastern part of the city were constructed for defence, a maze “calculated to baffle any enemy”, with the capacity for the waters of the town’s reservoirs to be let loose to drown invaders (Collis 1923: 246). A new look at these earth banks, using maps (Burma One Inch 84 H/2), aerial photographs (thanks to Dr Elizabeth Moore for supplying a rare copy of a World War II aerial photo of Mrauk-u from the Williams-Hunt Collection at SOAS) and satellite imagery (LandSat 2000 and IKONOS 1 metre) suggests rather erratic planning if defence was the main aim (Figure 4). The earth banks of Mrauk-u cover an area of more than 20 square kilometres. They extend more than 6 kilometres to the northeast of the citadel, as far as the Lemro River. However to the southeast, they are effectively on the edge of the city, except for banks fronting low hills that encircle an alluvial plain. It does not really look like a militarily viable fortification. There seems to be at least one large gap through which invaders could comfortably march, along a stream between the northern and southern groups of earth banks and past the Koe-thaung and Pizi-taung pagodas (Figure 4). There is also the question of the structurally similar earthworks at Dhanyawadi. It is difficult to attribute a defensive function to the Dhanyawadi banks, which form a single reservoir backing on to a hillside catchment area (Figure 2). We might look to water management as a more likely reason for the construction of the earthworks at both sites, to keep the saline water from the surrounding tidal rivers and creeks at bay and permit rice irrigation. The Mrauk-u kings are described as building extensive bunds for water retention as far as the Lemro River in the mid-15th century (Smart 1917: 66-67). It must be admitted that in the case of an attack, the complex system of banks and tanks would have favoured the locals rather than the intruders. Fortified lookout posts, some with gun ports, remain on the hills around the town (Tun Shwe Khine 1992; Shwe Zan 1995). These fortifications, along with the walls, gates and different kinds of earth banks, are given individual names in the local tradition, a tradition which considers the earth banks to be fundamentally defensive (Department of Archaeology n.d.). Histories describe how King Minbin, in the 16th century, opened the sluices of the reservoirs to hold back Burmese/Peguan invaders (Harvey 1925: 140, 158). However wars are irregular occurrences, while agriculture is constant. The dramatic notion of drowning one’s enemies by flooding the city’s defences, a story told also of Beikthano (ASB 190506: 7), may be more appealing than the notion of a hydraulic engineering project, but the original construction of the earth banks should be seen as a creative approach to a difficult problem of water management that helped bring to Mrauk-u the prosperity that made it attractive over the years to adventurers from both inside and outside the society. This is not to suggest that the banks may not have been useful in the defence of the city, rather that any defensive advantage they may have given would have been the fortunate consequence of earlier decisions regarding water management. 3

References. ASB 1906-1965 Report of the Superintendent, Archaeological Survey of Burma. Rangoon, Office of the Superintendent, Government Printing. Aung Myint 1998 Site Characteristics of Pyu and Pagan ruins. A Comparative Study of the Dry Areas in Southeast Asia: International Seminar, Kyoto, Japan. Berliet, Ernelle 2004 Geographie Historique et Urbanisation en Birmanie et ses Pays Voisins, des origines (IIe s. av JC) a la fin du XIII siecle. Doctoral Thesis. Ecole Doctorale Sciences Humaines et Sociales, Universite Lumiere, Lyon II Collis, M.S. 1923 "The Golden City of Mrauk-U." Journal of the Burma Research Society 13(3): 244-256. Department of Archaeology n.d. A Brief Presentation of Cultural Heritage Sites in Rakhine State (undated mimeograph), Department of Archaeology, Ministry of Culture, Union of Myanmar. Forchhammer, E. 1892 Report on the antiquities of Arakan The Archaeology Department, Burma. Gutman, Pamela 1976 Ancient Arakan, with special reference to its cultural history between the 5th and 11th centuries. PhD thesis. Australian National University, Canberra. http://thesis.anu.edu.au/public/adtANU20050901.112732/index.html Gutman, Pamela 2001 Burma's Lost Kingdoms: Splendours of Arakan Orchid Press. Gutman, Pamela & Bob Hudson 2004 The Archaeology of Burma, from the Neolithic to Pagan. Southeast Asia: from prehistory to history. I. Glover & P. Bellwood. Abingdon & New York, Routledge Curzon: 149-176. Harvey, G. E. 1925 History of Burma Frank Cass & Co Ltd, London. Johnston, E.H. 1944 "Some Sanskrit inscriptions of Arakan." BSOAS 21: 357-385. Kyaw Zan 2004 Report on Excavation at Ancient Dhanyawadi (Interim Departmental Report, in Burmese), Ministry of Culture, Department of Archaeology, Archaeological Branch Office, Rakhine State. Kyi Khin 2004 Excavation Report on Mounds VSL-6, VSL-7 and VSL-8 in Ancient Vesali (Interim Departmental Report, in Burmese), Ministry of Culture, Department of Archaeology, Archaeological Branch Office, Rakhine State. Nyunt Han 1984 The Study of Ancient City Vesali (Departmental Report, in Burmese). Yangon, Department of Archaeology. Shwe Zan 1995 The Golden Mrauk-U, an ancient capital of Rakhine U Shwe Zan, Yangon. Smart, R.B. 1917 Burma Gazetteer, Akyab District, Volume A Superintendent, Government Printing and Stationery, Union of Burma, Rangoon. Thin Kyi, Daw 1970 "Arakanese capitals: a preliminary survey of their geographical siting." Journal of the Burma Research Society 53(2): 1-13. Tun Shwe Khine 1992 A guide to Mrauk-U, an ancient city of Rakhine, Myanmar U Tun Shwe. Tun Shwe Khine 1994 A guide to Mahamuni: the highly venerated golden image of Buddha with authentic long history Rakhine Book Series.

Figure 1: location of the old settlements of Arakan mentioned in the text, plus some major landmarks (LandSat 2000 satellite image).

Figure 2: Dhanyawadi.
Figure 3: Vesali.
Figure 4: Mrauk-u.
# Maps are by the author. # This research is supported by a grant and Postdoctoral Fellowship from the Australian Research Council.