The Sigiriya Royal Gardens Analysis of the Landscape Architectonic Composition

Dr. Nilan Cooray

PDF Link

https://www.sigiriya.site/publicaitons

The Sigiriya Royal Gardens Analysis of the Landscape Architectonic Composition Nilan Cooray Delft University of Technology, Faculty of Architecture, Department of Urbanism

The Sigiriya Royal Gardens Analysis of the landscape architectonic composition

Dit proefschrift is goedgekeurd door de promotoren: Prof. dr. ir. C. M. Steenbergen Prof. dr. E. A. de Jong Samenstelling promotiecommissie: Rector Magnificus, Voorzitter Prof. dr. ir. C. M. Steenbergen, Technische Universiteit Delft, promotor Prof. dr. E. A. de Jong, Universiteit van Amsterdam, promotor Prof. N. de Silva, University of Moratuwa, Sri Lanka Prof. ir. E.A.J. Luiten, Technische Universiteit Delft Prof. dr. ir. P.H. Meurs, Technische Universiteit Delft Prof. dr. ir. A. van den Brink, Wageningen Universiteit Dr. ir. W. Reh, Technische Universiteit Delft abe.tudelft

Ontwerp: Sirene Ontwerpers, Rotterdam ISBN 978-1480030978 ISSN 2212-3202 © 2012 Nilan Cooray

Contents (concise) Abstract 9 Acknowledgements 11 1 Introduction 23 2 Context for the Study 41 3 Previous Research and Interventions 85 4 The Methodology 119 5 The Analysis 131 6 Conclusions 223 7 Perspectives 251

Abstract Besides the efforts that are of a descriptive and celebrative nature, studies related to Sri Lanka’s historical built heritage largely view material remains in historical, sociological, socio-historical and semiological perspectives. There is hardly any serious attempt to view such material remains from a technical-analytical approach to understand the compositional aspects of their design. The 5th century AC royal complex at Sigiriya is no exception in this regard. The enormous wealth of information and the material remains unearthed during more than 100 years of field-based research by several generations of archaeologists provide an ideal opportunity for such analysis. The present study aims, therefore, to fill the gap in research related to Sri Lanka’s historical built heritage in general, and to Sigiriya in particular. Therefore, the present research attempts to read Sigiriya as a landscape architectonic design to expose its architectonic composition and design instruments. The study, which is approached from a technicalanalytical point of view, follows a methodological framework that was developed at the Landscape Design Department of the Faculty of Architecture at the Delft University of Technology. The study reveals that the architectonic design of Sigiriya constitutes multiple design layers and multiple layers of significance with materialspatial-metaphorical-functional coherence, and that it has both general and unique landscape architectonic elements, aspects, characteristics and qualities. The richness of its composition also enables the identification of the landscape architectural value of Sigiriya, which will help reshape policies related to conservation and presentation of Sigiriya as a heritage site, as well as to its protection and management as a green monument. The positive results of the study also underline that the methodology adopted in this research provides a framework for the study of other examples of historical gardens and landscapes in Sri Lanka, which will eventually provide insight into the typological aspects of a possible Sri Lankan tradition of landscape design.

Acknowledgements

The present research is the outcome of work carried out in the field, archives and the drawing studio; the analysis of such material was carried out in 1995, 1996, 1999 and 2003 in Delft. It was the late Prof. dr. ir. Frits van Voorden, Chair, Architectural and Urban Conservation of the Faculty of Architecture of the Delft University of Technology, the Netherlands (TUD), who supervised and guided my research before his untimely demise. In him I found a great mentor who made me look at historical built heritage differently from the conventional art-historical perspective. Prof. dr. ir. Clemens Steenbergen, Chair, Landscape Architecture of the same faculty, who was the co-supervisor during the initial stages of my research, agreed to be my supervisor after the demise of Prof. van Voorden. I am most privileged to be associated with and guided by Prof. Steenbergen, who is one of Europe’s innovative researchers on landscape design, and whose research work on West European gardens and landscapes was a great source of inspiration for me. Prof. Steenbergen devoted much of his valuable time to discussing my study, reading the text and drawings meticulously and giving critical remarks to improve the quality of this research work. When Prof. Steenbergen in 2011 retired as Chair, he requested Prof. dr. Erik de Jong, with his great expertise in the history of landscape architecture, also to be a supervisor of this project. He devoted much time to going through the text of my study and gave his specific input to improve its quality. It has been an enviable pleasure to carry out this research under their expert guidance, and I respectfully record my deepest gratitude to these three teachers. I also take this opportunity to pay tribute to Dr. Roland Silva, the former Director General of the Central Cultural Fund (CCF) and under whom I had the privilege of developing a carrier in heritage conservation, for persuading me and providing institutional assistance to carry out this doctoral research in the Netherlands and giving his expert comments at various stages of this study; to Prof. dr. Senake Bandaranayake, under whom I had the opportunity to work at Sigiriya, and to be inspired by his exceptional knowledge of, and scholarly research on, the site, and for giving me all the facilities to carry out the study during his tenure as the Director General of the CCF; to Dr. Nandana Chutiwongs, the former Curator of the Leiden Museum of Ethnology, and Prof. dr. K. R. van Kooij, the former Chair, South Asian Art and Archaeology, Kern Institute, Leiden University, for not only introducing me to Prof. van Voorden of TUD, but making valuable comments during my research work; to Dr. ir. Wouter Reh of the subdepartment of Landscape Architecture of TUD for giving his expert advice, especially on the methodological aspects of the study; to Prof. dr. P. L. Prematilleke, former head of the Department of Archaeology of the University of Peradeniya, Sri Lanka, for giving his scholarly advice during the research work; and to Prof. Nimal de Silva, former Dean of the Faculty of Architecture, University of Moratuwa, Sri Lanka, who was one of my teachers during my architectural training at the same university, for devoting his valuable time to read the text and giving his constructive criticism to improve it. I feel fortunate to have obtained such assistance from an array of distinguished experts. I am also grateful to Dr. Gamini Wijesuriya for his valuable comments and the encouragement given at different stages of the study; to Dr. Ron van Oers for accompanying me to European gardens during field work, commenting on the text and also helping me in diverse ways to make my stay in the Netherlands a pleasant one; to architect Ismeth Raheem for his valuable advice during the archival research; and to architect Jayatissa Herath, my colleague since our undergraduate days, for going through the text and drawings and making valuable suggestions to improve them. I wish to acknowledge all the assistance given by the non-academic staff of the subdepartments of Architectural and Urban Conservation and Landscape Architecture of TUD, especially Ms. Will Hoogendijk-Hagen and Ms. Margo van der Helm, for providing logistical support during my study periods at TUD. My study in Delft on several occasions was made possible through the scholarships offered to me by the Netherlands National Commission for UNESCO, and the balance financial support given by my employer, the CCF. I am grateful to both these institutions for the kind contributions made to engage and complete this study. Messrs. S. Sivanantharaja and S. M. J. S. Samarasinghe of the GIS Division, Department of Survey, were cooperative in obtaining the Department’s GIS data covering the Sigiriya region for the morphologic analysis related to the study. Mr. Kusumsiri Kodituwakku of the CCF’s Sigiriya Project was helpful in discussing some of the questions which I was confronted with during this research. Messrs. Mohammed Sabet of TUD, Sarith Karunaratne of Jayatissa Herath Associates and Chanaka Bogahawatte of the Department of Survey helped me in producing most of the drawings, while Ms. Dilanthi Atapattu and Ms. Rovina de Soyza of the CCF assisted me in the word processing. My brother-in-law, Mr. Linus Fernando, volunteered in editing the language of the text and Ms. Judith Waters Pasqualge undertook the copyediting of the final text. To all these institutions and individuals and many others not mentioned, I wish to express my heartfelt thanks. Finally, I wish to extend my deep sense of gratitude to my wife, Dilinie, my daughters, Nayanamalie, Divyanie and Navyangie, who, no doubt, appreciate my work despite much inconvenience caused to them throughout this study.

Nilan Cooray Colombo, September 2012

To the late Prof. dr. ir. Frits van Voorden Former Chair, Architectural and Urban Conservation, Delft University of Technology, the Netherlands

Contents (extensive)

1 Introduction 23

1.1 Sri Lanka’s Historical Built Landscapes: General Overview 23

1.2 The Central Issue and Main Objective 34

1.3 The Secondary Objectives 36

1.3.1 Positioning Sigiriya in the Global Landscape Design Context 36

1.3.2 Conservation, Presentation and Management of the Heritage Site of Sigiriya 38

1.4 The Approach, Method, Scope and Limitations 38

1.5 The Structure of the Thesis 40

2 Context for the Study 41

2.1 General Context 41

2.1.1 Environment, Topography and Society 41

2.1.2 Attitude towards Nature 44 \

\ 2.1.3 The Cosmology 45

2.1.4 Discourse of Kingship 49

2.1.5 City Planning and Architecture 51

2.1.6 Irrigation Engineering 56

2.1.7 Pleasure Activities of Royalty 58

2.1.8 External Contacts 59 \

2.2 Sigiriya: A Basic Introduction 61

2.2.1 Historical Context 61

2.2.2 Physical Characteristics 63

3 Previous Research and Interventions 85

3.1 Previous Studies and Interpretations of Sigiriya 85

3.1.1 Function and Meaning 86

3.1.2 Design and Technical Aspects 94

3.1.3 Observations 102

3.2 Previous Conservation, Presentation and Development Activities 104

3.2.1 Sigiriya Proper 104

3.2.2 Visitor Infrastructure and Facilities 112

3.2.3 The Surrounding Hinterland 114

3.2.4 Observations 118

4 The Methodology 119

4.1 Survey of Existing Methodologies 119

4.2 Methodology of the Delft Tradition of Research 121

4.3 Adaptation of the Methodology for the Present Study 123

5 The Analysis 131

5.1 Characteristics of the Natural and Agricultural Landscape 131

5.1.1 Geomorphology and Topography 131

5.1.2 Natural, Topographic Configurations 142

5.2 Interaction of the Formal Layout with the Natural and Agricultural Landscape 149

5.2.1 Lines of the Formal Layout 149

5.2.2 Location and Orientation 151

5.2.3 Axial Arrangement 156

5.3 Geometry of the Scheme 160 17 Contents (extensive)

5.4 Symmetry and Asymmetry 165

5.4.1 Symmetry 165

5.4.2 Asymmetry 173

5.5 Incorporation and Organization of Spatial Entities 181

5.6 Treatment of Panorama/Horizon 184

5.7 Spatial Depth 190

5.8 Scenography of Movement 191

5.9 Visual Structure 203

5.9.1 Visual Layers 203

5.9.2 Hydraulic Elements 206

5.9.3 Natural and Built Elements 212

5.10 Activities and Functional Elements 215

5.10.1 Programmatic Domains 215

5.10.2 Pleasure Program 216

5.10.3 Functional Relationship of Spaces 218

6 Conclusions 223

6.1 Landscape Design Characteristics of Sigiriya 223

6.1.1 The Basic Form 223

6.1.2 The Spatial Form 229

6.1.3 The Metaphorical Form 232

6.1.4 The Programmatic Form 233

6.2 Sigiriya and Other Landscape Traditions: A Comparison of Landscape Designs 235

6.3 Sigiriya as a Landscape Architectonic Composition 246 18 The Sigiriya Royal Gardens

7 Perspectives 251

7.1 Conservation, Presentation and Management 251

7.2 Scientific and Methodological Aspects of the Research 259

7.3 Sigiriya in the Local and Global Design Context 262

7.3.1 A Sri Lankan Tradition of Landscape Design? 262

7.3.2 Landscape Design in the Global Context 262

List of Figures 265 Summary (Samenvatting)

271 Curriculum Vitae 277

Reference List 279

1. Introduction

1.1 Sri Lanka’s Historical Built Landscapes: General Overview

Historical references in literary sources indicate that Sri Lanka’s earliest built landscapes were the royal parks and woods that existed in the ancient royal center of Anuradhapura during the pre-Buddhist period (before the 3rd century BC), which were later offered to Buddhist monks by transforming them to monastic use. Therefore, the pre-Buddhist Mahameghavana and Nandana (later Jotivana) gardens at Anuradhapura could be regarded as such woods and parks of royalty.1 As per the commentaries of the chronicle Mahavamsa (chapters XI:2-3, XV:1-3, 7-9) these woods and parks would have consisted of thick foliaged fruit and flowering trees, aromatic plants and streams providing shade and coolness. The chronicle Culavamsa (chapter LXXIX:2-3, 4-5) mentions that the park called Nandana at Polonnaruva (12th century AC) was adorned with hundreds of fruits and blossoms, and indicates that the park called Lakkhuyyana, also at Polonnaruva, was planted with thousands of trees of every variety. The offering of a royal lodge within Mahameghavana at Anuradhapura to Arahant Mahinda, who led Buddhist missionary activities to Sri Lanka from India, for his night stay indicates that built structures were also elements of the landscape design of such parks and woods (Mahavamsa, chapter XV:11-13). An inscription of Nissankamalla (1187-1196) found at the pavilion named Priti-dana-mandapa at Polonnaruva, where the king distributed alms for the poor and forbade the picking of fruits from the surrounding orchards, indicates that this pavilion was also a built structure of a park (Prematilleke and Karunaratne, 1993:111). The references in the Mahavamsa (chapter XIV:1-10) to the meeting of King Devanampiya Tissa and Arahant Mahinda at Mihintale, 13 kilometers east of Anuradhapura, during a hunting expedition of the king also indirectly indicate that hunting parks were not unusual, at least during Sri Lanka’s pre-Buddhist era.2

1 The chronicle Mahavamsa (chapter I:21-23) also mentions a garden called Mahanaga that existed during the 5th century BC at Mahiyangana in central Sri Lanka as a customary meeting place of Yakkhas, one of the early inhabitants of the island. For more information on gardens referred to in ancient literary sources, see Dissanayake (2003:5-62).

2 Geiger (1960:62) mentions that hunting was a sport to which the king and noblemen were devoted. Adithya (1981:9), referring to the Mahavamsa, points out that hunting was an acceptedsport, despite Buddhismbeing the state religion, and killing was resorted to by royalty.

Figure1.1

1 Ranmasu Uyana 2. Isurumuniya 3. Vessagiriya 4. Tisavava Reservoir

Location of royal and monastic landscapes in relation to Tisavava reservoirat Anuradhapura (courtesy: Bandaranayaka 1993a)

The most interesting built landscapes, however, were the pleasure gardens (magul uyan) for the sensual enjoyment of royalty that continued over the centuries with a possible pre-Buddhist origin.3 As seen at Ranmasu Uyana4 (7th or 8th century

AC5), located below the earthen bund of Tisavava, a gigantic man-made reservoir at Anuradhapura (figures 1.1, 1.2), and at Dip Uyana (literally: Promontory Park, 12th century AC), located between the citadel and Parakramasamudra (figure 1.3), the great man-made reservoir at Polonnaruva, the pleasure gardens were elaborated with moatedisland pavilions and summer palaces, ornamental baths and swimming pools, artificial waterfalls and cascades.6 Ranmasu Uyana also showcases that natural boulders were used as caves and to build garden structures upon them.7 However, the most famous garden is at Sigiriya (5th century AC), which is the subject of this study. The provincial royal complex at Galabadda (11th century AC) in southern Sri Lanka, which is still largely uninvestigated, also falls into this category. The royal precinct of Kandy (19th century AC), the last royal capital of Sri Lanka, with the vast sheet of water of the man- made lake in the foreground and the green-forested mountain in the background, wasdesigned to give an impression of an abode,floating in the sky8 (figure 5.60).

3. Studies on Sinhala literature also indicate that water sports and water gardens were important aspects of Sri Lankan courtlylife (Gooneratne, 1983).

4. See ASCAR (1940-1945:18-22) and Paranavitana (1944:193-209) for detailed accountsof Ranmasu Uyana.

3 Van Lohuizen (1979:340) assignsthe date of the rock-cutreliefs of elephants on either side of a royal bath house at Ranmasu Uyana to the late 7th or early 8th century AC. However, Dissanayake (2003:55) argues that the garden was established during the periodfrom 3rd to 1st centuryBC.

4 The Culavamsa (chapter LXXIII:98-100) gives avivid description of the different varieties of trees and structures within the pleasure gardens at Polonnaruva. For a detail description of the variety of trees grown within the historical built landscapes and their use, see Dissanayake (2003:224-230). Also see Ashton, et al. (1997:391- 399) for the use of trees in Sri Lanka’s history, and De Silva, N. (1998) for indigenous trees associated in the traditional landscapes of Sri Lanka.

5 Upavana Vinodaya (1964:21) prescribes creating artificial caves and archeswithin the gardensby means of vegetation, and this indicatesthat caves are important featuresof the built landscape.

6 According to tradition, King Sri Vickramarajasingha (1798-1815) commissioned the royal architectDevendra Mulacariya to transform the city of Kandy into a celestial city. After seven days, the royal architect explained the designconcept to the king thus: “Your Majesty,I imagine that the thick green Udawattakele (the forested

mountain) behind the palace buildingis Neela Megha, the blue clouds of the sky; and that in front of the palace building,it is possible to createcloud-walls in white.Then by transforming the paddy fieldsinto a lake, one will see the reflection of cloud-walls and the palace buildings in water, and no doubt, it will appear like a city floating in the sky” (cited by De Silva,N. 1993:159).

Figure 1.2

Ranmasu Uyana at Anuradhapura

Figure 1.3

Dip Uyana at Polonnaruva (photo:Roelof Munneke)

Such architectural elements as walakulu bemma (the cloud drift parapet wall, as depicted in temple paintings) running at the level of the base of the palace buildings, and diyareli bemma (the wave swell parapet wall) over the lake, also heighten the above impression and reinforce the dynamic composition of the scheme.9 Although not yetinvestigated from a landscape design point of view, the rock-associated royal capitals of Yapahuwa (13th century AC), Dambadeniya (13th century AC) and possibly Kurunegala (14th centuryAC) could also be significant built landscapes (figure1.4).

1. Rock

2. Inner City

3. Outer City

4. Monastic Precinct

5. Moat

6. Rampart

Figure 1.4 Layout, rock associated royalcapital at Yapahuwa

9 For further detail on walakulu bemmaand diyareli bemma,see Seneviratne (1983:86) and Duncan (1990:101- 107).

The cave/rock-associated Buddhistmonastic sites, which were meant to concentrate on the religious ideals of the monks, were the earliest monastic landscapes10 of the island, datingfrom the 3rd century BC. As seen at Mihintale, Dambulla and numeroussuch sites (almost all dating from the 3rd or 2nd century BC), natural and/or partly excavated cave shelters on rocky mountain peaks or slopes were utilized with minimum disturbance to the naturalsetting. Sigiriya (3rd/2ndcentury BC) and modern Vessagiriya (ancient Issaramana, 3rd century BC, figures 1.1, 1.5) at Anuradhapura were also cave-associated monastic landscapes, before they were altered in the latter half of the 5th centuryAC.11

Figure 1.5

Vessagiriya at Anuradhapura

10 For more description of Sri Lankan monastic landscapes, see De Silva, N. (2009).

11 Kasyapa (477-495) absorbed the monastic landscapeof Sigiriya into his royal garden, while that of Vessagiriya was relaid as a ‘pancavasa’ type of monastery and renamed after his two daughters and himself as ‘Bo-Upulvan- Kasub-giri vehera’. For furtherdetail, see Silva (1988:222, 228-229).

The modern Isurumuniya at Anuradhapura, betweenRanmasu Uyana and Vessagiriya, provides a different example, with moats and island structures arranged in an axial layout with the backdrop of a towering rock mass12 (figures 1.1, 1.6, 1.7). As seen at Puliyankulama at Anuradhapura (10th century AC) and numerous other sites in the island, the Pancavasa Monasteries presents a highly formal layout with a series of concentric squares of monks’ cells, so ordered as to circumscribe an elevated, central ritual quadrangle. These monasteries are usually enclosedby a moat with axial avenues orientedto the cardinal directions leadingup to the central quadrangle13 (figure 1.8).

Figure 1.6 Layout, Isurumuniya at Anuradhapura (courtesy: Bandaranayake 1993a) 1. Tisavava Reservoir 2. Earthen Bund 3. Moat/Pond 4. Island Structure (?) 5. Rock 6. Earthen Embankment

12 Due to the secular sculpture found at the site and to its location in relation to Ranmasu Uyana, the very existence of Isurumuniya as a monastic establishment is questionable. Bandaranayake (1993a:32) suggests that together with Ranmasu Uyana, they were likely to have come under royal and monastic use collectively or separately at various stages of history. However, van Lohuizen (1979:340)attributes the secular rock-reliefs at Isurumuniya to the same period as those at Ranmasu Uyana (late 7th or early 8th century AC). Judging by the decorations at the entrance to the rock-cutshrine, Paranavitana (1953:173), on the other hand, suggeststhat the site was a Buddhistcenter from about the 10th century AC, or later.

13 See Silva (1995:215-273) for the planningsystem of the Pancavasa Monasteries.

The forest monasteries of Ritigala (9th century AC) and Arankele (9th-10th century AC), with well-defined pathways that traverse the thick wooded natural topography, and subtle positioning of the monastic structures with restricted views, present different characteristics to those described above14 (figures 1.9, 1.10). However, the monasticlandscape associated with Kaludiya Pokuna (literally: Dark Water Pond, 8th- 10th century AC) at Mihintale is unique and a rare example which demonstrates the integration of man-madegeometrical forms with natural elementssuch as rocks and water. The visual integration of the natural landscape beyond the monastery proper is achieved throughits large man-madepool (figures 1.11, 1.12).

Figure 1.7

General view, Isurumuniya at Anuradhapura

14 For a detailed description of theforest monasteries, see Wijesuriya (1998).

Figure 1.8

Layout, Pancavasa Monastery, Puliyankulama, Anuradhapura

1. Central Ritual Quadrangle

2. Concentric Squares of Monks’ Cells

3. Moat

4. Service Buildings

Figure 1.9

Pathway, Ritigala forest monastery (photo Lakshman Nadaraja; courtesy: Fernando, S. 2009)

Figure 1.10

Winding pathway, Arankeleforest monastery (photo:Lakshman Nadaraja; courtesy: Fernando, S. 2009)

Figure 1.11

Layout, flaludiya Pokuna Monastery at Mihintale

1.Pool

2.Stupa

3.Monastic Buildings

4.Island Structure

5.Entrance Porch

6.Axial Avenue

Figure 1.12

General view, flaludiyaPokuna Monastery at Mihintale

1.2 The CentralIssue and Main Objective

The site of Sigiriya celebrated 100 years of archaeological activity, research and interpretation in 1994. Many scholarly studieshave been carried out on various aspectsduring this period. With exposure of the remains of garden structures at the western precincts of the complex in the late 1940s, archaeologists began to interpretSigiriya also as an example of built landscape that can be dated to the 5th century AC (ASCAR 1952:18). Up to the commencement of the UNESCO-Sri Lanka Project of the Cultural Triangle at Sigiriya in the early 1980s, scholarly consideration of the historical built landscapes of Sri Lanka had mainly focused on several isolated examples, such as the royal gardens at Sigiriya, Ranmasu Uyana at Anuradhapura and the monastic gardens of AlahanaParivena at Polonnaruva (Paranavitana, 1944, 1955; Bandaranayake 1976).With the commencement of the above project at Sigiriya, much interest was generated not only to study various aspects of the built landscape itself, but also to understandthe process behind its creation. In this regard, Silva (1984) speculates that earlier monastic concepts of the arama or the park must have set the pace to design norms at Sigiriya, while Bandaranayake (1993a:8)declares,

‘The Sigiriyagardens are the survivals of a fairly recently identified Sri Lankan traditionof garden-art, of which there are few other surviving examples, some historical and literary documentation, and traces and fragments at nearly every site of the historical period. The gardens at Sigiriya are a concreteand mature expression on a grand scale of these various strands and traditions, which we see at other sites and in literary descriptions, in a fragmentary form.’

Based on the physical elements and characteristics, Bandaranayake (1993a:8-25) goes on to identifythree distinct types of ‘landscape traditions’ at Sigiriya: symmetrical water gardens, organic boulder gardens and stepped tiers or hanging gardens; he remarks that ‘each of these [traditions] has clear antecedents and successors within the Sri Lankantradition itself.’

As such, Silva attempts to hypothesize that Sigiriya was not a sudden occurrence in history, but that the already existing landscape design knowledge played a role in its landscape design. Bandaranayake goes one step further and hypothesizes that Sri Lanka had a tradition of landscape design and Sigiriyawas the climax of this tradition. Therefore, the above two statements focus on two important and interconnected theoretical issues that are central to research on historic built landscape in Sri Lanka: Did Sri Lanka possess a distinctive landscape tradition of its own?; Are historical examples, including Sigiriya,an outcome of this tradition?

There is no doubtthat a number of factors, such as ideological, political, socio- cultural and technical contexts of the period, would have influenced this process. Although Bandaranayake (1990a, 1993a) briefly touches upon these issues from a historical point of view, there have been no serious and in-depthstudies carried out in this regard. Therefore, one of the tasks for researchers of all disciplines related to historic built landscapes in Sri Lanka is to contribute to addressing these issues. From a landscape design point of view, the above-mentioned theoretical issues lead to several important and again interconnected research questions that are related to the design of historical built landscapes in Sri Lanka:Do Sigiriya and other historical built landscapes belong to one family of common characteristics (of landscape design)?; Which work(s) provided design ingredients to the designersof Sigiriya?; Was Sigiriya a prototype where new architectural inventions were experimented with?; or Was it a synthesis of many elements of Sri Lankan landscape tradition?; or Was it both?; What role in turn was played by Sigiriya in this tradition?Inorder to answer such researchquestions, one has to first carry out separate studies on the landscape design of Sigiriyaand other significant historical built landscapes that have survived in order to see by what (architectural) means the elementsare arranged/organized in their composition and so to understand their characteristics of landscape design. The comparison of such characteristics of these various historical built landscapes will then provide insight into answering the above researchquestions. Since it is not feasibleto undertake such a vast scope in a singlestudy of limitedduration, as a point of departure the present study will focus on Sigiriya, which is the relatively best preserved, much explored and well-documented historic built landscape in Sri Lanka.

Therefore, the main objective of the present study is to read Sigiriya as a landscape architectonic design with special emphasison the architectonic means (designtools, principles/rules, techniques) employed by the designers to arrange/organize the elements in the composition and thereby to understand its design characteristics.

1.3 The Secondary Objectives

Apart from partially contributing to the answers of the research questions mentioned in 1.2 above, the present study has the followingsecondary objectives that have either scientific or practical significance.

1.3.1 Positioning Sigiriya in the GlobalLandscape Design Context In order to position Sigiriya among other international landscape traditions, several criteria could be proposed. Antiquity coupled with the degree of preservation is one of them. As far as antiquity is concerned, only age is involved as a factor, and as shown in Figure 1.13 there were numerous gardens belonging to Egyptian, Babylonian, Chinese, Assyrian, Persian and Roman landscape traditions that predate Sigiriya. If the degree of preservation of its physical fabric is brought into the equation, Sigiriya ranks only second to the gardens of Rome, such as the private and public gardens of Pompeii and Herculaneum and the imperial Villa Adriana at Tivoli (Bandaranayake, 1993a:25). Despite articles published at international level (for example, Bandaranayake, 1993a, 1986b, 1997; Bopearachchi, 1993; Cooray, N., 1997), and despite attempts to find international parallelisms and correspondence with the ‘three types of garden traditions’ at Sigiriya by Bandaranayake (1993a:25-26), many international publications on builtlandscapes suggest that

Sigiriya is still almost unknown in the global landscape architectural design context.15 There is hardly any serious study so far that comparesSigiriya with otherinternational gardens (and traditions) in terms of a landscape design point of view. Therefore, one of the scientific objectives of the present study is to position Sigiriya in the global landscape designcontext through a comparison of its landscape design with those

of other international landscape traditions. However, as will be pointed out later, the present attempt is confinedto a comparison of the landscape design of Sigiriyato those of the West European landscape tradition.16

Figure 1.13

Schedule showingthe antiquarian valueof Sigiriya in relation to other international garden traditions

15. For instance, there is no mentionof Sigiriya in The Landscape of Man (Jellicoe and Jellicoe, 1995),which is a concise global view of built landscapes from the prehistoric to present eras, and is regarded as the standard work on the subject, not only among landscape architects, but other academics interested in the subject. The International Symposium on Gardens held in Leiden,the Netherlands, in 1990 (Fat and De Jong, 1991), as a project of the World Decade of Cultural Development of the UN and UNESCO, is another instance where there was no mention of Sigiriya in the discussions and deliberations on gardens of Europe, the Middle East and Far East. One exception is Holmes (2001:12-13), in which Sigiriyais mentioned at a globallevel among other historic gardens.

16. For the justification, see section 6.2.

1.3.2 Conservation, Presentation and Management of the HeritageSite of Sigiriya

Sigiriya has been conserved, presented and managed as a heritage site since the 1890s.17 The degree of scientific understanding of the site since then has no doubt influenced the shaping of the site’s conservation, presentation and management policies. As will be pointed out in Chapter3, the dominance of the archaeo-historical studies on Sigiriya would have also contributed to giving prominence to antiquarian values in deciding such policies. On the other hand, the increase in cultural tourismto the site since the 1980s has necessitated facilities for a betterinterpretation

of the site, as well as the establishment of such facilitiesas toilets, parking and visitor access within the site, and a proper road network, restaurants, guest houses and hotels within its hinterland. Therefore, such demands also have an impact on policies regarding the presentation and management of the site. Since the present study is expected to provide scientific knowledge on the landscape designof Sigiriya, and thereby to underline its landscape architectural value, one of the practical objectives is to re-examine the validity of some of the interventions carried out so far at Sigiriya,and to discuss policy issues and make suggestions for the consideration of heritage professionals in reshaping the site’s future conservation, presentation and management.

1.4 The Approach, Method,Scope and Limitations

Since the main objective of the present study is to read Sigiriya as a landscape architectonic design,the study will best be approached from a technical-analytical point of view.18 In such a study the availability of architectural theories, philosophy and the approach of the original designer, together with other works of the same designer, are vital sources. Moreover, the client’s functional requirements and

his aspirations are major considerations in the design process. Access to such information makes a study very much easier. However,in the study of Sigiriya,a historical example of the distant past, the original designer is unknown. Unlike designers of the recent past, any clues to the designer’s architectural theories, philosophies and approachin the form of letters,diaries and sketchesare also

17 See section 3.2 for details.

18 See Chapter 4 for the justification of a technical-analytical study.

absent. Further, there are no records of other works attributed to the same designer.19 Although the client for Sigiriya and his aspirations, at least during the major constructional phase, are brieflyindicated in the chronicles, it is difficult to rely on such information due to the biased nature of recording by the chroniclers. Such information may be useful only to cross-examine the results of the presentstudy.

Therefore, all these suggest that the only option availableto carry out the presentstudy is to analyzeand interpret the existing materialremains at the site.

As will be pointed in Chapter 3, Sigiriya is a multiperiod site with monastic phases preceding and following its royal phase. However, Sigiriya’s dominant identity is a royal complex, and hence this study will be limitedto analysis of the materialremains related to the royal phase. Therefore, analysis of the material remains related to other phasesis not within the scope of the present research.

Being a site that has been overrun by nature for several centuries, and due to abandonment, the superstructure of almost all the built features has disappeared. As far as the original plantingscheme is considered, it has gone through fastercycles without leaving any physical trace of its original character.20 In its present form, therefore, the site does not provideall the data required for this study. Unlike in

the western tradition there are no engravings of Sigiriya. Thus, the present study is limited to the material remains and the data that have been uncovered during the last 100 years of archaeological activity. This means that the outcomeof such a study cannot be treated as final, and will no doubt be elaborated on in the future with theexposure of more findings, employing advanced archaeological methods that will be developed in the future.

With regard to the research questions raised in 1.2 above, finding answers to such questions will only be possible after identifying the landscape design characteristics of Sigiriya and other historical built landscapes in Sri Lanka.Since a detailedstudy of all these built landscapes to identify such characteristics is an enormous academicexercise, which is not feasible in a single study, the present study is limited to indentifying the design characteristics of Sigiriya, and hence becomesthe first

step of this long process. The next stepsto be followed will be to study separately

19 As per the inscriptional evidence at Vessagiriya in Anuradhapura recordingthat Kasyapa has relaid the site as a monastery. De Silva, N. (2012) however opines that due to the use of Lapis Lazuli, a semi-precious stone imported from Afganistan, as a blue pigment in paintings found only at Sigiriya, Vessagiriya and Ranmasu

Uyana and the similarities in design conceptsfound at these three sites (‘landscape driven architecture’ insteadof ‘architecture driven landscape’) are factors that can be considered as evidence to attribute all three of these works to the same architect or the same design schoolserved during Kasyapa’s reign.

20 Although there have been a few attempts to investigate the archaeological and palaeo environment of Sigiriya and plant micro-fossil remains,such as pollen and hydroliths, especially from pond sediments, with a view to identifythe plants of the originalSigiriya garden (Bandaranayake, 1994d), the resultsof such studies do not provide insightto the present study.

other historical built landscapes to identify their individual design characteristics

– subjects of a series of studies of a similar nature. The comparison of such characteristics of these various historical built landscapes will ultimately provide answers to the research questions raised in 1.2 above. Hence, answering the research questionsis not within the scope of the present study.

1.5 The Structure of the Thesis

Chapter 2 provides a general cultural-geographic, technicaland historic context for the present study. Chapter 3 inquires into two aspects that are relevant to the presentstudy: first, the approaches to and conclusions of previous studiesand interpretations on Sigiriya by various scholars, with a view to see if such approaches are capable ofproviding a methodological framework for the present analysisof material remains

at Sigiriya; and second, the policies and approaches of conservation, presentation and management activities carried out so far at the site. The intention of Chapter 4 is to derive a methodological framework for the present study. It also involves a survey of the existingmethodologies, analytical toolsand keys employedand developed

by various researchers to study landscape design around the world. Considering the situation of Sigiriya, an appropriate methodology together with tools and keys will then be adapted to read the landscape design of Sigiriya. Chapter 5 deals with an analysisof the landscape design of Sigiriya by applying the methodological frameworkadopted in Chapter 4, and hence becomes the core chapterof the study. Based on

the above analysis, Chapter 6 will derive specific statements with regard to the design characteristics and also make a comparison with other world garden traditions in order to position Sigiriya in the global design context to read Sigiriya as a landscape architectonic composition. Chapter 7 dealswith the perspectives that may arise

out of the present study, in order to discuss policy issues and make suggestions to help reformulate overall policy and programs for the conservation, presentation and management of Sigiriya, and also to discuss the scientific and methodological aspectsof the present researchitself, based on the resultsof Chapter 5. This final chapter

is also expected to suggest other possible future studies that may arise out of the present study.

2 Context for the Study

2.1 General Context

2.1.1 Environment, Topography and Society

Situated approximately 50 kilometers off the southerntip of mainland Asia, Sri Lanka is an island in the Indian Ocean, measuring 430 kilometers north to south and 225 kilometers east to west (figure 0.1). The topographyconsists of a lowland coastal belt, an intermediate upland towards the interior, and highland in the center. The highland, which consists of a mountain mass, reaches a maximum elevation of about 2,500 meters abovesea level (figure2.1). Due to its locationin the Indian Ocean, the climate is influenced by the tropicalmonsoon system. This wind patterntogether with the central highland, which acts as a barrier more or less at right angles to the direction in which the monsoon blows, divide the country into two distinct climatic zones. The western, southwestern and central regions, which constitute the wet zone, receive a heavy rainfall, especially through the southwestmonsoon from May to September. Hot humid weather prevails in the lowlands, while it is cooler and more wholesome in the central highland. The dry zone, which constitutes about 70% of the total land area, in contrast, receives rain onlyfrom the northeast monsoon from mid-October to January, and suffers an annual severe drought for the rest of the year. Dry and harsh weather prevails in the north central, eastern and south central dryzone, while it is semi-arid in the northwest and southeast, and arid in the further north. The topography of the dry zone predominately consists of intermediate upland and lowland. The gently undulating terrain with several isolated erosion remnants and rocky hills in the north central plain gives way to monotonous, featureless flat landscape further north. The natural vegetation is characterized by rain forests in the wet zone with drier variants in the dry zone. Several rivers and streams that emerge from the central highland(the island’s hydrographic hub) flow in a radial pattern across both zones into the Indian Ocean. The rivers across the wet zone are perennial, while those across the dry zone are mostly seasonal in their flow with a tendency to dry out during droughts.One of the notable exceptions is the Mahaveli(the ‘Great Sandy

River’), which flows in a northeasterly direction across the dry zone and reaches the sea at Trincomalee on the easternseaboard. Apart from the wet dolines (villu) to the northwestand northeast, naturallakes are lacking.However, the coastalbelt is characterized by severallagoons and beachesof scenic beauty.

Figure 2.1

Map of Sri Lanka showingclimatic zones, topography, wind and the drainage pattern

A Wet Zone

B Intermediate Zone

C Dry Zone

D Arid Zone

Contemporary research (Deraniyagala, S., 1992:686, 2001:54) reveals that by about 125,000 BP there were prehistoric settlements in the island. During the Prehistoric Iron Age (circa 900-600 BC), the inhabitants had already settleddown to communal living to farm the land (Deraniyagala, S., 1992:709).21 The geographic position of Sri Lanka in relation to mainland Asia also made it one ofthe terminal points of constantimmigration from the subcontinent. In or around 500 BC, therefore, the culture of the indigenous inhabitants was overwhelmed by immigrants from India, who predominantly spoke an Indo-Aryan dialect (Paranavitana, 1967:1).The two cultures merged, and it was this amalgamation that laid the foundation to form a distinctive island civilization in the succeeding

centuries. These settlements were basically concentrated in the harsh dry zone. As at present rice was the staple food, and successful cultivation of rice in the dry zone depended upon proper storage and managementof rainwater. Initiallyexisting streams and rivers were dammed to construct reservoirs of modest scale to irrigatethe land during the dry season. These reservoirs thus became the first significant man-made features in the dry zone landscape.

During the latterpart of the first millennium BC, initial signsof urbanization emergedat Anuradhapura, located in the north central plains. Buddhism was introduced to the island from its birthplace in northern India during the reign of Asoka, emperor

of India in the 3rd century BC. Anuradhapura quickly became the center of both political power and the Buddhist religion. The new religion provided a serene philosophy of life, which served as an enduringsource of inspiration for creativity in art, architecture and literature. The social structure was based on the caste system, with the king at the apex and Buddhistmonks acting as the spiritual guardians of the nation, and as advisers to the king. Such features of state formation as the use of a script for writing, participation in external trade, emergence of Buddhist monasticestablishments, technological advances and the resultant social stratification contributed to make Sri Lanka a distinctworld civilization during the first millennium AC, with Anuradhapura as its great urban center.

Sri Lanka’s close proximity to the South Asian mainland on the one hand and its isolation due to the ocean on the other– as Japan is to China, and England to the European mainland – has made it considerably influenced by the mainland and atthe same time has preservedits distinct individuality.

21 The latest research indicates that the farming of barley and oats was practiced at least from about 8,000 BC in central Sri Lanka (Premathilake, 2000).

2.1.2 Attitude towards Nature

Ancient Sri Lankans, like those of many South Asian countries, did not distinguish themselves from nature, but believed that they were part of it. This attitude made them pay immense respect towards nature. The religio-cultural belief system of society alsoreflects this attitude. For instance, it is believed that deities (wruksha devata) live in association with trees,particularly those with large canopies, and the practiceof lighting oil lamps undersuch trees as a mark of respectto the deity is continued by villagers even today (de Silva, N., 1996:6). Buddhistteachings further promotedrespect towards nature.The three important events of the Buddha – birth, attainment of spiritual enlightenment and passing away – were directly associated with trees. The Bodhi-tree under which Buddha attained spiritual enlightenment and supreme wisdom is sacred to Buddhists. A sapling (southern branch) of this sacred tree at Bodhgaya in India, brought to Sri Lanka in the 3rd century BC and planted at the park of Mahameghavana at Anuradhapura, is still venerated by Buddhists as a supreme relic of the Master, and it is considered to be the world’s oldest, recorded, living historical tree. The cutting of trees was discouraged unlessfor a genuine purpose. The rituals to be performed before cutting a tree (for example, lighting an oil lamp under the selected tree for three nights to respectfully inform the associated deitythat the intention is for a good purpose,and to request the deityto leave the tree) also illustrate this point (de Silva, N., 1996:6). The consideration of ancient SriLankans towards treesis also verywell reflected in identifying man-made reservoirs, villages, and so on, in association with an individual tree or a grove of trees. The concept of a mountain cult of socio-religious life of pre-Buddhist Sri Lanka has also led to a respect for mountains, by assigning them some degree of numinous power (Basnayake, 1983:1). The association of pre-Buddhist deity Sumana Saman with the mountain peak Sri Pada is an example (Paranavitana, 1958). Special attention to mountainsis reflected in identifying rocky mountains with the shape of animals, such asAtu-gala (in the shape of an elephant), Ibba-gala (tortoise), Anda-gala (eel), Vandura-gala (monkey), Sigiriya (lion?), and so on. Since the natural environment was not harsh,but conducive to outdoor living,nature was not treated as an enemy of human existence or a rivalto be conquered or subordinated, but as a guardian.This attitude alsogave rise to a specific living pattern centeredon outdoor living. Therefore, most family activities of ancient Sri Lankans, such as cooking, eating, chatting and even entertaining outsiders, took place in outdoor spaces,while indoor spaceswere used, for example, for protection from rain and the privacy of female members. Hencetheindoor space of a house was very muchsmaller than the outdoor space.22 1 Also see De Silva, N. (1990:2) and Lewcock (1998:9-18).

Apart from providing food, the trees gave shade for outdoor living under tropical climatic conditions. This attitude towards nature also promoted the least human intervention in the naturallandscape and the incorporation of natural featuresinto the built environment without significant alteration. Moreover, popular beliefs in the worship of nature were fully appreciated in the art of building construction, both secular and religious, as reflected by the contents of various ancient Silpa texts (Manjusri Vastuvidyasastra 1995:4). § 2.1.3 The Cosmology The ancient Sri Lankan view of the physical nature of the universe is best reflected in Buddhist cosmology and cosmography. Although adapted from Hinduism, the Buddhist cosmology has no creation myth. However, some aspects of Hinduism were absorbed into Buddhism, and such Hindu gods as Brahma and Indra (Sakra of the Buddhist tradition) were incorporated into Buddhism, not as world creators, but as devout followers of the Buddha (Adikaram, 1946:145) who were pious men during previous lives. According to Buddhist cosmology, world systems are destroyed and recreated inkalpas (cycles or great ages). The world system that exists during a kalpa has several planes of existence in the vertical sequence divided into several worlds: underworlds (including various hells, world of departed beings, demons, animal kingdom, and so on), world of humans, and heavenly abodes of the gods (sagga-loka) on top of the MountMeru, the cosmic mountain. All these planes are subdivisions of the world of sensation (laukika), and higher than theseplanes is the world of form (rupa-loka), the abodes of the Brahmas who have a material body. Above these planes is the fourfold arupa-loka, or the world of no-form, the abodes of the Brahmaswho do not possess a material body (Adikaram, 1946:153-154). Mount Meru, the mythical cosmic mountain which lies at the center of the universe is thought to be the axis of mundi joining the earth with abodesof the Brahmas at the highestplanes. The heavenly abode of god Sakra is on the summit of MountMeru. This cosmic mountain sits upon three peaks, the Trikuta, while a thick forest of silk cotton trees covers the mountainslopes. The palacesof the four guardians of the world adorn the slopes of the Meru. Below the Meru on Trikuta is the underworld. Mount Meru is surrounded by seven annular seas, which are in turn separated from each other by several mountain ranges. Beyond the last of these ranges lies an ocean containing four continents, one each at the cardinal directions (Duncan, 1990:42-48). Therefore, this cosmography gives rise to a bottom-up vertical sequence of cosmic landscape with increasing divine habitation above the layers of the human world. The belief system of the pre-Buddhist conceptof mountain cult, in whichmountain tops were the realmsof gods who reside abovehumans, would have also reinforced this cosmography.

This cosmic landscape was a great inspiration to artists, sculptors and architects throughout the historical period.23 It is well reflected in the interior arrangement of the relic chambers of Buddhist stupas. These relic chambers, embeddedat the center of the masonry work of the totally inaccessible hemispherical dome of the stupa, functioned as repositories of the relicsof the Buddha, which signifies the presence of the Master withinthe stupa. A square pillarof stone, whichforms the central object of the relic chamber, represents the mythical Mount Meru. The small chambers below the Meru stone containthe objects connected with the world of Nagas (serpents) representing the underworld. The stone pillar sits upon three smaller stones representing the Trikuta. Placed on top of the pillar is the casket containing the relics of the Buddha,symbolizing the Buddha seated on the abode of god Sakra. The series of seven horizontal bands carved on the faces of the stone pillar represents the seven ranges of the mountains that surround Mount Meru (figure 2.2). In some cases, the northern, southern, eastern and western sides of the pillar are painted yellow, blue, white and red, respectively, which tallies with the accepted conventions about Mount Meru, where the northern, southern, eastern and western sides of the cosmic mountain are believed to be gold, blue sapphire, silver and coral, respectively (Paranavitana, 1946:20-24; Silva 2004:21-22; Seneviratne, 1991:366-376). The internal walls of the chamber in some instancesare painted with figures of gods of heavenly realms moving among clouds. Seneviratne (1991:366) concludes that the relic chamber therefore represents the middle world containing the earth and atmospheric region. The spatial divisions within the interiorof the painted Buddha imagehouses also symbolically re-createthe Buddhist cosmography representing the whole universe, where the lowermost painted register is divided vertically into panels depicting thevarious hells and underworlds, while ceiling paintings depict higher cosmicrealms (Bandaranayake, 1986a:22-23). Byciting several studies,Duncan (1990:48-49) remarks,

23 Also see Aryawansa (1958:61-70) for a description of the mythicalHimalayan region.

‘… the myth of Mount Meru became a paradigm for the spatial organization of state, capital and temple in much of Southeast Asia. Terrestrial space was structured in theimage of celestialspace. Many royal cities were exquisitely built to represent the cosmos in miniaturized forms, with the central part of the city representing the celestial city of the gods, high upon the cosmic mountain. These cities were built as a square or rectangleand fixed at the cardinaldirections. The squareform of the city was actually conceptualized as lying within a mandala, a circular cosmic diagram fixed at the fourcardinal directions and anchored by a fifth point in its center.By paralleling the sacred

shape of the mandala, these cities were transformed into microcosms of the cosmos. The king, by situating his palace at the centerof this mandala, occupied the center of the universe and the summit of the Mount Meru, and hence maintained the liminal statusof a god on earth. By occupying this position at the center of the cosmos, he became a cakravarti who could controlthe world throughthe magical power of parallelism.’

Figure 2.2

Relic chamber of a stupa at Mihintale on display at the site museum showing the mythical cosmic mountain (MountMeru) sitting upon three smallervertical stone pillars(partly covered) representing Trikuta. Also note the seriesof seven horizontal bands carved on the faces of Mount Meru representing the seven ranges of mountainsthat surround the mythical cosmic mountain and the paintings on the internalwalls of the chamber.

Angkor Thom in Cambodia (12th-13th century AC) is identified as the clearest example that set out to re-create the cosmic landscape (Kastof, 1991:172-173). Duncan (1990:53-58) argues that such principal politicalcenters in Sri Lanka as Anuradhapura (4th century BC-10th century AC), Polonnaruva (11th-13th century AC) and Kandy (17th-19th century AC) also have varying degrees of cosmic modeling and speculates that Sigiriyais perhaps the clearest exampleof a cosmic city in early Sri Lanka.

Another aspect of cosmology is the cosmic ocean of milk, with water as the creative agent. Accordingly, the high up in heaven Ahas Ganga, the starry river of milk (the Milky Way), falls into Lake Anotatta on top of Mount Meru. At each cardinal direction,this cosmic lake, the most sacred of lakes in Buddhist literature, has an outlet in the shape of the mouth of an animal.The northern, southern,eastern and westernoutlets are shaped like the mouths of a lion, bull, elephant and horse, respectively. The four streamsthat emerge from these outletsthen flow around the lake three times,and

at the eastern end of the lake become the Ahas Ganga, which drops into the ocean to the south. It is amrita, the fluid of creation, which circulates through the universe.

The water containing amrita possesses the magical propertyof cleansing humansand making them fit for divine association (Duncan 1990:45).

The 18th century, painted, Buddhist cave shrine at Kottimbulvala in Ratnapura District offers one of the best representations where the overalltheme of paintingsis inspired by Buddhist cosmological ideas. The major part of the rock ceiling of the central cave depicts the mythical Himalayan landscapecentralized around the

cosmic lake Anotatta.The variant inhabitants of the mythicalregion and its physical features such as lakes, streams, rocks, trees and plants are represented in the liveliest manner. The paintings on the wooden ceiling of the verandah are completewith hells, the human world consisting of different continents and major rivers, heaven with presiding divinities and many mythical Himalayan scenes (Chutiwongs, Prematilleke and Silva 1990c:41).24 The elaborate landscape composition featuring the cosmic lake is also well illustrated in the 18th century wall paintings of the Buddhist cave shrine at Dambulla (figure 2.3). Here, the waters issuing from the mouths of the directional animals are seen flowing out in four directions.25 The one to the east flows over a mountain, enters Lake Anotatta and thereafter divides into five great rivers before finally enteringthe ocean (Bandaranayake, 1986a:160, 178-179; Chutiwongs, Prematilleke and Silva, 1990b:78-79).

24 Commenting on the Buddhist cosmology reflectedin the paintings at Kottimbulvala, Chutiwongs, Prematilleke and Silva (1990c) declarethat “such a remarkable monochromic scheme is rarelyfound in the paintings of Sri Lanka.”

25 The conventional directions of the bull and horsehave been interchanged in the painting, perhaps due to a misconception of the artist (Chutiwongs, Prematilleke and Silva, 1990b:78-79).

Figure 2.3

Cosmic lake of Anotatta, a painting from Dambulla cave shrine, 18th century AC (photo: M. W. E. flarunaratne)

Being an agro-based society using irrigatedagriculture, this cosmic pattern represents the hydrologic cycle: the ocean, lakes, rivers and rainfallthat enable the earth’s fertility.Therefore, it reflects the ecological characteristic that is so essential for agrarian communities, where water is the essential commodity. Duncan (1990:53-54) argues that there is a link between the cosmic waters and the practical irrigation projects of the ancient Sri Lankan kings in modelingcosmic cities.

Duncan (1990:38) identifies two major discourses within the larger discursive field pertaining to kingship in ancient Sri Lanka, the Asokan and Sakra. The Asokan model, which came with Buddhismduring the MauryanBuddhist missionary activities of India during the 3rd centuryBC, was based on the Mauryan EmperorAsoka, who was looked upon as an ideal Buddhist king. According to this model, the king should be pious, righteous and devoted to the fostering of Buddhism and to the welfare of the people (Duncan, 1990:5). This favored immensely the creation of a landscape dominated by religious structures and public works. As seen at Anuradhapura, during the time of the laying out of Sigiriya, the kings constructed monumental religious structures, such

as stupas,26 that dominate the landscape. These stupas not only express the growing confidence and stability of the nation, but a determination to place a mark on the natural landscape (figure 2.4). Mahathupa or Ruvanvalisaya (2nd century BC, original height: approximately 90 meters), Abhayagiri (1st century BC/1st century AC, original height: approximately 106 meters) and Jetavana (3rd-4th century AC, original height: approximately 120 meters), the three mega stupas at Anuradhapura, are not only the largest monuments of their kind in the entire Buddhist tradition, but are still the tallest brick structures in the world, being surpassedin height only by the two stone-built great pyramids of Gizeh, Egypt. In the same way, kings devoted themselves to constructing massive man-made reservoirs at Anuradhapura, such as Abhayavapi (modern Basavakkulam, 4th century BC, 100 hectares in extent), Tissavapi (modern Tissavava, 3rd century BC, 160 hectares) and Nuwaravava (or City Tank, circa2nd century AC, 1,200 hectares), as public works to irrigate land to sustain agriculture. These towering stupas with large sheets of water (man-made reservoirs) in the foreground, symbolize the material and spiritual heights reached by ancient Sri Lankans. The landscape model of Asokan discourse on kingship was, therefore, simultaneously religious and utilitarian.

The Sakran model, on the other hand, is based on Hinduism, where God Sakra (Indra of the Hindu pantheon) is considered the king of gods. With regard to this model, Duncan (1990:40) comments,

‘In the Sakran, as in the Asokan discourse on kingship, the king was also expectedto be just, pious, caringand attentive to the needs of the citizens. However,the former view stressed the glorious and divine qualityof kingship. The king was seen a cakravarti,

a universal monarch who rules over his people and other kings just as the king of the gods, Sakra, rulesover the thirty-two gods in the Tavatimsa heaven.The Sakran modelof kingship stressedthe building of palaces, citiesand lakes that glorify the god-king. Theselandscapes were modeled upon textual descriptions of the cities of the gods in heaven on the top of Mount Meru.’

26 The stupa is an important and essential ritual edifice of Sri LankanBuddhist worship. Originally, it was meantto enshrine the relics of the Buddha,but later becamea representation of Him. The striking and dominant

element of the stupa is its hemispherical dome, which generallystands on an elevated square terrace. The dome is surmounted by a cubical structure and a cone. Constructed out of solid masonry, these elements of the stupa are arranged in a single vertical axis. The volume of each element gradually decreases as the eye travels upwards along this axis. The relics are enshrined withinsmall chambers of the solidmasonry of the hemispherical

dome. The contrasting combination of square and circularplan forms and of hemispherical, cubical and conicalvolumes gives a dramatic form to this edifice.

Although Buddhism incorporated God Sakra as a devout follower of the Buddha, the Sakran model, which promotedthe glorification of the king through buildingheavenly abodes, became oppositional to the Asokan one, and hence Buddhism during this period did not sanction the devotion of great expense to sacral kingship. However, Duncan (1990:53) points out that some aspects of the landscape model based on Sakran discourse were used by kings duringthis period as reflected in the erection

of storied monasticmansions, such as Lohapasada (chapterhouse), in honor of the religion ratherthan as palaces for the glorification of the king himself.

Figure 2.4

Mahathupa (Ruvanvalisaya), a colossal stupa at Anuradhapura, rising from the bank of an artificial reservoir

2.1.5 City Planning and Architecture Due to close cultural contacts with mainland India, the ancient Indian standard texts and treatises on city planning and architecture, such as Kautiliya Arthasastra, Manasara, Milindapanha and Mayamata, were known to the master builders of Sri Lanka from a very early date (Silva, 2000:50-51). The introduction of Buddhism to Sri Lanka from India during the 3rd century BC would have also provided an opportunity to associate with counterpart professionals in Indiaand to adapt prototypes of Buddhist ritualstructures that were developed in India.

According to Pali literature and other Buddhist texts, the cities in India during Buddhist times (3rd centuryBC) were usuallysquare and defendedwith walls on all four sides. There were four gates, one in the middle of each wall facing the four quarters, and four main streets led from these gates to the center of the city. With

regard to the fortifications, there were three successive moats: the ‘water moat,’ ‘mud moat’ and ‘dry moat’ (Geiger, 1960:53, 58). These descriptions also agree with the Kautiliya Arthasastra, accordingto which three moats, each narrower than the other, must surrounda royal residence with the gates located at the four cardinaldirections (Kangle, 1965:244). These treatises prescribe vastupurusa-mandala (figure 2.5), the gridded centralized diagram, as the basis of site planning(Mayamata, 1985:15-22). Manasara(1946:124-125) introduces 32 variants of vastupurusa-mandala that take the square or rectangular plan.

As per this system, which prescribes the organization of the citadel, the royal palaceis to be located at the center,and several concentric precincts fortified by walls and

entered through gateways facing cardinal directions are to be assigned for various other functions in a decreasing order of hierarchy(figure 2.6).27 Manjusri Vastuvidyasastra28 (1995) also prescribes the rules governing the selection of sites, locationof buildings according to a mandala concept, orientation, rules governing measurements,

method of selectingtrees, and so on, for the laying out of Buddhist monasteries and construction of buildings.29

By the time of the laying out of Sigiriya in the 5th century AC, the royal city of Anuradhapura had an organically evolved and well-developed city plan with the citadel as its nucleus, which is surrounded by a series of concentric rings of monastic and suburban settlements (Silva, 2000:51-62). Although the citadel was not a true square,30 the city wall and the moat beyond with gates facing cardinal directions, the streets running in north-south and east-west directions, dividing the citadel into four quarters and connecting the entrances and extending beyond, are in accordance with the city planningprinciples prescribed in the Indiantexts. The royal palace including the household was located withinthe walled citadel.

27 Also see Mayamata(1985:38-44) for a description of the towns.

28 Manjusri Vastuvidyasastra, a unique manuscript of a Silpa text dealingwith Buddhist architecture, is attributed to the 14th-15thcentury period, but indicates the continuity of earlier compilation of technical texts (1995:16).

29 Also see De Silva, N. (1988, 1989).

30 De Silva, N. (1997:7)suggests that “the city of Anuradhapura is in the form of ‘Murudange’, the drum shapewhich is considered as a non-Aryancity form.”

Figure 2.5

Vastupurusa-mandala, the griddedcentralized diagram (courtesy: Mayamata 1985)

Figure 2.6

Bird’s-eye view of an ideal royal palaceby P. fl. Acharya as per the descriptions given in Manasara(1946)

Also located withinthe citadel was the templeof the scared Tooth Relic of the Buddha and alms halls for the Buddhist monks. The five great Buddhist monasteries formed the inner ring outside the city walls. Established and patronized by royalty, as part of their program of fostering Buddhism, and developed from the 3rd century BC up to about the 5th centuryAC, each monastery had an extensive layout. The largestmonastery is

about 200 hectares in extent, which is larger than the citadel itself (about 160 hectares). Located at the center of each monastery was a colossal stupa as a distinct architectural focal point. The highest reaching 120 meters, these stupas dominated not only the respective monastery, but the entire skyline of Anuradhapura. The artificial reservoirs, including the agricultural settlements located beyond these monasteries, formed

the second ring of the city plan.The existing pre-Buddhist reservoirs of this ring were further developed by constructing additional reservoirs or by enlarging the existing ones by successive kings, again as part of their programto construct publicworks for the welfare of the people.In addition to serving a utilitarian purpose,these reservoirs were dominant elements in the cityscape, which also gave enchanting scenic beauty to the city. Therefore, Anuradhapura’s city plan and its morphology clearlyreflect the Asokan discourse on kingship as well as the socialhierarchy of the period. The location of Buddhist monasteries immediately around the city also portrays the influential role played by Buddhist monks as the advisors to the king in state affairs31 (figure 2.7).

Although the chroniclesmention that numerous secular buildingsexisted from a very early date in Sri Lanka, the state of preservation of such buildings at present does not provide information to understand the character of secular architecture at the time of the laying out of Sigiriya. On the other hand, the religious architecture of ancient Sri Lanka shows that it is essentially a mixture of deeply rooted local buildingtradition and the naturaldevelopment of the forms introduced from mainland India

during the Mauryan Buddhist missionary activity of the 3rd century BC (Paranavitana, 1967; Bandaranayake, 1974;Silva, 1988). Withina few centuries of the Christian era, architecture reached a very mature stage at Anuradhapura.The architectural designs of such Buddhist ritualstructures as stupas,vatadages and Bodhi-tree shrines show that although distinctive individuality had been acquired during the process, no profound modifications had taken place to obscure their origin. In India itself, the Buddhist architecture in vogue in the period during which the Buddhist religionwas introduced

31 The final phase of the city planning of Anuradhapura occurredwhen a certain section of Buddhist monks of the Great Monasteries of the inner ring preferred to be away from the busy urban environment and to lead a life based on meditation. The result was the establishment of several monasteries that were located beyond the ring of reservoirs and agricultural settlements. Therefore, these monasteries formed the outer ring of the city plan adjacentto the forests that were beyond. Thesemonasteries seem to have reliedon patronage from the

middle ring of the agricultural community rather than on royaltyand the urban elite in the city center. This phase, beginning in around the 5th centuryAC, culminated in around the 10th century(Silva, 2000:62, 2008:231-232).

had, by about the 5th century AC, been submerged by extraneous influences and developments within. The elaboration of detail at the expense of architectural form gradually becamethe rule in mainland India,both among Brahmanical sects as well as Buddhists. Sri Lanka, on the otherhand, preferred the ethical simplicity and monastic puritanism of the Theravada, the orthodox Buddhist church, which ceased to be of any influence in medieval India, to the mythological exuberance and metaphysical subtletyof Mahayana Buddhism and later Brahmanism. The keynote of religious architecture throughout the Anuradhapura period (which includesthe laying out of Sigiriyain the 5th century AC) was, therefore, its simplicity, harmonious proportions and dependence on the overall built form rather than on ornamentation and decoration to create an effect (Paranavitana, 1955:77).

Figure 2.7

City plan of Anuradhapura

1 Citadel

2 Abhayagiri Stupa and Monastery

3 Mahathupa

(Ruwanvalisaya) and Monastery

4 Jetavana Stupa and Monastery

5 Mirisaveti Stupa and Monastery

6 Dakkhina Stupa and Monastery

7 Ranmasu Uyana

8 Isurumuniya

9 Vessagiriya

10 Toluvila (Pancavasa Monastery)

11 Pacinatissa (Pancavasa Monastery)

12 Puliyankulama (Pancavasa Monastery)

13 Pankuliya (Pancavasa Monastery)

14 Vijayarama (Pancavasa Monastery)

15 Kiribath Vehera (PancavasaMonastery)

16 Halmillakulama (Reservoir)

17 Bulankulama (Reservoir)

18 Basavakkulama (Reservoir)

19 Tisavava (Reservoir)

20 Nuvara-vava (Reservoir)

21 Puliyankulama (Reservoir)

22 Malvatu Oya (Stream)

23 Western Monasteries (Tapovana)2.1.6

Irrigation Engineering

Being agriculturists in a system based on paddy cultivation, the early settlements were concentrated on riverbanks of the dry zone. Although the Indo-Aryan colonizers brought with them a basic knowledge of irrigation, the need to combat the prolonged droughtsof the dry zone demandeda much more organized artificial irrigation system

than before. In the pre-Christian era, village reservoirs of modest size to store water for irrigation became the characteristic feature of the agricultural and cultural landscape. From this basic understanding of water storage, a vast methodology of hydraulic engineering was developed within a short period. This consisted of building up a complexsystem of reservoirs, damming rivers, and constructing canals and anicuts.

Thekings, influenced by the Asokandiscourse on kingship,initiated massive irrigation projects as public works for the sustenance of agriculture, not only within the principal political center of Anuradhapura, but in the whole of Sri Lanka’snorth central plain.

Water was diverted along artificial canals to irrigate paddy fields. King Vasabha of the 1st century AC is regardedas the first significant builderof reservoirs and canals of great dimension. King Mahasena (276-303AC) was the greatest builderof tanks of colossalproportions. The Kalavapi of King Dhatusena (459-477 AC), father of the founder of Sigiriya, carriedwater to Tissavapiin Anuradhapura throughthe

86-kilometer-longartificial canal calledJaya Ganga.32 The country,which lacked inlandnatural lakes, was gradually transformed into a landscape dotted with hundreds of reservoirs of different sizes, which were interconnected by an intricate network of canals of different gradients traversing the gentlyundulating terrain of the dry zone.

The philosophy of water conservation and management of this agro-based society was that ‘every drop of water that falls on the island must not flow into the ocean without serving its people’ (Culavamsa chapter LXVIII:8-10).33

Taking full advantage of the differences in the natural contours of the topography, reservoirs were organized into a linear cascading sequence (figure 2.8) to allow efficient watermanagement (Madduma Bandara,1985).34 The water management for

32 Also see Parker (1909),Brohier (1934-1935) and Fernando, D. N. (1980) for detaileddescriptions of irrigation works.

33 Ellepola (1990:173) comments that although this famous statementwas made by King Parakramabahu in the 12th century,other Sri Lankanmonarchs were probablyless articulate; yet the philosophy of water conservation was traditional from very earlytimes.

34 Artificial reservoirs are linearly connected forming cascades, allowing surplus water from upstream reservoir(s) and return flow from upstreamcommand area(s) to reach the reservoir immediately downstream. This facilitates the reuse of water in the commandarea of the downstream reservoir, and in effectincreases the water available for irrigation.

cultivation was developed with great skill using many expedients. Among these, the well-known cistern (Biso-Kotuwa) played a singular role. A valve-pit locked the waters of the tank, which were released through sluices in an ingenious method. Hydraulic engineering was thus one of the greatest skills acquired by the people. Consequently, there was a considerable agricultural surplus to provide a basis for a stable economy and sustain a large non-agrarian community of monks, artisans, craftsmen, and so on, in addition to the royal court and military personal. This surplus, on the other hand, allowed kings to construct monumental religious structures such as stupas, which dominate the landscape. Water was the most preciousand treasured commodity

in this hydraulic civilization. This fact is vividly reflected in the famous Dhatusena- Kasyapa story as recorded in the chronicle Culavamsa (chapter XXXVIII:103). The response given by former King Dhatusena to his son Kasyapa, when the latter, in captivity and facing death, demanded treasuresof the state, was to show nothingbut the waters of Kalavapi, which he had constructed during his reign as the wealth of his prosperous kingdom. Therefore, at the time of the laying out of Sigiriya by Kasyapa in the 5th century AC, the country possessed a vast knowledge of hydraulic engineering with centuries of practical experience of water management (Ellepola,1990:175).

Figure 2.8

Organization of reservoirs into a linearcascading sequence (courtesy: Abeywickrama, 1991)

2.1.7 Pleasure Activities of Royalty

Despite making indirect reference to water festivals associated with royalty35 and having the physicalremains of pleasuregardens at Anuradhapura, such as RanmasuUyana, chroniclers do not specifically record the nature of the activities associated with pleasure gardens. Since such activities were connected with sensual pleasure,Dissanayake (2003:50-51, 2009:28) suggests that the author of the Mahavamsa deliberately omittedsuch activities as they were against the Buddhist philosophy.

The insignificant location of Ranmasu Uyana in relation to the overall city plan of Anuradhapura (figure 2.7) also indicates that the Asokan discourse on kingship would have also discouraged an emphasis on the pleasureactivities of royaltyas a major item of the officialroyal program of kings.

Although there are no records available in the chronicles on the nature of the pleasure activities of royalty, studiesof the Jataka stories and other Sinhalaliterary works, mostlydating from the 10th century onwards (Gooneratne, 1983; Dissanayake, 2003:231- 242), show that garden sports (uyan-keli) and water sports (diya-keli), which took the form of royal ceremony, were part of Sri Lankan courtly life,36 and the royal pleasure garden was the main theater for such activities. The pleasure gardens were not only meant for the king and his immediate family,but for a multitude including the nobility and harem as well. These studiesfurther indicate that these pleasureactivities were held for a significant duration of the day, and the water sports were held after thegarden sports, suggesting that water sportswere the climaxof the pleasure program

of royalty. The rock-cut reliefs of elephantsdallying among lotuses and squirtingwater over themselves and each other, on either side of a royal bath house at Ranmasu Uyana, and the liveliness of the movementof the elephants carrying lotusesin their trunksand engaging in water sports,as depicted in a landscapecomposition painted on a rock ceiling at Kotiyagala, very well express the mood of water sports of the period.37 Moreover, water sports mainly involved women,and these literaryworks give vivid descriptions of the specificgarments, female make-upincluding the hair dress

35 The Mahavamsa mentions that King Devanampiyatissa (3rd century BC) arranged a water festival for dwellers in the capital before he set forth to the MissakaMountain (chapter XIV:1-2),and Prince Dutugemunu (2nd century BC) made a tank near Kasa Mountain near Anuradhapura and held a water festivalduring his campaignagainst King Elara of Anuradhapura (chapter XXV:50-51), and another with women in the harem for a whole day after his consecration ceremonyat Tisavava (chapterXXVI:6-10).

36 Also see Geiger (1960:62).

37 See Jayasinghe, G. and Rassapana (2003:26) for the reproduced paintings of elephants at Kotiyagala.

and decoration, and the sensual mood associated with water sports. A painted female figure at Kotiyagala,38 in the attitude of decorating herself with flowers, also shows this.

Gooneratne (1983:6) observesthat in most literary works the women participating in water sports are describedas wearing red dresses, while in some it is said that they wore red and blue garments. Very often, the Sinhala poet saw women engaged in water sports as lightningaccompanied by dark rain clouds,symbols associated with rainmaking (Gooneratne, 1983:7). Godakumbure (1970) suggests that water sports themselves could be considered as part of a fertility cult. Since ancient Sri Lankan society mainly depended on irrigated water, Godakumbure attempts to support his argument by considering the connection of the king with rainmaking, the role of the

Rain-God and the relationship betweenwater, fertility and various Sri Lankan Buddhistfestivals and ceremonies. The connection between the pleasure parks themselves and rainmaking is furtherstrengthened by an incident recordedin the Mahavamsa (chapter XI:2-3), when a great cloud gathered over the site selected for the laying of a pleasure park at Anuradhapura by King Mutasiva(307-247 BC) and poured forth non-seasonal rain. Due to this incident the park was named Mahamegavana (the park of the great cloud). Dissanayake (2003:238) also suggests that it was customary for kings as the prosperity makers of the kingdom to engage in different activities associated with pleasuregardens to bring forth rain, and hence the pleasure gardens were considered as symbolsof rainmaking and prosperity.

2.1.8 External Contacts

The island of Sri Lanka occupies the southernmost position of mainland Asia in the Indian Ocean,and latitudewise it is almoston the Equator, where the annual monsooneffects and navigational winds blow for three months of the year in one direction and change direction for another threemonths. It was also the halfway point between

the two great empires of Rome and China. Due to the uniqueness of Sri Lanka’s geographical position, the island became a continuous navigational hub in relation to ancient and medieval trade between the Mediterranean and the China seas along the silk route of the sea39 (figure 2.9). Therefore, besides cultural contact with India due to close proximity to mainland Asia, the island has a long history of international exposure and transoceanic communication with otherparts of the ancient world,

38 See Jayasinghe, G. and Rassapana(2003:25) for the reproduced female figure at Kotiyagala.

39 Also see Huzayyin (1942).

through international trade and commerce. The most valuable merchandise from Sri Lanka were pearls and gemstones, hence the island name of Ratnadipa (Island of Gems). Referring to the 6th century AC vivid account(Topographia Christiana XI)

by the Greek writer Cosmas Indicopleustes, in which he refers to the chief Sri Lankan port at Mantaion the northwestern coast as the ‘greatemporium’, Silva (1988:1-

4) suggests that Sri Lanka was the port-of-return in the region for maritimetrade vessels from Chinese ports in the Far East and from Roman and other Red Sea ports in the West, and an ideal international center for barter and exchange.In the wake of such trade activity, the envoys of King Bhatika Abhaya (19 BC-9 AC) were sent to theRoman Empire (Weerakkody, 1990:158). Silva (2006:3-4) therefore argues that the financialresources for such massive undertakings as the construction of great stupas at Anuradhapura came equally from the income derived from exporting preciousitems and custom duties and other levies on international trade and commerce,and from agriculture.40

Figure 2.9

Ancient maritime trade routes betweenRome and China (based on Huzayyin, 1942)

40 Also see Kiribamune (2000).

It is mentioned in 5th and 6th century foreign records41 that Arab merchants and a Persian community were living in the city of Anuradhapura, which was, no doubt, due to such international trade activity. Many items of trade from the Far East and the West, such as stoneware and porcelain from China, and glazed ceramics and glass vessels from West Asia and Rome, have been unearthed during excavations at the seaport

of Mantai as well as at Anuradhapura (Ratnayake, 1990; Prickett-Fernando, 1990). The archaeological excavations conducted at Sigiriya have also unearthed Roman and Indo-Roman coins as well as earthenware utensilsbelonging to the Sassanian dynasty(222-651 AC) of ancient Persia (Codrington, 1924:32-33; Bandaranayake, 1984:17,

1993a:25-26; Bopearachchi, 1990, 1996:70-71). Graffiti numbers 219,221 and 230 on the MirrorWall of Sigiriyamention silk, whilein verse number399 “Chinese silk”is clearly mentioned (Paranavitana, 1956:134, 135-136, 141, 248).

Therefore, in this context it could be assumed that international maritime trade andcommerce contributed considerably to raise funds to lay out the massive complex at Sigiriya in the 5th centuryAC.

Sigiriya:ABasicIntroduction

Historical Context

Located in the dry zone of Sri Lanka’s north central uplands, Sigiriya is approximately 160 kilometers northeast by road of the present capitalcity of Colombo.The expansion of archaeological activityat Sigiriya in recent times reveals that the first occupant of this site was Mesolithic man (circa 3rd-2ndmillennium BC). Rock shelters, mainlyto the east of the central rock, were used with an occupational sequence starting nearly 5,000 years ago (Adikari, 1994; Karunaratne, P. and Adikari, 1994). However, the first significant interventions to the natural landscape occurred during the early historic period (3rd/2nd centuryBC up to the 1st century AC), with the establishment of a rock-

41 The accounts of a visiting Chinese scholar monk of the 5th century AC (A Record of Buddhistic Kingdoms) and Greek writer CosmasIndicopleustes of the 6th centuryAC (Topographia Christiana XI), as citedby Silva (2000:59)

shelter associated with Buddhist monasticsettlement on the western slopesadjacent to the main rock. At least 30 rock shelters in this area were adaptedas dwellings of the monastery. These earlymonastic cave dwellings are marked by a drip-ledge cut along the brow so as to prevent rainwater flowing into them. Some of the caves contain donor- inscriptions carvedjust below thedrip-ledge. The site became nationally famous due to a significant event of the island’s history, when Kasyapa (477-495), a son of King Dhatusena (459-477) by a wife of unequal birth, and hence not the lawful heir to the throne, assumed the kingship by a palace conspiracy, which ultimately led to the execution of his own father by having him walled in alive. Moggalana, his half-brother, born of the anointed queen and heir apparent to the throne, fled to India. According to the chronicle Culavamsa (chapter XXXIV:1-19), Kasyapa, fleeing the inevitable return of Moggalana, sought refuge in the inaccessible strongholdof Sigiriya (about 60 kilometers southeast of the then capital Anuradhapura). The chronicle further mentions that Kasyapa built a palace on the summit of the gigantic rock of Sigiriya by defending it with walls, and engineered a staircase in the form of a lion, which gave the name of the site Sihagiri(Lion Rock). Most of the remains exposedat the site are attributed to the major constructional phase of the 5th century AC (Kasyapan period), considered as the brightest age of Sigiriya (Bandaranayake 1993b:114). However, the capital was short-lived (only during the reign of Kasyapa), endingabruptly with the return of Moggalana from India and the subsequent death of Kasyapa by slashinghis throat with his own dagger on the battlefront. Moggalana, not interested in his rival brother’s royal center, handedover the site of Sigiriyato Buddhist monks. During the post-Kasyapan period commencing from the late 5th century to the 13th century AC, parts of the Sigiriya complex were converted for Buddhist monastic use by altering or modifying earlier structures to accommodate the needs of the monks. These latter interventions, however, did not affect the overall original physical layout of the Kasyapan period (Bandaranayake 1993a:24). Historical sources continue to record that there were two incidents of further assassination of royalty at Sigiriya, of Samghatissa II and Moggalana III at the beginning and end of the 7th century AC, respectively, indicating that the site did not entirelylose its politicalsignificance even afterits partial conversion for monasticuse (Basnayake, 1983:5-6). The graffiti inscribed at the site by numerousvisitors to Sigiriya, which are dated to a period from the 6th to the 13th or 14th centuries AC, also demonstrates that after the abandonment of the site as one of the major political centers at the end of the 5th century, it was visited purely for its secular and aesthetic value (Bandaranayake 1993b:122). Sigiriya has, therefore, a long-standing record of beinga destination of cultural tourism commencing from the 6th century. After the 13th or 14th century AC, the site was almost abandoned, and in the 16th and 17th centuries it was a distant military outpost of the Kandyan kings. With the final abandonment of the site after the 17th centuryAC, Sigiriya was overrun by the ever-advancing nature. The structures and numerous other creations, once viewed in elegance and splendor, were completely in ruins or buried in debris when Sigiriya came into focus as a site of antiquarian value in about the 1830s. Archaeological activity has been conducted from

1894. Mentioned in the chronicles as a hideout of a patricidal king, who seized the throne from his father,the primary attention of early archaeologists was to explorethe site as a stronghold (ASCAR, 1895:10). Although the evidence of a ‘pleasure garden’ was also noticed during this period (ASCAR, 1899:7), the unreachable rock summit,fortified by a series of moats and ramparts, would have undoubtedly encouraged early archaeologists to proceed on the abovelines. However, it was only from the late 1940s that archaeologists began to interpret the site as an example of built landscape, with the

exposure of a water-associated pleasuregarden to the west of the rock (ASCAR, 1953:15-19; Paranavitana, 1972a:40).

Sigiriya has been inscribed as the 202nd site on UNESCO’s World Heritage List in 1982 under Criteria II, III and IV,42 due to its outstanding universal values.43Physical Characteristics

Since the physical elements of Sigiriya have been described in great detail in several publications (ASCAR, 1894-1900, 1902, 1905, 1949, 1951-1953, 1965;

Paranavitana, 1950, 1956, 1961, 1972a; De Silva R. H., 1976; Bandaranayake, 1984, 1990a, 1993a, 1993b, 2005; Dissanayake, 2003, 2011), what is attempted below is not to reproduce or summarize such information, but to provide an overall picture of the physical characteristics to form a background for the presentstudy.

As it appears today, Sigiriya is centered on a monumental rock, a residual of denudation, which rises abruptly to a height of about 180 meters above the surrounding plain or 360 meters above the mean sea level (MSL), with sheer cliffs on all sides. Like a giant pebble, the rock sits on a natural hill, whose escarpments are dotted with natural boulders of varying size and picturesque appearance. The outer limits of this hilly terrain, which roughly follow the 200-meter contour line, is defined by a massivewall of roughlydressed stone, faced with brick in the west, and

42 The definitions of the criteria whichwere current and in use at the time of the nomination of Sigiriya for inscription on the World HeritageList are as follows:

Criterion II : Have exertedgreat influence, over a span of time or withina cultural area of the world, on developments in architecture, monumental arts, or town planningand landscaping.

Criterion III: Bear a unique or at least exceptional testimony to a civilization which has disappeared. Criterion IV: Be an outstanding example of a type of structure which illustrates a significant stage in history.

43 The Outstanding Universal Value means culturaland/or natural significance which is so exceptional as to transcend national boundaries and to be of common importance for present and future generations of all humanity (UNESCO, 2008:14).

by high earthen ramparts in the northeast and southeast (figures 2.10, 2.11, 2.12). This walled-in hilly terrain is roughly elliptical in plan and about 15 hectares in extent. Extending to the east and west beyond it are two inner rectangular precincts, each fortifiedby an earthen rampart and a moat in succession. The inner eastern rectangular precinctmeasures about 500 meters north to south, and 700 meters east to west, while corresponding measurements of the inner western rectangular precinct are 900 meters by 800 meters. The inner western rectangular precinct is further fortified by abrick-built (middle) rampart beyond the inner moat. An outer earthen rampart, which forms a large and elongated rectangle, encompasses the central rock, the walled-in hilly terrain and two rectangular precincts to the east and west. This rectangle, which measures about one kilometer north to south and three kilometers east to west, thus creates two outer precincts to the east and west beyond the inner eastern and inner western precincts. Just inside the outer rampart on the western limit of the outer western precinctis another wide moat. All the rectangular enclosures are symmetrically laid out on a single east-west axis, which cuts across Sigiriya rock at its center. The principal gatewaysof the complex in the east-west directionare also set on this axis. Extending further southwards from the outer limits of the hilly terrain is the partly

man-made and partlynatural Sigiri Mahavava(great Sigiriya reservoir), which is formedby an earthen bund nearly eight kilometers long. A rocky fortification, presently known as Mapagala (Rock of the Viceroy), borders the Sigiri Mahavava to the west, as well as the outer earthen rampart to the south. This fortification is linked to the hilly terrain at the south throughthe earthen bund of Sigiri Mahavava (figure2.13).

Figure 2.10

Satellite image of Sigiriya and its immediateenvirons (courtesy: Google Earth)

Figure 2.11

Semi-aerial view of Sigiriya from the northwest (photo: Dominic Sansoni/ThreeBlindMen)

Figure 2.12

Plan, Sigiriya

1. Palace on Rock Summit

2. Walled-in Hilly Terrain

3. Inner Western Precinct

4. Outer Western Precinct

5. Inner Eastern Precinct

6. Outer Eastern Precinct

7.Sigiri Mahavava

Figure 2.13

Greater Sigiriya showing Mapagala, Pidurangala, Ramakele and Mahavava

The inner western rectangular precinct (average elevation: 195 meters above MSL) has gateways on the north, south and west across its inner moat and through the rampart (figure 2.14). An axial pathway traverses the whole east-west length of this precinct from the westerngateway to its eastern limit,where the main gateway to the

walled-in hilly terrain is placed. The central zone of this precinct is dominated by water- associated structures, such as reflecting pools and ponds, fountains and serpentine streams, bathing pools and changing rooms, moated island structures and summer palaces, arranged in a symmetrical order along its east-west axis (figures 2.15, 2.16,2.17). These water bodies are fed by a network of surface and subsurface hydraulics, which operates on the principle of gravity and pressure. A rectangular compound associated with a ‘four-quartered’ landscape feature is located transversely across

the east-west axis within this precinct (figure2.18). This landscape feature originally had a central pavilionat the intersection of the main axis and the transverse axis. This landscape feature is reminiscent of the well-known garden form found within the ‘paradise gardens’ of ancient Persia, of which the Sigiriya version is one of the oldest surviving examples(Bandaranayake, 1993a:24-25, 1993b:123).

Figure 2.14

Plan, western precincts

1. Outer Rampart

2. Outer Moat

3. Middle Rampart

4. Inner Moat

5. Inner Rampart

6. ‘Four-quartered’ Feature

7. Miniature Water Garden

8. Summer Palace

9. Western Approach

10.Northern Gate

11.Southern Gate

Figure 2.15

Plan, central zone of the western precincts

1. Four-quartered Feature

2. Outer Compartment

3. Miniature Water Garden

4. Summer Palace

5. Octagonal Pond

6. AxialPathway

Figure 2.16

Central zone of the inner western precinctfrom the rock summit

Figure 2.17

Plan, Miniature Water Garden

Figure 2.18

Plan, four-quartered feature

1. Moated Central Pavilion

2. L-shaped Pond

3. Northern Outer Compartment

4. Southern Outer Compartment

5. Pavilion

Dotted with natural boulders of varying size and picturesque appearance, the hilly terrain is formed into a series of ascending terraces towards the base of the rock (figures 2.19, 2.20). These terraces are linked to each other by means of flights of steps. The cut-marks on nearly all the boulders indicate that they had a structure erected on top while the natural overhanging at their base served as rock-shelters (figures 2.21, 2.22). These rock-shelters were originally utilized by Buddhist monks of the pre-Kasyapan monastic phase. A few constructional works associated with the interiors of the rock-shelters and some retaining walls are also attributed to this period (Bandaranayake, 2005:10-15). The soffits of these rock-shelters below their

drip-ledges were plastered and painted duringthe Kasyapan phase and have close similarities to those of the main rock face. The ceiling painting on the rock-shelter popularly known as the Cobra-hood Cave is of a different type showing a geometric pattern (figure 6.1). Winding pathways are laid out through the arches, courtyards and alleys formed by these boulders (figures2.23, 2.24). Two main access routes to the rock summit in the form of ascending flights of steps traverse the western and southwestern escarpments to convergeat a brick-built gatehouse on the southwestbase of the main rock (about 250 meters above MSL).

1. Western Entrance to Hilly Terrain

2. Western Route

3. Southwestern Route

4. Pathway associated with Mirror Wall

5. Lion Plateau

6. Main Rock

Figure 2.19

Plan, western escarpment

Figure 2.20

Semi-aerial view of the westernescarpment (photo: R. Swathe Inc.; courtesy: Bandaranayake, 1993b)

Figure 2.21

Boulder showing the cut-marks on the top to receivethe brick masonryof a structure

Figure 2.22

A rock-shelter at the base of a boulder

Figure 2.23

Boulder arch No. 01

Figure 2.24

Boulder arch No. 02

An elevated pathway, punctuated by a short flight of steps and set along a collar-like declivity of the westernrock face, at an averageheight of 15 meters abovethe base of the rock,runs northwards fromthis gatehouse. Thispathway is protected by a highbrick-built parapet wall on its western edge, and by the natural overhang of the rock on top (figure 2.25).Popularly known as the MirrorWall, due to its highlypolished plaster on the internalsurface, this parapet wall contains poems and other graffiti inscribed by ancient visitors

to the site from the 6th to the 13th or 14th centuries AC, who recorded their emotional expressions of the site (Paranavitana, 1956; Priyanka, 1990, 1994, 2010). The drip-ledge cut at the brow along the entire length of the western andpart of thenorthern rock faces prevented rainwater flowing along the surface intothis pathway. Thisprotected pathway winds around the northwest corner of the rock and continues alongthe northern face

as rather a steep stairwaybuilt on firm ground, to lead on to an elevated and elongated plateau.This plateau (about300 meters above MSL), which protrudes northwards from the northernrock face, measuresabout 64 metersnorth to southand 32 meters east to west (figures 2.26, 2.27). Built on this plateau against the northern vertical face of the rock and facing north is a staircase-house in the form of the forepart of a colossallion. At present, only the massive forepaws of the so-called Lion-Staircase-House and an internal passage with flights of steps leading up towards the rock summit have survived (figure 2.28). The final ascent to the summit of the rock from this so-called Lion Plateau is, therefore, firstlythrough this Lion-Staircase-House and then along a protectedzigzag

stairway above it, built on the northern vertical face of the rock, of which only the crevices cut on the rock surface to support the masonry work of the stairway have survived.

Figure 2.25

The Mirror Wall and the pathway

Figure 2.26 Plan, Lion Plateau 1. Lion-Staircase- House 2. Pathway/Stairway associated with Mirror Wall 3. Main Rock 4. U-shaped Structure

Figure 2.27

Semi-aerial view of the Lion Plateauwhich protrudes from the northernrock face (photo:Dominic Sansoni/ ThreeBlindMen)

Figure 2.28

Remains of the Lion-Staircase-House (note the crevicescut on the upper part of the rock surfaceto support the masonry work of the zigzag stairwayabove the Lion-Staircase-House to reach the rock summit)(photo: Dominic Sansoni/ThreeBlindMen)

The area belowthe horizontal drip-ledge cut high up along the length of the westernand northern rockfaces was originally plastered and paintedwith the famousSigiriya Ladies (Sigiri landun), only 19 of which have survived on adjacent depressions of the western rock face above the protected pathway44 (figure 2.29). This painted band was originally about 140 meters long and, at the widest, about 40 meters high (Bandaranayake, 1986a:26). According to some graffiti scrawled on the Mirror Wall, there had probably been approximately 500 such figure paintings originally. This would have been, according to John Still (1907:43), a British civil servant of the early 20th century, perhaps the largest picture in the world.In this painting composition, only the upper partof the body of the ladies, cut off about 20 centimeters below the waist by clouds, are shown floating among the clouds.45 Similar paintings of the ladies are also found on one of the rock-shelters

44 Three more depressions, higher up on the western rock face, also contain plaster and pigments and fragments of a paintedfigure.

45 The terracotta figurines unearthed in excavations at the western escarpment of the hilly terrain display close similarities in general form and conceptto the ladies painted on the rock face. Bandaranayake (1993c) dates theseexquisite figurines to the periodbetween the 7th and 10th centuries, and believes that they may have been replicasof the painted ladies, used as souvenirs to be carriedaway by visitorsto Sigiriya.

situated on the hilly terrain, the only difference being that theyare not cut off belowthe waist by clouds, and hence are full figurerepresentations with legs bent in a flyingposture. A paintedlady and pigments discovered on the plastered outer surface of the Mirror Wall in2004 (Wagaachchi, 2005)indicates that thewhole of theouter surface of the MirrorWall was also paintedin a similar manner to that of the rockface. However, the surviving paintedfigure indicates thatit is also a fullfigure representation verymuch similar to the paintedfigures of the rock-shelters of the hillyterrain.

Figure 2.29

Remnants of the plasterband on the rock face showing the Sigiriya Ladiesffoating among clouds(photo: Gamini Jayasinghe; courtesy: Bandaranayake 1986a)

Roughly elliptical in plan and about 1.5 hectares in extent, the summit of the rock is occupied by palace buildings, servicestructures, rock-cut and brick-built pools and water retainingstructures as well as gardensand terraces (figures2.30, 2.31,

2.32). Another interesting feature is an east-facing rock-cutseat which was originally providedwith a canopy (figure 2.33). Running at the middle and entire north-south length of the rock summit is a limestone-paved pathwaywhich links variousspaces on either side (figure 2.34).The entire rock summit was originally surrounded by a

parapet wall rising from the slopes below the edge, on almost the total periphery of the rock (figure 2.35). Originally concealed within the parapet wall surrounding the rock summit, a north-south running drain of considerable cross-section cut on the western cliff was used to collect the rainwater run-off of the western sector of the rock summit. A vertical drain cut on the southwestface of the rock carriedthe water down to a cistern built at the base of the rock.

Figure 2.30

Plan, rock summit

1. Main Palace

2. Other Palace Structures

3. Upper Palace Garden

4. Lower Palace Garden

5. Pool

6. Rock-cut Seat

7. Limestone-paved Central Pathway

Figure 2.31

Rock summit (viewfrom southeast) (photo:Dominic Sansoni/ThreeBlindMen)