ANCIENT INDIA Buddhist, Hindu, Jain by Susan L. Huntington with contributions by John C. Huntington Contents Preface xi Acknowledgments and Credits xiii A Note on Pronunciation and Transliteration of Sanskrit xix Introduction xxiii Geographical Considerations, xxiii * The Problem of Dating, xxv * Art and Culture, xxvi Part One, Foundations of Indie Civilization: The Prehistoric and Protohistoric Periods 1 Antecedents of Indie Civilization 3 Stone Age Painting and Sculpture, 3 • Early Neolithic Art, j * Conclusion, 8 2 The Indus (or Harappa) Civilization (ca, 2300 to 1750 B.C.) 9 The Cities and Towns, to - Sculpture, 12 ■ Seals, 18 * Pottery, 24 • The Eclipse of the Indus Civilization, 25 3 The Vedic and Upanisadic Periods (ca. 1500 to 450 b.c.) 26 The Indo-Aryans, 26 • Literary Evidence: The Vedas (ca. 1500 to 800 b.c.), 27 • Literary Evidence: The Upanisads (ca. 800 to 450 b.c.), 28 * Other Literary Evidence: The Puranas and Epics, 30 - Mahavira, Sakyamuni Buddha, and the Rise of Magadha, 31 ■ Archaeological Evidence, 33 ■ Other Indigenous Trends: Megalithic Remains of Southern India, 34 * Conclusion, 36 Part Two. Period of the Early Dynasties 4 The Maurya Period (ca. 323 to 185 b.c.) 41 Edicts and Pillars, 43 ■ Rock-cut Architecture, 48 * Other Asokan Monuments, 30 * vii Viii CONTENTS Maury a-period Sculpture from Pataliputra, 31 • Maurya-period Terra-cotta Sculpture, 54 * Conclusion, 55 5 The Sunga Period and Related Developments (ca, Second Century to First Century b.c.) 5^ Vidisa in the Sunga Period, 57 * The Mathura Region During the Sunga Period, 60 * -Buddhist Art of the Sunga Period: Free-standing Architectural Monuments, 61 • Rock- cut Architecture of the Sunga Period: The Western Deccan, 74 • The Eastern Deccan: The Andhra Pradesh Region, 83 * Suhga-period Terra Cottas, 88 - Conclusion, 89 6 Regional Developments (ca. Late First Century B.c. Through First Century a.d.) 90 The Vidisa Region: Sand, 91 ■ Buddhist Rock-cut Architecture of the Western Deccan: Bedsa, 100 * Eastern India: Khandagiri/Udayagiri, 103 • Conclusion, 108 7 The Saka and Parthian Kingdoms in the Indie Sphere (ca. First Century B.c. to Mid-First Centurya .d.) 109 Introduction to the Bactro-Gandhara Region, 109 - The Kapisa Region, no • The Gandhara Region, 116 * The Swat Valley (Ancient Uddiyana), 119 • Northern India (The Mathura Region), 122 • Conclusion, 123 8 The Northwest and Northern Regions Under the Kusanas (ca. Late First Century to Third Century a.d,) 123 Royal Shrines, 126 ■ The Bactro-Gandhara Region: Architecture, 130 • The Bactro- Gandhara Region: Sculpture, 133 • Northern India: Mathura and Related Sites, 150 - Non-Buddhist Sculpture at Mathura, 139 ■ Conclusion, 162 9 Regional Developments in the Deccan (ca. Second and Third , Centuries) 163 The Western Deccan Caves, 163 - The Eastern Deccan: The Andhra Pradesh Region Under the Later Satavahanas and Iksvakus, 174 ■ Conclusion, 183- Part Three. Dynasties of the Middle Period ,10^The Gupta Period {Fourth to Sixth Centuries) - 187 Hindu Art of the Early Gupta Period, 188 ■ Buddhist Art of the Fifth Century: North- Central India (Sand), 196 • Buddhist Art of the Fifth Century: North India (Mathura and Sarnath), 200 • Buddhist Art in the Northwest, 203 * Hindu Temple Architec­ ture, 206 * Brick Temples and Terra-cotta Art, 213 * Metal Images, 218 • Conclusion, • 1 r The Gupta Aftermath 220 The Disintegration of the Empire, 220 ■ Outgrowths of the Gupta Idioms (ca. 550 to CONTENTS ix 700), 222 • Eastern India, 223 • Buddhist Art in the East, 223 • Hindu Art in the East, 227 • Western Indian Developments, 229 * Conclusion, 236 12 Buddhist Cave Architecture (Fifth Through Seventh Centuries) 23$ Ajanti, 23g * Bagh, 260 • Kanheri, 262 * Aurangabad, 263 • Ellora, 268 • Conclusion, 274 13 Hindu Rock-cut Architecture of the Deccan (Kalacuri and Early Western Calukya Phases) 273 The Early Kalacuri Period, 273 * Caves of the Early Western Calukyas, 282 • Conclusion, zgo 14 Southern Developments Under the Pallavas and the Pandyas 291 The Pallavas, 291 • The Early Pandyas, 319 * Conclusion, 321 15 The Early Western Calukya and Related Schools of the Deccan 322 Structural Temples of the Early Western Calukyas, 322 • The Eastern Calukyas, 337 * The Nolambas, 338 • Conclusion, 340 16 Hindu Rock-cut Architecture of the Western Deccan 341 Ellora (Rastrakuta Phase), 341 * Conclusion, 330 Part Four, Later Northern Schools 17 Kasmir and Related Schools 333 Pre-Karkota Remains, 334 * The Karkota Period (ca. 625 to 855), 337 • The Utpala Dynasty (ca. 855 to 939), 363 • The Two Lohara Dynasties and the Last Hindu Kings (Tenth to Fourteenth Centuries), 368 * Kasmiri Ivories and Metal Images, 368 • The Art of Adjacent Regions: Western Himalayan Foothills and Western Tibetan Cultural Region, 374 * Conclusion, 385 vr8 Bihar and Bengal Under the Pala and Sena Kings 387 Buddhist Art, 389 * Hindu Art, 407 * Conclusion, 413 19 Orissa and Related Regions 413 Srlpura (Sirpur) and Rajim: Madhya Pradesh, 413 * Hindu Art and Architecture of Orissa, 421 ■ Buddhist Art of Orissa, 444 * Conclusion, 448 20 North-Central and Northwestern India: The Art of the Rajput Clans 449 The Gurjara-Pratiharas of Kanauj (ca, 730 to 1027), 432 * The Haihayas (or Kalacuris) of Tripun, 462 * The Candellas of Bundelkhand, 466 ■ The Paramaras of Malwa, 480 • X CONTENTS The Solankis of Gujarat (ca. 950 to 1304 [961-1244?]), 483 • The Gahadvalas of Varanasi (ca. 1075 to 1200), 499 * Manuscript Painting* 300 • Conclusion, 302 Part Five, Later Schools of the Deccan and the South 21 The Cola and Related Schools of the Tamil South (Mid-Ninth to Thirteenth Centuries) 509 Conclusion, 538 22 Later Deccan Schools 340 The Western Gangas of Talakad (Tenth Century), 341 • The Later Calukyas of Kalyani {973 to ca. 1189), 343 * The Yadavas of Devagiri (1191 to 1311), 348 • The Kakatlyas of Hanamkonda and Warangal (ca, Mid-Eleventh Century to ca, 1325), 349 * TheHoysalas of Southern Karnataka (ca. 1006 to 1346), 333 • Conclusion, 572 23 The Vijayanagar Period (ca. 1336 to 1565) 373 Conclusion, 586 24 The Nayak Period 5#7 Conclusion, 600 25 The Kerala Region 601 Conclusion, 613 Afterword Notes 619 Select Bibliography 639 Glossary ^13 List of Maps 733 Illustration Index ^^ Index j4y Color Plates appear following pages 162} 402, and 394. A Note on Pronunciation and Transliteration of Sanskrit For the sake of simplicity and consistency, wherever appropriate, deity names, religious concepts, and other technical terms have been given in the Sanskrit (Samskrta) language in this book. A Sanskrit transliteration and pro­ nunciation guide is given, below. Transliteration tor all other Sanskritic languages (Hindi, Panjabi, etc.) follows the same system; Dravidian lan­ guages (Tamil, Malayalam, etc.), are trans­ literated according to this system whenever possible. Pronunciation of these languages varies considerably and no attempt has been made to provide a pronunciation guide for them. Other languages (Persian, Tibetan, etc.), are transliterated according to standard systems. Chinese words have been given according to the Wade-Giles system in the text, but the pinyin romanization is provided in parenthesis following the word in the index. Originally described in a systematized manner more than 2,500 years ago by ancient Indie grammarians, including the well known Panini Detail of 18.13* (ca. fourth century B.c.), Sanskrit is the first language known to have been analyzed linguis­ tically and to have been given a method of some figures of the Tan trie tradition in both writing strictly based on a scientific description Buddhism and Hinduism (see Fig. 18.13, a of actual pronunciation. Careful description of detail of which appears above). Although the pronunciation was desirable, for, if prayers were pronunciation system was intended to provide to be effective, the pronunciation of deity names a uniformity to the spoken language, it never and ritual incantations, including btjas and really extended beyond the true Sanskritic mantras, had to he correct. From a religious and speakers and was probably not wholly consistent artistic point of view, it is interesting to note even among them. that the sounds of the Sanskrit system are enu­ The Sanskrit grammarians recognized that merated as the “fifty sounds” which comprise there were five places of articulation of sound, the “garland of skulls” (kapalamala) worn by and classified the sounds used in the language, XlX XX A NOTE ON SANSKRIT and thereby the characters which represent bral), and dental. Today, Sanskrit is most these sounds, accordingly. The Sanskrit system commonly written in the Devanagari script and thus groups sounds into the following categories: the characters are arranged and transliterated in guttural, palatal, labial, lingual (retroflex/cere- this order: Table i a a i X u w “true” vowels r f 1 } e ai 0 au “mixed vowels” mjm nasal “modifier” (anusvara) h aspirate “modifier” (visargd) ka kha Sa gha Jia guttural consonants ca cha P jha ha palatal consonants ta tha da dha na lingual consonants ta tha da dha na dental consonants pa pha ha bha ma labial consonants ya ra la va semivowels sa sa sa spirants ha aspirate [*] [aspirate “modifier” (upadhmaniya)] A fuller classification of the sounds of the language is usually given thus: Table 2 t T3 ^ “Cr £ s > $ 'k <? £ & £ .s g & g *P £H&r £ & & P#SP * $ §* &■£ ■So' O-J7 Gutturals k kh g gh h h bl a a e afi Palatals c ch j Jh h y s i l Linguals t th 4 dh T s r f Dentals t th d dh « l s l 1 Labials p ph b bh m V m u u 0 tfW4 Note: The anusvara (m) is not traditionally included in this classification system. 1. Visarga 3. “E” and “at” are guttural and palatal. 2. Upadhmaniya 4. “O” and “aw” are labial and guttural. A NOTE ON SANSKRIT XXI The characters for the consonants, semivowels, as they are in English. However, kh} gh, th} spirants, and aspirate h are pronounced as and ph are given a noticeable aspiration, in syllables containing the inherent vowel sound contrast to English, Th is not pronounced as in of a, and, as in the first table above, are properly the English word think but rather as in goaf- transliterated with an a following them. The hcrd, Ph is not as in pfeone but is pronounced inherent vowel is transcribed following these as in stop-him. C is pronounced almost exactly characters in the transliteration of whole words as is the ch in church, while Sanskrit ch is given unless otherwise indicated. an even stronger aspiration. In early translitera­ The vowels a, i, i, u, u} e, ai, o, and att, are tions, this led to ch and chh being used to repre­ pronounced as in either German or Italian. The sent the two letters. However, the desire for short a is a neutral vowel and is pronounced as parallelism in transliteration has superseded the the u in but. Thus, sattva is pronounced “suttvu.” attempt to suggest pronunciation through En­ In traditional Sanskrit phonetics, both e and o glish spelling. Accordingly, candra is pronounced are considered diphthongs, and, regardless of the “chundru” and Sanskrit words with ch, such as form written, are pronounced as long vowels. chattra, are given even more aspiration. There is evidence that the vowel r was at one The distinction between the retroflexes and time pronounced as the er in butter, but at least dentals is extremely difficult for the untrained as early as the beginning of the Christian era it ear to hear, and, for all practical purposes, may came to be pronounced as uri” (as in river). be overlooked in pronouncing words. The Thus, Krsna is pronounced “Krishna.” The retroflex t and d resemble t and d in English, vowel l was originally pronounced as the le in while the dental t and d are essentially similar bottle but is now pronounced as “lri” (as in to the t and d in Italian. Technically, in the axl^r/m). It is rarely met with and the long pronunciation of the retroflex, the tongue vowel, | is actually a grammarian’s artifact, touches the gums just above the teeth, while invented for symmetry, and does not exist in for the dental, as its name implies, the tongue the language. squarely hits the back of the teeth. The modifiers include the anusvara (m or m) The nasals of each category, n, n, nn, and and the visarga (h). Rather than representing m, offer no special problems. N is pronounced independent sounds, these symbols indicate some as ng in son£ and n is pronounced as ny in carcyon. modification in the realization of the character Two spirants, s and s, were originally distinct with which they are written. The anusvara sounds, but merged in later times. The written indicates a nasalization of a preceding vowel so symbols for them are sometimes used inter­ that, for example, samara is pronounced “sun- changeably in inscriptions. Both are pronounced saru.” The anusvara is also often used, though sh. Thus, Sesa is pronounced “Shesha.” incorrectly, to replace any of the five nasals Due, in large part, to extensive publication in (h, n, n, m) when they are followed by English on the Indie religions, many South consonants of their own class. The visarga (h), Asian words have made their way into English- occurring at the end of a word or syllable, is a language dictionaries. For example, abhiseka noticeable exhalation of breath, frequently occurs as “abhiseka” and, although the pro­ followed with a slight echo of the preceding nunciation guide suggests how it be spoken vowel, much like a very soft grace note at the correctly, the likelihood is that the English close of a musical phrase. Another aspirate, also reader, seeing the word in context, will pro­ transcribed as h, is known as the upadhmanlya; nounce it without the sh sound; a more readable since it is a variant of the visarga> it is often English spelling might have been “abhisheka.” omitted from the list of characters and has been Because of problems of tills sort, it was decided, placed in brackets in these tables, (A third variant, for the most part, to ignore the English-language the jihvamuliya, is not usually included in charts dictionaries with regard to South Asian words of the script.) in this book. Thus, technical terms in Sanskrit Most of the consonants are pronounced much and other South Asian languages will be treated XXii A NOTE ON SANSKRIT as foreign words. As there are many more such that pattern (hence, dharmacakra has been pre­ words in this volume than occur in the English- ferred to dharma cakra). When the first of two language dictionaries, this also allows for Sanskrit words ends in an a or a and the second consistency of usage. “Karuna and upaya as in the series begins with a or a, we have joined components of bodhi/’ is jarringly inconsistent the words to deter pronunciation of the final and, “karuna and upaya as components of and initial vowels as distinct sounds. Thus, we bodhi ” has been preferred. Moreover, this have given padmasana instead of padma asarta, allows an emphasis on the technical words that but dharmacakra mudra remains split. Ill the are a necessary part of the study of South Asian case of place names and other geographic terms, art and permits usage of these words according diacritics have been used whenever possible, to their meanings in the Indie context, and not with the exception of Pradesh (not Pradesa), as popularized or limited by English practices. which appears in several modem state names. The notable exceptions have been the use of When authoritative spellings could not be found, Buddha (not btiddha) and bodhisattva (not ho- the terms have been given without diacritics. dhisattva). Normally, we have chosen to split No diacritics were used on the map names, nor compound Sanskrit terms into separate words. was it possible to standardize the spellings of However, when the joined version is more map words which reflect the many languages of commonly seen in English, we have followed South Asia, according to a single system.