Copyright by Matthew David Milligan 2010 The Thesis Committee for Matthew David Milligan Certifies that this is the approved version of the following thesis: A Study of Inscribed Reliefs within the Context of Donative Inscriptions at Sanchi APPROVED BY SUPERVISING COMMITTEE: Oliver Freiberger Janice Leoshko Supervisor: A Study of Inscribed Reliefs within the Context of Donative Inscriptions at Sanchi by Matthew David Milligan, B.A. Thesis Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of Master of Arts The University of Texas at Austin August 2010 Dedication In memory of Dr. Selva J. Raj, and for all of my teachers, past, present, and future. v Acknowledgements I’d like to begin by thanking my two co-supervisors, Dr. Oliver Freiberger and Dr. Janice Leoshko. Their comments, insights, and--most of all, patience--have given me the opportunity to learn much during this process. I am grateful to other professors at the University of Texas at Austin who have taught me much these past few years, including the arduous task of teaching me Sanskrit. For this, I am indebted to Dr. Edeltraud Harzer, Dr. Patrick Olivelle, and Dr. Joel Brereton. I extend much appreciation to those in India who have helped me research, travel, and learn Prakrit. First, I thank Dr. Narayan Vyas (Retd. Superintending Archaeologist, ASI Bhopal), for helping to arrange my research opportunity at Sanchi in 2009. Also at Sanchi, S.K. Varma--who may yet prove to be the incarnation of Emperor Aśoka-- assisted a great deal, as well as P.L. Meena, and the kind staff of the ASI Archaeological Colony and the Gateway Retreat. From a previous visit to Sanchi, I must thank Subodh Kumar of Bodh Gaya who willingly accompanied me on my first trip. I am grateful to Rajashree Mohadikar and Vimal Bafana for their language assistance while I studied Prakrit on the AIIS Pune program. Elsewhere, I thank Dr. Lars Fogelin for kindly nudging me in the right directions years ago. More recently, I thank Dr. Julia Shaw who helped me gain access to Sanchi. At home, I could not have done this without Amanda Boundy--who has read much more of this thesis than she probably would have liked--and all of my family and close friends. Finally, I must say thank you again to everybody for illuminating my experience. 8/13/2010 vi Abstract A Study of Inscribed Reliefs within the Context of Donative Inscriptions at Sanchi Matthew David Milligan, M.A. The University of Texas at Austin, 2010 Supervisor: Oliver Freiberger Inscribed relief art at the early Buddhist archaeological site of Sanchi in India exhibits at least one interesting quality not found elsewhere at the site. Sanchi is well known for its narrative reliefs and reliquaries enshrined in stūpas. However, two inscribed images of stūpas found on the southern gateway record the gifts of two prominent individuals. The first is a junior monk whose teacher holds a high position in the local order. The second is the son of the foreman of the artisans of a king. Both inscribed stūpa images represent a departure from a previous donative epigraphical habit. Instead of inscribing their names on image-less architectural pieces, these two particular individuals inscribed their names on representations of stūpas, a symbol with a multiplicity of meanings. In this thesis, I use two perspectives to analyze the visual and verbal texts of these inscribed reliefs. In the end, I suggest that these donations were recorded as part of the visual field intentionally, showing the importance of not only inscribing a name on an auspicious symbol but also the importance of inscribing a name for the purpose of being seen. vii Table of Contents List of Tables ......................................................................................................... ix List of Images ..........................................................................................................x List of Figures ........................................................................................................ xi Chapter 1 Introduction ............................................................................................1 1.1 The Sanchi Landscape and Geography ...................................................3 1.2 Sanchi's Archaeological History .............................................................6 1.3 Viewing Stūpa Images ...........................................................................10 1.4 Towards a Grammar of Representation ................................................16 Chapter 2 Reading the Southern Gateway ............................................................18 2.1 Structure of a Gateway ...........................................................................19 2.2 Location of the Stūpa Images ...............................................................27 2.3 Inscriptions in the Reliefs .....................................................................35 2.4 Conclusion .............................................................................................39 Chapter 3 Reading the Inscriptions at Sanchi .......................................................41 3.1 The Anatomy of a Donative Inscription ................................................42 3.2 Architecture and Donation .....................................................................49 3.3 The Placement of a Donative Inscription...............................................57 3.4 Gifting Power and Monastic Buddhists .................................................64 3.5 Conclusion .............................................................................................68 Chapter 4 The Stūpas on the Gateways ................................................................70 4.1 The Features of the Stūpas .....................................................................71 4.2 Types of Stūpas ......................................................................................78 4.3 Conclusion .............................................................................................89 Chapter 5 Conclusion ............................................................................................92 viii Appendix 1: All Thirty Stūpas and their Features Discussed in Chapter 4 ..........96 Appendix 2: Stūpas on the Four Gateways ...........................................................97 Bibliography ........................................................................................................127 Vita .....................................................................................................................131 ix List of Tables Table 1.1: Construction Periods of the Sanchi Hill..............................................8 Table 2.1: Location of the Stūpa Images ...........................................................28 Table 2.2: Location of the Stūpa Images according to Gateway .......................29 Table 2.3: Architectural Composition of Mānushi Buddha Scenes ...................31 Table 3.1: Palaeographic Groupings of the Inscriptions at Sanchi ....................41 Table 3.2: Donor Communities on the Ground Balustrade of Stūpa no. 1 ........47 Table 3.3: Number of Donations Compared to Number of Donors ...................47 Table 3.4: Types of Architectural Pieces with Donor Frequencies ...................49 Table 3.5: Architectural Pieces Compared to Volume and Inscribed Number ..49 Table 3.6: Inside Facing Architectural Pieces....................................................59 Table 3.7: Outside Facing Architectural Peices .................................................60 Table 3.8: Local vs. Non-Local Inscriptions Facing Inside ...............................64 Table 3.9: Local vs. Non-Local Inscriptions Facing Outside ............................64 Table 3.10: Major Non-Local Towns Compared to Donated Pieces ...................64 Table 3.11: Monastic Titles and Donors ..............................................................65 Table 3.12: Types of Monastic Donors................................................................69 Table 4.1: The Location of the 30 Stūpa Images ...............................................83 Table 4.2: Mānushi Themed Stūpa Images ........................................................84 Table 4.3: Parinirvāṇa Themed Stūpa Images ..................................................85 Table 4.4: Historical Themed Stūpa Images ......................................................86 Table 4.5: Unidentifiable Stūpa Images .............................................................88 x List of Images Image 1.1: Sanchi Stūpa no. 1 and Gateway from the North................................1 Image 1.2: Inscribed Stūpa Image and a Non-Inscribed Image on the South Gateway ..............................................................................................2 Image 1.3: Parinirvāṇa Themed Stūpa from the North Gateway .......................11 Image 1.4: An Unidentifiable Stūpa Image from the East Gateway ...................16 Image 2.1: Location of the South Gateway's Narratives .....................................20 Image 2.2: Three Architraves on the Outside Face from the South ....................22 Image 2.3: Location of the West Gateway's Narratives ......................................25 Image 2.4: Location of the North Gateway's Narratives .....................................26 Image 2.5: Location of the East Gateway's Narratives .......................................27 Image 2.6: Worship of the Headdress with Ivory-Workers Inscription ..............36 Image 2.7: Aśoka's Visit to the Rāmagrāma Stūpa with Balamitra's Inscription ........................................................................................37 Image 2.8: Central Mānushi Buddha Stūpa with Ānanda's Inscription ..............37 Image 3.1: Five Crossbars, Three Railpillars, and One Copingstone from Stūpa no. 1 ........................................................................................50 Image 4.1: Top Architrave with Mānushi Buddha Theme .................................79 Image 4.2: Parinirvāṇa Themed Panel ...............................................................80 Image 4.3: Bottom Architrave Depicting a Legend Scene .................................81 Image 4.4: An Unidentifiable Scene from the East Gateway .............................82 xi List of Figures Figure 4.1: Sections of a Stūpa ............................................................................72 Figure 4.2: Aṇḍa Measurements of Stūpa Images ...............................................76   Chapter 1 – Introduction A popular form of early Indian Buddhist ritual focused on one important type of monument: a stūpa (Image 1.1). Typically, stūpas are hemispherical reliquary mounds meant to either enshrine relics for ritual engagement, or commemorate the Buddha or the Buddha’s parinirvāṇa. This thesis investigates depictions of stūpas carved on the upper part of gateways at Sanchi1 in central India. These represented stūpas share placement along the 1st century C.E. gateways of stūpa no. 1 at Sanchi. They also share many common iconographic features, designs, and physical locations along the crossbars of the gateways. However, two of these representations display a unique characteristic of Buddhist art in the region: they are inscribed with donors' names. By studying these inscribed stūpa representations, I hope to demonstrate their utility in finding a grammar to read early Buddhist stūpa imagery and early Buddhist donation. Image 1.1: Sanchi Stūpa no. 1 and Gateway from the North   1 Throughout this thesis, I do not use diacritics for modern placenames, like Sanchi, to keep with modern conventions. In the original language, Sanchi, of course, was not Sanchi, but Kākanāva during the time studied by this thesis. For further discussion, see P. H. L. Eggermont, "Sanchi-Kakanada and the Hellenistic and Buddhist Sources," in Deyadharma: Studies in memory of Dr. D. C. Sircar, ed. G. Bhattacharya (Delhi: Sri Satguru Publications, 1986): pp. 11-27.   Image 1.2: Inscribed Stūpa Image and a Non-Inscribed Image on the South Gateway From these inscribed reliefs, I ask the following questions: 1.) who are the donors? What is their relationship to the Buddhist community at Sanchi? 2.) What are these two stūpas? Are they artistically unique in any way? 3.) Is there any connection between the stūpas’ artistic scenes and their donors? 4.) How do these two stūpas and their inscriptions fit into the Buddhist gateway art at Sanchi? And lastly, 5.) What is their significance in the development of artistic expressions of Buddhism during the ancient period? In short, this thesis attempts to add a new perspective to the study of ancient Buddhist material remains through the close examination of two unique artistic and epigraphic examples.   At the heart of this thesis is a grammar of representation. I explore some relationships among three elements: 1.) stūpa depictions--the visual texts; 2.) the inscriptions--the verbal texts; 3.) and the surrounding built landscape. Using one theory from archaeology2, my approach views material culture as a text to be read using a carefully designed grammar, built word by word, sentence by sentence, from the ground up. By studying the placement and content of the donative inscriptions and depictions of stūpas, I hope to be able to uncover something of their meaning when the two types of texts occur together. 1.1 The Sanchi Landscape and Geography Designated a UNESCO World Heritage Site of India in 1989, Sanchi resides outside the heartland of Buddhist history. Located nearly 22 hundred kilometers from Sarnath in Uttar Pradesh, and some 2,800 kilometers from Bodh Gaya in Bihar, Sanchi is an unlikely location much of early Indian Buddhist history. Nevertheless, Sanchi gained prominence early in the extant historical record. Preserved as a historic landmark because of its lengthy history, the site provides a large cache of artistic and inscriptional material to investigate. Emperor Aśoka's (c. 273-236 B.C.E.) patronage may be one possible reason for the site's early growth and construction of monumental structures. According to legend, before he became emperor, he accepted the position of Viceroy of the Mauryan Empire. His headquarters was at Vedisha in central India. Vedisha served as a focal point for the Mauryan Empire because it was a large urban center positioned along a known trade   2 See C Tilley, Material Culture and Text (London: Routledge, 1991).   route. Northern Black Polished Ware associated with the city’s ancient rampart shows the city’s earliest urban occupation occurred around the time of the Mauryas. As viceroy in Vedisha, Aśoka met his wife3 and remembered the area as auspicious and critical to the imperial agenda.4 Later, as a convert to Buddhism, he famously opened seven of the eight original stūpas erected over the bodily relics of Śākyamuni Buddha. He distributed the relics and built 84,000 stūpas across his empire.5 One such stūpa might be no. 1 on the Sanchi hilltop, as an Aśokan pillar, near its south gateway, bears an inscription warning expulsion to dissident monks. According to local legend, to honor his beloved wife, and presumably, to provide seclusion for Buddhist monks, Aśoka founded Sanchi.6 Besides Aśoka's history with the region, Sanchi also sat between the large urban, trade centers Vedisha and Ujjain.7 The overflow of wealth passing between Vedisha and Ujjain undoubtedly gave the religious community a great advantage in seeking donations, evidenced by the numerous donations recorded in stone at Sanchi. The establishment of a religious center on Sanchi’s hilltop may have been as much economically motivated as it was spiritually.8   3 See the Mahāvaṃsa 13.6-11 for a historical reference to Aśoka meeting Devi in Vedisagiri. For the English, see W Geiger, Mahavamsa: Great Chronicle of Ceylon (New Delhi: Asian Educational Services, [1912] 2003): pp. 88-9. 4 D Mitra, Sanchi (New Delhi: Archaeological Survey of India, 2001): p. 5. 5 See J Strong, The Legend of King Aśoka, (Princeton: Princeton University Press, 1983): p. 219 for a translation of the relevant portion from the Aśokāvadāna. 6 It is certainly possible that any number of religious communities, including Buddhists, already occupied the Sanchi hilltop and/or region. 7 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 2. 8 However, Sanchi was undoubtedly not without competition, as at least one pre-existing tradition already discovered the fortunes of residing between Vedisha and Ujjain. Sometime soon before most stone monuments were built at Sanchi, the Heliodorus pillar in Vedisha was raised. This freestanding monolithic pillar records the erection of a garuḍa-dhvaja, or “Garuḍa emblem,” by Heliodorus the Greek from   Additionally, Sanchi’s fertile landscape warranted the construction of several water tanks and dams.9 The local agricultural community may have relied on the water stored at Sanchi to grow crops and maintain their lifestyles through a mutually symbiotic relationship with the Buddhist monastic community.10 Dams and tanks dating to the last centuries B.C.E. were key features in the relationship between the monastic Buddhists on the hilltops and the farmers below. Irrigation canals were built for distribution. 11 Put simply, the monks could have provided religious services and water as the laity provided donations, food, and labor.12 Therefore, a mutually dependent relationship naturally formed between the monastic Buddhists on top of the hill and the laity below. Sanchi was not the only major religious center in the region. Located in a radius of about 15 kilometers around the Sanchi hilltop are innumerable Buddhist and non- Buddhist sites. Cunningham discovered four large Buddhist sites before the 1854 publication of The Bhilsa Topes.13 These sites are now known as Satdhara, Sonari, Andher, and Morel Khurd (previously Bhojpur). Each large subsidiary site resembles the      Takṣaśilā. Heliodorus was a bhagavata, or Brahmanical devotee, sent by the mahārāja Antialkidas. This early Brahmanical inscription clearly shows that Vedisha was already associated with the Vasudeva, the devadevasa, or “god of gods.” The Heliodorus pillar is clear evidence of Vedisha’s non-Buddhist importance before, or at the same time as, the widespread creation of stone Buddhist monuments on the Sanchi hilltop. On paleographic grounds, the Heliodorus pillar is assigned an approximate date of c. 150 B.C.E.. 9 J Shaw, Buddhist Landscapes in Central India (London: The British Academy, 2007): p. 233. She says that those from the ancient period are quite distinguishable from more recent village tanks. 10 Ibid., pp. 252-3. Shaw discusses “service villages” (aramīkagāma) in the Cūlavaṃsa (v. 46.115). They provided labor to monasteries and met the nutritional needs of its inhabitants. 11 Ibid., pp. 239-40. Regarding control structures, Shaw says: “The dams are usually pierced by a stream channel at their deepest point…the natural drainage point for the dam catchment…masonry remains, attesting to some kind of monumentalized control structure, have been found in the feeder streams of the four highest dam sites [of Sanchi, Devrajpur, Morel kala, and Ferozpur].” 12 See L Fogelin Archaeology of Buddhism (Oxford: Altamira Press, 2006). In his study of Thotlakonda monastery, some of the local population was employed by the monastic community to perform a number of services. 13 A Cunningham, The Bhilsa Topes (London: Smith, 1854).   Sanchi hilltop: one centralized, major stūpa with smaller stūpas and temples in proximity.14 1.2 Sanchi’s Archaeological History According to J.A.S. Burgess15, General Taylor of the Bengal Cavalry was the first British officer to record a visit to Sanchi. In 1818, during a campaign against the Pindharas, he noticed that three large gateways were standing and that the southern gateway had fallen. The dome of stūpa no. 1 was largely untouched and even had many portions of the balustrade in-situ. Stūpa no. 2 was also undisturbed. The dome of no. 3 was in good standing condition; however, its lone gateway had fallen. Taylor saw eight other stūpas but he did not record their condition. Burgess suspects that Taylor believed the monuments were undisturbed for many years. Mr. Herbert Maddock, Political Agent at Bhopal, obtained permission from the government in 1822 to “dig” into the two large stūpas. Seeking treasure, Maddock and a Captain Johnson, the Agent’s assistant, dug into stūpa no. 1 “from the top to what he believed to be the bottom of the foundation.” 16 They claimed to not find any open spaces. Stūpa no. 2 was “also half destroyed by the same bungling amateur antiquaries…they also probably also completed the ruin of the other minor monuments previously unnoticed by the few visitors.” Later, after these amateur blunders, a number   14 This basic pattern is deceiving. Stūpa no. 2 is located partially down the side of the hill. As further archaeology has shown, there were indeed other monuments—specifically stūpas—built on the sides of the hill. 15 J.A.S. Burgess, "The great stupa at Sanchi-Kanakheda" in Journal of the Royal Asiatic Society of Great Britain and Ireland (January, 1902): pp. 29-45. 16 Ibid., p. 34.   of serious observers recorded numerous plates and sent them to James Prinsep, coin- assayer for the East India Company, for analysis.17 In 1849, the Government of India ordered Lieutenant FC Maisey to Sanchi. He prepared an illustrated account of the stūpas, sculptures, and known inscriptions. In 1850, Maisey met Alexander Cunningham, Major General in the British army and then- amateur archaeologist. He corroborated with Cunningham and visited for the first time in 1851. During his seven-week stint with Maisey, Cunningham began repairs on stūpa no. 3, which was wrecked in 1822. In their repairs, they found stone boxes, inscribed with “ma” and “sa,” referencing the famous Maudgalyāyana and Śāriputra from Buddhist literature. They sunk a shaft into stūpa no. 2 and found an inscribed stone box enclosing four steatite inscribed caskets with the names of famous early Buddhist saints and teachers from the area.18 The two also sunk a shaft into stūpa no. 1 but, as their predecessors had discovered, nothing was there.19 Three years later, Cunningham published The Bhilsa Topes.20 His book was the first useful description of the Sanchi region, but is of limited use in terms of its theories. Cunningham worked with Georg Bühler on re-translating the known inscriptions. Between 1881 and 1912 H.H. Cole and others undertook minor restoration and clearing of vegetation. Nevertheless, in 1912, John Marshall began the largest excavation and   17 Brian H. Hodgson in 1824 sent two to Prinsep. Dr. Spilsbury sent him a drawing of a gateway sculpture in 1835. In 1837, Captain E. Smith copied and sent Prinsept twenty-five inscriptions and Captain W. Murray sent more drawings, specifically of the lower architrave of the south gateway. 18 See M Willis, Buddhist Reliquaries from Ancient India (London: British Museum Press, 2001) and M Willis, "Buddhist Saints in Ancient Vedisa," in Journal of the Royal Asiatic Society 11.2 (2001): pp. 219- 228. 19 Given the sketchy nature of the early endeavors to recover relics and the shadowy nature of their so- called reports, we cannot leave out the suggestion that Maisey and Cunningham’s predecessors found remains, removed them, and sold them for a profit. 20 A Cunningham, The Bhilsa Topes (London: Smith, 1854).   restoration project at Sanchi. Despite the many blunders of the early visitors, Marshall's work was quite successful for its time. Marshall published a three-volume set, The Monuments of Sanchi that remains the most comprehensive and authoritative work on the Sanchi main site. Marshall developed a six-phase sequence beginning the third century B.C.E. and continuing until the twelve century C.E., shown in Table 1.1. Phase # Approximate Years Period Phase 1 300 – 200 B.C.E. Mauryan Phase 221 200 B.C.E. – 100 C.E. Post-Mauryan / Śuṃgan Phase 3 100 C.E. – 300 C.E. Satavahana/Kshatrapa Phase 4 500 – 600 C.E. Gupta Phases 5+ 600 C.E. – 1200 C.E. Post-Gupta Table 1.1: Construction Periods of the Sanchi Hill In the same volumes, epigraphist N.G. Majumdar wrote a chapter on all the known inscriptions from Sanchi and its aforementioned subsidiary sites.22 To date, Majumdar’s study of the inscriptions serves as the basis for nearly all scholarly works concerning Sanchi’s epigraphy.23 Marshall’s volume one contains a description of the monuments, up to when he was writing. Alfred Foucher discusses and interprets Sanchi’s sculptures. Majumdar’s   21 Phase 2 is the Early Historic Period in which this thesis works. The ground balustrade of stūpa no. 1 dates to the middle of Phase 2, while the four gateways are slightly later. 22 Included in his list are the Aśokan pillar, reliquary inscriptions, donative inscriptions from the balustrades, and Gupta-period land grants. 23 See V Dehejia “Collective and Popular Basis of early Buddhist Patronage” in B. Miller (ed.) The Powers of Art (Delhi: Oxford University Press, 1992), U Singh, "Sanchi: The history of the patronage of an ancient Buddhist establishment", in Indian Economic and Social History Review 33.1 (1996): pp. 1-35, and M Willis, Buddhist Reliquaries from Ancient India (London: British Museum Press, 2001) as examples.   chapter on inscriptions is at the end of volume one. He labeled the inscriptions according to their location. In volumes two and three Marshall published numerous plates of all the gateways, balustrades, miscellaneous fragments, and rubbings of the Brāhmī inscriptions. Since Marshall, there have been several serious attempts at excavation and survey. In 1936, Hamid uncovered a large monastery directly west of stūpa no. 1.24 In 1995-6, the Archaeological Survey of India cleared a cluster of small stūpas southwest of stūpa no. 1, outside the designated tourist boundary. A stairway built into Building 8 was recently uncovered.25 S.B. Ota cleared other sections east of stūpa no. 1 and revealed paving stones and other small features. P.K. Mukherjee unearthed a seventh century monastery cluster. Sadly, both Ota and Mukherjee’s excavations have yet to be published and are only available in the ASI's Bhopal office.26 British archaeologist Julia Shaw did the most significant recent work. She began the Sanchi Survey Project (SSP) in 1998. She aimed to “move beyond” the ritual landscape to “an examination of the archaeological landscape as a whole.” She did not see sites in the same geographical region as existing in isolation; rather, they were interconnected insofar as they shared resources, populations, and goals. Her massive survey project stretched from the Sanchi hill proper to sites nearly 25km away. In sum, over 750km^2 were surveyed. She reports that 35 new Buddhist sites, 145 settlements, 17 irrigation works, and over 1,000 sculpture and temple fragments were documented   24 See M Hamid, "Excavation at Sanchi," in Annual Report Archaeology Survey of India, 1936-37 (Delhi: 1940): 85-7. for details. 25 J Shaw, Buddhist Landscapes in Central India (London: The British Academy, 2007) in chapters 9 and 11 believes this previously misunderstood building was used as a viewing platform to see the other hilltop sites in the area. Similar platforms have been found at the other corresponding sites. 26 Ibid., she briefly discusses these recent findings on p. 21. However, they were explicitly shown or discussed when I visited the site in person.    during the two six-month seasons between 1998 and 2000.27 Shaw's work postulates an early Indian Buddhist landscape where monks, nuns, farmers, local patrons, merchants, and others, were economically linked through large and ever-expanding interdependent exchange networks.28 1.3 Viewing Stūpa Images A most unusual feature of early Buddhist material culture in South Asia—and specifically at Sanchi—is the clear absence of an image of the Śākyamuni Buddha until the Common Era. A fundamental question is, of course, “why?” Because this thesis is an investigation of a common symbol of Buddhism, the stūpa, it is worthwhile to understand where this symbol fits in early Buddhist art. Is the stūpa an aniconic representation of the Buddha? Is the stūpa a commemorative mark for the Buddha’s parinirvāṇa?   27 Ibid., p. 20. 28 Following in the footsteps of Julia Shaw and the SSP, there have been several other attempts at surveying Buddhist landscapes in South Asia (See L Fogelin, Beyond the Monastery Walls, Dissertation. University of Michigan (2003), L Fogelin Archaeology of Buddhism (Oxford: Altamira Press, 2006), and J Hawkes, "The sacred and secular contexts of the Buddhist stupa site of Bharhut," in Buddhist Stupas in South Asia, ed. by J Hawkes (New Delhi: Oxford University Press, 2009). The general trend of landscape archaeologists has been to shift the focus from the Buddhist monuments to the Buddhist archaeological landscapes. For Sanchi, this means looking away from the Sanchi hilltop and its carved remains and towards the surrounding region, connected through various exchange networks. By considering the wider archaeological contexts, new questions may be asked to improve the ongoing academic dialogues regarding the ancient Buddhist saṃgha. Viewing Buddhist sites broadly removes their restriction as being sole repositories of monumental architectural, epigraphic, and sculptural evidence. Instead, Hawkes (p. 146-7) says that viewing the wider archaeological contexts expands our understanding of the relationship between Buddhism, the state, and social and economic structures through their mutual involvement in trade, urbanism, and agricultural practices. Image 1.3: Parinirv In 1985, Susan Huntington published survey of Indian religious art. Indian civilization—and by association, Indian art peoples during the Neolithic period. presenting Indian art as a continuous, monolithic, linear progression towards a classic period of Buddhist and Hindu art. yields considerable pre-historical development towards later post Buddhist iconographic representations. She refers to a “continuum of ideas” that was “important in the formulation of both the Indus and Mesopotamian civilizations number of sculptures from the Indus  29 S L Huntington, The Art of Ancient India  āṇa Themed Stūpa from the North G The Art of Ancient India29 The book introduces the reader to an evolutionary view of —beginning with the proto The book is organized into consecutive periods, For Huntington, study of the Indus Valley civilization -Maurya Hindu and “may have had religious associations as suggested (Boston: Weatherhill, 1985). ateway , an accessible -historic .” A   by its popularity in later Buddhism,” 30 referring to the Bearded Man with a shouldered garment (Buddhist robes?) and an interesting Seated Figure who could be in a meditative posture. In 1990, she published an article that further develops her thesis regarding the Buddha image, which she had first proposed in The Art of Ancient India.31 A relief from Bharhut (100 B.C.E.) depicting the Buddha’s descent from Trayastriṃśa heaven at Sankasya demonstrates her primary argument: relief scenes where the Buddha is absent are scenes of worship, or reenactments.32 In this relief, she argues that “the figures appear to move as if in a clockwise procession around the ladders in the nearly ubiquitous circumambulation ritual used in Buddhism” leading her to conclude that “Sankasya had become a major pilgrimage center and that an actual set of stairs—perhaps the very ones depicted in this relief—were the focal point of worship.” Sankasya’s status, in her view, was simultaneously both a sacred spot associated with the Buddha’s life and a sacred location for reenacting, through devotion, the famed event.33   30 Ibid., p. 13. 31 K Karlsson, Face to Face with the Absent Buddha (Uppsala: Uppsala University Press, 1999): p. 48 believes that the intense modern debate surrounding the aniconic phase started with Huntington’s discussion of Buddha’s descent from Trayastriṃśa heaven (Figure 3). S L Huntington, The Art of Ancient India (Boston: Weatherhill, 1985): p. 73 says “this relief and others often given an aniconic interpretation are not aniconic. Rather, it might be suggested that the Buddhological message of many subjects depicted in early Buddhist art was not an emphasis on Śākyamuni Buddha or his life but to other aspects of the religion.” For further discussion in The Art of Ancient India, see pp. 98-100. 32 S L Huntington, “Early Buddhist Art and the Theory of Aniconism,” in Art Journal 49 (1990), p. 404: “However, the theory that I am proposing—that reliefs like [Buddha’s descent from Trayastrimasha heaven at Sankasya] portray a place but not an event in the Buddha’s life—allows another interpretation that perhaps better accounts for the elements depicted in the relief.” 33 Ibid. p. 404.   Vidya Dehejia-- critical of Huntington--34 proposes a less-rigid method of looking at relief images where the Buddha is absent. For Dehejia, the three elements of time, space, and protagonists contribute to a narrative quality of early Buddhist art.35 From these three elements, she derives seven modes to visually narrate a tale from Buddhist literature, each corresponding to a distinct method of storytelling.36 Dehejia uses the modes of visual narration to counter Susan Huntington’s theory of “pageantry.”37 With a discussion of the same Bharhut relief, Dehejia looks at the entire Ajātaśatru pillar, including all three registers, rather than just the Sankasya panel. Each register emphasizes “being in a state” of acknowledging the great wisdom of the Buddha.38 She identifies many emblems in early Buddhist art that “may be read as aniconic   34 See V Dehejia “Aniconism and the Multivalence of Emblems,” Ars Orientalis 21 (1991): pp. 45-66. Her criticism goes as far as to not give Huntington credit for contributing to her own work. 35 R L Brown “Narrative as Icon” in J Schober Sacred Biography in the Buddhist Traditions of South and Southeast Asia (Honolulu: Hawaii University Press, 1997): pp. 64-109 challenges the assumption that all scenes are intended to be narratives, but rather argues that they function as icons (pp. 64-5). He describes Dehejia’s reading of Figure 4 as a “puzzle” and believes Dehejia has ignored the element of space in formulating her elements of narration. He says that the “organization of the three scenes [in Figure 4] has … little to do with the narrative as a story told through time” as, spatially, the descent from preaching in the Trayastriṃśa heaven by the Buddha is actually below the location where the preaching took place (namely the high heaven) (p. 67). He furthers his argument by discussing the difficult physical location of many scenes, as scenes in caves were visible only to those with an oil lit lamp. Even then, one would have to go through the painstaking effort of having the scenes interpreted, perhaps, by a monastic specialist (as Dehejia proposes). Other locations are high into the air (as at open-air stūpa complexes) like the Sanchi gateways. 36 See V Dehejia, “On Modes of Visual Narration in Early Buddhist Art,” in The Art Bulletin, 72.3 (1990): pp. 374-392. Her seven modes are: 1.) easily identifiable monoscenic narratives, 2.) culminating monoscenic narratives, 3.) synoptic narratives, 4.) conflated narratives, 5.) continuous narratives, 6.) linear/sequential narratives, and 6.) narrative networks. For a detailed discussion of each mode specifically, see her lengthier book, V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997): pp. 10-32. In her chapter “On Modes of Visual Narration” she revises her original modes from the 1990 article and adds a semiotic factor to her theory. 37 Dehejia repeatedly uses this term in V Dehejia “Aniconism and the Multivalence of Emblems,” Ars Orientalis 21 (1991): pp. 45-66 to refer to Huntington’s theory of reenactment of famous scenes from the Buddha’s life at sacred pilgrimage sites, such as Sankasya. It is unclear as to whether or not the term is meant to draw an academic parallel to pageantry in Christianity or if it is used as a derogatory term. 38 V Dehejia, “On Modes of Visual Narration in Early Buddhist Art,” in The Art Bulletin, 72.3 (1990): p. 380.   presentations of the Buddha.” In the end, she calls for the need to “recognize, accept, and even admire the multiplicity of meanings apparent in early Buddhist sculpture and painting, in which the artist reminded the viewer of the manifold religious interpretations that may be suggested by any single emblem.” 39 Also critical of Huntington's arguments, Rob Linrothe points out hermeneutical inconsistencies with the argument. He criticizes Huntington's over-reliance on “eyewitness accounts” of the Chinese pilgrims from the mid-to-late centuries C.E. (namely Faxian) to argue that Sankasya had become a major pilgrimage site in the years after the Buddha’s death. He rightly claims it is problematic to rely on texts that post- date the pieces in question by hundreds of years.40 He also examines Huntington's use of the Lotus Sutra (saddharma-pundarika). Despite the textual encouragement of the making of images, the earliest known redaction of the text occurs in the late third century C.E. Though the early stratums of the text date to the first century C.E., this is still not earlier than the extant Indian Buddha images and at least a century earlier than known Chinese Buddha images.41 Furthermore, although the Lotus Sutra may emphasize merit associated with making and venerating Buddha images, its familiarity with the image custom clearly draws on pre-established image traditions. Making images, Linrothe argues, was not a new phenomenon: the “innovation [was] of making images of the   39 Ibid., p. 45. 40 R Linrothe “Inquiries into the Origin of the Buddha Image: A Review,” East and West 43 (1993), pp. 245-6. He quotes Tucci saying, “We cannot say that his information is always exact; in this kind of writings we cannot expect to find everywhere that historical preciseness of details we demand from modern authors.” Faxian and other later Chinese pilgrims, then, should not be “demanded too much of.” 41 Ibid., p. 243.   Buddha” (emphasis his).42 Regarding the theory of reenactments, Linrothe aptly finds it curious that nobody, in Huntington’s pageants, plays the character of the Buddha in these reenactments.43 In this thesis, I accept that stūpas can be aniconic representations of the Buddha. However, representations of stūpas at Sanchi are much more than images within narrative scenes. To expand on the meaning of stūpa images, I use a basic grammar of representation to investigate the variation in scenes where stūpas occur, the different parts of a stūpa image, and the physical context in which the stūpa appears. Essential to this perspective are architectural elements of the gateways, the written text appearing in the two inscribed stūpas, and the similarities and differences between stūpas and stūpa scenes on each of the gateways. Reducing the stūpa image to aniconism disallows a sophisticated reading of complicated imagery. Too many fundamental units comprise a stūpa image to ignore the sum of the parts.   42 Ibid., p. 244. 43 Ibid., p. 249: “why is it, if there was no disinclination to represent the Buddha, that no one plays the part of the Buddha himself [in the pageants]?” Image: 1.4: An U 1.4 Towards a Grammar of Representation In this thesis, I dissect the inscribed reliefs of Sanchi to understand the form, function and meaning of a stūpa image, I aim to evaluate the nature of donation during the time of the inscribed images. Of the nearly 900 balustrade and a mere 11 are relationship between the few donated images and the image's composition facilitates a new way of understanding Sanchi's artistic landscape imagery on the southern gateway  44 In Chapter 4 I will discuss the different categories of represent the commemoration of the Buddha's history or literature.  nidentifiable Stūpa Image from the East G stūpa image. Simultaneously, analyzing these aspects of a donative inscriptions at Sanchi, 349 belong to the gro on the four gateways. Acknowledging the fundamental and the stūpa image is completely different from the donated sculpt stūpa images at length. This particular parinirvāṇa or it may be a specific scene from later Buddhist ateway44 und . The donated ure on stūpa may   the other three gateways, as the only donated images at Sanchi appear on the southern gateway. The inscribed stūpa images also do not appear in the same types of visual fields as the other material with donative inscriptions. The inscribed reliefs are on the architraves of Sanchi's southern gateway. Many other representations of stūpas found at Sanchi are also on architraves--with one main exception, on the north gateway one stūpas image is on the western pillar facing east. I argue that the consistency in which the stūpa images appear in specific locations on the gateways is not random. Rather, the artistic program was made of patterns linking basic visual elements, such as a hierarchy of fields, inside versus outside the balustrade and gateway, and the different components that make an architrave or balustrade. Using this principle of patterned visual placement of important imagery-- stūpas or inscriptions--I study the 30 stūpa images from the gateways and the slightly earlier 349 donative inscriptions appearing on the ground balustrade. To begin my investigation and formulation of a grammar of representation in which to begin contemplating how to approach the large body of data, I take a cue from several different sources. First, I look to Ananda Coomaraswamy and his “Early Indian Architecture” series. According to one recent reviewer, Coomaraswamy tirelessly worked to establish “a taxonomy and morphology of historically documented monuments.”45 His chief goal was to discuss around 130 Indian architectural terms. He favored historical details and had a genuine concern for a material, or archaeological, study of the constructed environment. He defined the fundamental units of buildings   45 P Wagoner, "Ananda K. Coomaraswamy and the Practice of Architectural History," in The Journal of the Society of Architectural Historians, 58.1 (1999): p. 63.    through forms and meanings. Underlying the fundamental units was a formal logic, or “grammar.” This grammar “governed [the combinations of units] into full-fledged architectural statements.”46 Part of the effectiveness of Coomaraswamy’s approach rests in his determination not to rely exclusively on either visual or written texts. Similarly, for this thesis, I will be using two sources: visual and inscriptional. Updating Coomaraswamy’s approach, I adopt archaeologist Christopher Tilley’s approach towards the Stone Age rock paintings at Nämforsen, in northern Sweden. His book, Material Culture and Text47, outlines one attempt to find a grammar of representation. He accuses previous researchers of “reading” without any grammatical system. In this way, without a formal grammar, the sentence “The cat sat on the mat” may be equally read “mat the on sat cat the” or “the mat the sat on cat.” Each sentence means the same thing to Nämforsen's previous researchers. Without a grammar, they miss meaningful combinations of words and distinctions between verbs, adjectives, and nouns. In short, the early researchers possessed no understanding of signifiers, signifieds, or phonemes.48 To locate a grammar that makes sense of the visual sentences, first “we must at least be able to recognize where one word ends, the next begins, and be able to distinguish differences between words.”49 The problem eventually becomes, however, how do I read the text once the sentences and pages have been identified? To read the “book,” we must re-write it by translating the visual imagery into a language we   46 Ibid., p. 64. 47 C Tilley, Material Culture and Text (London: Routledge, 1991). 48 Ibid., pp. 11-12. 49 Ibid., p. 16.    understand. Through this “editing,” the pages begin to become organized into meaningful units. Tilley investigates associations between different types of designs on rock carving surfaces. He asks: “What kinds of combinations occur? What are the permissible limits of design combinations?”50 From here, he is able to begin reading simple combinations and eventually gains a refined interpretation of Nämforsen's rock carvings. Once developed, the grammar of representation in this thesis discusses the form of a stūpa image and several possible interpretations. Like Coomaraswamy in his essays on early Indian architecture, I am attempting to comprehend the underlying formal grammar between fundamental units of a structure and how it governs the units’ combination(s) into full-fledged architectural statements. How do all the components of a stūpa’s anatomy define a stūpa? What combinations and patterns of units are meaningful? Re-reading Sanchi stūpa no. 1 in this manner may provide an insightful look at the intertextuality between stūpa representations, actual stūpas, and the architectural landscape encompassing the stūpa complex.   50 Ibid., p. 30. 18 Chapter 2 -- Reading the Southern Gateway Among the carvings, on the southern gateway, image meets text. In general, the monumental gateways are known for their sculptures, relief narratives, and Buddhist themes. The southern gateway is similar to the other gateways in this way. However, besides the sculptures and relief narratives, there are three inscribed reliefs--two stūpas and one other scene--which are curiously not present in any form on the other gateways. In this chapter, through an analysis of the fundamental units comprising a gateway, I suggest reading the inscribed reliefs together to better understand the unique artistic vision present at the south entrance to Sanchi stūpa no. 1. To begin, I ask what are the general architectural units of a gateway? What type of imagery is present and where? Next, I focus on the depictions of stūpas on the four gateways at Sanchi. I seek to answer the questions: is there a recurring pattern of stūpa representations? In what kinds of scenes do the stūpas appear? Lastly, after describing the three inscribed reliefs from the southern gateway, I question: where exactly do the inscriptions appear? What is significant about the placement of text within an image? In the end, I argue that the inscribed reliefs and their physical placement are revealing of the donors' unique status at Sanchi. Vidya Dehejia has called the southern gateway a “medley of themes that bear no integral relationship one to the other.”1 Although I do not attempt to unite all the fragmented scenes, through studying the gateway’s known inscriptions with the associated imagery, my intention is to conduct an expanded reading. Removing Dehejia's 1 V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997): p. 129. 19 label, I suggest that several parts of the gateway could be and should be read together using inscribed reliefs as a basis. 2.1 Structure of a Gateway According to Mitra2, the building sequence of Sanchi's four gateways was the south gateway first, then the north, then the east and west toraṇas. Two reasons are given as to why the south gateway is the proper entrance into the circumambulatory path. First, the south entrance is where the large, now-fragmented Aśokan column resides. Secondly, the stairs to the upper circumambulatory path are just inside the entrance of the south gateway. These two features not only give the southern entrance functional prominence, but show that the accompanying gateway was perhaps the first constructed to establish the initial construction phase of the circumambulatory path. In sum, a proper circumambulatory gesture may be to enter through the south, walk clockwise around the stūpa until one reaches the south again, entering into the stairway going up into the upper path, and circling the stūpa once more, clockwise, then leaving down the stairs exiting the stūpa ritual area altogether through the western opening. Keeping in mind that a gateway functionally serves as an entrance into the circumambulatory path between the ground balustrade and the stūpa, a gateway was an elaborate construction, with many different parts. For a long time the principle focus of study was the narrative reliefs on the bottom pillars or architraves. However, pillars and architraves are separate architectural units. Pillars hold the entire gateway upright and architraves, high in the air, extend horizontally between two ends. The same can be said 2 D Mitra, Sanchi (New Delhi: Archaeological Survey of India, 2001): p.18. 20 also for the other basic components of a gateway. There are differences in both types of art depicted and the components which makeup a gateway's basic structure (see Image 2.1 for Dehejia's rendering of the southern gateway). Image 2.13: Location of the South Gateway's Narratives 3 V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997): Appendix 3, p. 283. 21 Beginning at ground level, two pillars work in conjunction to raise the rest of the gateway. Typically two sides of a pillar--the front and side faces--are carved with relief panels, containing various scenes. Many scenes have been identified with the Buddha's final life as Gautama, or well-known previous lives. A vedika pattern separates rectangular panels. Most pillars have four to five panels, depending on the gateway. For example, the west face of the east pillar of the southern gateway, shown in Image 2.1, has three small panels and a large dvārapāla figure in the bottom panel resting on the ground. Moving vertically, on top of the gateway pillars are capitals. On all four of the gateways, capitals are above the top of the ground balustrade of stūpa no. 1. Three different types of capitals are used.4 The east and north capitals each consist of four elephants with riders. Four plump dwarves hoist up the top portions of the western gateway. Lastly, the south gateway, reminiscent of king Aśoka and the broken capital immediately to the east, has four lions holding the above gateway on their backs. All the capitals are transitional markers between the vertical uprights and the horizontal architraves which appear just above. The last component of a gateway is the architrave. Each of the Sanchi gateways have three horizontal architraves. Although the architraves look like a solid stone crossbar high in the air, they are actually separate units by themselves. The main portion is the crossbeam, extending between two separated ends. Between each of the three crossbeams are dies, square panels which interrupt the horizontal flow of the architraves. Small, separated vertical struts appear between the middle architrave and the top and 4 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 138-41. 22 bottom crossbeams respectively. Each portion of all three architraves are carved with different scenes. Every member of the described units may have individual scenes, or carry on a scene continuing from the horizontal crossbeam. For example, ends of the architraves sometimes are part of the larger scene being depicted on the crossbeam. A famous example is the Vessantara Jātaka on the north gateway. This Jātaka scene is so large that it continues on the back and front faces of the bottom architrave. Image 2.25: Three Architraves on the Outside Face from the South 5Southern Gateway, a view. In ARTstor [database online]. [cited August 6, 2010]. Available from ARTstor, Inc., New York, New York. Negative number: A36.63 of the American Institute of Indian Studies Collection. 23 Image 2.2 shows three architraves of the front, outside face of the south gateway. Each of the architrave components can be seen clearly. However, several of the components have not been preserved. The Archaeological Survey of India (ASI) replaced the missing pieces with solid blank ones. Replaced were the eastern end of the middle architrave, the western end of the bottom architrave, and all six vertical struts between the middle architrave and the top and bottom architraves.6 Interestingly, despite the reconstructive efforts of the ASI, the top architrave and the bottom are reversed.7 As they stand now, the view is incorrect (as shown in Image 2.2). Instead, on the outside face of the gateway, a queen Māyā birth or nativity scene appears on the top. In its place should be the Mānushi Buddha scene with the inscribed central stūpa. Likewise, the bottom architrave nature scene with flowers, dwarves, lotuses and garlands on the bottom should depict the siege of Kuśināgara, or more commonly known as the “war of the relics." Although much has been lost from the south gateway, it is generally representative of the standard form of a typical gateway at Sanchi. A general classification of all the relief imagery on the gateways may fit into five categories based on subject: 1.) scenes portraying various events in the life of the Buddha, such as the great departure (east gateway, outside face)8; 2.) Jātaka stories showing previous lives of the Buddha, such as the Chaddanta Jātaka (southern gateway, inside face) or Mahākāpi Jātaka (west gateway, south pillar of the front face); 3.) scenes 6 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 155-6 speculates about what has gone missing at Sanchi. 7 J Marshall, The Monuments of Sanchi vol. 2 (Delhi: Swati Publications, 1982 reprint): pl. 10. 8 These types of scenes appear in two different forms in the reliefs, large and small. For example, a single image, such as a cakra, may be an aniconic reference to the Buddha’s teachings at Sarnath. On the other hand, a very large, complicated architrave depicts the Buddha’s great departure. 24 which recount famous Buddhist history, such as the visit to the Rāmagrāma stūpa by Aśoka; 4.) scenes representing the seven earthly (Mānushi) Buddhas in the form of trees and stūpas; 5.) miscellaneous scenes and decorations such as an abundance of plant life, mythological creatures such as dwarves, and animals such as the goose.9 Images 2.3, 2.4, and 2.5 diagram the reliefs from the other three gateways for the reader's sake. The diagrams are helpful to envision where scenes appear generally. For example, on the lower pillars are many life scenes from the Buddha's final life as Gautama. Mānushi Buddha scenes exclusively appear on the top and middle architraves. All but one of the Mānushi Buddha scenes appear on the outside face of the gateways. Similarly, all but one of the scenes with stūpas are on the architraves and not on the pillars. The only representation of a stūpa on a pillar is on the north, facing east. In the next section, I examine the location of the stūpa representations closer and determine if there is a recurring pattern and/or meaning in their placement on the architraves. 9 I have adopted five of Karlsson’s six categories of carvings from K Karlsson, Face to Face with the Absent Buddha (Uppsala: Uppsala University Press, 1999): pp. 97-102. He states: “The first three categories cannot be found on stūpa 2. The fourth category existed on stūpa 2 but in a simpler form. The different bodhivṛkṣas and stūpas are on stūpa 1 connected with specific Buddhas. All three categories 1, 4 and 5 have been called aniconic, but it is to the first category that the main aniconic signs belong and it must be examined more closely.” The sixth (fifth in Karlsson’s numbering) category which I have omitted includes wheels or tridents. 25 Image 2.310: Location of the West Gateway's Narratives 10 V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997): Appendix 3, p. 284. 26 Image 2.411: Location of the North Gateway's Narratives 11 Ibid., p. 285. 27 Image 2.512: Location of the East Gateway's Narratives 2.2 Location of the Stūpa Images At Sanchi, there are 30 representations of stūpas on the four gateways. Five of those stūpas images appear on the southern gateway. Three are on the top architrave, 12 Ibid., p. 286. 28 which originally faced outwards, one stūpa on a west pillar die that Mitra identifies as a parinirvāṇa scene in connection with the other scenes of the Buddha’s life. The last is the Rāmagrāma stūpa being attended to by Nāgas and visited by king Aśoka on the front, outside face of the middle architrave. In comparing these five stūpa images to the other 25 at Sanchi, are these locations typical in the grand scheme of stūpa imagery at Sāñcī stūpa no. 1? Of the 30 representations of stūpas, what locational patterns emerge between the gateways and their stūpa images? Table 2.1 lists where all the stūpa images are found. The 29 stūpa images included in the table are divided according to the architrave units explained in in the section above. A majority (66%) of the stūpa images appear facing outwards, away from the real stūpa. Many of the images occur in the top architraves. Table 2.2 takes apart all 29 stūpa images and lists them by which gateway they appear. Although the southern gateway hosts two inscribed stūpa representations, the other three gateways have more scenes with stūpas. All Gateways Top arch. Mid. arch. Bot. arch. Ends Dies TOTAL Outside (front face) 12 1 0 6 1 20 Inside (front face) 0 0 1 0 8 10 12 1 1 6 9 2913 Table 2.1: Location of the Stūpa Images Several patterns emerge by comparing the gateways and their occurring stūpa images. All four gateways have one set of three stūpa images that appear on the top architrave of the outside, front face. Three of the gateways have two stūpa images that 13 One stūpa which is omitted from this table is the aforementioned image on the west pillar of the north gateway. This image appears below the architraves and is thus not part of the components analyzed in this table. 29 appear on the ends of those same outside architraves. The south gateway does not have any stūpa images on its ends. The same three gateways which have stūpa images on the ends also have at least two stūpa images depicted in dies on the front, inside face of the architraves. The recurring placement of stūpa images on the three other gateways formulates a tangible, recognizable difference between the southern gateway and the others. South Gateway Top arch. Mid. arch. Bot. arch. Ends Dies TOTAL Outside (front face) 3 1 0 0 1 5 Inside (front face) 0 0 0 0 0 0 5 West Gateway Top arch. Mid. arch. Bot. arch. Ends Dies TOTAL Outside (front face) 3 0 0 2 0 5 Inside (front face) 0 0 0 0 3 2 8 North Gateway Top arch. Mid. arch. Bot. arch. Ends Dies TOTAL Outside (front face) 3 0 0 2 0 5 Inside (front face) 0 0 0 0 2 2 7 East Gateway Top arch. Mid. arch. Bot. arch. Ends Dies TOTAL Outside (front face) 3 0 0 2 0 5 Inside (front face) 0 0 1 0 3 4 9 Table 2.2: Location of the Stūpa Images according to Gateway The same trend continues when studying the scenes as well, not just the stūpas' physical placement. A majority of stūpa images can be found in Mānushi Buddha scenes. Each Mānushi Buddha scene contains seven elements, which are made up of either stūpas or trees representing each of the earthly Buddhas. Six of the seven Buddhas 30 represented are the immediate predecessors of Gautama Buddha. Gautama is, of course, the seventh Buddha in these representations, as he is the most recent. The six predecessors are: Vipaśyin, Śikhin, Viśvabhū, Krakuchchhanda, Kanakamuni, and Kāśyapa. In top architrave of the southern gateway, there are four trees and three stūpas. From left to right the trees have been identified as: sirīśa (acacia sirissa) of Krakucchanda Buddha; the udumbara (ficus glomerata) of Kanakamuni; the nyagrodha (ficus indica) of Kāśyapa (the banyan fig tree leaves well marked and also on middle lintel); lastly, ficus religiosa of Śākyamuni, whose leaves are carefully drawn.14 The Mānushi Buddha scenes are usually depicted with alternating stūpas and trees, placed side by side. On occasion only trees seem to be represented, such as on the front side of the middle lintel of the north gateway. Mitra views the nāgapuṣpa tree on the top lintel of the west gate as representing the future Buddha Maitreya.15 In table 2.3, I systematically list the number of elements in each scene. To form a proper Mānushi Buddha scene there must be seven elements. For the west and south gates, there are three stūpas and four trees. However, for the north and east gateways there are five stūpas and two trees shown. Perhaps artistic vision may be the relationship between the Mānushi Buddha scenes and these two gateways. The same guild or program coordinator could have been in charge of both gateways and thus his method of depicting the Mānushi Buddhas was three stūpas with four trees. The other two gateways may have had either a separate guild or a separate program coordinator who envisioned the scenes slightly differently. Temporally, the presence of Balamitra’s donative 14 J Marshall, The Monuments of Sanchi vol. 2 (Delhi: Swati Publications, 1982 reprint): pl. 15. 15 D Mitra, Sanchi (New Delhi: Archaeological Survey of India, 2001): p. 41. 31 inscriptiosn on the west and south gateways functions as an additional connection between these two gateways which share the same number of trees and stūpas in their Mānushi Buddhas scenes.16 South Gateway Left End Center arch. Right End TOTAL ELEMENTS Stūpas 0 3 0 3 Trees 0 4 0 4 7 West Gateway Left End Center arch. Right End TOTAL ELEMENTS Stūpas 0 3 0 3 Trees 1 2 1 4 7 North Gateway Left End Center arch. Right End TOTAL ELEMENTS Stūpas 1 3 1 5 Trees 0 2 0 2 East Gateway Left End Center arch. Right End TOTAL ELEMENTS Stūpas 1 3 1 5 Trees 0 2 0 2 All Gateways Left End Center arch. Right End TOTAL ELEMENTS Stūpas 2 12 2 16 Trees 1 10 1 12 Table 2.3: Architectural Composition of Mānushi Buddha Scenes 16 Balamitra is also mentioned as a donor of a lion sculpture on the north gateway. However, unlike the donative inscriptions on the west and south gateways, the northern gate donative inscription does not specify that Balamitra is the pupil of a senior monk. Although it seems likely that this is the same Balamitra from the other inscriptions, it cannot be said for sure without additional information. 32 The south gateway is unique in its depiction of the Mānushi Buddhas in at least one other way. The south gateway’s scene is entirely inclusive on the top center architrave. The other gateways, in order to form the proper number of seven, must include either stūpas or trees on the ends of the top architrave as well. In the case of the north and eastern gateways, the ends contain stūpas, although the west gate, previously associated with the south gate based on its stūpa to tree ratio, contains trees on its ends instead of stūpas. Thus, the north and eastern gateways may be further associated with each other based on the close similarity in their Mānushi Buddha scenes. The southern gateway's Mānushi Buddha is also distinct because of the inscribed stūpa in the central portion of the top architrave. The donor mentioned in this relief is Ānanda, son of the foreman of the artisans of king Śātakarṇī. Interestingly, only one other stūpa image is inscribed. Just below Ānanda's Mānushi Buddha architrave on the south gateway is a large narrative scene. The front face middle architrave depicts the Rāmagrāma stūpa being visited by a king and protected by a number of Nāgas. The donative inscription is written inside of the stūpa in the middle and attributed to Balamitra, pupil of Aya Cūḍa. However, the inscription seems to have no relationship with the scene itself, as the donor usually has no known relationship to the scene, perhaps suggesting that it was Balamitra’s own choice to have the scene represented, as he may have contributed the proper amount of funds to its cost. The location of the Rāmagrāma stūpa scene on the middle architrave is interesting, as the other 28 stūpas come either on the top architraves, in panels, or in dies, not in narratives. 33 The stūpa in the scene has been verified by a number of scholars as being the Rāmagrāma stūpa from Buddhist legend. According to Mitra, Of the eight original stūpas, Aśoka is said to have opened up seven with the intention of distributing the relics contained therein among innumerable stūpas erected by himself. He failed to secure the relics from the stūpa of Rāmagrāma, zealously guarded and worshipped by the nāgas. To the right of the stūpa is Aśoka with his retinue, and to the left are the nāgas with their families.17 Dehejia, being one of the most recent scholars to write on the topic, agrees.18 Writing more than a century ago, Ferguson recognizes that it could be the Rāmagrāma stūpa, but suggests one alternative reading. Because of the popularity of the legend, he states that it might refer to the desire of locating and dividing the relics rather than the actual event in Buddhist history: “It by no means follows that the Dagoba [stūpa] here represented is that at Rama Grama [sic], but the action is the same, and may have been traditionally related of fifty other places.”19 J. Vogel in Indian Serpent Lore or The Nāgas in Hindu Legend and Art, describes the scene as follows: …we find the other version of the legend depicted in a very convincing manner. On the proper right side of the panel the serpent-demons are shown in their watery home, surrounded by forest-trees, the lower part of their body concealed by the waves. Then we see the Nāgas approaching with their offers the stūpa, which occupies the place of honour in the centre of the tableau. The male Nāgas, as usual, are distinguished by means of a five-headed snake-crest, whereas their female counterparts exhibit but a single cobra issuing from behind their head. On 17 Ibid., p. 39. 18 V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997): p. 129. She also believes it is a monoscenic narrative representation of the Rāmagrāma stūpa and king Aśoka. She uses the Aśokāvadāna as her primary source (p. 164): “A variation on the tale contained in the Aśokāvadāna tells us that the nāgas took Aśoka into their nether regions and showed him how deeply they honored the relic casket. It is this scene that the artist has chosen to portray [in the Amaravati medallion Fig. 144], surrounding the relic casket with figures of adoring nāgas and nāgiṇīs, confident that viewers will thereby recall the entire sequence of events connected with the legend.” 19 J Fergusson, Tree and Serpent Worship (London: Asian Educational Services, [1873] 2004): p. 118. 34 the other side of the sacred monument a royal cortege consisting of a horsemen, a chariot, and mounted elephants in drawing near. We may assume that the personage standing on the chariot (he is attended by a chowrie-bearer) represents the great king Aśoka.20 Marshall and Foucher describe the scene almost identically and are equally sure that the royal personage is Aśoka, along with this viceroy (uparāja).21 For centuries this architrave has been interpreted almost unanimously one way. Although alternative interpretations may be possible, considering the evidence, this interpretation seems to be the most plausible. From my discussion of the where the stūpa images appear, as well as my description of Mānushi Buddha scenes and the solitary Rāmagrāma scene on the south gateway, in this section I made several clear distinctions between the patterns of stūpa image placement. Although stylistically the southern gateway's Mānushi Buddha scene is related to the one on the western gateway, the presence of two stūpa images on the ends of the western gateway separate it from the south. Overall, the southern gateway's depictions of stūpas breaks the pattern found in the other three gateways. The placement of the images is only semi-consistent with the other gateways and suggests that the southern gateway's artistic vision was intentionally different, or perhaps an early experiment. 20 J Vogel, Indian Serpent Lore or The Nagas in Hindu Legend and Art (London: Arthur Probstain, 1926): pp. 126-7. 21 See J Marshall, The Monuments of Sanchi vol. 2 (Delhi: Swati Publications, 1982 reprint): pl. 11. Although the middle section of the southern gateway is fairly conclusively the Rāmagrāma stūpa, or, at the very least, a stūpa of similar status in Buddhist lore, on the eastern gateway another lone depiction of a stūpa has caused some controversy and has been mislabeled as also being the Rāmagrāma stūpa. The southern gateway’s Rāmagrāma stūpa likely stems from the Aśokāvadāna, or a similar story, given its imagery. But from where does the eastern gateway’s so-called Rāmagrāma stūpa scene draw its inspiration? 35 2.3 Inscriptions in the Reliefs The most unique characteristic of the reliefs from the southern gateway are the three donative inscriptions appearing inside the visual field.22 In each inscription, the donor’s written agency seeps into the scene and is part of the scene. Other donative inscriptions from the other gateways are not found in any scenes. The southern gateway's three different donors were: Balamitra the monk, Ānanda the son of the foreman of the artisans, and the ivory worker guild from a nearby town. This evidence distinguishes the south gateway from the other three and warrants its need for greater study. Beginning with the lowest inscription (Image 2.6), found on the western pillar, facing east, the inscription reads: “[this] carving was done by the ivory-workers of Vidisha” (no. 400).23 Because the inscription appears very similarly to the other two readable inscriptions from the southern gateway—that is, within the architecture of the scene itself—the inscription is assuredly donative also, despite not containing the usual formula. The inscription states that at least some of the southern gateway’s stone was carved—if not donated—by a local guild. The inscription’s word for “ivory” (daṃta) indicates that the guild specialized in working with very hard materials and was familiar with carving reliefs. Paired with the other two donative inscriptions located on the gateway proper far above, at least three entities were involved in creating the content of the southern gateway. 22 A fourth inscription appears on the south gateway but is illegible. 23 Majumdar read: "Vedisakehi daṃtakārehi rupakaṃmaṃ kataṃ" in J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 342. 36 The inscription occurs just above the figures, on the roof portion of the architecture, as part of the scene itself. The other two readable inscriptions have the same physical relationship with their inscriptions, as the two donative inscriptions are found inside two representations of stūpas, showing a stylistic similarity between the physical locations of each donative inscription. Image 2.6: Worship of the Headdress with Ivory-Workers Inscription The Rāmagrāma stū (Image 2.7). The inscription reads: “The gift of Balamitra, a pupil of the Preacher of the Law Aya-Cūḍa” (no. 399). Image 2.725: Aśoka's V Image 2.8: Central The last inscription lies on the erroneously restored top architrave, in the center the Mānushi Buddha scene 24 Majumdar read: "1.) aya-cū Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint) 25 From M Benesti, Stylistics of Buddhist Art in India vol. 2 2003): pl. IV. 37 pa scene has the next inscription, on the middle architrave 24 isit to the Rāmagrāma Stūpa with Balamitra's Mānushi Buddha Stūpa with Ānanda's I (Image 2.8). This scene, as previously described, is unique in ḍasa dhamakathikasa; 2.) atevāsino balamitrasa dāna : p. 342. (New Delhi: Aryan Books International Inscription nscription stūpa of ṃ" in J Marshall, The , 38 its different portrayal of the Mānushi Buddhas. The inscription reads: “The gift of Ānanda, son of Vāsiṭhī, the foreman of the artisans of the rājan ṣirī Śātakarṇī” (no. 398).26 The mentioning of king Śātakarṇī has been traditionally used to date the construction of the gateways to c. 25 C.E.27 Both Balamitra and Ānanda's inscriptions fall well within the visual field of their respective scenes, the same as the ivory-workers guild inscription. Because the inscriptions occur inside each of the stūpas' aṇḍa, or "shell" outer casing, there is a noticeable absence of garlands, drapery, or any other kind of adornment. Every other stūpa image not found on the southern gateway has adorning features, flower garlands draped across the aṇḍa and/or dangling from the chattras. In this way, the ivory-workers guild inscription, Balamitra and Ānanda's inscriptions are large parts of the scene. No other donative inscriptions found on the gateways encroaches on the relief like these inscriptions. Other donative inscriptions on the gateways are located on the following: a lion (north gateway), two elephants (north), crowning triratna (east), and pillars (north, east and west). The lion, elephants, and triratna are all three-dimensional sculptures.28 However, none of the other inscriptions appear on architraves. Nevertheless, the placement of the inscription on these sculptures does clearly demonstrate that the donors 26 Majumdar read: "1.) rāño sirisātakaṇisa; 2.) Āvesanisa Vāsiṭhīputrasa; 3.) Ānaṃda dānaṃ" in J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 342. 27 Śātakarṇī likely refers to Śātakarṇī the first of the Sātavāhana line. According to the short chronology, he gained power in roughly 11 C.E. However, Vincent Smith, who Karlsson states as impossible to follow, suggests something beginning the 3rd century B.C.E., which is very improbable. Instead, K Karlsson, Face to Face with the Absent Buddha (Stockholm: Elanders Gotab, 1999): p. 96 relies upon an alternative date. J Shaw, Buddhist Landscapes in Central India (London: The British Academy, 2007) follows Marshall’s suggestion, agrees with this timeline as the sequence of construction clearly shows that the ground balustrade predates the gateways by a generation or two. 28 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 341-2. 39 intended to gift the sculptures themselves, somewhat similar in form to Balamitra and Ānanda's inscriptions. Therefore, it may be reasonable to assume that Ānanda and Balamitra, with their names etched inside of the stūpas of the images, also intended to donate not only the stūpas within their individual scenes, but the entire scene, exactly like the ivory worker guild. If this possibility is in fact true, then Ānanda and Balamitra possessed a considerable amount of gifting power and influence on the south gateway’s artistic program. The donative inscriptions from the southern gateway could have served as early attempts to place donative inscriptions in reliefs or sculptures at Sanchi. Their effect on the relief images itself may have been unintentional, as the later gateways display a different artistic vision for this feature by including ornamentation. 2.4 Conclusion In this chapter, I sought to connect the dots between the various units of the south gateway. Using images and inscriptions, I related the south gateway to its three counterparts and shown how it is both similar and different. Throughout, I presented some data to draw attention to several similarities and differences within the context of the two inscribed stūpas on the southern gateway. First, through describing the various types of scenes present at each of the gateways, I outlined how the south gateway’s version of the Mānushi Buddha scene represents a different artistic vision from the other three gateways. One example is the inclusive nature of the scene’s trees and stūpas on one continuous architrave rather than extending onto the end panels. Further evidence of this distinction is the donative inscription upon the central stūpa, donated by Ānanda. Ānanda, son of the foreman of 40 the artisans of a king, enacted a similar constructive power as the local ivory workers who are also mentioned on the southern gateway. Next, I emphasized the Rāmagrāma stūpa scene and discussed the donor’s prominence not only on the south gateway but at Sāñcī as a whole. Balamitra donated no less than two times—with a possible third sculpture on the north gateway—each an individual sculpture or image. On the south gateway, his heavy hand dictated an entire architrave, like Ānanda on the top architrave, and the ivory guild on the eastern pillar below. As indicated his inscription, Balamitra was probably the solitary financier and instigator of the Rāmagrāma scene. Further, the Rāmagrāma scene is one of only several unique scenes upon the gateways. Finally, through an articulation of the inscriptions present on each of the gateways, I tried to show that the south gateway's inscriptions within the visual field led to a forced omission of decorations along the stūpa image's shell. Located within architecture of the reliefs, the inscriptions, along with their donors, are very much a part of the scenes themselves. Between the unique composition of the southern gateway’s inscribed scenes and the unique placement of the gateway’s inscriptions, the south gateway's artistic vision separates it from the other three. To return to Dehejia’s assessment of the south gateway as “medley of themes that bear no integral relationship one to the other,”29 I would contend that the south gateway's convergence of image and text enables some new information to be gleaned from the old reliefs. 29 V Dehejia, Discourse in Early Buddhist Art (New Delhi: Munishiram Manoharlal Publishers, 1997):p. 129.    Chapter 3 -- Reading the Inscriptions at Sanchi At Sanchi, 848 inscriptions date from the Early Historic Period (c. 200 B.C.E. – 200 C.E.).1 In this chapter, I use the many donative inscriptions found on the ground balustrade of stūpa no. 1 (c. 100 B.C.E.) to learn about donor tendencies at Sanchi. I utilize these tendencies to illuminate the nature of donation during the time of Balamitra, the monk who donated the central architrave of the southern gateway (c. 25 B.C.E. - 25 C.E.). From this study, I suggest that donors maintain agency throughout, from the time of the ground balustrade to the time of Balamitra and the gateways. To study the inscriptions, I adopt several perspectives. The first perspective looks at the inscriptions' anatomy--that is, the common text of the inscriptions. Within this perspective, I ask, what are the common units of an inscription's written text? What information may be gleaned from studying these units separately and then together? The second perspective considers the architectural pieces inscriptions are placed upon. What are the fundamental pieces inscriptions are inscribed upon and are there differences between the pieces? Next, I examine the relationship between the written text and its architectural piece. Are there any visible patterns in interpreting these relationships? Lastly, I attempt to determine Balamitra's place amongst the donors at Sanchi. What type of donor was Balamitra, and what tendencies do similar donors exhibit?   1 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 263-396.    For this study, I compiled a table of the inscriptions and amended Majumdar’s list of inscriptions in The Monuments of Sanchi2 as needed. However, in this thesis, I study only inscriptions Marshall nos. 15 to 404, 3 which are the inscriptions found on the ground balustrade of stūpa no. 1 (nos. 15-388), the four gateways of stūpa no. 1 (nos. 389 to 404), and miscellaneous ground balustrade fragments no longer situated in-situ. I chose the ground balustrade around stūpa no. 1 to inform me about donor tendencies for many reasons. First, the ground balustrade contains the largest inscriptional deposit in the ancient Indian Buddhist world. Next, the sheer number of inscriptions yields a credible amount of testable data. Part of the attraction of studying this balustrade is that it is mostly in-situ4, like the accompanying gateways. During reconstruction, several crossbars were flipped upside-down. The information provided by the inscriptions' written text yields names, occupations, families, monastic lineages, and non-local villages. Because of their similar written formulas, I consider the inscriptions on the ground balustrade as precursors for the slightly later inscriptions on the four gateways. 3.1 -- The Anatomy of a Donative Inscription Donative inscriptions record gifts made to the Buddhist community at Sanchi. 5 This phenomenon occurs at many sites throughout India.6 The inscriptions are in the   2 In brief, I added occupational, monastic lineage, and/or origin markers to certain inscriptional records when it is clear that the original record matches an individual listed elsewhere. 3 The numbers following the citation refer to the inscription number within Ibid., pp. 287-396. 4 Nos. 364-388 were found in miscellaneous places on the ground and nos. 405-462 are pavement slabs, which are also not extant. 5 At least 52 inscriptions show that a donor gifted more than once. However, because of the Sanchi inscription’s corroded or irregular written script, or unsystematic spelling standards, it is often difficult to determine finite numbers. These 52 inscriptions are included into the 349 and not subtracted. The grand    Aśokan Brāhmī7 script and typical northern Indian epigraphical Prakrit language. Of these 349 inscriptions, 15 are unreadable, or in very fragmented form. Although several of these inscriptions are unusable, most have some data to contribute to this study and are calculated in final totals.8 Over time, there have been numerous attempts to date the Sanchi donative inscriptions, but generally, Majumdar’s work at the end of Marshall’s volume 1 is the standard.9 Table 3.4 below outlines the groups and approximate dates, in linear order, of the different inscriptions found at Sanchi. For comparison’s sake, I included Ramprasad Chanda’s slightly earlier analysis. Ramprasad Chanda published his findings in the      total is the total number of inscriptions and not the total number of donors. The total number of donors may be estimated, but not accurately assessed. 6 Although by no means an exhaustive list, the most famous open-air stūpa sites comparable to Sanchi are Bharhut, Amaravati, Nagarjunakonda, Mathura, and Bodh Gaya. See H Luders “A List of Brahmi Inscriptions” in Epigraphia Indica 10 (1912) for a neatly outlined early attempt to group some of these donative inscriptions. Donative inscriptions can be found in significantly lesser volumes at eastern Deccan cave sites such as Karle, Bedsa, etc. 7 We owe the decipherment of Brāhmī script to James Prinsep, who was the first to identify the oldest forms of the script. In J Prinsep "Note on the Facsimiles of Inscriptions from Sanchi near Bhilsa" in Journal of the Asiatic Society of Bengal VI (1)(1837): pp. 160-1, Prinsep remarked on the frequently recurring Brāhmī letters “da” and “nam”: “I was struck at [the inscriptions’ terminations] with the same two letters…it immediately occurred that they must record either obituary notices, or more probably the offerings and presents of votaries, as is known to be the present custom in the Buddhist temples…” At the end of his article, he presented the alphabet as he knew it, entirely correct with the exception of the vocalic ṛ, which is actually jha, and five others which he was unable to locate (gha, ṅa, jha, ña, and o). See R Salomon, Indian Epigraphy (Oxford: Oxford University Press, 1998): p. 207 for a more comprehensive discussion. 8 Not only are there problems in reading the inscriptions due to handwriting irregularities, but the stone has worn down considerably in over two-thousand years of weathering. In some cases, the inscriber themselves were inaccurate. For example, one inscription (no. 180) records the gift of an outside crossbar of the monk Dhamarakhita from the town of Kacupatha, while another, the next inscription (no. 181), records the gift of an outside crossbar of the nun Dhamarakhitā, also from Kacupatha. In my database of inscriptions, I have recorded these two as different persons, giving two different gifts. However, theoretically it is possible that the engraver simply made a mistake and added the long ā mātra to no. 181, or that for no. 180 he simply left out the long ā mātra. In either case, assuming there is a mistake, these two donors would be one donor, separated in my archive because of a written mistake by the engraver. 9 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 301-396.    Memoirs of the Archaeological Survey of India10 in 1919, while Majumdar’s edition is from 1927. Nevertheless, Majumdar relies heavily upon Chanda’s considerations. Inscriptions Marshall’s groups Chanda’s dates11 Aśokan Edicts 1; 250 B.C.E. 1; 250 B.C.E. Stūpa 1 balust.12 2a; 175-125 B.C.E. 4a; 175 B.C.E. Stūpa 3 reliquary 2a; 175-125 B.C.E. -- Stūpa 2 balust. / reliquaries 2b; 125-100 (?) B.C.E. 4b; 150 B.C.E. Stūpa 1 misc. fragments 2a-c; 125-70 (?) B.C.E. -- Stūpas 1/3 gateways 3; 100 B.C.E. – 15 CE13 7; 75 B.C.E. – 25 B.C.E. Stūpa 3 balust. 3; 100 B.C.E. – 50 B.C.E. -- Kushāna inscriptions 4; 100-150 C.E. 9; 100 C.E. Guptan inscriptions14 5; 600 - 800 C.E. -- Table 3.1: Palaeographic Groupings of the Inscriptions at Sanchi Palaeographically, the gateways15 belong to a later generation of Brāhmī characters. Majumdar assigns them to group 3 (as opposed to group 2a for the balustrade).16 The primary reason for dating the inscriptions of the gateways to an entirely later period lies in the south gateway’s reference to King Śātakarṇī. At the time in which Majumdar and Marshall were writing, the debate regarding the king’s exact time   10 R Chanda, "Dates of the Votive Inscriptions on the Stupas at Sanchi," in Memoirs of the Archaeological Survey of India no. 1 (Calcutta: Indological Book Corporation, 1977): 1-15. 11 See Ibid., pp. 14-15. He includes the Nagarjuni Hill (group 2) cave inscriptions of Dasharatha, the Besnagar Garuḍa (3 and 5a for the later maharaja bhāgavata inscription) pillar inscriptions, the Nanaghat cave inscription (5b), and the Hathigumpha inscription (6) of Kharavela into his chronologically arranged groups. 12 Notable here is that Marshall dates Temple 40 to the same period. Their primary evidence rests with an individual named Data-Kalavaḍa, who donated portions of the ground balustrade of stūpa no. 1 (nos. 353- 355) and a pillar from Temple 40 (no. 790). See J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 269 for Marshall's discussion of palaeographic similarities. 13 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): pp. 272-279. The very large date discrepancy in Marshall is a result of two contemporaneous scripts appearing on various parts of the gateways. 14 Two well-known Gupta era inscriptions are nos. 833 and 834. Nos. 835, 839 also date to roughly the same period. No. 842 is an important inscription that mentions rulers of Mahāmālava, also known as Mālwā, the central region where Sanchi is located. 15 It is obvious that the gateways themselves all belong to a contemporary period because certain content reappears. For instance, the imprecatory inscriptions are duplicated on the North, East, and West gateways. In addition, names of patrons also reappear. Balamitra, the disciple of Aya Cūḍa, apparently donated on not only the south gateway, but also on the east and north gateways. Similarly, the banker Nāgapiya, of Kurara village, donated on both the east and west gateways. 16 Ibid., pp. 274-5.    was contested.17 Cunningham, using a Purāṇic list, placed Śātakarṇī in the first quarter of the first century C.E. Bühler placed Śātakarṇī as early as the middle of the second century B.C.E. in contrast.18 Majumdar’s final assessment regarding Śātakarṇī is that he is Śātakarṇī II from the Hathigumpha inscription, thus dating the gateways to around the middle of the first century B.C.E. As first noted by Chanda19, there are at least two forms of writing appearing on the gateways.20 The imprecatory inscriptions bear what Majumdar calls an “ordinary” appearance while the inscriptions placed inside reliefs as part of the relief images or, at least, very near the relief images, are of an “ornate” style. Majumdar calls this ornate script stylistically beautiful and symmetrical.21 Majumdar describes the ornate script as possessing broadened knobs at the tops of letters, almost like a serif.22 The ornate style eventually becomes the Brāhmī of the northern Kṣatrapas and Kushāna inscriptions. The a, ka, cha, ta, da, va, and sa letters all show the tendency to serif. In the end, however, the stylistic differences are not alphabetical differences. For sake of dating, all of the gateways from stūpa no. 1 and the solitary gateway from stūpa no. 3 date to about the same period based on palaeography.   17 Ibid., pp. 275-8. 18 See ibid., pp. 276-7 for a discussion on this point. 19 R Chanda, "Dates of the Votive Inscriptions on the Stupas at Sanchi," in Memoirs of the Archaeological Survey of India no. 1 (Calcutta: Indological Book Corporation): pp. 4-5. 20 A Cunningham, The Bhilsa Topes (London: Smith, 1854): pp. 272-3. He noticed a difference in epigraphic style also, and that, on the northern entrance, one inscription was hidden by a later balustrade extension. 21 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 273. 22 The ornate style can also be found at Bharhut and Mathura.    I begin to study the inscriptions by placing the inscriptions into several distinct groups based on information provided.23 The most basic group is simple. Inscriptions in this group state only the donor’s name: Dhamadināya dānaṃ. The gift of Dhamadinā (no. 188). Other inscriptions may indicate more, such as the donor’s place of residency: Pusagirino Nāvagāmakasa dānaṃ. The gift of Pusagiri, an inhabitant of Navagāma (no. 183). Each of the above inscriptions are gifts of non-monastic Buddhists. On the contrary, monastic Buddhists identify themselves in their inscriptions: Dhamarakhitasa bhichuno dānaṃ. The gift of the monk Dhamarakhita (no. 187). Some other inscriptions indicate the donor’s precise relationship to the Buddhist community: Vudināye upasikāye dānaṃ. The gift of the female lay-sister Vudinā (no. 19). A few of the inscriptions indicate the donor’s mercantile status:   23 I have carefully annotated Majumdar’s original list in my table because his list contains a lack of conformity and thus a lack of surety. For instance, inscriptions nos. 343, 344, 345 read “gift of Data Kalavāḍa from Vedisā.” However, no. 107 reads “gift of Datakalavaḍa” only. All four donative inscriptions are seemingly referring to the same person but contain obvious differences. The trio of inscriptions read the person’s name as “Data Kalavāḍa”, containing a long ā vowel and a retroflexed ḍ consonant. The trio also clearly states that the person is from the local town of Vedisā. All three of these inscriptions are located on crossbars facing outside, away from the stūpa. Inscription no. 107, also an outside crossbar, has a slightly different variation to the person’s name, “Data Kalavaḍa,” possessing the retroflexed ḍ consonant but missing the long ā vowel and the town in which the donor was from. Although it is likely that there are two different persons at work within these four inscriptions, it is more likely that all four of these inscriptions refer to the same person, and that person is from the town of Vedisā. Therefore, on my list, I attribute all four of these inscriptions to the same donor, regardless of the minor inscriptional differences. For no. 107, I will cautiously add that he is, like the other trio, from the town of Vedisā, which we can confirm as the modern town of Vedisha, just north of modern Sanchi. In this way, I have included only a single “Data Kalavāḍa” in my calculations of multiple donors, which remains at 52.    Kaṃdaḍigāmā seṭhino dānaṃ. The gift of the banker from Kaṃdaḍigāmā (no. 43). Several inscriptions indicate a familial relationship: Ujeniyā kalūraputasa bumūsa dānaṃ The gift of Bumū, son of Kalūra from Ujjain (no. 53). Accordingly, donor communities are identified based on the information the donor contributes. Sanchi Buddhist Community # % of total Monastic Donors 147 42% Lay Donors 175 50% Mercantile Donors 27 8% 349 100% Table 3.2: Donor Communities on the Ground Balustrade of Stūpa no. 1 (c. 100 B.C.E.) Donations Donors Donations per donor Monastic 147 121 1.21 Lay 175 140 1.25 Mercantile 27 17 1.59 TOTAL 349 278 1.255 Table 3.3: Number of Donations Compared to Number of Donors Monastic Buddhists are the easiest to identify. Of the th349 inscriptions from the ground balustrade, 147 (42%) identify themselves as members of the monastic community. Similarly, 175 (50%) are likely part of an unidentifiable lay community. Interestingly, amongst all donors, only approximately 4% of donors on the ground balustrade identify themselves as part of the lay community, clearly stating their status as an upāsika, upāsaka, gahapati, etc. While it is tempting to divide the laity into the “official and unidentifiable” partitions, for the purposes of this study, I included both the     very small community of upāsikas, upāsakas, gahapatis, etc., with the large mass of donors who record no affiliation whatsoever.24 Most donors gifted only a single architectural piece and are representative of the general donor population. Some donors, such as Samika the merchant, donated multiple times on multiple architectural pieces. Samika, along with his son Siripāla, donated three consecutive crossbars (nos. 200, 201, 202). Others, such as the monk Dhamarakhita from Kachupatha, donated multiple crossbars in different locations (nos. 180 and 187)25. It is not entirely clear what incentive a donor might have in donating consecutive pieces. So, why would Samika, along with his son Siripāla, donate three crossbars when it may have been simpler to gift a single rail-pillar or copingstone? One of several possibilities is that donors who donated more than once, like Dhamarakhita, likely visited or were solicited for dāna at different times, resulting in multiple architectural pieces appearing at different locations.26 Or, equally as likely, when the balustrade was pieced together, both of Dhamarakhita’s crossbars, instead of being placed together, were placed near each other   24 It may be worthwhile in the future to explore this curious feature of the inscriptions further. However, for the moment, the position that the official lay community at Sanchi was actually quite tiny is untenable. A future study based on this particular data anomaly would begin by contrasting the 4% official lay community with the monastic community (42%). Shifting the numbers would position the monastic community as the overwhelming majority. A brief analysis would suggest that the stūpa cult was primarily supported by the regional monastic community. 25 Dhamarakhita’s two donations occur in the same area, both facing outwards, and both are crossbars. Both crossbars are part of the eastern portion of the balustrade. 26 In some instances, records indicate that a donor may have been responsible for donations at multiple stūpas (such as at contemporaneous stūpas nos. 1 and 3), or across multiple landscapes (such as the ground balustrade of stūpa no. 1 and the berm balustrade on the same stūpa, constructed at different time periods). For example, because the berm balustrade of stūpa no. 1 dates to the same palaeographic period as the ground balustrade, it is safe to assume that the nun Achalā from Nadināgara (no. 465) on the berm balustrade is the same nun Acalā from Nadināgara (no. 170) on the ground balustrade. Another non- monastic case is Idadata the pāvārika (cloak) merchant (nos. 131 and 472) and Isidatā, wife of Sakadina (nos. 142 and 500).     rather than side-by-side, simply by happenstance or consequences of the construction process. These are the fundamental units within the written text. The inscriptions as a whole are a wealth of vital information that can be approached from many directions. The written text has these individual units to examine. Similarly, the visual text can be split into distinct units. Weaving the two texts, the names, relationships, and other information provided in the inscriptions generates a sophisticated understanding of how the stūpa's many parts come together. 3.2 -- Architecture and Donation Railpillars are the fundamental pieces available for donation. Two uprights hold together three crossbars. Copingstones are at the top of the railpillars and cap the balustrade. There is a visual hierarchy between these three pieces. Copingstones may be the on top physically and are the largest, heaviest sections, but the standing railpillars are the most important functionally. Table 3.6 lists the number of donations of each balustrade piece. The position of the inscription on these different architectural pieces may indicate several important qualities of the donors. Crossbars Railpillars Copingstones Monastic 36% 22% 79% Lay 52% 55% 18% Mercantile 5% 23% 3% Percentage of Total 69% 21% 9% Major Donor Group Laity Laity Monastic Table 3.4: Types of Architectural Pieces with Donor Frequencies Inscribed # Total # % Inscribed Volume of Piece Crossbars 242 362 67% 6,944 in3 Railpillars 74 132 56% 35,626 in3 Copingstones 33 60 55% 63,360 in3 Table 3.5: Architectural Pieces Compared to Volume and Inscribed Number.     Visually, the crossbars of the balustrade are the most effective at blocking a person’s gaze inside or outside the circumambulatory path. Although the crossbars are by far the most common piece, only 67% of the original 362 original pieces remain, if we are to assume that each original piece had a donative inscription. Before 19th century Western investigation, the site was likely pillaged for useful stone. Other pieces were lost due to the time. Image 3.1:Five Crossbars, Three Railpillars, and One Copingstone from Stūpa no. 1 Whatever the case may be, the remaining inscribed crossbars are a vital source of information. The same goes for the preservation of the railpillars and the copingstones. In short, only 63% of the total architectural balustrade pieces remain in-situ or preserved    by the Archaeological Survey of India. This potentially affects calculating definite numbers or percentages. However, because a majority of the balustrade does remain, it is still possible to formulate useful methods for reading inscriptions because the majority is a reflection of the general pattern. As shown in Table 3.4, the laity is responsible for a majority of the common crossbars. This is no surprise given that the laity is also the largest donor group. One expects the largest group to donate the majority of the most common pieces. The laity similarly donated the majority of railpillars. However, note that members of the mercantile class have a greater frequency of railpillar donations than they do crossbars or copingstones. One reason for this could be that mercantile donors, perhaps wealthier than the average person, could afford the larger piece more frequently than less-affluent donors of other affiliations could. Fewer copingstones survived than other pieces. Out of the 60 original copingstones needed to complete the entire balustrade, only 33 remain. The Archaeological Survey of India replaced the missing ones. Because copingstones are the largest and rarest, they may have been the most expensive, assuming these donative records are records of financial transactions. Similarly, the inscribed and elaborately carved gateway architraves where Balamitra and Ānanda’s inscriptions appear are comparable to the. On the other hand, visually the copingstone is not as important as the upright railpillar. Even Balamitra and Ānanda's architraves are more akin to crossbars visually, despite their large size. Where the copingstone functions as a cap, a railpillar supports the entire balustrade. The same is true for gateway pillars.    However, there may be little significance in seeking a hierarchical value between railpillars and copingstones. In her book on giving in early Buddhism, Ellison Banks Findly suggests that the conjunction of Buddhism and the newly emergent householder category of the era led to “patrons of the [Buddhist] religion prosper[ing] socially in terms of their status and reputation, for dāna teachings tell potential donors that the more one gives the greater … their reputations.”27 This system allowed a donor’s worth to be based on merit and not on birth. The merit or reputation acquired through a donation at Sanchi might not have depended on railpillars or copingstones. Donors of any social background may enhance their reputation among all sectors of society by gifting to renunciants who are similarly from all social origins.28 Gifting is particularly apt in a religious community where, not only can your reputation increase from donation, but also where some sort of intangible, theological or soteriological merit is simultaneously acquired. The gift of a large railpillar or copingstone, monumental architectural pieces, could have served as a substantial enhancer of reputation as donations only, not as markers of the amount donated. The use of the word dāna is nearly unanimously viewed   27 E.B. Findly, Dāna: Giving and Getting in Pali Buddhism. (Delhi: Motilal Banarsidass, 2003): p. 17. 28 Ibid., p. 17. On p. 38, Findly notes that “…the market-oriented culture, in which Buddhism emerges, reflects a shift away from the valuation of traditional duty and obligation and a greater celebration of individual choice. This shift is based on the increased freedom brought about by social and economic changes, and allows for individual initiative and creativity. It also means, however, in the case of renunciant petitioners, that householders are not obligated by preset affiliation to support them, as they are in Vedic settings.” Therefore, the conscious choice to donate to the saṃgha at Sanchi seemed to allow for a freedom in gift choice, as the donor was not forced to give something specific. There appears to be a wide range of possibilities to choose from, depending on the type of gift one sought to donate, which was at least partially dependent on socio-economic status as well as devotion to this particular religious community.    as an early synonym of deyadhaṃma, a religious gift29, and perhaps both railpillars and copingstones were gifts of some repute. Looking at Sanchi's inscriptions, a copingstone is never mentioned directly, while several inscriptions deliberately mention the gift of a "pillar." For instance, on the ground balustrade, no. 102 records the gift of a pillar by all the relatives of the monk Nāgila. Significantly, the other occurrences where a pillar is named specifically in the inscription are located on the gateways. No. 397, on the east gateway, left pillar, the banker Nāgapiya gifts a pillar. Nāgapiya's donation is duplicated in no. 403 on the west gateway, left pillar. Across from Nāgapiya's inscription is one of Balamitra's inscriptions. In addition to identifying himself as the same Balamitra from the south gateway's Balamitra, the inscription notes that Balamitra gifted a pillar. Together, these four pillar gifts contribute much significance to the value of railpillars. Nevertheless, the infrequency of copingstone inscriptions may be a testament to their value as well. Findly’s notion that gifting in early Buddhism increased social reputation is supported by the stratification of the available number of gifts. Just 33 of the total inscriptions are copingstones, a mere 9% of the total. The copingstones are also by far the heaviest pieces and may have been a great example of gifting power. Their weight is nearly twice that of a railpillar and almost six times that of a crossbar. If the funds for the donation of a copingstone were not for the symbolic pride of donating the biggest, rarest piece of a religious structure, then the sheer cost of transportation of the   29 In the western Deccan cave sites the term deyadhaṃma (Pkt: “religious/meritorious gift”) is frequently used in similar kinds of Buddhist donative inscriptions.    massive stone would require significant funds to finance.30 Whoever, then, could gift such a rare item undoubtedly received the invisible benefits associated with the ability. Accordingly, the monastic community is responsible for most of the copingstone donations. 79% of them were donated by members of the monastic community, an overwhelming percentage when comparing to other donor frequencies. It may be justified to say that the monastic community had a pre-determined pursuit to donate copingstones, whether they were the most soteriologically auspicious pieces, the most expensive pieces, or purely just the largest and most symbolic of giving power. The same cannot be said about the railpillars, as the laity were the major donor group but at only 55%, not as overwhelming of a majority as the monastic group was for copingstones. In the same vein, the inscriptions themselves may provide information about the relative values of the architectural pieces. Each piece in the examples below may be compared to the very large architraves on the southern gateway where artistic scenes show representations of stūpas. Inscription no. 308 may hint at the price of a single gift on the ground balustrade: Vejajasa gāmasa dānaṃ. The gift of the Vejaja village. If it takes the accumulated funds from one single village, of which there is not another single donation, then the donation of a balustrade piece was expensive. Alternatively, the village of Vejaja was relatively poor, or the village itself was disinterested in giving to   30 It is unlikely that the labor cost was simply absorbed as overhead by the saṃgha.    the Buddhist community and could raise only a few donations. Whatever the case may be, the comparison between this inscription, on a single railpillar, to others is worthwhile to consider. One comparison is to a set of three consecutive donations by the merchant Samika and his son Siripāla (nos. 200-2). They read simply: Samikasa vānikasa putasa ca sa Siripālasa dānaṃ. The gift of three crossbars by Samika, the merchant, along with his son Siripāla. Samika, being a pious and wealthy merchant, probably desired to donate a set sum of funds towards the construction of a stūpa balustrade. However, his available funds were perhaps not enough to acquire a railpillar (as the accumulated funds of the Vejaja village, was, in contrast). One speculation is that instead of giving just one crossbar, Samika was determined to gift his entire sum, earning him three crossbars. These identically inscribed crossbars were assembled and placed into position at the same time. Samika probably did not visit the site more than once to donate (or, alternatively, were solicited more than once), but gave a set sum, more than enough for one crossbar but not quite enough for a railpillar. Thus, three crossbars, all lined up in a row, are in his name. In contrast, the donor Balamitra donated twice on the gateways heavily carved with artistic scenes. If Samika donated in the time of Balamitra, he might have preferred just one donation with art carved upon it, like Balamitra.31   31 The ground balustrade was undoubtedly a collective endeavor. Individuals like Samika contributed as a group to the founding of a monument. V Dehejia “Collective and Popular Basis of early Buddhist Patronage” in B. Miller (ed.) The Powers of Art (Delhi: Oxford University Press, 1992): pp. 35-45 discusses this ideal at length. However, by the time of Balamitra and the construction of the gateways, the noticeable lack of donative inscriptions may point towards a different model of monument founding other than “collective and popular” patronage as Dehejia suggests. On p. 44, she goes on to awkwardly state that    Another inscription helps establish relative value of the donated architectural pieces. Inscription no. 175 reads: Subāhitasa Gotiputasa rāja-lipikarasa dāna. The gift of the royal scribe Subāhita, son of Goti(puta) (no. 175). It compares to the southern gateway inscription by Ānanda, donor of the top architrave. Ānanda is the son of the foreman of the artisans for King Śātakarṇī. Subāhita’s inscription is marked on a railpillar, the same as the Vejaja village inscription. It is the gift of a royal (rāja) scribe, an entirely unique mercantile title. There are no other royal inscriptions on the ground balustrade, although several other inscriptions reflect donations of other scribes. Lastly, Subāhita's inscription is written in the same genitive case as Gotiputa, a monastic figure known from reliquary inscriptions. Gotiputa was the teacher of all of the so-called Hemavata teachers in this region and bears the epithet sapurisa,32 or saint (literally, “good man”). There could be several ways to translate sapurisa, but Majumdar has taken Subāhita, the scribe, as being the son of Gotiputa. Literally, perhaps the inscription could be translated as “The gift of the royal-scribe Subāhita, who is [born] of Gotiputa.” This rendering would maintain the genitive case      “the patronage of religious art was not the prerogative of the merchant and the banker. Apparently, the wealth necessary to indulge in such a luxury belonged also to persons of humbler professions like the ironmonger and stone mason, the gardener, and the fisherman.” What Dehejia is implying here is uncertain, for two of the donations upon the gateways are of the same banker, a profession Dehejia says is not concerned with the patronage of religious art. It is unclear by the time the gateways were constructed if Samika would have been able to donate religious art of not, based on a slightly altered or new model of patronage. 32 E.B. Findly, Dāna: Giving and Getting in Pali Buddhism. (Delhi: Motilal Banarsidass, 2003): p. 192. In Findly’s discussion of a sappurisa (sapurisa in the Sanchi Prakrit), the good person “gives a gift respectfully, with his own hand, with consideration, in purity, and with a view to the future.” The sappurisa serves as a model for the proper use of wealth. The use of the title at Sanchi seems to fit accordingly with the model described by Findly. The sapurisas of old, namely those whose relics are enshrined in stupa no. 2, acted for the benefit and welfare of their whole community, as their titles suggest.    while still also maintaining the reference to the sapurisa Gotiputa. It could be equally possible to translate the compound “Gotiputa” as a genitive tatpuruṣa: “son of [a] Goti.” However, based on the other inscriptions containing this man’s name, Gotiputa appears to be the full rendering. Majumdar, being the unsurpassed expert on these inscriptions, agrees.33 Considering inscription no. 175’s status as middling between a royal mercantile occupation and that of being the son of a famous monastic teacher in the area, Subāhita was probably not part of the anonymous masses of donors, but rather existed as a rather affluent member of the community. Subāhita’s considerable status supports the idea that railpillars, in addition to copingstones, held status beyond crossbars. 3.3 -- The Placement of Donative Epigraphy One meaning of a donative inscription can be derived by looking beyond its economic value. By reading between the reliefs and looking at an inscription’s physical presence on the architectural piece in which it is inscribed I can extrapolate much information. In this section I focus on three primary physical attributes of a donative inscription and ask two main questions: 1.) What is the affect of placing a donative inscription either inside facing the stūpa or outside facing away from the stūpa? ; 2.) how does the donor’s territorial residence affect patterns of donation? Combining the written text with the physical placement of the same text on an object, I emphasize that the significant value of images and inscriptions lies in their relationship to not only each other but also to their architectural and artistic context.   33 J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 295.     The Inside/Outside Distinction Inscription no. 24 lies on the outside face of a crossbar. The donor is a man named Oḍaka from the village of Vāḍīvahana. His village shows that he was either solicited in his home town for contributions or that he made a trip to the Sanchi area and contributed his gift locally. Either way, his inscription is a simple crossbar and the text is only readable from outside the circumambulatory path. On the opposite side of Oḍaka’s, inwards, facing the stūpa, is inscription no. 25. This inscription is only readable from within the circumambulatory path, as its text faces the stūpa. Inscription no. 25 was the gift of a man named Vajiguta and on a railpillar rather than a crossbar. Based on the donated piece alone, Vajiguta, who does not indicate a village that he is from, seems to have contributed something more than Oḍaka. Looking closer at the inscriptions' placement, one of them faces outwards and is readable to non-circumambulators while the other faces inwards, towards the stūpa, and is only readable by those participating in the pradakṣiṇa ritual. In this section I suggest that the some inscriptions on the inside faces of fragments are not only rarer, but possibly soteriologically more valuable and consequently perhaps even more expensive. Table 3.6 outlines the different inside facing inscriptions. CB stands for crossbar, the most common piece. RP refers to the railpillars, and CS refers to the copingstones. Henceforth, an architectural piece will be abbreviated not only with these distinctions but will also contain either an I or an O to indicate whether the inscription is located inside or outside on the specific piece. At stūpa no. 1, there are only 53 total inscriptions residing     inside the circumambulatory path. Only 53 (15%) of the 349 known donative inscriptions face inwards--a very small percentage. As shown in the table, the ratio of ICBs to IRPs and ICSs continues in the same statistical manner. ICBs make up sixty percent of the total inside inscriptions, IRPs 28% percent and the ICSs just 11%. To put it in a larger context, the six ICSs inscriptions are just six out of 349, meaning they are just 1.7% of the inscriptions on the ground balustrade. A noteworthy pattern emerges when comparing Table 3.6 and Table 3.7 (which lists the outside architectural pieces). Monastic Buddhists are responsible for a majority of the inscriptions residing on the inside faces. On the other hand, the lay community is the major donor group of outside facing inscriptions. An exception are the rare OCSs, of which the monastic Buddhists are the majority. Significantly, the monastic community, despite being the minority community, donated far more frequently the rarer and more prestigious pieces (coping stones). Similarly, they were also the majority group of inside facing inscriptions. In a soteriological context, the monastic community ought to be more concerned than the laity with inscriptions which bear their name and face towards the stūpa, a symbol for the Buddha. ICB IRP ICS Inside Total Monastic 16 8 5 29 Lay 16 6 1 23 TOTAL 32 1534 6 53 Percentage 32/53 = 60% 15/53 = 28% 6/53 = 11% Major Group -- Monastic (53%) Monastic (83%) Monastic (55%) Table 3.6: Inside Facing Architectural Pieces   34 Includes one donation by a member of the mercantile class. This chart does not include those donations except for in the “grand total” tally. The “Inside total” column does reflect this one donation.     OCB ORP OCS Outside Total Monastic 80 8 21 109 Lay 111 35 5 151 TOTAL 210 59 27 296 Percentage 210/296 = 71% 59/296 = 20% 27/296 = 9% Major Group Laity (53%) Laity (59%) Monastic (78%) Laity (51%) Table 3.7: Outside Facing Architectural Pieces In an article where he discusses donative inscriptions in great detail, Gregory Schopen argued that a donative inscription is a donor’s permanent “presence”.35 Looking at several texts describing rituals for the dead, Schopen concludes that inscribed donor’s names are “something more than a mere record.” They represent the person, even if they are no longer living. The donors, including the elite monastic Buddhists, intended to leave their essence in proximity to “another, more powerful presence".36 If a donative inscription itself is a living presence of the donor, then the balustrade piece might be thought of as an extension of that permanent presence. Once inscribed with the donor’s information, a balustrade piece becomes part of the stūpa complex. Curiously, as Schopen37 documents, these tiny markings carved into monumental stone structures tend to be high into the air—six meters or higher—hand- written in a rugged Brāhmī script in the Prakrit language that few people presumably could read, let alone see. Instead of functioning as a signpost for visitors, Schopen argues that the inscribed presences continuously engaged in ritual activity with the stūpa. The stūpa, on the other hand, contained—or was at least thought to contain—relics of the Buddha. Therefore, the stūpa itself imbued the presence of the Buddha like donors   35 G Schopen Buddhist Monks and Business Matters (Honolulu: University of Hawaii Press, 2004): pp. 382-392. 36 Ibid., p. 392. 37 Ibid., pp. 386-9.    presences infused into inscriptional materials. For Schopen, proximity to the Buddha— or at least the Buddha’s material presentation—is paramount. Archaeologist Lars Fogelin privileges another aspect of early Buddhist post- mortem "essences." Fogelin studies mortuary practices at Thotlakonda, a site in Andhra Pradesh. Buddhists there built small funerary cairns.38 These cairns perpetually engage the Great Stūpa in the afterlife; ritually and magically accumulating merit for the donor. Fogelin shows that 75% of funerary cairns—small rock mounds covering a burial site— were outside of the Buddhist monastery were in locations possessing a clear view of the monastery itself. Visibility, for him, versus Schopen’s emphasis on proximity, is more a factor than proximity, as not a single funerary cairn, containing "essences" Buddhists also, was found on opposite sides of hills where sight of the monastery was impossible. With the distinction between inside facing inscriptions and outside facing inscriptions, visibility could also be a factor at Sanchi. Fogelin39 previously argued that within the circumambulatory paths of stūpas a ritual practitioner is cut-off from view of anything but the stūpa and the path. A balustrade restricts an individual’s sightline—made of the pieces which donor’s had their names inscribed upon. The balustrade also restricts non-ritual practitioner’s visualization of the person or persons walking along the circumambulatory path. Hence, the individual practitioner acts alone or in small groups. Ritual space outside the circumambulatory   38 L Fogelin, Beyond the Monastery Walls, Dissertation. University of Michigan (2003): p. 301. 39 L Fogelin, "Ritual and Presentation in Early Buddhist Religious Architecture," in Asian Perspectives, 42.1 (2003): pp. 129-154.    path is not individual, contemplative worship but probably communal or corporate ritual guided sometimes by a ritual specialist.40 If a donative inscription placed on a balustrade fragment is a living presence of the donor, Fogelin’s method of analyzing what the ritual practitioner can and cannot see from inside their ritual space may apply. A given balustrade piece has two distinct, opposing sides. One side faces inwards towards the stūpa while the other faces outwards towards something else, most likely the landscape of the region, a town, or a temple. Because inscriptions permanently facing the stūpa are, like the individual circumambulators, in constant visualization of the stūpa, or, actually, the Buddha, those inside a piece, such as the inside a crossbar, may have been more soteriologically desirable than inscriptions on the outside, earning merit through both proximity and visualization. At the same time, these inscriptions also could have been more desirable for those wanting their reputation and/or donor power to be known to those actively circumambulating inside of the circumambulatory path, as there they would be readable for those who were literate. While the cairns at Thotlakonda are mostly within view of a stūpa, very few of the inscriptions on the ground balustrade of stūpa no. 1 at Sanchi are on the inside, facing the stūpa. Balustrade pieces, such as a crossbar, rail-pillar, or copingstone are different than cairns, which are piled rocks. One possible implication is that while only the wealthy or extremely devout could donate any balustrade piece, only the wealthiest, most   40 Ibid., p. 134.    influential, or most in need of enhanced reputation could inscribe their name, and thus their presence, on the inside piece.41 The Local/Non-Local Distinction The inside/outside distinction is also useful when considering where the donors came from. Returning to the duo of nos. 24 and 25, Oḍaka in no. 24 indicates he is from the village of Vāḍīvahana. Therefore, Oḍaka is a non-local donor to the Sanchi area. In contrast, some donors seem to omit their village deliberately, perhaps indicating that they came from a nearby location or even Sanchi proper. In my calculations, I considered donors like Vajiguta locals. Tables 3.8 and 3.9 split the donors into two categories, local and non-local, and separate them into those whose inscriptions face inwards and those whose inscriptions face outwards. Interestingly, the pattern which emerges shows that non-local donors were more likely to have their names inscribed on the inside facing the stūpa. A majority of donors listed on the outside do not specify a village. Amongst donors of the major architectural pieces, the railpillars and copingstones, donors who did specify a location donated inside railpillars and inside copingstones at a higher frequency than those who did not indicate a location. Non-locals were also the dominant donors of railpillars and copingstones when the inscription faces outside. In total, of the 109 railpillars and copingstones with accompanying inscriptions, 81 of them, 74%, were donated by donors who included their village.   41 Because these inscriptions are hidden from everyone but those circumambulating the stūpa, it remains possible that the inside facing inscriptions were actually cheaper. However, the sacred nature of the inside space, as it is closest to the stūpa itself, means that it is more likely that the inside space was a restricted space, reserved for those actively engaging in ritual with the sacred object, the stūpa.    ICB IRP ICS INSIDE TOTAL Local 11 (34%) 5 (33%) 3 (50%) 19 (36%) Non-local 21 (65%) 10 (66%) 3 (50%) 34 (64%) Table 3.8: Local vs. Non-Local Inscriptions Facing Inside OCB ORP OCS OUTSIDE TOTAL Local 146 (70%) 17 (29%) 3 (11%) 166 (56%) Non-local 64 (30%) 42 (72%) 24 (89%) 130 (44%) Table 3.9: Local vs. Non-local Inscriptions Facing Outside Town # Monastic Lay OCB ORP OCS ICB IRP ICS Local 103 45 40 63 17 3 12 5 3 Vedisa 16 8 6 10 3 2 0 0 1 Ujjain 49 15 33 26 8 3 11 0 1 Kurara 14 7 7 5 5 4 0 0 0 Nadinagara 13 12 1 6 1 4 2 0 0 Mahisati 10 2 8 9 0 0 0 1 0 Kuraghara 12 3 7 6 4 1 0 1 0 Bhogavaḍhana 6 0 6 6 0 0 0 0 0 Kaṃdaḍigāmā 5 0 1 4 1 0 0 0 0 Table 3.10: Major Non-Local Towns Compared to Donated Pieces Taking the two distinctions together--the inside/outside distinction and the local/non-local distinction--it is clear that an effort was made to relate the donor, the architectural piece, and the placement of the inscription upon that architectural piece. This analysis has attempted to demonstrate that some type of meaning was associated with placement. For Balamitra and Ānanda's inscriptions on the southern gateway, the placement of their inscriptions inside the artistic field was also meaningful, as the inscribers, at this time, had some history in thinking about where inscriptions do and do not occur. 3.4 -- Gifting Power and Monastic Buddhists Looking closer at Balamitra's inscription, inside the Rāmagrāma stūpa representation on the southern gateway, some new important information is visible. Balamitra has the title    antevāsin meaning disciple, while his teacher has the two titles aya (noble) and dhamakathika (preacher of the Law).42 The titles listed in Table 3.11 are in addition to the standard designations of “monk” (bhikṣu) or “nun” (bhikṣuṇī) as they occur on the earlier ground balustrade.43 Title Meaning # of Donors Numbers in Marshall aya Noble 6 14444, 190, 209, 26545, 294, 303 antevāsin Pupil 11 52, 85, 118, 214, 229, 265, 267, 269, 270, 348, three hundred and forty nine sapurisa Good-man, or saint 1 28846 bhadata Reverend 2 102, 206 thera Elder 1 303 sutatika Versed in the sūtras 1 30447 pacanekayika Versed in the five scriptures 1 24248 2149 Table 3.11: Monastic Titles and Donors The first term, aya, is probably the Prakrit version of the Sanskrit title “arya,” simply meaning noble. Six donors from the ground balustrade have this title. Additionally, many donors call teachers or companions aya. At least 20 individuals at Sanchi have the title aya. One solitary man is simply just called “aya” in nos. 632, 634, 654, and he should probably be included in the group as well. Invariably, I considered an   42 This translation is according to Majumdar in J Marshall, The Monuments of Sanchi vol. 1 (Delhi: Swati Publications, 1982 reprint): p. 342. 43 Generally, there does not seem to be any gender distinction between these advanced titles. 44 Aya Pasanaka also donated crossbars in nos. 148 and 149. 45 Aya Kāṇa is himself referred to as a pupil of Aya Bhaṃduka. 46 It is very curious that the title sapurisa should be used in this situation. The only other instance at Sanchi when this title is used is to refer to the famous teachers whose relics are enshrined in stūpa no. 2, as well as at several of the subsidiary sites. It is very likely Bhadiya was a very esteemed individual, as he also possesses the title yugapaja, or “pathfinder of the age” according to Majumdar’s translation. It possible that yugapaja is not a title at all, but until a more apt context for its use is found I rely on Majumdar’s interpretation. 47 The nun Avisinā donated another crossbar, no. 305. She is the only known nun to have received any kind of honorific spiritual title aside from bhikkuni. 48 Devagiri is elsewhere referred to as an aya, but in no. 242, where he is called a pacanekayika, he is said to be accompanied by his pupils. 49 Excludes those such as Aya Kāṇa who have two or more titles.    individual with the title aya a monastic Buddhist, as well as their pupils. In eleven cases donors refer to themselves as antevāsin, or pupils. Usually they are pupils of teachers also possessing a title, whether it is aya or one of the others on the list, as with the case of Balamitra. The most interesting title that appears is sapurisa (Skt: satpuruṣa), meaning “saint” or “good man.” No. 288 refers to Bharadiya, who also has the title yugapaja. While yugapaja is an unclear term that Majumdar leaves untranslated50 (and I have as well), the title of sapurisa only occurs in one other context at Sanchi. Inscription nos. 3- 14 refer to reliquary caskets found inside of stūpas no. 2 and 3.51 Each of the inscriptions on the reliquaries from stūpa no. 2 refer to sapurisas whose relics are enshrined there. Inscription no. 3 reads: “(Relics of) sapurisa Kāsapagota, teacher of all the Hemavatas.” These titles refer to monastic luminaries of the region – early prominent teachers who were responsible for the widespread propagation of Buddhism in the region. The fact that Bharadiya, a solitary donor on the ground balustrade, possesses that title means that the tradition as started by the teachers enshrined in stūpa no. 2 was still thriving generations later. One of the teachers whose relics are enshrined at stūpa no. 2 is the aforementioned sapurisa Gotiputa. Gotiputa is the only one of those teachers who is referred to again through donors on the ground balustrade. Inscription no. 175 names a donor Subāhita, son of Gotiputa. In nos. 171, 172, and 173, Majhimā is mentioned as being the wife of Subāhita. Subāhita himself is part of the mercantile class, having the   50 Ibid., p. 328. 51 Ibid., p. 295.    title of “royal scribe.” Another donor, from no. 290, is Bhaṃduka, a local monk who donated an IRP.52 Bhaṃduka elsewhere is referred to as having the title aya53, and is the teacher of a number of other monks who were part of the donation network at stūpa no. 1. One of the luminaries from the early days at Sanchi, Gotiputa, continued his lineage through several generations. The number of monastic and lay Buddhists who are either his familial relatives or spiritual descendants is quite large, and could include the likes of Balamitra by the time the gateways were inscribed. While not within the scope of this thesis to elaborate upon Gotiputa’s monastic lineage, his lineage's success shows that elite monastic Buddhists may have possessed a significant, and specific, type of donor power. Aya Bhaṃduka, son of sapurisa Gotiputa, donated many times at Sanchi. Within the corpus of his donations, one inscription is inside a rail pillar (IRP) and another inscription on the inside of a copingstone (ICS). Bhaṃduka’s brother—or at least spiritual brother—Subāhita, brought with him his wife to donate at Sanchi. Subāhita also donated one rail pillar himself and his wife three crossbars. Just from these two members of Gotiputa’s lineage there are at least 12 (and very likely many more) donations, including some significant ones which warrant further study. My above brief discussion of one prominent monastic Buddhist lineage typifies the gifting power and tendency of the monastic community at Sanchi. Although the roster seems quite large, given the three hundred and forty nine individual donations, upon closer examination it becomes very clear that between the donors who have donated   52 In ins. no. 307, the same man gifted an ICS. 53 Ins. Nos. 265 and 267. Two of his pupils are Aya Kāṇa and Dhamadata, two monks, one of which himself possesses the prestigious aya title.     multiple times and the donors who are related to one another in some way or another, the community seems rather small. Similarly, on the gateways, only seven donative inscriptions remain, two of which are by Balamitra (nos. 399 and 402), who is almost assuredly the same man as he is in both instances referred to as the pupil of the same titled teacher, who is called “the Preacher of the Law.” Another two of the donative inscriptions are also by the same man, Nāgapiya, the sethin from Achāvaḍa (nos. 397 and 403). Just two donors are responsible for over half of the donative inscriptions on the gateways (four of seven total). The non-monastic donor, Nāgapiya, has the occupation of a sethin, or banker. From the ground balustrade, many of the sethin donors gave multiple large pieces, just like Nāgapiya. From the instances of Balamitra and Nāgapiya, it is evident that even at the time of the inscribing of the gateways that there appears to be a relationship between a donor’s position within the community and the type of gift given. 3.5 -- Conclusion In this chapter, I utilized a number of techniques to study the numerous donative inscriptions residing upon the ground balustrade at stūpa no. 1. I analyzed the architectural pieces themselves, the placement of the inscriptions upon the architectural pieces, the relationship between the inscriptions and the communities from which the donors come from, and, last but not least, the relationship between the architectural pieces, the donors, and the monastic community. The conclusions and insights I yielded from this large amount of data from the ground balustrade helps to begin to synthesize the relatively little inscriptional data that exists upon the Sanchi gateways. However, there is not a sufficient amount of testable     data resting upon the gateways. Therefore, I categorize the data from the ground balustrade into Types of donors. These types of monastic donors were grouped using the analyses outlined in this chapter. There are several criteria used in group creation. First, is the donor a monastic Buddhist? Second, does the donor possess a title? Is the donor local or non-local? And lastly, has the donor donated a railpillar, copingstone, or is the inscription located on the inside? Depending on how each donor fits these criteria, several groups emerge (Table 3.12). Type % of Monastic Community Criteria 1 .7% Monastic, Title, Local, Major Piece 2 2.2% Monastic, Title, Non-Local, Major Piece 3 2.2% Monastic, Title, Local 4 3% Monastic, Title, Non-Local 5 3.7% Monastic, Local, Major Piece 6 17% Monastic, Non-Local, Major Piece 7 7.4% Monastic, Local 8 26% Monastic, Non-Local Table 3.12: Types of Monastic Donors I applied the criteria to Balamitra's inscription. He fits all of the criteria of a Type 1 donor. Through this informative statistical study of all of the balustrade donative inscriptions, Balamitra is part of a group that is elite amongst the Buddhist community, a point only hypothesized when thinking about each of the anatomical parts of his inscription on their own. When taking each of the criteria together, it becomes apparent that only .7% of other donors are similar to him.   Chapter 4 – The Stūpas on the Gateways In chapter 2, I described the southern gateway of Sanchi stūpa no. 1. In this description, I outlined all a gateway's fundamental units: the two pillars, two capitals, three architraves, two ends of architraves, struts between the middle architrave and the top and bottom architraves, and the four dies separating each architrave. On the fundamental architectural units of the gateways are the 30 representations of stūpas, 29 of which are located above the ground balustrade, and one, on the northern side, below the gateway capitals. 12 of the 30 are on the top architraves, six are on architrave ends, and nine on dies (see Tables 2.1 and 2.2). Five of the 30 stūpas occur on the southern gateway, eight on the north, eight on the west, and finally nine on the east. Just two of these stūpas are inscribed. Both inscribed stūpas are on the outside face of southern gateway. In this chapter, I discuss the iconography of the 30 stūpas, exploring their various forms and the range of their forms on the four gateways. With the data I collected at Sanchi, I compare and contrast the inscribed stūpas on the southern gateway with the others 28 non-inscribed stūpas. I look closely at all 30 stūpas' physical features to determine similarities and differences. Through the description and subsequent analysis of the data, I aim to review where the two inscribed stūpas from the southern gateway fit within the corpus of stūpa representations at Sanchi and try to determine if the similarities and differences are significant. Before gathering this data, I generated several questions that I hope to be able to answer in my study. Do stūpas’ iconographic similarities fit the thematic similarities?   For instance, do all the stūpas in Mānushi Buddha scenes conform to a distinct iconographic style? Even more specific, do the inscribed stūpas exhibit any unique characteristics that set them apart from not only other stūpas of a similar kind, but from the other stūpas from other gateways? And from the southern gateway? Finally, the most important question regarding these inscribed stūpas, presuming their individuality: do the inscriptions make the stūpas unique or more significant, or do their iconographic features? Throughout my study, I also speculate whether or not the stūpas images were carved from a standard (or “cookie-cutter”) image and ask if there were different molds for different types of stūpas. My aim, at the very least, is to present and analyze new data in several different ways to begin to understand the image of a stūpa at Sanchi. 4.1 The Features of the Stūpas Six features of a stūpa representation are important to my study. Figure 4.1 shows each of the features and only the balustrade (vedika) is not labeled, but can be distinguished visually as the repeating row of railpillars between entrances marked by the gateways (toraṇa). I derive these features from ones appearing on large actual stūpas, such as Sanchi stūpa no. 1. The features appearing on the depicted relief stūpas directly mimic those found on their larger three-dimensional brethren.   Figure 4.11: Sections of a Stūpa The first feature is the balustrade. In the iconography of a stūpa image, the balustrade is similarly depicted to how it appears at Sanchi stūpa no. 1, with several crossbars connecting two rail pillars. On top, across the entire balustrade are the copingstones. In the representations of stūpas, the balustrade may either have two connecting crossbars between each pillar, or three, depending on the depiction or scene. Some of the stūpas also possess, like Sanchi stūpa no. 1, multiple balustrades, representing a ground circumambulatory path and an upper one. Usually the balustrade is the architectural element that separates the circumambulatory path from the other portions of the stūpa, but in the stūpa images it is doubtful a circumambulatory path was intended. Instead, it is a symbolic feature. The difference can be clearly seen when comparing the very large, three-tiered stūpa representation from the west pillar of the north gateway (#12 on the master list of stūpa representations, Appendices 1 and 2) with any other stūpa representations that have balustrades. In the aforementioned stūpa   1 Adapted from L Fogelin, Beyond the Monastery Walls, Dissertation. University of Michigan (2003): p. 13.   representation, there are numerous figures clearly inside the balustrade, although in the standard stūpa scene nobody ever stands within the balustrade.2 The next feature is the aṇḍa, or hemispherical “egg” portion of the stūpa. Aṇḍas from the time of Aśoka onwards were made of varying sizes of bricks. For stūpas containing relics, the aṇḍa encloses the reliquary containers as a protective shell. In some iconographic measurements later in this chapter the aṇḍa serves as the primary object of study, as it is large and easily measurable, thus an ideal candidate to determine size--and perhaps scale--of stūpa representations. On some stūpas, aṇḍas emerge from a raised platform, known as a drum. Drums appear to serve one primary function: to raise the aṇḍa above the ground. While each of the reconstructed stūpas at Sanchi have drums,3 it is unclear as to whether or not they originally had them. According to attending archaeologists at the site, Stūpa no. 3 was reconstructed to match the appearance of a stūpa on its gateway.4 On the gateways of stūpa no. 1, only a few of the stūpa images possess drums. Thus, because of the total reconstruction of numerous stūpas from piles of bricks, the frequency of actual drums built at open-air monastic sites such as Sanchi is up for debate. The use and significance of depicted stūpas to reconstruct the “real” stūpas remains an ongoing discussion that may be discussed in a future study. In the reliefs, drums are a decidedly unique feature used in only a few instances.   2 Actually in several of the stūpa scenes, figures are actually standing on top of the balustrade performing a devotional ritual. 3 Here I am referring to the ASI’s reconstruction of stūpas nos. 2 and 3 specifically. While the balustrade of stūpa no. 2 was largely found undamaged, none of the balustrade of stūpa no. 3 remained in-situ. Stūpa no. 3 is the only other stūpa at Sanchi that had a toraṇa, which stands in its original place even today. 4 Additionally, they have used this, and other stūpa images, as models for the reconstruction of many stūpas in the area, and not just no. 3.   Topping the aṇḍas of stūpas are two features: the harmikā, a small ornamental balustrade, and a chattra, or parasol. Depending on the type of harmikā, they usual surround the one or many chattras. A chattra sometimes occurred with three circular disks on top of each other. However, in the images of stūpas in the reliefs, chattras occur either as a singular disk, or as separate disks peaking from the harmikā at different locations. As many as five chattras occur in the Sanchi toraṇa figural stūpas. Finally, the last feature of a stūpa is the toraṇa. Toraṇas are arched gateways attached to the ground balustrade. At Sanchi, only stūpas nos. 1 and 3 have gateways. In the representations of stūpas, there is only one known toraṇa in the Sanchi reliefs. On the east-face of the west pillar on the north toraṇa at stūpa no. 1, there is a peculiar stūpa with three balustrades (an irregular occurrence by itself). At the center of the bottom balustrade is a lone toraṇa with two architraves. It is unlikely this represents anything at Sanchi because of the singular toraṇa and the only two architraves. At Sanchi, each toraṇa has three architraves. In the analysis below, I will not discuss the iconographic representations of toraṇas because there is only the singular occurrence in the Sanchi reliefs.5 Beyond these six features are several other quantifiable aspects used in my analysis. First, I measure the ratio of width and height. I measured each stūpa image and came up with a ratio number to represent the width to height. For height, I measured from the top of the aṇḍa to the bottom of the drum or bottom balustrade, as shown in Image 4.2. For width, I measured from end to end the widest part of each aṇḍa. The   5 At Mathura and Amaravati there are a number of relief images containing stūpas with toraṇas. However, a discussion of these images is beyond the scope of the present study.   final ratio is a simple equation: width divided by height (W/H). The resulting number shows one of three things about the represented stūpa. First, if the number is exactly one- -which, for most of the stūpas I measured it was--the width and height are nearly the same. However, for a number less than 1, the height is greater than the width. For instance, ratios of .5 shows that the height of the stūpa is double that of its width. The significance here shows that the stūpa is accentuated for unknown reasons.6 If the ratio is 1.5, then the width is double that of the height, indicating that the mass of the stūpa’s aṇḍa is being “implied” and appearing to be much more than it actually is to the naked eye.7 For this thesis, the ratios are only considered insofar as they are related to the stūpa’s iconographical relationship with other stūpas. Next feature I look at the number of balustrades per stūpa. Some stūpas have up to three balustrades while others have just one. The balustrades in these images may have two or three crossbars to every railpillar. The variability between the occurrences of two crossbars to that of three crossbars appears to be insignificant currently. I have not outlined it in any table. However, the number of balustrades per stūpa is related to the overall study in this chapter. Adornment of the stūpas in the images is a simple feature to compare, as nearly all the stūpas in the gateway relief art possess decorations of some sort. Usually these decorations are garlands placed, or in the act of being placed, on the drum, by   6 L Fogelin, "Material Practice and the Semiotic Metamorphosis of a Sign: Early Buddhist Stupas and the Origin of Mahayana Buddhism," Unpublished Manuscript, 2009, will discuss the terms “accentuation” and “implied mass” in great detail. He argues that the creators of such stūpas were attempting to create visual tricks, or illusions, to imply that the some stūpa are larger than in reality. Fogelin uses evidence from rock- cut stūpas from the Western Deccan to argue his point. 7 Ibid., pp. 24-6. worshippers. Other decorations may include unique patterns traced into the face of the drum or aṇḍa, or drapery dangling from the garlands draped around the width of the aṇḍa. In some of the to worship at the stūpa place garlands omission of ornamentation on any of the this is an artistic choice or if there was something unique about the scenes on the southern gateway. On the two inscribed of each stūpa, it may be ar Figure 4.2 At the tops of the stū chattras are typically reserved for divine figures and kings, as depicting kings or aniconic representations of the Buddha, such as a  Adapted from L Fogelin, Beyond the Monastery Walls 13.   chattras. The main ornamentation aṇḍas. They dip back and forth, up and down across the scenes, mythological flying figures or people coming on the aṇḍa. Amongst the stūpa stūpas on the southern gateway. It is unclear if stūpas, because the inscription is carved across the middle tistically impossible to display a garland. : Aṇḍa Measurements of Stūpa Images8 pa images, I quantify the number of chattra shown in other reliefs Bodhi Tree , Dissertation. University of Michigan (2003): p. s are the s, there is an odd s. Because , the   number of chattras is likely significant. One theory is that the more chattras there are, the holier the site or person whose relics are enshrined in the stūpa. However, the relationship between the number of chattras and the apparent significance of the stūpa within the scene is unclear. The number of chattras may be left to the artist. Alternatively, it could be a random insertion that depends on unavoidable environmental factors, such as space on the architrave or die. For those stūpas that do possess a drum, and hence multiple balustrades, the height of the drum is measurable. The height of each drum is almost uniformly nearly half that of the total height of the entire stūpa. Therefore, the height of the drum is, at Sanchi, an artistic constant. No other feature is as uniform as the drum height. The last quantifiable element is scale. In all the renderings of stūpas, the stūpas do not appear alone but are shown with trees, mythological creatures or people. In 29 of the 30 stūpas, at least one person is of measurable height to the stūpa. Comparing the stūpas’ heights to the people in the scenes, there are four possible scales: 1.) the height of the stūpa equals the height of the person(s); 2.) the stūpa is the about height of two persons; 3.) the stūpa is the height of approximately three-fourths of a person; 4.) lastly, the stūpa may be up to the height up a person and one-fourth the height of another person. Tables 4.2-3 include this measurement. The lone stūpa that does not have a comparable person in its scene has a height relatable to the height of an elephant. On the inside of the east toraṇa a number of elephants worship a single stūpa. This stūpa is about the height of one elephant.    Together, these quantifiable aspects of the stūpas allow me to compare and contrast each of the 30 stūpas with one another and look for patterns of consistency or lack thereof. Through these comparisons, I suggest several ways to read certain stūpas based on the quantifiable criteria outlined above. 4.2 Types of Stūpas From a close study of the content outlined above, three thematic categories emerged from the 30 scenes with stūpas. Stūpas in Mānushi Buddha scenes exemplify the first type. Table 2.3 from the second chapter outlines where and how many times the Mānushi Buddha scenes occur. All the Mānushi Buddha scenes at Sanchi with stūpas are on the outer face of the top architraves of each of the four toraṇas. A proper Mānushi Buddha scene must contain seven elements, a combination of trees and stūpas. The Mānushi Buddhas are the earthly Buddhas, six of which are the immediate predecessors to Gautama. The scenes may consist of alternating elements, such as tree, stūpa, tree, etc., or be exclusively all trees. One clear example from the southern gateway (Image 4.1) shows the different elements, the alternating pattern of tree, stūpa, tree, etc. Between each element, though, are little scenes consisting of devotees who flank the trees or stūpas. The devotees come in different heights, although some may be mythological figures while others are perhaps merely human. Each figure is doing a unique activity in the little scene--some are folding their hands and others are folding their hands in reverence. Others perform different kinds of actions depending on their relationship with the tree or stūpa. In total, there are six Mānushi Buddha scenes containing 16 stūpas in those scenes.    Image 4.1: Top Architrave with the Mānushi Buddha Theme The parinirvāṇa stūpa is the next identifiable type (Image 4.2). There are three signs closely related to important biographical events in the Buddha’s life: the bodhivṛkṣa with the Enlightenment, the cakra with the First Sermon, and the stūpa for the parinirvāṇa.9 These three signs are often represented together in certain contexts and many scholars interpret them as aniconic signs of the Buddha, referencing his life.10 Although Susan Huntington formulated a theory of reenactments11, Linrothe wonders why nobody plays the character of the Buddha in these reenactments.12 Vidya Dehejia stresses multivalency, emphasizing multiple possible interpretations for a single image.13 At Sanchi, context may be used to deduce parinirvāṇa scenes, where in several locations along the gateways, such as the outside face of the south gateway architraves, there is the consistent and well-known pattern of signs referencing the Buddha’s biography. One such stūpa from the south gateway, appears on the die between the middle and lower   9 K Karlsson, Face to Face with the Absent Buddha (Uppsala: Uppsala University Press, 1999): p. 174. 10 See ibid., pp. 174-186 for a lengthy discussion of aniconism and stūpas. 11 Susan Huntington has challenged the frequency of aniconism. In S L Huntington, “Early Buddhist Art and the Theory of Aniconism,” in Art Journal 49 (1990), pp. 401-8 she proposes that the symbols in question depict later worship at Buddhist sites and the beginnings of pilgrimage to those sacred sites. She earlier mentioned this concept in S L Huntington, The Art of Ancient India (Boston: Weatherhill, 1985). 12 R Linrothe “Inquiries into the Origin of the Buddha Image: A Review,” East and West 43 (1993): p. 249: “why is it, if there was no disinclination to represent the Buddha, that no one plays the part of the Buddha himself [in the pageants]?” 13 V Dehejia “Aniconism and the Multivalence of Emblems,” Ars Orientalis 21 (1991): p. 45. architraves on the west side. On the direct opposite die is a birth scene. that, between the middle and top architraves, is a enlightenment of Gautama. Directly below the dharmacakra on a pillar, symbolizing the First Sermon at Sarnath. pattern, it seems likely that this the same contextual criteria, I identified at least three such parinirvāṇa at Kuśinagara . Image 4. The last thematic type consists of legendary or historical scenes from Buddhist literature. These scenes typically take place after the related to the dissemination of Buddhism or important turning points in Buddhist history. The first and most important of these scenes is the previously described scene (Image 2.7 from Chapter 2) toraṇa, it seems the character of  Bodhi Tree symbolizing the stūpa, below the false capital, is a Fitting the contextual stūpa could be a representation of the parinirv stūpas that could . 2: Parinirvāṇa Themed Panel parinirvāṇa of the Buddha and are . On the middle architrave, outside face of the south Aśoka approaches the Rāmagrāma stūpa On the die above āṇa. Using represent the Rāmagrāma with the intention of opening it to obtain relics for the sake of splitting them up to create more stūpas. Of course, as aforementioned in chapter 2, worshipped the stūpa and preve Image 4.3, may not be identifiable, but are very likely scenes known to the creators of the image, as they would have been drawing Image 4.3: Another identifiable legendary scene is found on the west pillar, upper panel on the inner face of the northern toraṇa, is exceptionally unique in many ways. It is also the only the false capitals of the gateways. Marshall representing the Mallas of parinirvāṇa. While Marshall and Foucher’s interpretation may be disputed, the intention is clear—it is some scene, either known or unknown, taking place after the Buddha’s death and is part of the subsequent history of Buddhism.  14 J Marshall, The Monuments of Sanchi vol. 1 15 L Fogelin, "Ritual and Presentation in Early Buddhist Religious Architecture," in 42.1 (2003): p. 147. 16 The identification of this stūpa involve the stūpa at Kuśinagara, is due to its grand scale and disassociation with the surrounding panels.  nāgas and nāginīs guarded and nted Aśoka’s intended actions. Some legend scenes, as in on some inspiration. Bottom Architrave Depicting a Legend Scene toraṇa. The stūpa here, with three balustrade stūpa 14 , and others15, identify the scene as Kuśinagara, honoring the stūpa of the Buddha after his 16 Other stūpa (Delhi: Swati Publications, 1982 reprint) Asian Perspectives as a “legend scene” and not as a parinirvāṇa scene, even though it might s and a to appear below s in other scenes : pl. 8. , may not be identified as part of any known scene, such as the inside face of the lower architrave on the east tora categorized that scene as a “legend scene." The last thematic group consi as parinirvāṇa scenes, as they are not contextually part of the Buddha’s biography, and are not similarly on the same grand scale as the previously mentioned legend scenes. These “unidentifiable” (Im as being either parinirvāṇa or an iconographic ambiguity, 30 of the stūpas along with their location and thematic scene identification. Image 4.4: An  All three of the parinirvāṇa scenes that I have highlighted in this chapter are all contextually part of a large thematic scheme involving several other depictions from the Buddha’s biography. This particular scene stands alone and is truly unique in its  ṇa, where the elephants are paying homage to the sts of stūpas in scenes that are not easily categorized age 4.4) scenes are seven in number and could be scenes or legend scenes, but because of a thematic ambiguity they are not classifiable like the others. Unidentifiable Scene from the East Gateway  size, iconography, and placement. stūpa. I also categorized Table 4.1 lists all    Location Scene KEY 1 SO, T, ML Mānushi S South 2 SO, T, MC Mānushi N North 3 SO, T, MR Mānushi E East 4 SO, M, C Legend scene W West 5 SO, M, L Parinirvāṇa O Outside 6 NO, T, L Mānushi I Inside 7 NO, T, ML Mānushi T Top arch. 8 NO, T, MC Mānushi M Mid arch. 9 NO, T, MR Mānushi B Bottom arch. 10 NO, T, R Mānushi P Pillar 11 NI, M, R Unidentifiable L Left side 12 NI, P, L Legend scene C Center 13 NI, P, R Parinirvāṇa R Right side 14 EO, T, L Mānushi M Middle 15 EO, T, ML Mānushi ML Mid left 16 EO, T, MC Mānushi MC Mid center 17 EO, T, MR Mānushi MR Mid right 18 EO, T, R Mānushi 19 EI, T, L Unidentifiable 20 EI, T, R Unidentifiable 21 EI, M, C Legend scene 22 EI, P, R Legend scene 23 WO, T, ML Mānushi 24 WO, T, MC Mānushi 25 WO, T, MR Mānushi 26 WO, B, L Unidentifiable 27 WO, B, R Unidentifiable 28 WI, T, L Unidentifiable 29 WI, T, R Unidentifiable 30 WI, M, R Parinirvāṇa Table 4.1: The Location of the 30 Stūpa Images I now turn to a detailed comparison of the features of each thematic type. Table 4.2 lists all the stūpas from the Mānushi Buddha scenes and their accompanying quantifiable features. Each column uses the previously described markers. At the bottom of each column, where relevant, I have included the average number for that criterion. A   comparison of Tables 4.2-4 shows that the averages, where indicated, are important factors in the different types of thematic stūpas. Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 1 SO, T, ML Mānushi 1 2 No 1 0.57 3/4 person 2 SO, T, MC Mānushi 1 2 No 5 0.53 1 person 3 SO, T, MR Mānushi 0.88 2 No 1 0.48 3/4 person 6 NO, T, L Mānushi 1.11 1 Yes 1 No drum 1 person 7 NO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 8 NO, T, MC Mānushi 1 1 Yes 1 No drum 1 person 9 NO, T, MR Mānushi 1 1 Yes 1 No drum 1 person 10 NO, T, R Mānushi 1 1 Yes 1 No drum 1 person 14 EO, T, L Mānushi 1 1 Yes 5 No drum 1 person 15 EO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 16 EO, T, MC Mānushi 1.17 1 Yes 1 No drum 1 person 17 EO, T, MR Mānushi 1.2 1 Yes 1 No drum 1 person 18 EO, T, R Mānushi 1.19 1 Yes 4 No drum 1 person 23 WO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 24 WO, T, MC Mānushi 1 1 Yes 1 No drum 1 person 25 WO, T, MR Mānushi 1 1 Yes ? No drum 1 person AVG 16.55 AVG 1.19 AVG 1.625 Table 4.2: Mānushi Themed Stūpa Images The 16 Mānushi Buddha stūpas express a great amount of standardization among some of their features. For example, the width to height ratio is nearly uniformly around 1, with only a few exceptions. This means that the width and height of the stūpas are   almost exactly the same. Similarly, the scale of the scenes are uniformly nearly the height of one person in each scene. Scale in the small scenes within the Mānushi Buddha architraves appears to be relative. For example, figures approaching the tree part of each scene are, in general, smaller in scale than figures approaching the stūpas. The discontinuity of scale between the small scenes within the larger, architrave-wide Mānushi scene shows the complexity in a single themed architrave. Comparing the multi-scaled Mānushi Buddha scenes with the other types of scenes, in other scenes the scale is uniform within the whole architrave (legend scenes) or die (parinirvāṇa scenes). This different by itself could warrant a future study of the Mānushi Buddha architraves. Table 4.3: Parinirvāṇa Themed Stūpa Images Looking at the other features of Mānushi Buddha scenes, the number of balustrades is also consistently just one per stūpa. The greatest deviation from these standards is the south toraṇa Mānushi Buddha scenes, which are different in a variety of ways. As I discussed in chapter 2, the south gateway has a distinct expression not seen elsewhere at stūpa no. 1. One possible interpretation of the standardization of the Mānushi Buddha scenes is that they were constructed, or at least designed, in the same instance. Their uniformity, from these iconographic characteristics and their consistent Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 5 SO, M, L Parinirvāṇa 1.28 1 No 5 No drum 1 person 13 NI, P, R Parinirvāṇa 1.25 0 Yes 3 No drum 1 person 30 WI, M, R Parinirvāṇa 1 1 Yes 3 No drum 1 person AVG 1.18 AVG .66 AVG 3.67   placement on the outside face of the top architraves shows that these stūpas were possibly planned and designed together. Table 4.3 shows the three parinirvāṇa stūpas. The location of the parinirvāṇa scenes are all on dies, not architraves. Comparing the placement of the different types of scenes, the parinirvāṇa scenes on dies are limited in the amount of depicted activity. These scenes are self-contained and a direct contrast to Mānushi Buddha scenes that have several small scenes included within the long architrave. The specific features of parinirvāṇa stūpas tend to show some degree of uniformity, similar to the Mānushi Buddha scenes. Their width to height ratio is consistently above one, meaning that their widths are always greater than their heights. Comparing these ratios to the Mānushi Buddha ratios, there is a marked contrast amongst the physical dimensions of the stūpas. The number of chattras is also a visible point of difference and distinctiveness. The average number of chattras among these three parinirvāṇa stūpas is 3.67, which is far above any of the other averages in any of the other groups. Between the width to height ratio and the average number of chattras, the parinirvāṇa scenes exhibit a certain degree of iconographic conformity as well. Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 4 SO, M, C Legend scene 0.88 2 No 1 0.56 2 persons 12 NI, P, L Legend scene 0.62 3 Yes 3 0.56 2 persons 21 EI, M, C Legend scene 1 2 Yes 1 0.47 1 elephant 22 EI, P, R Legend scene 1 1 Yes 0 0.38 1.25 person AVG .875 AVG 2 AVG 1.25 Table 4.4: Historical Themed Stūpa Images   Stūpas represented in relief scenes showing historical events known from legends after the parinirvāṇa are found on a variety of gateway locations. The Rāmagrāma stūpa scene is on the middle architrave of the southern gateway and utilizes nearly every available space to narrate its events. Similarly, the stūpa scene with elephants from the eastern gateway narrates the journey of the elephants to the stūpa, carrying flower and branches to the stūpa as offerings with their trunks. Both architrave legend scenes show movement among the characters, all migrating towards the center of the architrave where the stūpa rests. Another legend scene, on the pillar from the north gateway, utilizes much space to illustrate not only the worship of a stūpa, like the other scenes, but the elaborate form of the central stūpa, which has multiple balustrades and even a two-arched toraṇa. This panel on the side of the pillar makes use of enough space to compare to any two other pillar reliefs. A third legend scene is located on a die from the eastern gateway. It's composition and use of space compares to parinirvāṇa, however its stūpa is stylistically similar to the others in legend scenes. Like the previous two types, the legend scene type displays its own unique iconographic characteristics (Table 4.4). First, on average, the width to height ratio is below one. Generally, these stūpas are taller than they are wide and have accentuated peaks to highlight their prominence. The average number of balustrades on these stūpas is also the highest amongst all the types of stūpas, similarly demonstrating the artists’ desire to accentuate height over width. This group also is the only group where each of the stūpas possesses a drum. Drums, as I previously described, hold the aṇḍa up. Again, the intention is to highlight the height of the stūpas. Looking at the scale of these stūpas,   the scale is always above one person, presuming that the height of an elephant, as in stūpa #21, is greater than the height of a person. In each iconographic way possible, these stūpas from the legends of Buddhism are portrayed as taller, and presumably, larger than any of the other stūpas in any of the other scenes. Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 11 NI, M, R Unidentifiable 0.715 1 Yes 1 0.52 1 person 19 EI, T, L Unidentifiable 0.75 1 Yes 1 0.51 1 person 20 EI, T, R Unidentifiable 0.91 1 Yes 1 0.41 1 person 26 WO, B, L Unidentifiable 1 1 Yes 1 No drum 1.25 person 27 WO, B, R Unidentifiable 1 1 Yes 1 No drum 1.25 person 28 WI, T, L Unidentifiable 1 1 Yes 1 No drum 1 person 29 WI, T, R Unidentifiable 0.9 1 Yes 1 No drum 1 person AVG .9 AVG 1 AVG 1 Table 4.5: Unidentifiable Stūpa Images The last group, the unidentifiable group of seven, does exhibit distinct trends amongst their iconographic traits. They appear on architrave ends and dies, as individual, unconnected panels. The width to height ratio of the stūpas is just below one on average, showing they are usually slightly taller than they are wide, a characteristic common of the legend scenes. However, unlike the legend scenes, this group averages only one balustrade per stūpa, a strong difference from the legend scenes, but yet a strong similarity to the parinirvāṇa scenes. Nevertheless, the contrast between the average number of chattras, which is just one in this group, is a strong disparity from the   parinirvāṇa scenes, which averages nearly four chattras per stūpa. Additionally, the scale of one person may fit either the parinirvāṇa scenes or the Mānushi Buddha scenes. Without any more evidence, it is impossible to place any of these stūpas in other categories based on their iconographic consistencies or differences. Even stūpa nos. 26 and 27, with scales greater than one person, do not fit the mold of legend scene stūpas because they do not have drums or multiple balustrades. The width to height ratios on average is below one, which is also a strong divergence from the parinirvāṇa scenes. In the end, this unidentifiable iconographic type of stūpa, though consistent with one another in many ways, does not clearly fit into any of the other types. Therefore, they could represent a wholly other type of stūpa that we are thematically unable to identify accurately or imagine currently. 4.3 Conclusion The purpose of this chapter has been to outline several iconographic characteristics of representations of stūpas on the Sanchi toraṇas. Beyond describing and analyzing these characteristics, where do the two inscribed stūpas from the south toraṇa fit into this data? To begin, the Mānushi Buddha stūpa of Ānanda, #2 in the Tables, stylistically conforms to the stūpas of other Mānushi Buddha stūpas, possibly demonstrating that Ānanda, along with whomsoever may have planned and carved the architrave, wanted to create a stereotypical Mānushi Buddha stūpa. Alternatively, considering that the southern gateway was the first of the four, Ānanda's Mānushi Buddha stūpa may have been amongst the first of its kind, thus setting the standard of Mānushi Buddha stūpas on the Sanchi architraves. Within the cluster of Mānushi   Buddha stūpas on the top architraves, Ānanda's stūpa has a nearly perfect width to height ratio of one and a scale of one person. Both quantifiable features are consistent with the group of Mānushi Buddha stūpas as a whole. The only noteworthy difference is that Ānanda's stūpa does not possess any sort of adornment. This may be explained by the fact that none of the stūpas on the south toraṇa have any decoration.17 One possibility is that because of the two stūpas’ inscriptions, no ornamentation was allowed on the aṇḍa of the stūpas. Consequently, to unify all the features of the south toraṇa’s stūpas, none of them possessed any garlands or decorations.18 The other inscribed stūpa, #4 in the Tables, also fits into the iconographic stereotype of legend scenes. The inscribed Rāmagrāma stūpa, a gift of Balamitra, is taller than it is wide, has multiple balustrades, a single chattra, a drum, and is two persons high in its scale—all established stylistic norms of this type of stūpa. The only remarkable aspect of this stūpa is its inscription. Unfortunately, comparing the iconographic characteristics does not shed much light on to the fact that it is inscribed and the others are not. Returning to the questions brought up in the first section of this chapter, I have provided some evidence to suggest that there does exist an iconographic relationship within the different thematic types of stūpas. Each thematic type of stūpa is visibly   17 If true, this explanation would shift all of the focus on the central inscribed stūpa of Ānanda. Because the inscription takes up the entire face of the stūpa, no garland was possible. All other stūpa representations would thus conform to this appearance, only without the inscription. Other than its centrality to the architrave and the inscription, this would be an extraordinary amount of emphasis placed on one stūpa. 18 Other reasons are possible as well. For instance, the supervisor of the south gateway may have wanted to make the southern gateway different, and therefore the deliberate omission of garlands in this way is a unique choice.   different the others, given a careful eye. However, subtle differences among stūpas of the same type show that there was not a “cookie-cutter” stūpa from which all stūpas were artistically conceived and subsequently carved. Different types of stūpas were noticeably intentionally different and were constructed as such. Thinking about each of the types of stūpas and their relationships with the gateways, the only apparent connection within the data is that the stūpas from the south gateway are slightly more unique with their iconography. None of them are decorated, all but one have drums (a rare feature otherwise), and all but one have two balustrades. The appearance of the two donative inscriptions on this unique set of stūpas enhances their appearance because the writing in the center of the aṇḍas functions almost as an ornamentation itself—a written decoration perhaps—and gives these two stūpas a special place within the corpus of donated architecture and art at Sanchi.    Chapter 5 -- Conclusion Throughout this thesis, I studied several inscribed reliefs from Sanchi's southern gateway. In my research, I accumulated two bodies of data to which I asked research questions. First, I looked at the images on which donative inscriptions appear. Two images are depicted on the south gateway's top and middle architraves. In these scenes, at the center of each architrave is an image of a stūpa. The top scene shows three stūpas and four trees, all associated with the seven earthly (Mānushi) Buddhas. In the center of this scene is a stūpa donated by Ānanda, the son of the foreman of the artisans of king Śātakarṇī. The scene on the middle architrave shows what is perhaps the Rāmagrāma stūpa being approached by Emperor Aśoka. The name of the monk Balamitra was inscribed on the face of this stūpa, exactly the same as the stūpa donated by Ānanda. Both the Rāmagrāma stūpa of Balamitra and the Mānushi Buddha stūpa of Ānanda are unique artistic occurrences at Sanchi because they have donative inscriptions in the center of their visual fields. To obtain a better understanding of the stūpa images, I compared these two representations of stūpas to the other 28 that appear on the four gateways. The 349 donative inscriptions located on the ground balustrade of Sanchi stūpa no. 1 was the second body of data I examined. These inscriptions provided a rich source of information regarding the epigraphical habit of donors during the 1st century B.C.E. In this large amount of information, I found patterns of donation by analyzing the basic units of a donative inscription, such as a donor's name, occupation, village, or familial or monastic lineage. Beyond the written text of the inscription, I studied where the inscriptions were recorded on the ground balustrade and if there was any significance to    their placement. Joining the written textual data of the inscriptions with the physical placement of the inscriptions, I discovered several trends which were informative regarding donor habits at Sanchi. I then applied the donor habits from the ground balustrade to the two inscribed stūpa images found on the southern gateway. The donative inscriptions of the two donors of the stūpa images, Balamitra and Ānanda, exhibited several qualities that fit into the donor patterns from the ground balustrade. I emphasized several important aspects of the inscribed stūpas images. First, the two scenes where the inscribed stūpas images appear are unique in that they have no counterpart on the other three gateways. The Mānushi Buddha scene from the top architrave is similar to the other Mānushi Buddha scenes, but does not conform to the exact compositional form set by the other scenes. Balamitra's inscribed Rāmagrāma stūpa similarly contains a donative inscription within its visual field and is a solitary representation of that scene at Sanchi. In both cases, the inscribed stūpa appears visually significant, as the inscription seems to function as a type of visual, or written, decoration, clearly distinguishing these two stūpas from the other 28. From the content of Balamitra's inscription on the face of the Rāmagrāma stūpa, I matched him with a particular type of donor occurring at Sanchi. Balamitra exhibits the qualities associated with only .7% of the total donors at Sanchi. An examination of both the visual text and the verbal text enhances our general understanding of Ānanda and Balamitra. Looking at either of the texts alone is not enough, as they were not intended to be seen separately, but rather as one image containing two parts. Thinking about    Ānanda and Balamitra's distinct presences and their link to the southern gateway's relief art may lead to a better comprehension of gateways and gateway art more generally. From my analyses, I suggest several conclusions. One, that the features of the inscribed stūpa images are unique, if not very original. Two, the location of the inscribed stūpa images is significant. And three, the donative inscriptions' content and placement within the visual field infers a connection between visual and verbal texts. In the same vein, from these conclusions, I argue that two different epigraphical habits reveal themselves. The first, appearing on the ground balustrade, displays a mass collective patronage of monastic Buddhists, laypersons, members of the mercantile community, and others. The names of these donors appear in brief, often-unseen inscriptions on functional architectural pieces such as crossbars, railpillars, and copingstones. In contrast, Balamitra and Ānanda donative inscriptions epitomize a custom of inscribing names within donated relief imagery. Donations were recorded as part of the visual field, showing the importance of not only inscribing a name but also the importance of inscribing a name for the purpose of being seen. With this observation, many future directions are possible for research. In the beginning of this thesis, I set out to answer several questions. Although I am able to present some data to infer a connection between the images and their donative inscriptions, more data would go a long way in checking my conclusions on a broader scale. To gather more data, I might analyze Bharhut with its cache of donative inscriptions, relief images, and inscriptional labels. Not only are the sites similar in arrangement--with balustrades encircling a major stūpa--they are also located within the    same general region. A simple question to begin with might be: what does Bharhut add to my discussion of inscribed reliefs at Sanchi? Does it modify my conclusions or support my early findings? One way to approach these questions would be to apply them to a series of separate studies. For instance, at first it may be useful to study only the donative inscriptions together and then only the relief art together. After exploring the fundamental similarities and differences of the two groups it may be possible to read them together, focusing on the points of intersection between the visual and verbal texts, similar to my method in this thesis. Utilizing this avenue in future studies may offer valuable insight into the process of inscribing names onto images and architecture. Returning to the large questions I asked in Chapter 1, how do these two inscribed stūpas from the southern gateway fit into the Buddhist gateway art at Sanchi? And, what is their significance in the development of artistic expressions of Buddhism during the ancient period? There is still much to explore at Sanchi and outside Sanchi. My brief study of stūpa relief images attempted to find a grammar of representation through the analysis of several fundamental units. That is, the architectural context of these inscribed reliefs shows that the location of the images is important, as they consistently appear in only several spots, above the ground balustrade. The features of a stūpa image are also important, as, when read together in several ways, the style of individual stūpa image indicates much about the scene where the stūpa image occurs. Lastly, these two aspects of the visual text provide a starting point from where to read the verbal text. On these inscribed relief images, it is impossible to read the visual text without reading the verbal text, as they each have been written together to make a stūpa image.   Appendix 1:All Thirty Stūpas and their Features Discussed in Chapter 4 Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 1 SO, T, ML Mānushi 1 2 No 1 0.57 3/4 person 2 SO, T, MC Mānushi 1 2 No 5 0.53 1 person 3 SO, T, MR Mānushi 0.88 2 No 1 0.48 3/4 person 4 SO, M, C Legend scene 0.88 2 No 1 0.56 2 persons 5 SO, M, L Parinirvāṇa 1.28 1 No 5 No drum 1 person 6 NO, T, L Mānushi 1.11 1 Yes 1 No drum 1 person 7 NO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 8 NO, T, MC Mānushi 1 1 Yes 1 No drum 1 person 9 NO, T, MR Mānushi 1 1 Yes 1 No drum 1 person 10 NO, T, R Mānushi 1 1 Yes 1 No drum 1 person 11 NI, M, R Unidentifiable 0.715 1 Yes 1 0.52 1 person 12 NI, P, L Legend scene 0.62 3 Yes 3 0.56 2 persons 13 NI, P, R Parinirvāṇa 1.25 0 Yes 3 No drum 1 person 14 EO, T, L Mānushi 1 1 Yes 5 No drum 1 person 15 EO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 16 EO, T, MC Mānushi 1.17 1 Yes 1 No drum 1 person 17 EO, T, MR Mānushi 1.2 1 Yes 1 No drum 1 person 18 EO, T, R Mānushi 1.19 1 Yes 4 No drum 1 person 19 EI, T, L Unidentifiable 0.75 1 Yes 1 0.51 1 person 20 EI, T, R Unidentifiable 0.91 1 Yes 1 0.41 1 person 21 EI, M, C Legend scene 1 2 Yes 1 0.47 1 elephant 22 EI, P, R Legend scene 1 1 Yes 0 0.38 1.25 person 23 WO, T, ML Mānushi 1 1 Yes 1 No drum 1 person 24 WO, T, MC Mānushi 1 1 Yes 1 No drum 1 person 25 WO, T, MR Mānushi 1 1 Yes ? No drum 1 person 26 WO, L, L Unidentifiable 1 1 Yes 1 No drum 1.25 person 27 WO, L, R Unidentifiable 1 1 Yes 1 No drum 1.25 person 28 WI, T, L Unidentifiable 1 1 Yes 1 No drum 1 person 29 WI, T, R Unidentifiable 0.9 1 Yes 1 No drum 1 person WI, M, R Parinirvāṇa 1 1 Yes 3 No drum 1 person  Appendix 2: Location Scene Width to Height Ratio 1 SO, T, ML Mānushi 1  1 The stūpas are in order as they appear in the table in Appendix 1.  Stūpas on the Four Gateways Stūpa #1 Balustrade # Decoration Chattra # Drum Height 2 No 1 0.57  1 Scale 3/4 person  Location Scene 2 SO, T, MC Mānushi Inscription: 1 rāño sirisātakaṇisa 2 āvesanisa Vāsiṭhīputrasa 3 anaṃda dānaṃ Translation: “The gift of Anaṃda, son of V Śātakarṇī.”  Stūpa #2 Width to Height Ratio Balustrade # Decoration Chattra # 1 2 No 5 āsiṭhī, the foreman of the artisans of the  Drum Height Scale 0.53 1 person rājan ṣirī  Location Scene Width to Height Ratio 3 SO, T, MR Mānushi 0.88  Stūpa #3 Balustrade # Decoration Chattra # 2 No 1  Drum Height Scale 0.48 3/4 person  Location Scene 4 SO, M, C Legend scene Inscription:2 1 aya-cūḍasa dhamakathikasa 2 atevāsino balamitrasa d Translation: “The gift of Balamitra, a pupil of the Preacher of the Law Aya  2 The inscription is not readable in this photograph.  Stūpa #4 Width to Height Ratio Balustrade # Decoration Chattra # 0.88 2 No 1 ānaṃ -Cūḍa.”  Drum Height Scale 0.56 2 persons  Location Scene 5 SO, M, L Parinirvāṇa  Stūpa #5 Width to Height Ratio Balustrade # Decoration Chattra 1.28 1 No 5  # Drum Height Scale No drum 1 person  Location Scene 6 NO, T, L Mānushi  Stūpa #6 Width to Height Ratio Balustrade # Decoration Chattra # 1.11 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 7 NO, T, ML Mānushi  Stūpa #7 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 8 NO, T, MC Mānushi   Stūpa #8 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1 person     Stūpa #9 Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 9 NO, T, MR Mānushi 1 1 Yes 1 No drum 1 person  Location Scene 10 NO, T, R Mānushi   Stūpa #10 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 11 NI, M, R Unidentifiable  Stūpa #11 Width to Height Ratio Balustrade # Decoration Chattra # 0.715 1 Yes 1  Drum Height Scale 0.52 1 person    Stūpa #12 Location Scene Width to Height Ratio Balustrade # Decoration Chattra # Drum Height Scale 12 NI, P, L Legend scene 0.62 3 Yes 3 0.56 2 persons  Location Scene 13 NI, P, R Parinirvāṇa  Stūpa #13 Width to Height Ratio Balustrade # Decoration Chattra # 1.25 0 Yes 3  Drum Height Scale No drum 1 person  Location Scene Width to Height Ratio 14 EO, T, L Mānushi  Stūpa #14 Balustrade # Decoration Chattra # Drum Height 1 1 Yes 5 No drum  Scale 1 person  Location Scene 15 EO, T, ML Mānushi  Stūpa #15 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 16 EO, T, MC Mānushi  Stūpa #16 Width to Height Ratio Balustrade # Decoration Chattra # 1.17 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 17 EO, T, MR Mānushi  Stūpa #17 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1.2 1 Yes 1 No drum  Scale 1 person  Location Scene Width to Height Ratio 18 EO, T, R Mānushi 1.19   Stūpa #18 Balustrade # Decoration Chattra # Drum Height 1 Yes 4 No drum  Scale 1 person  Location Scene 19 EI, T, L Unidentifiable   Stūpa #19 Width to Height Ratio Balustrade # Decoration Chattra # 0.75 1 Yes 1  Drum Height Scale 0.51 1 person  Location Scene 20 EI, T, R Unidentifiable   Stūpa #20 Width to Height Ratio Balustrade # Decoration Chattra # 0.91 1 Yes 1  Drum Height Scale 0.41 1 person  Location Scene 21 EI, M, C Legend scene  Stūpa #21 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 2 Yes 1 0.47  Scale 1 elephant  Location Scene 22 EI, P, R Legend scene  Stūpa #22 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes 0 0.38  Scale 1.25 person  Location Scene 23 WO, T, ML Mānushi  Stūpa #23 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes 1 No drum  Scale 1 person  Location Scene 24 WO, T, MC Mānushi  Stūpa #24 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes 1 No drum  Scale 1 person  Location Scene 25 WO, T, MR Mānushi  Stūpa #25 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes ? No drum  Scale 1 person  Location Scene 26 WO, L, L Unidentifiable  Stūpa #26 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1.25 person  Location Scene 27 WO, L, R Unidentifiable  Stūpa #27 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes 1 No drum  Scale 1.25 person  Location Scene 28 WI, T, L Unidentifiable   Stūpa #28 Width to Height Ratio Balustrade # Decoration Chattra # 1 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 29 WI, T, R Unidentifiable   Stūpa #29 Width to Height Ratio Balustrade # Decoration Chattra # 0.9 1 Yes 1  Drum Height Scale No drum 1 person  Location Scene 30 WI, M, R Parinirvāṇa   Stūpa #30 Width to Height Ratio Balustrade # Decoration Chattra # Drum Height 1 1 Yes 3 No drum  Scale 1 person 127 Bibliography Bhattacharya, Gouriswar. Deyadharma: Studies in Memory of Dr. D.C. Sircar. Delhi: Sri Satguru, 1986. Bénisti, Mireille. Stylistics of Buddhist Art in India. New Delhi: Aryan International, 2003. Brown, Robert L. "Narrative as Icon." Sacred Biography in the Buddhist Traditions of South and Southeast Asia. By Juliane Schober. Honolulu, HI: University of Hawai'i, 1997. Burgess, J.A.S. "The Great Stupa at Sanchi-Kanakheda." Journal of the Royal Asiatic Society of Great Britain and Ireland (1902): 29-45. Chanda, Ramaprasad. Dates of the Votive Inscriptions on the Stupas at Sanchi. New Delhi: Indological Book, 1977. Coomaraswamy, Ananda Kentish., Michael W. Meister. Essays in Early Indian Architecture. New Delhi: Indira Gandhi National Centre for the Arts, 1992. Cunningham, Alexander. The Bhilsa Topes. London: Smith, 1854. Dehejia, Vidya, and K. B. Agrawala. Unseen Presence: the Buddha and Sanchi. Mumbai: Marg Publications, 1996. Dehejia, Vidya. "Aniconism and the Multivalence of Emblems." Ars Orientalis 21 (1991): 45-66. Dehejia, Vidya. Discourse in Early Buddhist Art: Visual Narratives of India. New Delhi: Munshiram Manoharlal, 1997. 128 Dehejia, Vidya. "On Modes of Visual Narration in Early Buddhist Art." The Art Bulletin 72.3 (1990): 374-92. Fergusson, James. Tree and Serpent Worship. London: Asian Educational Services, [1873] 2004. Findly, Ellison Banks. Dāna: Giving and Getting in Pali Buddhism. Delhi: Motilal Banarsidass, 2003. Fogelin, Lars. Archaeology of Early Buddhism. Lanham: AltaMira, 2006. Fogelin, Lars. Beyond the Monastery Walls. Diss. University of Michigan, 2003. Fogelin, Lars. "Material Practice and the Semiotic Metamorphosis of a Sign: Early Buddhist Stupas and the Origin of Mahayana Buddhism." Unpublished Manuscript, 2009. Fogelin, Lars. "Ritual and Presentation in Early Buddhist Religious Architecture." Asian Perspectives 41.1 (2003): 129-54. Geiger, Wilhelm. The Mahavamsa: Great Chronicle of Ceylon. New Delhi: Asian Educational Services, 2003. Hamid, M. "Excavations at Sanchi." Annual Report Archaeology Survey of India (1940): 85-87. Hawkes, Jason. "The Sacred and Secular Contexts of the Buddhist Stupa Site of Bharhut." Buddhist Stupas in South Asia. Delhi: Oxford UP, 2009. Huntington, Susan L., and John C. Huntington. The Art of Ancient India. New York: Weatherhill, 1985. 129 Huntington, Susan L. "Early Buddhist Art and the Theory of Aniconism." Art Journal 49 (1990): 401-08. Karlsson, Klemens. Face to Face with the Absent Buddha. Uppsala: Uppsala University, 1999. Lüders, Heinrich. A List of Brahmi Inscriptions. Calcutta: Superintendent Government Printing, 1912. Leoshko, Janice. Sacred Traces: British Exploration of Buddhism in South Asia. Aldershot, Hants, England: Ashgate Pub., 2003. Linrothe, Rob. "Inquiries into the Origin of the Buddha Image: A Review." East and West 43 (1993): 241-56. Marshall, John, N. G. Majumdar, and Alfred Foucher. The Monuments of Sanchi. London: Swati Publications, 1983. Mitra, Debala. Sanchi. New Delhi: Archaeological Survey of India, 2001. Prinsep, James. "Note on the Facsimiles of Inscriptions from Sanchi near Bhilsa." Journal of the Asiatic Society of Bengal 6.1 (1837): 451-63. Salomon, Richard. Indian Epigraphy: a Guide to the Study of Inscriptions in Sanskrit, Prakrit, and the Other Indo-Aryan Languages. New York: Oxford UP, 1998. Schopen, Gregory. Bones, Stones, and Buddhist Monks. Honolulu: University of Hawai'i, 1997. Schopen, Gregory. Buddhist Monks and Business Matters. Honolulu: University of Hawai'i, 2004. Shaw, Julia. Buddhist Landscapes in Central India. London: British Academy, 2007. 130 Singh, Upinder. "Sanchi: The History of the Patronage of an Ancient Buddhist Establishment." Indian Economic and Social History Review 33.1 (1996): 1-35. Strong, John. The Legend of King Aśoka: a Study and Translation of the Aśokāvadāna. Princeton: Princeton UP, 1983. Tilley, Christopher. Material Culture and Text. London: Routledge, 1991. Vidya, Dehejia. "Collective and Popular Basis of Early Buddhist Patronage." The Powers of Art: Patronage in Indian Culture. By Barbara Stoler. Miller. Delhi: Oxford UP, 1992. Vogel, Jean. Indian Serpent Lore. London: Probsthain, 1926. Wagoner, Phillip. "Ananda K. Coomaraswamy and the Practice of Architectural History." Journal of the Society of Architectural Historians 58.1 (1999): 62-7. Willis, Michael D. "Buddhist Saints in Ancient Vedisa." Journal of the Royal Asiatic Society 11.2 (2001): 219-28. Willis, Michael D., Joe Cribb, and Julia Shaw. Buddhist Reliquaries from Ancient India. London: British Museum, 2000. 131 Vita Matthew David Milligan was born in Michigan in 1985. He completed high school at Brandon High School and the OSMTech Academy in Michigan in 2003. He then attended Albion College from 2003-2005, earning his B.A. in Religious Studies and Anthropology. In 2004, he traveled to Bodh Gaya, India on the Antioch Buddhist Studies program. During the fall of 2007, Matthew entered Graduate School at the University of Texas at Austin. In 2009, he participated in the American Institute of Indian Studies Prakrit language program in Pune, India. Email address: mattdmilligan@gmail.com This thesis was typed by the author.