Ajiṇṭhā Caves, Part I

in Article
Published on: 27 October 2016

Rajesh Kumar Singh

Rajesh Kumar Singh is an art historian with specialization on the subject of Ajanta caves. Many of his publications on the same include 'An Introduction to the Ajanta Caves'.

AJIṆṬHĀ caves, part I

Some aspects of the historical and political background

Dr. Rajesh Kumar Singh

B1-264, Siddharth Bungalows

Sama – Savali Road, Vadodara 390022, Gujarat, India

Email: rksingh1970@gmail.com, +91-9429925036



Some Aspects of the Historical and Political Background


Anonymous ancient name. 2

Anonymous ancient province. 3

Unknown ancient quantum of edifices. 4

Architectural lexicons of Ajiṇṭhā.. 6

The age of the caves. 8

The earlier phase. 8

The later phase. 11

Mahārāja Hariṣeṇa.. 13

The debate on the contours of Hariṣeṇa’s dominions. 14

The Ṛiṣika and Aśmaka connection.. 17

The ‘golden age’ 20

Before Hariṣeṇa.. 22

After Hariṣeṇa: Ellora’s earliest phase. 23

During Hariṣeṇa: Revival of Ajiṇṭhā.. 24

Some nikāyas. 26

The abandonment. 29

Table: Vākāṭaka chronology - different versions. 32

Captions for figures. 33

Works Cited.. 34

Notes. 41


Anonymous ancient name

The renowned World Heritage—and the ancient Buddhist monastery—of Ajiṇṭhā (अजिंठा) or Ajantā (अजंता) has preserved some of the oldest specimen of paintings in the world dating back to the second or first century BCE. In spite of the significant amount of literature that has accumulated on the subject since the early nineteenth century little clarity has emerged about the complex background and the times of Ajiṇṭhā that created the monument. For example, do we know what the Ajiṇṭhā monastery was called in the ancient times? The answer is, no. Ajay Mitra Shastri attempted to explore the ancient name of Ajiṇṭhā (Shastri 1991). He has rightly identified that even the modern name of the site is problematic. It is generally believed—and we have no reason to doubt the belief—that the site is named after a nearby village called Ajiṇṭhā, some 14 km from the cave site. It is not, however, the nearest village, for the nearest village is actually लेणापुर (Leṇāpur i.e. the village of caves), which is less than one km from the site. An excavation was carried out at Leṇāpur that yielded Sātavāhana pottery (Ghosh 1967, 14). Doubtlessly, Leṇāpur was connected to the monastery (sanghārāma) in some way; perhaps some of the workers, or overseers, or supervising monks lived there. However, the point as to how Leṇāpur failed in the bid to lend a name to the ancient monastery is a mystery. In any case, Ajantā is not Ajiṇṭhā, not at least phonetically. We need no imagination to deduce that the spelling of Ajiṇṭhā—in the earliest reports, government files, and archaeological studies—relied entirely on the phonetics, since the British officers and European scholars pronounced Ajiṇṭhā as Ajantā. Consequently, the prevalent spelling and pronunciation were perpetuated and well established. Nonetheless, a point may be noted that the locals, the non-Anglicised Maharashtrians, and the Marāṭhī media and literature often prefer Ajiṇṭhā from Ajantā. The present author identifies himself more with the locals than the global. Therefore, as a symbolic gesture the present author prefers to use Ajiṇṭhā in the present series of articles with the risk of some reader’s regrettable inconvenience.

Anonymous ancient province

Do we know where Ajiṇṭhā was located, i.e. in which ancient province or country? The answer would have to be in the negative. Ajiṇṭhā is located today in the Aurangābād district of the Maharashtra state of India. Its coordinates in the centre of the scarp are latitude 20°33'9.62"N, longitude 75°42'0.68"E (Google Inc. 2013).[1] According to Walter M. Spink, Ajiṇṭhā in the ancient times was situated in the province called Ṛiṣika[2] (Figure 1).

Fig. 1:  ‘Map of Vākāṭaka sites’ Spink and Yaguchi (2014:xv) and Weiner (1977:11)

This belief is based on V.V. Mirashi’s identification of the Ṛiṣika country with Khāndeś (Mirashi 1963, 123; plate P). However, in that identification, Mirashi has not suggested that Ajiṇṭhā was located within the boundaries of the Ṛiṣika country. Actually, it is difficult to speak conclusively about it, because the caves are situated on the ghāṭs (slopes) of the Sahayādri range above which was the Mūlaka[3] or Aśmaka[4] country and below was the so-called Ṛiṣika country. Ajiṇṭhā was on the border of the two countries/provinces. There is no incontrovertible record to ascertain whether the bordering ghāṭs were included in the Ṛiṣika or Mūlaka or Aśmaka country.

Unknown ancient quantum of edifices

Do we know what the makers’ count of the total rock-cut edifices was? The answer again would have to be in the negative. The particular question is posed here, for the English word cave can be problematic at times. Consider, for example, the basics (without getting into details).[5] The Archaeological Survey of India numerates 30 caves,[6] and the same quantum and numbering is naturally perpetuated in the literature on the subject. The quantum and the numbering system is not, however, free from anomalies. Let us notice a few of them.

Fig. 2:   Ajiṇṭhā: Cave 26-complex


Fig. 3: Cave 26-complex (isometric plan): conjectural reconstruction by Spink and Ajeet Rao (artwork by Ajeet Rao and Nitin Veturkar; courtesy of Walter M. Spink and Samskara)

Take the case of the edifice(s) created by the Monk Buddhabhadra (Figure 7). Although Buddhabhadra is a well-known patron of Cave 26[7], a notion today is generally established that the monk created the Cave 26 only. The notion goes that the edifice called ‘sugatālaya’ (the abode of Sugata, i.e. the Buddha) in the inscription comprises of a stūpa temple and two lower adjuncts in front that flank the courtyard. In recent decades, Spink’s research, followed by the present author, reveals that the temple actually has four adjuncts flanking on either side[8] (Figure 11). The adjunts comprise of the caves 25, 27, and the lower wings duo (Figure 9). Earlier, caves 25 and 27 were not recognised as adjuncts; they were identified as independent edifices, independent caves. However, for the patron Buddhabhadra and other monastics, the temple (Cave 26) with the four adjuncts would obviously have been identified as one śailagṛiha, whereas we count them as three: Caves 25, 26 and 27. The problem is not just about numbering. It is about understanding the original architectural setting of a patron’s ‘कीर्त्ति’[9] (kīrtti); this was the word used for the temple and the monument—in Buddhabhadra’s donative inscription, which is right there, in the porch of the same temple.


Fig. 6: Ajiṇṭhā: Cave 26-complex in c. 480 CE, conjectural 3D model minus the details, from (Singh 2012a:61)

Let us take another example. A notion is well published that the two-storeyed Cave 6 is a single edifice[10] (Figure 7). It is also self-evident by the fact that only one number has been assigned to the two storeys. In latest research, however, the two floors are identified as two distinct edifices. New nomenclatures had to be adapted for them, i.e. Cave Upper 6 and Cave Lower 6. The two floors have very different floor plans (Figure 10), iconographic schemes, and designs. Little is common between them. It appears strongly that the upper floor was not even planned when the lower floor was well underway.[11]



Fig. 7: Caves Upper 6 (above) and Lower 6 (below), from (Burgess 1880, plate XXXII)

To cite a third example, there is a number given to a rock-cut cistern called Cave 18 (Figure 4), while the fact is that there are many other cisterns at the site that are anonymous; they have no number at all; they have not been identified and counted as caves. That is where the problem lies. The problem is with the English word cave. It relegates the original structure. Just anything that is hewn out in the monolithic context is treated as a cave. The word does not distinguish among a cistern, cell, hall, shrine, temple, or a sprawling complex with many adjunct units. There will be no such problems if the modern student adopted the lexicons that the makers of Ajiṇṭhā had used. Many architectural lexicons are found in the Ajiṇṭhā inscriptions. There is a variety of them ranging from the generic to specific terms.

Architectural lexicons of Ajiṇṭhā

The following is a non-exhaustive glossary of some architectural terms that appear in the Ajiṇṭhā inscriptions[12]: शैलगृहम्[13] śailagṛham = rock-cut architecture; कन्दरा kandarā[14] = natural or rock-cut cave; लयन[15] layana/ एकाश्मकम्[16] ekāśmakam = monolithic architecture; उपासय upāsaya[17] (Prākṛit) = उपाश्रय upāśraya (Snk.)= a monastic dwelling made of any material; वेश्म[18] veśma = dwelling; नागेन्द्रवेश्म[19] Nāgendraveśma = Nāgendra shrine; उववरक[20] uvavaraka (Prākṛit) = अपवरक apavaraka (Snk.) = a monk’s cell; मंडप[21] maṅḍapa = literally, a pavilion, but in the given contexts it connotes a Buddha temple; पसादा[22] pasādā (Prākṛit) = प्रासाद prāsāda (Snk.) = palace, large temple; विहार[23] vihāra = a monastery, but also a temple with or without monastic dwelling cells; मुनिराजचैत्यम्[24] munirājacaityam/ सुगतालयम्[25] sugatālayam/ चैत्यमन्दिरम्[26] caityamandiram/ स्तूपविहार[27] stūpavihāra/ गन्धकुटी[28] gandhakuṭī = a Buddha shrine or temple consisting of a stūpa, or Buddha image, or both; कीर्त्ति[29] kīrtti = literally, fame, but also a famous monument or monument for fame, in religious context it connotes a temple; गवाक्ष[30] gavākṣa = a window shaped like a leaf of pīpal (ficus religiosa);  निर्य्यूह[31] niryyūha = the summit or pinnacle(s) of a gateway, entrance, door, or gavākṣa; सुवीथि[32] suvīthi = picture gallery[33]; वेदिका[34] vedikā = pavilion, hall, alter, or shrine[35]; स्तम्भ[36] stambha = pillar; विभङ्ग[37] vibhaṅga = usually a staircase, but in the context of Cave 16, it seems to denote the extant tunnelled staircase; घरमुख[38] gharamukha = façade; भित्ति[39] bhitti = wall.

The age of the caves

Let us move to another question, again a fundamental one. When were the caves made? This question, unfortunately, is still being probed and debated. Even after nearly two centuries of studies, there is no precise clarity. What we have are different opinions. Some leading opinions are objectively presented here. No one seems to have a clear idea. After ascribing many phases and centuries stretching from the third century BCE to the seventh century CE,[40] experts started agreeing that Ajiṇṭhā developed through two different periods. Conventionally, the two periods have been called the Hīnayāna and Mahāyāna. This type of classification, periodization, and nomenclature is now found to be problematic, as the exact meanings of the two terms have come under intense review.[41] Untill the debate is settled, some scholars find it safer to use the terms ‘earlier’ phase and ‘later’ phase.[42] What is the period of the two phases? Even this is not absolutely clear.

The earlier phase

Untill recently, we ascribed some two to three centuries for the earlier phase: third century BCE to first century BCE.[43] However, an opinion is forming in current research that the five edifices of the earlier group, i.e. caves 9, 10, 12, 13, and 15A[44] (Figure 5), should be placed in the first century BCE.[45] M. K. Dhavalikar has currently expressed the view that ‘the early group (Hīnayāna) of Ajiṇṭhā caves were possibly carved during the reign of Sātakarṇī II in the latter half of the 1st century BC’ (Dhavalikar 2003, 4). The incised inscription on the rock-cut façade of Cave 10 records the donation of a wooden façade by ‘son of Vāsiṭhi.’[46] A donor of the same name is found at Sānchī too: ‘an inscription on the southern gateway, which is the most impressive, records that “it was the gift of Ananda, the son of Vasithi (Vasishthi, Sk.), the chief of the artisans (aveshanin) of king Siri-Satakarni” ’ (Dhavalikar 2003, 4). We do not know whether the two inscriptions point to the same person, but if they do, then Ajiṇṭhā’s earlier phase and that of the Sānchī gateways must be treated as contemporaneous. Monika Zin too is inclined to support this dating based on additional sources, such as art, style, and iconography (Zin 2003). Dieter Schlingloff has shown that there prevailed in the second century BCE a common pool of the narrative and iconographic repertoire amid Bhārhut, Amarāvatī, Nāgārjunikoṇḍā[47], Sānchī, Mathurā, Gāndhāra, and Ajiṇṭhā (Schlingloff 2013, vol. II, 4-13). Based on the iconographic concordances, Schlingloff has dated the earlier phase of Ajintha to the second century BCE (Schlingloff 2013, vol. II, 3).

At this time, the region of Ajiṇṭhā fell in the dominion of the Sātavāhana rulers. They had the capital at Pratiṣṭhāna (modern Paiṭhan), which is nearly 160 km from Ajiṇṭhā. Ancient highways connected the pilgrimage centres of the Buddhists, Hindus, and Jains. Ajiṇṭhā lay between the meeting point of Uttarāpatha and Dakṣiṇāpath.[48] Thus, the travelling carvans of traders, armies, and people would have halted at such sites that were near the highways with picturesque waterfalls.

The Sātavāhana rulers were great patrons of art and culture. They patronised Buddhist monuments as much as they patronised the Hindu monuments. However, there is no evidence to suggest that the Sātavāhanas had any direct involvement in the creation of Ajiṇṭhā. Epigraphic records indicate that the first phase of Ajiṇṭhā was developed by the virtue of collective patronage. Incised and painted inscriptions are extant in many of the caves that convey that different donors paid for the different parts of the edifices. Thus, there was no royal patronage during the first phase of Ajiṇṭhā. It was all done by merchants, traders, ordinary people, upāsakas (laity), and monks themselves.

Because no reference to royal involvement is found, it would be superfluous to speak of the political background or any dynasty for the first phase of Ajiṇṭhā. Therefore, let us move to the second phase of Ajiṇṭhā.


The later phase

The age of the second group of caves is no less problematic. We may quantify in very loose terms that some 25 caves were excavated in the later phase. Initially, several centuries were ascribed ranging from the third century CE to the seventh century CE.[49] This dating is so deep entrenched that UNESCO’s webpage on the Ajiṇṭhā caves still carries the old dating: ‘The first Buddhist cave monuments at Ajanta date from the 2nd and 1st centuries B.C. During the Gupta period (5th and 6th centuries A.D.), many more richly decorated caves were added to the original group’ (UNESCO 1992-2015). Archaeological Survey of India’s website is more ingenuous: ‘The caves were excavated in different periods (circa 2nd century B.C. to 6th century A.D.) . . . . The second phase of paintings started around 5th-6th centuries A.D. and continued for the next two centuries . . . . The specimen of these exemplary paintings of Vakataka period could be noticed in cave nos. 1, 2, 16, and 17’ (Archaeological Survey of India 2011). In other words, the ASI believes or it wants people to believe that the Vākāṭakas ruled up to eighth century CE!

Nowadays many scholars assign the age to the fifth century CE or late fifth century CE.[50] However, A. P. Jamkhedkar’s dating is one of a kind: ‘Ajanta rock cut caves are creations of about 700 years, roughly from 200 BC to AD 525’ (2009). Walter M. Spink has ascribed the age in rather specific terms. He says that all the caves of the later period were developed from circa 462 to 480 CE. He clarifies, however, that ‘For simplicity, I have used specific dates, although the reader should allow a margin of error of a year or two—probably not more for each assignment’ (W. M. Spink 2005, xi). Further, ‘. . . . the “specific” dates which we have applied to every feature of Ajanta can never claim to be absolutely precise, since the whole dated sequence depends on specifically dated termini which slightly pre-date and slightly post-date the period of Ajanta’s patronage. So one must order Ajanta’s development by reference to these outer limits’ (W. M. Spink 2005, 2). Spink’s ‘short chronology’ can be found in most of his publications. His latest revised chronological chart can be found in (Spink and Yaguchi 2014, xii). His dating was initially questioned.[51] However, some authors have recently accepted his dating including Hans Bakker,[52] Naomichi Yaguchi (Sahapedia.org 2015), and the present author[53]. Spink has proposed that all of the fifth-century caves were begun and largely completed during the reign of a single king within whose dominions the site of Ajiṇṭhā fell. The king’s name was हरिषेण (Hariṣeṇa). This king occupies a place only in the margins of India’s history.[54] But for Spink: ‘Harisena, with his power centered at ancient Vidarbha (eastern Maharashtra) was surely the greatest in India, and possibly in the entire world at the time. By the end of his reign, his sway extended over the whole of central India from the western to the eastern sea’ (W. M. Spink 2005, 7). Spink’s map of the Vākāṭaka Empire is reproduced here at Figure 3.

Two inscriptions at Ajiṇṭhā, one in the Ghaṭotkacha cave at Gulwāḍā, and the Thālner grant found in the neighbouring Dhule district, mention Hariṣeṇa’s rule. Ajiṇṭhā Cave 16 inscription, vs. 17: हरिरामहरस्मरेन्दुकान्तिर्हरिषेणो हरिविक्रमप्रतापः, harirāmharasmrendukāntirhariṣeṇo harivikkramapratāpah, ‘Then his son became king [. . .] Harisheṇa, who, in loveliness, resembled Indra, Rāma, Hara, Cupid, and the moon, and who was brave and spirited like a lion’ (Mirashi, Inscriptions of the Vākāṭakas, Corpus Inscriptionum Indicarum series 1963, 108, 110)]. Ghaṭōtkacha Cave inscription, vs. 13: अथ देवराजसूनुर्हरिषेणो […] हस्तिभोज, ‘atha devarājasūnurhariseno [. . .] hastibhoja, Then there is Harisheṇa, the son of Devarāja [...] Hastibhōja’ (Mirashi 1963, 117, 119). Cave 17 inscription, vs. 21: … वदनारविन्दचन्द्रे [।*] परिपालयति क्षितीन्द्रचन्द्रे हरिषेणे हितकारिणि प्रजानाम् [२१॥*], … vadanārvindacandre [|*] paripālayati kṣitīndracandre Hariṣeṇe hitakāriṇi prajānāṃ, ‘While that moon among the princes, Harisheṇa, whose face resembles a lotus and the moon, and who doeswhat is beneficial for (his) subjects [. . .] is protecting the earth’ (Mirashi, Inscriptions of the Vākāṭakas, Corpus Inscriptionum Indicarum series 1963, 126, 129). Concerning the reign of Hariṣeṇa, Burgess wrote: ‘Pandit Bhagwanlal is probably right in assuming that the Harisheṇa mentioned in line 21 is the Vākāṭaka prince whose name occurs in Ajanta [inscription] No. 3, and that the Vākāṭakas were the lords paramount whom these rulers [King Upendragupta II or Dharādhipa, the donor of Cave 17] obeyed’ (Burgess 1883, 128).

Mahārāja Hariṣeṇa[55]

It was around 460 CE when Mahārāja[56] Hariṣeṇa (महाराज हरिषेण) came to the throne[57]  of the Western Vākāṭaka Empire.[58] Hariṣeṇa appears to have become the greatest ruler of at least the Vākāṭaka dynasty (for comparative chronology, see Table 1). The dynasty had modest beginnings. However, Hariṣeṇa was ambitious. Within a year or two after coming to the power, he began a victory compaign raiding several countries in the various directions. His inherited kingdom would have been confined to a part of the central Deccan whose exact contours cannot be determined with certainty. Under him, the Vatsagulma branch of the Vākāṭakas excercised predominance over the Nandīvardhana branch. No attempt was made to eliminate King Narendrasena and King Prithivisena II who were the rulers of the other branch, for they were दायाद (persons of common ancestry). Mahārāja Hariṣeṇa’s ‘सचिव’ (secretary), nicknamed prime minister by some writers, was Varāhadeva who developed the Ajiṇṭhā Cave 16 and the Ghaṭōtkacha cave temples at Gulwāḍā, and donated them to the Buddhist sangha.[59] He has left lengthy incised inscriptions on the exterior and porch walls of the beautiful cave temples that proclaim his donations.[60] Together they provide the pedigree of Varāhadeva and Mahārāja Hariṣeṇa. Cave 16 provides a प्रशस्ति (eulogy) of Hariṣeṇa, as was customary in donative inscriptions. They are composed in poetic verses.

            Hariṣeṇa is important because he was obviously supportive of the Buddhist. In Spink’s view, he was directly involved in the patronage activity at Ajiṇṭhā. He postulates that ‘Cave 1 was sponsored by the great Vākāṭaka emperor Hariṣeṇa himself’ (W. M. Spink 2007, 7).

The debate on the contours of Hariṣeṇa’s dominions

Verse 18 of Varāhadeva’s donative inscription at Ajiṇṭhā provides the names of many ancient countries or provinces: ‘He [Hariṣeṇa] (conquered), Kuntala, Avantī, Kaliṅga, Kōsala, Trikūṭa, Lāṭa, Āndhra[parānta] . . . which, though very famous for valour. . . .’ [61] The place names have been identified[62] (Figure 2). However, the missing or illegible words have given rise to different opinions. A brief overview of the same is given below.

The verse intends to convey a point about the named countries. But the verb is missing. Buhler and Mirashi conjectured that the missing verb implied conquest, and they concluded that Hariṣeṇa must have conquered the named provinces. In later research, Spink accepted the conjecture as a fact. He says that Hariṣeṇa was ‘the greatest ruler of the mid-fifth century; by the end of his brief reign (c. 460-477 CE), his domains in central India stretched from sea to sea’ [ (Spink and Yaguchi 2014, 1) (W. M. Spink 2009, fig. 1)]. For Spink, there is no scope to doubt the claim of the inscription. But, for D. C. Sircar: ‘It is a vague claim, which may indicate some sort of hostile relations of the Vākāṭaka king with those countries. It is impossible to believe that the countries were completely subjugated by Hariṣeṇa’ (Sircar 1965, n. 2). Spink does not entertain the view of A. M. Shastri, a Vākāṭaka epigraphist, who somewhat echoes Sircar’s opinion:

It is possible that he [Hariṣeṇa] carried out hurried raids in some of these territories, which the poet turned into full-fledged conquests and added to them some other regions where his relations were ruling or with which he had nothing to do militarily. It would therefore be wrong to assert merely on the basis of an otherwise unsubstantiated statement of a court-poet that ‘Hari Shena’s supremacy was recognised throughout the Deccan extending from Malwa in the north to Kuntala in the south, and from the Arabian Sea in the west to the Bay of Bengal in the east’[63] or to assert that ‘he was the greatest ruler in the world’[64] in the late fifth century AD. What we can assert more reasonably is that he brought the process initiated by his father to a logical conclusion by establishing his authority over the territory of the other branch of the Vākāṭakas either by overthrowing Prithivi Shena II, its last known member, or his successor, if any, or automatically without much resistance due to the failure of an heir. We can also conclude justifiablty that for a time, howsoever short it may have been, he was a king who commanded a very high esteem among his contemporaries due as much to his great military strength as to his human qualities. [ (Shastri 1997, 205)]

Let us also observe the opinion of Hans Bakker whose views are in line with Sircar and Shastri:

Nowhere else in the available sources concernfing Hariṣeṇa, nor in those of his supposed antagonists for that matter, do we find any confirmation of this putative extraordinary military achievement, which is most unlikely ever to have taken place in historic reality. The verse rather seems to suggest that the Vākāṭaka king was outdoing or putting to shame all the surrounding kings.[65] [ (Bakker 1997, 34-35)]

Bakker’s idea of Hariṣeṇa’s dominions is obvious in the following remarks:

Although Hariṣeṇa’s imperialist exploits thus seem to lack sufficient epigraphic basis, the verse at issue underlines his expansionist ambitions, and one of the first victims thereof might have been his eastern kinsman Narendrasena… It [the verse] does not mention those territories for which we have some positive evidence that they had came under Hariṣeṇa’s sway, viz. Aśmaka and the eastern Vākāṭaka kingdom. Hariṣeṇa, or his minister in his stead, did not mention them, first of all because the inscription probably does not refer to his conquests, as we have just argued. Morover, to specify them as such may have been thought embarrassing, as it might have cast doubt on Vatsagulma’s pretension that they formed an obvious part of the kingdom. [ (Bakker 1997, 35)]

There is no end to Spink’s opponents. Another Vākāṭaka epigraphist, Brahmanand Deshpande has rejected the core of Spink’s constructions.[66]

However, some studies corroborate Spink’s broader conclusions. To cite a few, Shobhana Gokhale has studied the evidence of Hariṣeṇa’s expansionism (Gokhale 1992). Shastri (1997), L. S. Nigam (2004), and Donald M. Stadtner (2004) believe that Hariṣeṇa or his contemporaries raided, even if briefly, the ancient country of Dakṣiṇa Kōśala (Chhattisgaḍh). If the theories of Hariṣeṇa’s conquest of Kanherī and Dakṣiṇa Kōśala are accepted then the conquests of the other regions, as claimed in the same inscription, must be accepted too. Another relevant point is argued by Bakker. He says that Hariṣeṇa had already probably inherited Aśmaka, Ṛiṣika, and Mūlaka provinces so that there was no need to mention them in the list of the raided countries. We cannot go into all the details here. Recently, Spink has made a detailed analysis of all sources and provided a comprehensive overview of a potential picture of the political background especially during the end of the Vākāṭaka era (W. M. Spink 2005, 119-162).

The Ṛiṣika and Aśmaka connection

Ajiṇṭhā inscriptions make a mention of two specific provinces: ऋषिक (Ṛiṣika) and अश्मक (Aśmaka). The king of Ṛiṣika probably donated the caves 17-20 (Figure 6) as learned from the donative inscriptions in Cave 17[67] and Cave 20[68]. The contours of the Ṛiṣika country were largely the Khāndeś region, and the site of Ajiṇṭhā lay on the border of Ṛiṣika and the neighbouring country above the ghāṭs, that is, the Aurangābād region called Mūlaka by Mirashi (Mirashi 1963, plate V) and Aśmaka by Spink. The region to the south of Godavari comprising parts of Ahamadanagar and Bhir districts is identified as Aśmaka by Mirashi (1963, 123, 124) and Shastri (1997, 48), but for Spink, it is Mūlaka. Only further research will confirm which of these identifications is valid. The name of the king of the Ṛiṣika country who donated caves 17, 18, 19, and 20, and ‘adorned the earth with stūpas and vihāras’ has been interpreted variously. Cave 17 inscription, verse 9 mentions: ‘धराधिपारख्यां प्रथमो बभार दध्रे द्वितीयो [र*]विसाम्ब संज्ञाम ॥ [९ ॥*], dharādhipārakhyām prathamo babhāra dadhre dvitīyo [ra*]visāmba sanjām || [9||*], The elder (of them) bore the title of a king, while the second bore the appellation Ravisamba’ (Mirashi 1963, 125, 128). To Mirashi (Mirashi 1963, 122), the elder’s, i.e. the donor’s name is lost. Trusting Mirashi’s name-lost theory and relating with Cave 20 inscription, Spink has conjectured that the elder brother’s name is Upendragupta (W. M. Spink 2005, 203 ff.). However, Shastri identifies the name to be Dharādhipa. He has created a separate genealogy chart of the donor (Shastri 1997, 47) with the following explanation:

Mirashi observes that the name of the elder brother of Ravisamba, the last known member of the family, has not been preserved though ‘he bore the title of a king.’ The expression Dharādhip-ākhyam employed in connection with the first prince (kumāra) appears to denote ‘one bearing the name Dharādhipa’ even as Ravisāmba-saṁjṅām used for the second son and rightly taken to mean bearing ‘the appellation Ravisāmba’. The name Dharādhipa, ‘lord of the earth’, need not be regarded as unusual as names like Narendra, ‘lord of men’, and Bhūpati or Avanindra, ‘lord of the earth’ are known to have been borne by individuals. Utterly grieved at the untimely decease of Ravisāmba, [the] elder brother Dharādhipa developed an utter distaste for worldly pleasures and riches and pursued a pious life dedicated to Buddhism. He covered his kingdom with stūpas and vihāras and made charities to the suppliants. [ (Shastri 1997, 47-48)]

Bakker (Bakker 1997, 36) too calls him Dharādhipa: ‘The “name” of the elder brother is odd: Dharādhipa (i.e. “Lord of the Earth,” or “King”).’ Bakker rejects Spink’s reading on the following grounds:

In his treatment of the inscription in Cave XVII, Spink goes one step further than Mirashi when he identifies the elder brother called ‘King’ with the donor of Cave XX. In the latter cave a very fragmentary donative inscription is found [ (Chhabra 1952, 113)] which seems to feature the name Upendra, who, possibly, is the son of a man whose name might have begun with Kṛ. One of the ancestors (the great-great-great-great grandfather) in the pedigree of Ravisāmba and his anonymous brother is called Upendragupta; hence Spink conjectures that the elder brother might have been named after him, in which case the name of the father in the inscription of Cave XX, whose name might have begun with ‘Kṛi,’ should be read as ‘Kṛiṣṇadāsa.’ However, the inscription in Cave XX is in such a bad condition that hardly any reliable information can be derived from it; moreover, Cave XX is not Cave XVII or Cave XIX. It is difficult therefore to consider Spink’s theory as any more than ingenious speculation. [ (Bakker 1997, 38)]

Ancient highways connected Ṛiṣika, Aśmaka, and Mūlaka with the rest of India. The name of the Aśmaka province is mentioned in the donative inscription of Cave 26 donated by Monk Buddhabhadra. The donor, in fact, erected the monument in the honour of Bhavvirāj who was the minister of the Aśmakarājā, as we are told in verses 9-13:

अनेकजन्मान्तरबद्धसौहृदम् स्थिरम् कृतज्ञं सुधियम् विपश्चितम [|*] सुरासुराचार्यमतेषु कोविदम् महानुभावाश्मकराजामन्त्रिणं || [९ ||*] . . . तम भव्विराजम उद्दिश्य मातापितरमेव च [|*] भिक्षुणा बुद्धभद्रेण कारितः सुगतालयम् [१३ ।।*][69]

Aneka-janm-ātta(nta)ra-baddha-sauhṛidaṁ sthiraṁ kṛitajṅaṁ sudhiyaṁ vipaśchitam [|*] sur-āsur-āchāryyā(ryya)-mateshu kovidaṁ mahānubhāv-Āśmaka-rāja-mantriṇam || [9 ||*] . . .  Taṁ Bhavvirājam uddiśya mātā-pitaram-eva cha [|*] bhikṣuṇā Buddhabhadreṇa kāritaḥ Sugat-ā[layaṁ] [13 ||*]

The monk Buddhabhadra has caused (this) temple of Sugata to be made in honour of his parents as well as (that) Bhavvirāja who served the mighty king of Aśmaka as the latter’s minister, who was attached to him (the monk) in friendship through many successive births…’ [ (Chhabra 1952, 115-116, 118)]

As seen above, Monk Buddhabhadra goes on to record that he was attached to the minister through many successive births. Maybe, Buddhabhadra belonged to or hailed from the Aśmaka country. Thus, there is a direct evidence that the Ṛiṣika and Aśmaka provinces had a role to play at Ajiṇṭhā.

About the involvement of other provinces, or the people from other provinces, there is no such direct evidence. This is because the names of other principle donors are not known. However, we get the name of one माथुर (Māthur) in the pedestal of the shrine Buddha image of Cave 4[70], but it is uncertain whether he donated just the shrine part or the whole edifice. The doubt is there because of the unusual location and unusually short length for a donative inscription. Māthur is quite a popular surname in India, but here it is the first name. Frequently, many people’s names in India are still associated with the places of origin. We wonder whether Māthur hailed from Mathurā. For the rest of the caves, we can only conjecture based on stylistic, iconographic, and pilgrimage routes that other donors might have hailed from regions as far and wide as Āryāvarta[71], Greater Magadha[72], Anūpa, Kōśala, Avantī, Lāṭa, Mūlaka, Vidarbha, Kuntala, Aparānta, Āndhra, and Kalinga.

The ‘golden age’

It has not been possible to map the true extent and diversity of the patronage activity of the fifth-century India. The story of the Gupta patronage to art and culture is relatively well known. However, the sotry of the Vākāṭaka patronage remains a work in progress. It is now well understood to the Vākāṭaka historians that the ‘golden age’ of India stretched much beyond the dominions of the Guptas. Especially during the time of the fifth-century Ajiṇṭhā, the Guptas were confined to north India, and a large swathe of their erstwhile dominions in central India and the Deccan were in control of the two branches of the Vākāṭakas. Particularly, during the Hariṣeṇa-period, the Vākāṭakas had great expansionism that lasted only briefly, less than two decades. Altogether, the Vākāṭakas in the fifth-century were able to create or allowed for an impressive bulk of patronage activity. Ajiṇṭhā is only a tip of the iceberg. The extant archaeological, epigraphic, numismatic,[73] and art historical corpus are enough to suggest that patronage under both the branches of the Vākāṭakas was quite prolific and consistent. The consistency is equal in the Buddhist as well as the brāhmaṇical[74] patronage. It is only a distortion of history that Ajiṇṭhā and Hariṣeṇa have taken all the limelight, even in the current scholarship.

            It is outside the scope here to cast an account of the various known patronage activities of the two branches of the Vākāṭakas. But, an assessment of Ajiṇṭhā must not be made without comparing with the art historical and archaeological corpus of the patronage by the Nandivardhana branch. Hans Bakker has made an extensive survey of the comparatively lesser known art history of Mansar, Rāmagirī, Nagardhan, Mandhal, and Pravarapura.[75] Bakker has highlighted brāhmaṇical heritage of the Vākāṭakas. The students of art and archaeology must keep in mind that what we have today is never the complete picture, for the greater, the larger part, is perished or some may still be waiting to come to light in the future. Another certainty is with regard to the medium and material. The greater part of the extant and known art historical and archaeological corpus has come down to us by the virtue of the material: the stone, or rock, as the case may be. But, these were not the only media. In fact, it can be reasonably argued that these were never the most prevalent media, for a work in stone or rock is often more expensive and demanding than in other media and materials. Perhaps, the larger corpus in wood, brick, and other perishable media is irretrievably lost. The art historical map, therefore, is doomed to remain inadequate, incomplete, and imperfect. The available corpus, therefore, must be treated only as indicative of a larger, much wider scenario. Take, for example, the epigraph in Cave 17, Ajiṇṭhā, verse 22, which says: ‘He [the donor] . . . adorned the earth with stūpas and vihāras . . . .’[76] Judging by the contents of the whole inscription, it does not seem to be a false claim, even if it is obviously an exaggeration. We wonder, therefore, where those stūpas are.[77] The monasteries that were of the rock-cut type have been called ‘ekāśma,’ ‘layana’, or ‘śailagṛha’ in the Ajiṇṭhā inscriptions.

Before Hariṣeṇa

As noted earlier, there was a resurgence of patronage to the Buddhists in the Hariṣeṇa period. The sites included, among others, Ajiṇṭhā, Ghaṭōtkacha, Aurangābād, Banōṭī, Bāgh, and Kanherī. What was the scenario in the pre-Hariṣeṇa period? Advancement in research indicates that Hariṣeṇa was a Vākāṭaka different from the rest in his clan. Prior to his reign (c. 460-477 CE), there prevailed a lull with regard to major Buddhist patronage. In the period of the lull, a brāhmaṇical wave likely prevailed through the southern and middle kingdoms of India. The lull indicates that the Buddhists had taken a backseat in the cultural and theological map of the regions in context. Indeed, Buddhist archaeology of the Deccan, eastern, and western India does not show any major initiative of the establishment of new and large monasteries during the third to early fifth century CE. In this period, little signs of usage, worship, or habitation are noticed in the existing Buddhist sites. However, some older monasteries must have continued somehow to remain in function but fresh and elaborate developments are barely seen during the period in question. M. K. Dhavalikar in Late Hīnayāna Caves of Western India (1984) has documented many lesser-known and newly rediscovered sites of wesern India that he attributes to a ‘hiatus period’ from the third to mid-fifth centuries CE. The sites are Pītalkhōrā, Nāsik, Junnar, Karāḍ, Pohaḷe, Wāi, Nadsur, Nenāvaḷī, Shirwaḷ, Kuḍā, Mahāḍ, Shelārwāḍī, Khéḍ, and Kanherī. None of these sites is comparable to the scale and grandeur of Ajiṇṭhā, Ellora, Sānchī, Amarāvatī, Nālandā, etc. Apparantly, Dhavalikar’s late Hīnayāna caves were rather for the individual monastics or small groups of monastics. There is little, if any scope of the existence of the Buddhist sangha having been situated at these sites.

After Hariṣeṇa: Ellora’s earliest phase

In the post Hariṣeṇa period, the picture is rather worse. We do not see any substantial activity in the sixth century Buddhist India. The problem is compounded by the fact that the period of the earliest caves of Ellora is still far from settled. Especially, the period of the Buddhist caves of Ellora has been a tough nut to crack. Various dates have been ascribed ranging from the fifth to the eighth centuries CE.[78] On art, architectural, and iconographic grounds, the age of Ellora’s Buddhist temples cannot be placed at par with the Hariṣeṇa-period caves of Ajiṇṭhā. Evidently, Ellora picks up the threads of the art and architectural legacy of Ajiṇṭhā, and takes it to the next level; they are not contemporaneous, for nothing of Ellora is yet visible at Ajiṇṭhā. Obviously, Ellora started after the collapse of Ajiṇṭhā and the Vākāṭakas. Not only the art, architectural, and iconographic grounds give rise to such a conclusion, but the political and sociological geography also point to the same. Some observations on Ellora may be in context.

The first Buddhist phase of Ellora cannot be dated prior to the mid-sixth century CE. They could never have begun until several decades had elapsed after the death of Hariṣeṇa in c. 477 CE. We know of only one successor of Hariṣeṇa, his son Sarvasena III who did not rule for long (Table 1). After him, we do not hear anything of the Western Vākāṭakas.[79] The same is the case of Sarvasena III’s contemporary, Prithivisena II of the Eastern Vākāṭaka kingdom. That kingdom too is lost in oblivion around the close of the fifth century CE. A number of regional powers take control of the Vākāṭaka domains. The Early Kalacūris of Mahiṣmatī take control of the Ellora region in the sixth century CE. It is not clear what role the Early Kalacūris played at Ellora. There is no epigraphic record to guide us. It will be too much to expect that soon after taking control of the region they started sponsoring the Buddhist caves of Ellora. If a likely brāhmaṇical agitation and sociological upheaval were among the larger factors of the collapse of Hariṣeṇa’s kingdom—a point that is discussed elsewhere[80]—one cannot expect the smaller immediate successors to commit the same heresy that Hariṣeṇa in a way committed. What the Early Kalacūris did was to support the brāhmaṇical patronage. Spink believes that the rock-cut caves in the Mumbai region like Elephanta and Jogeshwari were excavated during this time (mid-sixth century CE). In addition, some brāhmaṇical caves were also begun at Ellora that was already a tīrtha due to the Ghṛiṣṇeṣvara jyōtirlinga. The first cave temples were from caves 27 to 17 near the waterfall, all brāhmaṇical.[81]

At the end of nearly half a century of their rule, the Early Kalacūris are supplanted by the Early Cālukyas of Bādāmī. This dynasty is already known to have patronised the rock-cut caves of Bādāmī. Art historically, many caves of Ellora have well demonstrated the stylistic and iconographic relationship with Early Cālukyan art and architecture. Thus, it appears more reasonable that the Early Cālukyas inaugurated the series of the Buddhist caves of Ellora.[82]

During Hariṣeṇa: Revival of Ajiṇṭhā

Fresh or renewed patronage activities at several rock-cut monasteries (and presumably several others in the perishable media of wood or brick) within the dominions of Hariṣeṇa happened shortly after he came to the throne. This fact, based on current research, implores us to gauge the scenario. In the lack of required proof to shed light on how this happened the reasoning and inference are the only options. How was the prolonged ‘hiatus’ of nearly three centuries broken no sooner than Hariṣeṇa’s enthronement? How would anyone have known Hariṣeṇa’s policies, unless it was announced in some way by Hariṣeṇa himself? How did so many patrons come together for sponsoring the establishment of new monasteries, or planning the renaissance of the older monastery of Ajiṇṭhā? How did the communication among patrons take place, especially when the patrons came from different spheres of society and administration, with some from the royal courts, others being high priests with friendship with ministers, the kings themselves, and wealthy merchants? As is known, not all the patrons were monks, not at least as wealthy or resourceful monks as Buddhabhadra was. In fact, a majority of the donors were from the non-monastic community. The Buddhist world had two broad sections. One, that of the monastics, and the other that of the laity. The laity served for the larger pool of patrons; it was true also of the Hariṣeṇa period.

The coming together of the various representatives or adherents of some larger or smaller sanghas, mainly belonging to the Mūlasarvāstivāda nikāya, would have added to the strength or the renewed formation of the sanghārāma at Ajiṇṭhā. This could not have happened without the overt or covert support from the Maharaja Hariṣeṇa. Monks travelled from one monastery to another that permitted the traffic of ideas. Such possibilities are corroborated by the varieties of idioms and styles of the art, architecture, and iconography of Ajiṇṭhā. Painting has not survived at other sites. So necessary comparisions cannot be made. But there is enough archaeological, sculptural, architectural, epigraphic, and iconographic material to show that the ideas at Ajiṇṭhā came from various regions. Technical expertise took time to develop as the excavations progressed year by year. There are numerous examples to show that the work in the initial years was less than perfect. Some serious errors were also committed. Senior and expert monks would have been adept to the principles and methods of art and architecture, but a majority of the workforce evidently comprised of such people who were less than experienced; there must also have been those who had never carved a cliff before. This is mainly because we do not find major rock-cut excavations in the preceding decades or centuries. As to how the work was organised, a picture drawn by Romila Thapar for c. 200 B.C. – A.D. 300 may hold good even for the case of Ajiṇṭhā in the Hariṣeṇa period:

Buddhism hovered in the background of most activities at this time [c. 200 B.C. – A.D. 300], also supported by the rich and the powerful. Buddhist texts such as the Milindapanha, Mahavastu, and Saddharma-pundarika were supportive of the ethos created by the mercantile community. It is therefore not suprising that monasteries were well endowed, with huge stupas being built or small ones renovated, and that the Buddhist Sangha became affluent and respected. Some monasteries had large endowments, employing slaves and hired labourers to work the land and labour on other enterprises. The excavating of these structures into hillsides, or the building of free standing ones, would have required the monks to be proficient in various kinds of management skills, collecting donations, gauging technical expertise, controlling labour, maintaining accounts, supervising construction, to say the least. Some were given special designations to carry out this work, but most, being untrained, would have merely supervised the labour. They would doubtless have been assisted by lay followers in these professions. The days when the Buddhist monks lived entirely on alms collected in the morning hours became a distant memory for those in the bigger and richer monasteries who ate regular meals in monastic refectories. [ (Thapar 2003, 270)]

Some nikāyas

As noted, the Hariṣeṇa-period Ajiṇṭhā was synchronically orchestrated by noted people of the times including high priests, wealthy merchants, rājās, ministers, and even the secretary (‘saciva’) of the Mahārāja Hariṣeṇa. The most outstanding monk known to us was Monk Buddhabhadra who had so much of money and political influence that he was able to donate a series of ambitious edifices the likes of which were unprecedented. To his credit, we owe the Caves 21, 23, 24, 25, and 27 (Figure 7). Monk Buddhabhadra has left a donative inscription on the porch wall of Cave 26. No such dedicatory inscriptions are found in the neighbouring caves. Spink’s detailed study of the caves’ sequential developments and structural interrelationships (W. M. Spink 2007) demonstrates that the neighbouring caves (with the exception of Cave 22) were patronised by the same patron who patronised Cave 26. Hence, Monk Buddhabhadra patronised not only Cave 26 but also the adjacent caves 21, 23, 24, 25, 27, and also possibly the highly incomplete Cave 28 on the farthest western end of the cliff. A careful reading of his donative inscription indicates that he might have been the head priest of a sanghārāma somewere in the Aśmaka country. Burgess had first deduced the point:

Buddhabhadra seems to have been no common monk (vs. 7)… the epithet ‘abhijānopapanna’ (vs. 16), which seems to mean that he was of noble family, indicate, too that he was more than a common begging friar. Perhaps, we shall not err, if we assume that he occupied a position analogous to that of a Jaina Sripuj and was the spiritual head of some Bauddha sect. The fact that he mentions ‘his striving for the welfare of the people’ (vs. 16), and ‘his having taken upon himself the care of the people,’ may be adduced in support of this view. [ (Burgess 1883, 133)]

Monk Buddhabhadra had an esoteric yet materially powerful connection with the Aśmaka country. His friendship with Bhavvirāja, the minister of Aśmakarājā, had continued from many previous existences, as we are told in his donative inscription in Cave 26, verses 9-10:

The monk Buddhabhadra has caused (this) temple of Sugata to be made in honour of his parents as well as in honour of (that) Bhavvirāja who served the mighty king of Aśmaka as the latter’s minister, who was attached to him (the monk) in friendship through many successive births…’ [ (Yazdani 1952, 118)]

Buddhabhadra might even have been the regional head of one of the Buddhist nikāyas. Was he a Bahuśrutiya[83] or a Mūlasarvāstivādīn? The recent updates on the efforts of identifying the themes of the painted murals indicate that a majority of the painted themes at Ajiṇṭhā relate most closely to the extant sources of the Mūlasarvāstivāda Vinaya.[84] On the paintings of the fifth-century CE, Schlingloff writes:

The majority of the narrative paintings are based on the vinaya of the Mūlasarvāstivādin[85], and on the poems of Āryaśūra[86], and Aśvaghoṣa[87]. All the works of literature connected with the paintings belong to the Hinayana and accordingly, and accordingly themes that are particularly Mahayanistic were not portrayed, with the possible exception of the Avalokitesvara devotional pictures (Schlingloff 1987, 175-80). The term ‘Mahayana-Caves’ for the later caves, which originates from the time when the literary background of the paintings was still unknown, can therefore only be used as a label; it would be a mistake to form conclusions about the ideology behind the paintings on the basis of this term. [ (Schlingloff, Ajanta: Handbook of the Paintings 2013, vol. I, 75) ]

However, in the saptamāṇuṣībuddha mural of Cave 22, a painted inscription records that the donor was either a follower of the Aparaśaila or Mahāyāna. Cohen’s edition is ‘Success! This is the religious donation of the Śākyabhikṣu Aparaśaila…’[88] (2006, 331). In note 3, the editor observes, ‘perhaps, it read sākyabhikṣo m aparaśailanikāyasya.’[89] However, N. P. Chakravarti’s edition had the last hemistich as ‘Success! This is the meritorious gift of the Śākya monk, a follower of the Great Vehicle…’[90] (1952, 112). Of the two editions, Cohen’s suggested reading probably holds the key. But, an inscription in the interior of Cave 10 suggests the presence of another nikāya at the site, that of the Chetika: ‘Vipaśvin, the Complete and Perfect Buddha. Belonging to Cetika ?rika’[91] (Cohen 2006, 304). Cohen elaborates in the subjoined notes:

The inclusion of the word cetika, in this record has been widely read as an indication that the present donor was a member of the Cetika nikāya, a sub-sect of the Mahāsaṅghika. According to Vasumitra, this sect’s name derives from the fact that its founder lived on Caitya-hill near Amarāvatī, and does not indicate anything about the sect’s doctrinal stand (Jiryo Masudo. “Origin and Doctrines of Early Buddhist Schools: A Translation of the Hsüan-Chwang Version of Vasumitra’s Treatise,” Asia Major. 2 [1925]: 15). In fact, Vasumitra stipulates that one of the Cetika nikāya’s characteristic tenets is, “Even if one makes offerings to a stūpa one cannot acquire great fruits” (Masuda: 38). Assuming this inscription referes to the Cetika nikāya, it is worth noting that the only other nikāya mentioned at Ajanta, the Aparaśaila of inscription #90, was also a sub-sect of the Mahāsāṅghika, also originally from the Amarāvatī region. [ (Cohen 2006, 304)]

From the above it cannot, however, be concluded that the main patrons of Ajiṇṭhā came from diverse nikāyas, for the timing of the aforementioned inscriptions ought not be missed. In both the instances, the murals and the inscriptions are datable to the Period of Disruption, c. 478-480 CE. It was a period when the original patrons had abandoned the site and many ordinary monks and upāsakas made use of the empty spaces for donating some painted and carved Buddhas and stūpas of their own. These were placed on the empty and prominent locations of the abandoned caves. Spink calls them ‘intrusive’ images because they intrude upon and violate the original layout and scheme of things.[92] Thus, in the two instances, it were the new donors—who were not the owners or original patrons of the edifices—that belonged to the Mahāyāna or Aparaśaila and Chetika nikāyas. Thus, even the above evidence does not help us to find out which schools the original patrons belong to. When epigraphy and the murals in question are of no help, we may profitably rely on the sources exposed by Schlingloff (2013) and Zin (2003). Their research on the precise identification of the painted themes reveals unambiguously that a majority of the iconographic sources were drawn from the Mūlasarvāstivāda Vinaya.

The abandonment

Spink has proposed that the history of the later phase of Ajiṇṭhā and that of the Vākāṭakas, particularly Hariṣeṇa, are inextricably linked. The death of Hariṣeṇa cast the death of the monastery. The last quarter of the fifth century proved to be fatal for the two branches of the Vākāṭakas as well as Ajiṇṭhā. Not only for Ajiṇṭhā, but also for the other Buddhist monasteries where patronage activities were going on. How did both the Eastern and the Western Vākāṭakas collapse together? The dominions of the Vākāṭakas are seen in the control of the successive rulers from the early sixth century CE onwards: the Kadambas, Early Kalacūris, Viṣṇukuṇḍīns, Nalas, Mūṇḍaputras, and Kumbhakarṇas.[93] What were the factors that wiped them out? There is no clarity. To add to the mystery is the archaeological, art historical, and epigraphic evidence preserved at Ajiṇṭhā, Banōṭī, Bāgh, Aurangābād, and Ghaṭōtkacha that collectively indicate that the developments in all of these monasteries that were situated in different provinces across the dominions of Hariṣeṇa abruptly ended. The on-site evidence of the rushed as well as immediate abandonment of the work is incontrovertible but so detailed that several volumes will be required simply to document them.[94] They consistently suggest that the work by the original patrons had discontinued unexpectedly and rapidly. At places, the works seem to have been abandoned overnight, which could never be resumed. There are signs as if a catastrophe had suddenly befell on a particular date. Spink has proposed the year to a conjectural c. 477 CE (Spink and Yaguchi 2014, fig. xii). Spink says that the date is not absolute but reconstructed to understand and explain the synchronous and intermittent development of the majority of the caves. He concedes that the proposed time bracket of Ajiṇṭhā’s development could be shifted a few years forward or backward based on further research. A detailed examination of the on-site evidence will indicate that Spink’s formulations are quite deductive. It will also become clear that fixing a conjectural or approximate date is helpful in understanding the chronological sequence of excavation and painting activities. It helps in learning how Ajiṇṭhā is closely related to Aurangābād, Banōṭī, Ghaṭōtkacha, Bāgh, and Kanherī, and how some of the ideas also came from larger monastic establishments, such as Sānchī, Sārnāth, Kanherī, Nāsik, Mathurā, Nālandā, and others in northern India, Pakistan, Afghanistan, and Central Asia.

The unexpected and simultaneous abandonment of as many as five monasteries of the Hariṣeṇa period warrants explanation. Since these monasteries are spread in different geographical locations, all within the kingdom of Hariṣeṇa, the abandonment will have to be attributed to a larger and common factor; it looks like a calamity or a catastrophe had suddenly afflicted the kingdom of Hariṣeṇa. What is more arresting; the situation also seems to have plagued the kingdom of the Eastern Vākāṭakas, since they too declined just as swiftly by the close of the fifth century CE. Therefore, it was a factor that was larger than a particular patron’s edifice; it was larger than a particular monastic establishment; it was larger than the kingdom of Ṛiṣika or Aśmaka; it was even larger than the kingdom of Hariṣeṇa. Within decades, it appears to have pulled the curtains on the Vākāṭaka dynasty as a whole. What was the factor? A likely answer may be related to sociology. But, based on certain contents of the eighth chaper of Danḍīn’s Daśakumāracarita as well as other related pointers, Spink postulates that the downfall of the Vākāṭakas was a handiwork of the Aśmakas who had created a confederation of some disgruntled rajas and attempted to usurp the empire. Spink’s volume one is devoted entirely to this theme (W. M. Spink 2005). In his theory, it is exclusively the political sphere where the theatre of unrest and insurrection is staged. Whereas in the opinion of the present author it is primarily the sociological sphere where the theatre of a potential ethnic strife is played out which resulted in political instability and anarchy.

Shall we ask then: Was there a sociological disturbance? Since Hariṣeṇa has been eulogized in the Ajiṇṭhā and Ghaṭōtkacha inscriptions,[95] it is clear that he was supportive of the Buddhists.[96] Could it be possible that he was probably murdered for the same reason? Were the Buddhists chased away under a widespread sociological upheaval? Only future research can answer the questions.




Table: Vākāṭaka chronology - different versions


Shastri (1997, 212)

Bakker (1997, 169-71)

Spink (2005, 166)

Undivided dynasty


Vindhyasakti I

250–275 CE



Pravira alias Pravarasena I

275–335 CE

275–335 CE



Nandivardhana branch


Gautamiputra I




Rudrasena I

335–355 CE

335-360 CE


Prithivisena I

355–385 CE

360-395 CE


Rudrasena II

385–395 CE

395-405 CE

385-390, allows for year 5 inscr.

Acting monarch Prabhavatigupta


405-419 CE

390-405, estimated 15 yr regency

Yuvaraja Divakarasena

395–410 CE




410–420 CE

419-422 CE

405-410, as brother’s reign was long

Pravarasena II

420–455 CE

422-457 CE

410-445, allows for year 32 inscr.


455–480 CE

457-461 CE

445-455, no evidence on length

Prithvisena II

480–500/505 CE

475-477 CE, subordinate to Vatsagulma

455-475, allows for year 17 inscr.


478-495, independent


Vatsagulma branch


Sarvasena I

325–355 CE



Vindhyasakti II alias Vindhyasena

355–400 CE

360-400 CE


Pravarasena II

400–425 CE

400-450 CE, subordinate to Nandivardhana


Sarvasena II

425–455 CE



455–480 CE

450-460 CE



480–510 CE

460-478 CE

460-477, Devasena ruling 458

Sarvasena III



478-483, died before 486

Compilation by R.K. Singh. The dates are approximate.




Captions for figures

Figure 1          ‘Map of Vākāṭaka sites’ from (Spink and Yaguchi 2014, xv) and (Weiner 1977, 11)

Figure 2          ‘Distribution of Vākāṭaka inscriptions and coins’ from (Shastri 1997, map)

Figure 3          ‘India in the age of the Guptas and Vākāṭakas’, from (W. M. Spink 2009, fig. 1), adapted from (Schwartzberg 1992, plate III.D.1)

Figure 4          Ajiṇṭhā: Cave 18, front view

Figure 5          Ajiṇṭhā: the Śrāvakayāna caves

Figure 6          Ajiṇṭhā: Caves 17-20

Figure 7          Ajiṇṭhā: Cave 26-complex

Figure 8          Ajiṇṭhā: Cave 26A (below Cave 26), rediscovered in 2012

Figure 9          Ajiṇṭhā: Cave 26-complex in c. 480 CE, conjectural 3D model minus the details, from (Singh 2012a, 61)

Figure 10       Caves Upper 6 (above) and Lower 6 (below), from (Burgess 1880, plate XXXII)

Figure 11       Cave 26-complex (isometric plan): conjectural reconstruction by Spink and Ajeet Rao (artwork by Ajeet Rao and Nitin Veturkar; courtesy of Walter M. Spink and Samskara)




Works Cited

Archaeological Survey of India. 2011. “World Heritage Sites - Ajanta Caves.” Archaeological Survey of India. Accessed November 9, 2015. http://asi.nic.in/asi_monu_whs_ajanta.asp.

Bakker, Hans T., ed. 2004. The Vakataka Heritage: Indian Culture at the Crossroads. Groningen: Egbert Forsten.

—. 1997. The Vakatakas: An Essay in Hindu Iconology. Groningen: Egbert Forsten.

Bronkhorst, Johannes. 2007. Greater Magadha: Studies in the Culture of Early India, HDO series. Leiden and Boston: Brill.

Burgess, James. 1883. Buddhist Cave Temples and Their Inscriptions, Archaeological Survey of Western India series. Vol. IV. London: Trubner & Co.

Burgess, James. 1880. “Cave Temples of Western India.” In The Cave Temples of India, by James Fergusson and James Burgess, 165-524. London: W. H. Allen.

Burgess, James, and Pandit Bhagwanlal Indraji. 1881. Inscriptions from the Cave Temples of Western India, Archaeological Survey of Western India series. Mumbai: Government Central Press.

Chakravarti, N. P. 1952. The Painted Inscriptions. Vol. IV, in Ajanta: Monochrome Reproductions of the Ajanta Frescoes Based on Photography, by G. Yazdani, 111-112. London: Oxford University Press.

Chhabra, B. Ch. 1952. The Incised Inscriptions. Vol. IV, in Ajanta: Monochrome Reproductions of the Ajanta Frescoes Based on Photography, edited by Gulam Yazdani, 114-118. London: Oxford University Press.

Cohen, Richard Scott. 2006. Appendix: Ajanta’s Inscriptions. Vol. II, in Arguments about Ajanta: Ajanta, History and Development -- Handbuch Der Oorientalistik (HDO) series, by Walter M. Spink, edited by J. Bronkhorst, 273-339. Leiden: Brill.

Collins, Mark. 1907. “Geographical Data of the Raghuvamsa and Dasakumaracarita.” Unpublished doctoral thesis, Philosophical Faculty, University of Leipzig, Leipzig.

DeCaroli, Robert. 1995. “An Analysis of Dandin's Dasakumaracarita and its Implications for both the Vakataka and Pallava Courts.” Journal of the American Oriental Society, 671-678.

Deshpande, Brahmanand. 2002c. Dasakumaracharita and the Date of the Ajintha Caves. Vol. I, in Studies in Epigraphy, by Brahmanand Deshapande, 80-91. Aurangabad: Kailash Publications.

—. 1992. “Dating of Ajintha Caves in the Context of Vakataka Decline.” Maharashtra Pathik, July: 8-14.

Deshpande, Brahmanand. 2002b. The Vakataka Rule in Khandesh. Vol. I, in Studies in Epigraphy, by Brahmanand Deshpande, 65-79. Aurangabad: Kailash Publications.

Deshpande, Brahmanand. 2002a. The Vakataka-Traikutaka Relations. Vol. I, in Studies in Epigraphy, by Brahmanand Deshpande, 54-64. Aurangabad: Kailash Publications.

Deshpande, Brahmanand. 1996. Vakataka Nrpati Harisena: Ek Punarmulyankana. Vol. II, in Shodhamudra, by Brahmanand Deshpande, 29-41. Aurangabad: Kailash Publications.

Dhavalikar, M. K. 1984. Late Hinayana Caves of Western India. Poona: Deccan College Postgraduate and Research Institute.

—. 2003. Sanchi, Monumental Legacy series. New Delhi: Oxford University Press.

Dutt, N., ed. 1947-50. [Mūlasarvāstivāda-Vinaya] Gilgit Manuscripts, The Kashmir Series of Texts and Studies, 71. Sanskrit edition. Vol. III. IV vols. Srinagar.

Fergusson, James, and James Burgess. 1880. The Cave Temples of India. London: W. H. Allen.

Ghosh, A. 1967. Ajanta Murals: An Album of Eighty-Five Reproductions in Colour. repr. 1987, 2004. Edited by A. Ghosh. New Delhi: Archaeological Survey of India.

Gnoli, Raniero, ed. 1977-78. [Mūlasarvāstivāda-Vinaya] The Gilgit Manuscripts of the Sanghabhedavastu, Being the 17th and Last Section of the Vinaya of the Mūlasarvāstivādin, Serie Orientale 49. II vols. Roma: ISIAO.

Gokhale, Shobhana. 1992. “Epigraphical Evidence on Harishena's Conquest from Kanheri.” In The Age of the Vakatakas, by Ajay Mitra Shastri, 269-278. New Delhi: Harman Publishing House.

Google Inc. 2013. Latitude 20°33'9.62"N, longitude 75°42'0.68"E. Vers. 7 October. Accessed November 13, 2015.

Goyal, Shankar. 2012. “An Investigation into the Feudal Nature of the Vakataka Economy: A Study of the Epigraphic Evidence.” Presidential Address: XXXVII Annual Congress of the Epigraphical Society of India (24-26 Feb 2012). Baroda: Epigraphical Society of India and Dept of Archaeology and Ancient History, M. S. University of Baroda. 1-31.

Indraji, Pandit Bhagwanlal. 1881. Inscriptions from the Cave-Temples of Western India with Descriptive Notes, &c. Repr. Indian India, Delhi, 1976. Bombay: Archaeological Survey of Western India, printed at Government Central Press.

Jamkhedkar, A. P. 2009. Ajanta: Monumental Legacy series. New Delhi: Oxford University Press.

—. 1991. “Some Reflections on Professor Spink’s Chronology at Ajanta.” Maharashtra Pathik, April.

Johnston, E. H., ed. 1928. The Saundarananda of Asvaghosa, Punjab University Oriental Publications series. Translated by E. H. Johnston. Vol. XIV. Lahore: The University of the Panjab. Repr. Tokyo 1972, Motilal Banarasidass 1975.

Kannal, Deepak H. 1996. Ellora: An Enigma in Sculptural Styles. New Delhi: Books & Books.

Kern, Hendrick J., ed. 1891. The Jātaka-Mālā, Harvard Oriental series. Repr. Cambridge, Mass. 1914. Vol. I. Boston: Harvard University Press.

Khandalavala, Karl. 1990. “The History and Dating of the Mahayana Caves of Ajanta.” Maharashtra Pathik, September: 18-21.

Khandalavala, Karl. 1991. The Chronology of Caves 16, 17, 19, 26, 1 and 2 at Ajanta and the Ghatotkacha Cave. Vol. I, in The Art of Ajanta: New Perspectives, edited by Ratan Parimoo, Deepak Kannal, Shivaji Panikkar and Jayaram Poduval, 105-129. Delhi: Books & Books.

Khoroche, Peter, trans. 1989. Once the Buddha was a Monkey: Aryasura's Jatakamala. Chicago-London: The University of Chicago Press.

Kulke, Harmann. 2004. “Some Thoughts on State and State Formation under the Eastern Vakatakas.” In The Vakataka Heritage: Indian Culture at the Crossroads, by Hans T. Bakker, 1-9. Groningen: Egbert Forsten.

Malandra, Geri H. 1993. Unfolding a Mandala: The Buddhist Cave Temples at Ellora. New York: State University of New York.

Mirashi, V. V. 1960. Historical Data in Dandin's Dasakumaracarita. Vol. I, in Studies in Indology, edited by V. V. Mirashi, 165-177. Nagpur: Vidarbha Samshodhan Mandal.

Mirashi, V. V., ed. 1963. Inscriptions of the Vākāṭakas, Corpus Inscriptionum Indicarum series. Vol. V. Ootacamund: Government Epigraphist for India.

Mitra, Debala. 1996. Ajanta. 11th. New Delhi: Archaeological Survey of India.

Nagaraju, S. 1981. S. Buddhist Architecture of Western India, c. 250 B.C. — c. A.D. 300. Delhi: Agam Kala Prakashan.

Nigam, L S. 2004. “Impact and Extension of Vakataka Art in Chhattisgarh (Daksina Kosala).” In The Vakataka Heritage: Indian Culture at the Crossroads, by Hans T Bakker, 143-156. Groningen: Egbert Forsten.

Raven, Ellen M. 2004. “Kings in Copper.” In The Vakataka Heritage: Indian Culutre at the Crossroads, by Hans T Bakker, 19-31. Groningen: Egbert Forsten.

Ruegg, David Seyfort. 2004. “Aspects of the Investigation of the (earlier) Indian Mahayana.” Journal of the International Association of Buddhist Studies XXVII (1): 3-62.

Sahapedia.org. 2015. Ajanta Caves: Dr. Rajesh Kumar Singh in Conversation with Prof. Naomichi Yaguchi. Video. Edited by Phani Damara. Sahapedia.org. Pune, 25 September. Forthcoming.

Schiefner, Anton. 1876-1878. “[Mūlasarvāstivāda-Vinaya] Indische Erzählungen.” Bulletin de l'Académie impériale des sciences de St.-Pétersbourg. Vols. XXI-XXIV. XXI: 433-93; XXII: 123-38; XXIII: 1-70, 529-65; XXIV: 449-508. Repr. Übersetzungen aus dem tibetischen Kanjur, Sinologica Coloniensia series, XXV, Otto Harrassowitz Verlag, Weisbaden, 2007.

—. 1882. [Mūlasarvāstivāda-Vinaya] Tibetan Tales Derived from Indian Sources, of the Kahgyur by A. Schiefner, Trübner's Oriental series. Translated by W. Ralston. Boston: James R. Osgood & Co. Repr. 1925, 1990.

Schlingloff, Dieter. 2013. Ajanta: Handbook of the Paintings. III vols. New Delhi: IGNCA and Aryan Books International.

—. 1987. Studies in the Ajanta paintings: Identifications and interpretations. Delhi: Ajanta Publications.

Schwartzberg, Joseph E., ed. 1992. A Historical Atlas of South Asia. 2nd. Oxford: Oxford University Press.

Shastri, Ajay Mitra. 1991. Names of Ajanta: Modern and Ancient--A Re-appraisal. Vol. I, in The Art of Ajanta: New Perspectives, edited by Ratan Parimoo, Deepak Kannal, Shivaji Panikkar, Jayaram Poduval and Indramohan Sharma, 67-70. New Delhi: Books & Books.

—. 1997. Vakatakas: Sources and History, Great Ages of Indian History series. New Delhi: Aryan Books International.

Singh, Rajesh Kumar. 2014. “Ajanta's Antiquity: Sources and Problems Relating to History, Epigraphy, and Early Chronological Developments.” PhD thesis, Department of Art History and Aesthetics, Faculty of Fine Arts, M.S. University of Baroda, Baroda.

—. 2012b. An Introduction to the Ajanta Caves: With Examples of Six Caves. 1st, paperback. Baroda: Hari Sena Press.

—. 2009. “Numerating the Sailagrhas at Ajanta.” Journal of the Asiatic Society of Mumbai, 122-126.

Singh, Rajesh Kumar. 2009. “Some Grey Areas in Fixing the Date of the Vakataka Phase of the Ajanta Caves.” Edited by R. D. Choudhury. Kala: the Journal of the Indian Art History Congress, 2008-2009 (Indian Art History Congress and Sharda Publishing House) XIV: 69-85.

—. 2012a. “The Early Development of the Cave 26-Complex at Ajanta.” South Asian Studies, 37-68.

—. 2015c. “The Times of Ajintha - Facets of Anthropological and Sociological Background.” Sahapedia.org. Forthcoming.

Sircar, D. C. 1965. No. 63--Ajanta Cave Inscription of the Time of Harishena. Vol. I, in Select Inscriptions, by D. C. Sircar, 449-455. Calcutta: University of Calcutta.

Speyer, J. 1895. The Jatakamala: Garland of Birth-stories of Buddha, Sacred Books of the Buddhists series. Vol. I. London: Henry Frowde. Repr. Delhi 1971, 1990.

Spink, Walter M. 2005-2014. Ajanta: History and Development series. Edited by J. Bronkhorst. VII vols. Lieden: Brill.

Spink, Walter M. 2006. Arguments About Ajanta, Ajanta: History and Development series – Handbuch der Orientalistik (HDO) series. Edited by J. Bronkhorst. Vol. II. VIII vols. Leiden: Brill.

—. 2007. Cave by Cave: Ajanta, History and Development series -- Handbuch der Orientalistik (HDO) series. Edited by J. Bronkhorst. Vol. V. VIII vols. Leiden: Brill.

—. 2009. Painting, Sculpture, Architecture Year by Year: Ajanta, History and Development series – Handbuch der Orientalistik (HDO) series. Edited by J. Bronkhorst. Vol. V. VIII vols. Leiden: Brill.

Spink, Walter M. 1992. “The Achievements of Ajanta.” In The Age of the Vakatakas, edited by Ajay Mitra Shastri, 177-202. Delhi: Harman.

—. 2005. The End of the Golden Age: Ajanta, History and Development -- Handbuch der Orientalistik (HDO) series. Edited by J. Bronkhorst. Vol. I. VIII vols. Leiden: Brill.

Spink, Walter M., and Naomi Yaguchi. 2014. Defining Features, Ajanta: History and Development series. Edited by J. Bronkhorst. Vol. VI. VIII vols. Lieden: Brill.

Stadtner, Donald M. 2004. “Vidarbha and Kosala.” In The Vakataka Heritage: Indian Culture at the Crossroads, by Hans T Bakker, 157-165. Groningen: Egbert Forsten.

Stietencron, Heinrich von. 2004. “Questions Addressed to Walter Spink.” Chap. 8 in The Vakataka Heritage: Indian Culture at the Crossroads, edited by Hans T. Bakker, 107-108. Groningen: Egbert Forsten.

Suzuki, T., ed. 1955-61. [Mūlasarvāstivāda-Vinaya] The Tibetan Tripitaka, Peking Edition. Tibetan trans. Vols. XLI-XLIV. Tokyo-Kyoto.

Thapar, Romila. 2003. The Penguin History of Early India from the Origins to AD 1300. New Delhi: Penguin Books India.

UNESCO. 1992-2015. “Ajanta Caves.” UNESCO: World Heritage Convention. Accessed November 9, 2015. http://whc.unesco.org/en/list/242.

Vaidya, P. L. 1959. Buddhist Sanskrit Texts. 2nd ed. 1999. Vol. XXI. Darbhanga: The Mithila Institute.

Weiner, Sheila L. 1977. Ajanta: Its Place in Buddhist Art. Berkeley: University of California Press.

Yazdani, Gulam. 1952. Ajanta: Monochrome Reproductions of the Ajanta Frescoes Based on Photography, with an Explanatory Text by G. Yazdani and Appendices on the Painted and Incised Inscriptions by N.P. Chakravarti and B.Ch. Chhabra. Vol. IV. IV vols. London: Oxford University Press.

Zin, Monika. 2003. A Guide to the Ajanta Paintings: Devotional and Ornamental Paintings. Vol. II. II vols. Delhi: Munshiram Manoharlal Pvt. Ltd.

—. 2003. Ajanta: Handbook der Malereien - Devotional and Ornamental Malereien. Vol. I. II vols. Wiesbaden: Harrassowitz Verlag.

Zin, Monika, and Dieter Schlingloff. 2007. Samsaracakra: Das Rad der Wiedergeburten in der Indischen Uberlieferung. Dusseldorf: Ludicium Verlag.







[1] Archaeological Survey of India’s coordinates are slightly imprecise in the official guidebook, “Lat. 20°32’N and Long. 75°45’E” (Mitra 1996, 1), and quite incorrect on the official webpage that provides “75°40’N; 20°30’E” (Archaeological Survey of India 2011).

[2] Spink writes, ‘. . . like the ancient province of Asmaka just to the south, and Anupa to the north, ancient Risika, in which Ajanta was included, was already part of the extensive domains that Harisena inherited when he came to the power’ (W. M. Spink 2005, 7).

[3] Mirashi identifies the Aurangabad region as Mūlaka (Mirashi 1963, 124).

[4] For Mirashi, ‘south of Mūlaka lay Aśmaka (modern Ahmadnagar and Bhir districts)’ (Mirashi 1963, 124).

[5] For details, see (Singh 2009).

[6] See, for example (Mitra 1996, 5).

[7] For Monk Buddhabhadra’s donative inscription, see (Chhabra 1952, 114-18).

[8] For Spink’s conjectural line drawing, see (Singh 2012a, fig. 26) and for Singh’s 3D sketch, see (Singh 2012a, figs. 35-37).

[9] Cave 26 inscription, verses 7-8 (Chhabra 1952, 115).

[10] Mitra writes: ‘This [Cave 6] is a double-storeyed monastery’ (Mitra 1996, 36). The italic is supplied by the present author.

[11] For a detailed study, see (W. M. Spink 2007, 83-113) and (Singh 2014, vol. I, 330-48).

[12] The English renderings represent the present author’s understanding of the lexicons as used in the given contexts.

[13] Cave 26 inscription, vs. 6 (Chhabra 1952, 115).

[14] Cave 16 inscription, vs. 27 (Mirashi 1963, 109).

[15] Cave 16 inscription, verses 29, 32 (Mirashi 1963, 109).

[16] Cave 17 inscription, vs. 24 (Mirashi 1963, 127).

[17] Cave 12 inscription, line 3 (Indraji 1881, 68) and (Burgess 1883, 116). Only upā…. is preserved and the last akṣaras (letters) are illegible. Indraji’s reading as upāsaya (Snk. upāśraya) has been accepted by later editors.

[18] Inscriptions in caves 16, vs. 22 (Mirashi 1963, 109) and Cave 26, verses 14, 19 (Chhabra 1952, 116).

[19] Cave 16 inscription, vs. 25 (Mirashi 1963, 109).

[20] Cave 12 inscription, line 3 (Burgess 1883, 116).

[21] Inscriptions in Cave 17, verses 24, 29 (Mirashi 1963, 127) and Cave 16, vs. 31 (Mirashi 1963, 109).

[22] Cave 10 inscription, line 2. Full inscription: धमदेवस . . . नस [।१] पसादा द[ा]नम् पवजितस [॥२], ‘dhamadevasa . . . nasa [|1] pasādā d[ā]naṃ pavajitasa [||2], The pasādā is the gift of Dhamadeva . . . for the renunciates’ (Cohen 2006, 296). The scribe has spelt पसादा pasādā, which is the Prākṛit of प्रासाद prāsāda in Sanskrit. Cohen in the subjoined notes has identified a problem in the donor Dhamadeva’s claim that he donated the pasādā. The claim can not be true, since the same temple has other donative inscriptions by other donors who donated some parts of the temple. It is, therefore, inconceivable that Dhamadeva donated the whole pasādā.

[23] Inscriptions in Cave 4, line 1 (Cohen 2006, 284) and Cave 17, vs. 1 (Mirashi 1963, 124).

[24] Cave 17 inscription, vs. 24 (Mirashi 1963, 127).

[25] Cave 26 inscription, vs. 13 (Chhabra 1952, 116).

[26] Cave 16 inscription, vs. 24 (Indraji 1881, 71) and (Mirashi 1963, 109). In note 9 of the same page, Burgess added: ‘This [the Chaitya-Mandira] is no longer found: it may have been structural and outside.’ The remark is representative of the modern historian’s misconception that the word caitya or stūpa means only the symbolic representation of the Buddha and not the anthropomorphic one. Even in the Cave 17 inscription, the usages of the words ‘munirājacaityam’ and ‘stūpavihāra’ obviously refer to the image shrine that is housed there. It is clear that for the scribes and makers of Ajinṭhā there was no discrimination, polarity, or antethesis between the symbol (that is, the so-called stūpa) and the image. Iconographically too the fact is most clearly corroborated, for in the caves 19 and 26 the stūpa and the image stand together in one synthetic unity.

[27] Cave 17 inscription, vs. 22 (Mirashi 1963, 126).

[28] Cave 17 inscription, vs. 27 (Mirashi 1963, 127).

[29] Cave 26 inscription, verses 7-8 (Chhabra 1952, 115, 117 n.7).

[30] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[31] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[32] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[33] The word suvīthi also means a long porch, a street. In the given context, Mirashi’s translation as given above is most appropriate.

[34] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[35] Vedikā means many things depending on the context: sacred altar, quadrangular spot in the courtyard of a temple, the portion of a building, which is covered by the walls, rail, balustrade, ledge, etc. Mirashi translates it as a ledge (1963, 111). There is neither any ledge nor any rails or balustrade to visible in the Cave 16. Even the eaves outside the porch cannot be interpreted as a ledge or vedikā. Therefore, the contextual meaning ought to be as given above.

[36] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[37] Cave 16 inscription, line 19, vs. 24 (Mirashi 1963, 109).

[38] Cave 10 inscription (Burgess 1883, 116).

[39] Cave 10 inscription (Cohen 2006, 296-97).

[40] With little previous research to depend on, James Burgess could hardly hide his pain while dating the caves, the dating that was to perpetuate for many decades: ‘We shall probably not be far wrong if we attribute the generality of the paintings in Caves I., II., XVI., XVII., &c. to the sixth century, which we may gather from the style of alphabetical character used in a few painted inscriptions and names of figures is the date of these paintings. The later pictures may then be attributed to the seventh century, and the earlier ones, in Caves IX. And X., may possibly date even as far back as the second—in the time of the Andhrabhritya kings, the great patrons of Buddhism in the first three centuries of our era’ (Burgess 1880, 285).

[41] For a recent overview, see (Ruegg 2004, 5-18, 28-31).

[42] Many publications in the last few decaes follow such a dating style. They are too many to cite.

[43] Nagaraju has dated the earlier phase to c. 220 BCE to 100 BCE (Nagaraju 1981, chart v). Yazdani classified the earliest activity in the second century BCE: ‘There are in fact specimens of painting in Cave X at Ajanta with inscriptions which, according to expert epigraphists have been unanimously assigned to the middle of the second century B.C. The art of painting even in these early examples is fairly well developed, and indicates practice and a continuity of tradition over hundreds of years.’ [ (Yazdani 1952, 2)] Dhavalikar once maintained, ‘It is generally assumed that the Hinayana activity in Western India came to a grinding halt by the close of the second century A.D. only to be resumed again by the Mahayanists in the latter half of 5th century, if we accept the chronology for the latter group of caves at Ajanta proposed by Walter M. Spink. What happened during the hiatus of about three centuries is not known.’ [ (Dhavalikar 1984, 1)] Schlingloff’s dating is clear from the titles of the two sections of his magnum opus on Ajanta: ‘1. The paintings of the Second Century BC. 2. The Paintings of the Fifth Century A.D.’ [ (Schlingloff 2013, 3-4)]

[44] Some of these earlier caves were reworked in the fifth-century phase with the so-called ‘intrusive’ paintings, scultpures, inscriptions, etc. For details on these, see (W. M. Spink 2006, 199-271).

[45] Of these, caves 9 and 10 are temples, and the rest (caves 12, 13, and 15A) are monastic dwellings. One more cave has recently come to light in 2012. This one is a modest cave without any painting, inscription, or carving. It is a cell that had been buried in debris near the riverbed below Cave 26 (Figure 9). It was first included in a PhD thesis (Singh 2014, vol. II, figs. 212-13), and the same is being published here for the first time. We do not know which phase it belongs to. Prima facie, it looks to belong to the earlier phase.

[46] ‘वासिठिपुतस कट हादिनो घरमुख दानम्’ (Burgess 1883, 116); the wooden facade is the donation of the son of Vāsiṭhi (author’s translation).

[47] Apart from Amarāvatī and Nāgārjunikoṇḍā, there are several other lesser known or newly rediscovered Buddhist sites of the ancient Āndhradeśa that belong to different ages. Due to inadequate studies, it is difficult to speak of the relationship of Ajinṭhā with these sites. Some of them are Allurū, Alugōḷū, Bāvīkoṇḍā, Bhāṭṭīprōḷū, Chinnāganjam, Chanḍāvāram, Ḍupaḍu, Ḍhuḷikaṭṭā, Garikapaḍu, Vaḍḍamānu, Neḷakōṇḍapaḷḷī, Guḍīvāḍā, Gajurabaṇḍa, Phanigirī, Uppuguṇḍuru, Peḍḍaganjam, Rāmā Tīrtham, Sankram, Salihuṇḍam, and Ṭhoṭlākōṇḍā.

[48] Northern India and northern highway were known as Uttarāpatha. Southern India and southern highways were known as Dakṣiṇāpatha.

[49] Yazdani thought that the development at Ajinṭhā was a continuous affair: ‘There are specimens in Caves IX and X which tell a continuous story of the development of technique from the second century B.C. to the fifth century A.D…. There is a complete absence of blue, which does not appear at Ajanta until the fourth or fifth century A.D.’ [ (Yazdani 1952, 2)] ‘.... This painting [Cave X, Ṣaḍadanta Jātaka] probably belongs to the third century A.D…. [Śyāma Jātaka] may be assigned to a period extending from the last quarter of the third century A.D. to the first quarter of the fourth, that is, to a time immediately preceding the sovereignty of the Vākāṭakas…’ [ (Yazdani 1952, 3)] The dating is found again on p. 1, ‘In the fourth and fifth centuries A.D. Ajanta, under the enlightened patronage of the Vākāṭaka kings, became an important centre of Buddhist religion and art…’ [ (Yazdani 1952, 1)] On p. 2, Yazdani speaks of a sixth century development, but rules out for any seventh century development: ‘Through lack of royal patronage, art at Ajanta began to decline rapidly in the latter half of the sixth century A.D., and there is no vihara or chaitya at Ajanta, which can be assigned with certainty to the seventh century A.D.’ [ (Yazdani 1952, 2)]

[50] For example, see (Schlingloff 2013, 3-4). Monika Zin writes, ‘The complex of monasteries at Ajanta, built in the 1st century B.C. and 5th century A.D., appears so unique to us today…’ (Zin 2003, 1).

[51] For questions and criticisms of Spink’s ideas, see Khandalavala [ (1990), (1991)], Deshpande [ (1992), (1996), (2002a), (2002b), (2002c)], Jamkhedkar (1991), Shastri (1997, xii, 209), Bakker (1997, 37 n. 127), and Stietencron (2004). Many of them are reproduced with Spink’s replies in (W. M. Spink 2005).

[52] See (Bakker 1997, 34).

[53] See, Singh [ (2009), (2012a, 37), (2012b, 28), (2014, 25)].

[54] A leading book on early Indian history (Thapar 2003) does not mention Hariṣeṇa.

[55] The present is merely a sketch of the political conditions during the fifth-century phase of Ajiṇṭhā. The contents are based on current research including the latest corpus of epigraphy, on-site archaeological evidence, numismatics, iconographiy, literary and textual sources, historiography, and deductive exploration of the relevant historical material. If any significant error is brought to the attention of the author, he is willing to reexamine and rectify the error where applicable.

[56] According to the Thālner grant of Hariṣeṇa, the king bore the title of ‘Mahārāja’ (Shastri 1997, 112). Notably, Vandhyaśaktī I bore the title of Samrāṭ and Dharma-Mahārāja; Pravaraṣeṇa I bore the titles of Dharma-Mahārāja and Hāritīputra. The rest of the kings including Hariṣeṇa are styled simply Mahārāja.

[57] According to Walter M. Spink, Hariṣeṇa reigned from c. 460 to c. 477 CE ( (W. M. Spink 2006); (W. M. Spink 2005-2014)). Hans T. Bakker has endorsed this: ‘We assume with Spink that Devasena’s son Hariṣeṇa ascended the throne of Vatsagulma in about AD 460’ (Bakker 1997, 34). Yazdani and Mirashi had proposed a slightly different dating: ‘King Harisheṇa, according to eminent historians (V. V. Mirashi, Hyderabad Archaeological Series, No. 15, p. 3), flourished in the period c. A.D. 475-500’ (Yazdani 1952, 1).

[58] In the Groningen Oriental Studies Colloquium on the Vākāṭakas (Groningen 6-8 June 2002), it was reconciled that the two branches of the Vākāṭakas may be conveniently termed as the Eastern Vākāṭakas and the Western Vākāṭakas, where the former had the capital at Nandīvardhana (Nagardhan, taluka Rāmṭek, district Nāgpur) and the latter had the capital at Vatsagulma (district Wāshim). See (Kulke 2004, 3).

[59] Some donative inscriptions at Ajiṇṭhā clearly mention that the donations were for the sangha. But most donative inscriptions use the phrase देय धर्म्मो deya dharmmo (donation to the Dharma). The Sangha and the Dharma are parts of the Triratna, i.e. the Buddhist trinity: the Buddha, the Dharma, and the Sangha. The image of the Triratna has been systematically depicted in Buddhist art and archaeology, Ajiṇṭhā being no exception. The epigraphic and iconographic depictions convey that there must have been the presence of the Sangha in some form at Ajiṇṭhā, a point that is ignored in scholarship. The point leads to a question. If there was a Sangha there, what role did it play in the development and functioning of the monastery?

[60] For the Varāhadeva’s inscriptions, see (Mirashi 1963, 103-119).

[61] स कुन्तलावन्तिकलिङ्गकोसलत्रिकूटलाटान्ध्र[परान्त] जानिमान [।] U - U - - U U शौर्यविश्रुतानपि स्वनिर्देशगुणाति - U – [18 ||]; sa kuntalāvantikaliṉgakōsalatrikūṭalāṭāndhra[parānta] jānimān [|] U – U — U - U śauryaviśrutānapi svanirdeśaguṉāti – U – [18||] [ (Mirashi 1963, 108)]. Sircar reads शौर्यविश्रुतान् as –U नैरधृतान् (1965, 453); परान्त parānta is supplied by Gokhale (Gokhale 1992).

[62] Kuntala = Kannada country, district round Banavāsi. Avanti = western Mālwā, capital Ujjain. Kalinga = the country between Mahānadī and Godāvarī; in the narrow sense = Purī-Ganjam region of Orissā. Kōsala = South Kosala = modern Chattisgaḍh and adjoining parts of Orissa, including Rājpur-Sambalpur-Bilāspur region. Trikūṭa = country to the west of Nāsik. Lāṭa = central and southern Gujarāt, between Mahi and Tāpi; Nausāri-Bharuch region. Āndhra = the Telugu speaking country to the south of Godāvarī. Aparānta = the northern Koṇkan. [ (Sircar 1965, 453, n. 2) and (Mirashi 1963, 107)]

[63] V. V. Mirashi in CHI, III, Part I, p. 144. See also CII, V, p. xxxi. Vide for a somewhat exaggerated account, Vākāṭaka Nṛipati āni Tyānchā Kāla, p. 59. [Note in original.]

[64] Walter M. Spink, “The Achievements of Ajanta”, AV, p. 117. [= (W. M. Spink 1992, 117); note in original.]

[65] On the subject of the disputed verse 18 of Varāhadeva’s donative inscription, Bakker offers a reconostructive or conjectural reading for supplying a sense for the missing words: ‘He [i.e. Hariṣeṇa], [stands above] these foreign kings of Kuntala, Avantī, Kalinga, Kōśala, Trikūṭa, Lāṭa, and Āndhra, renowned fro their valour and courage, their own names […]’. [ (Bakker 1997, 35)]

[66] See Deshpande [ (Deshpande 1992), (Deshpande 1996), (Deshpande 2002c), and (Deshpande 2002b)].

[67] For the donative inscription of Cave 17, see (Mirashi 1963, 120-129).

[68] For the donative inscription of Cave 20, see [ (Indraji 1881, 76-77), (Burgess 1883, 132)].

[69] Author’s transliteration in Devanāgarī.

[70] For the Cave 4 inscription, see (Cohen 2006, 284).

[71] For the changing contours of Āryāvarta and Greater Magadha, see (Bronkhorst 2007).

[72] For the contours of Greater Magadha, see (Bronkhorst 2007, 1-9).

[73] There is a dispute on whether the Vākāṭakas minted coins. In 1997, Shastri published a hoard of coins (1997, 136-146) that he identified as those of the Vākāṭakas. The hoard was further studied by Ellen M. Raven (2004). However, Shankar Goyal believes the hoard is fake (2012, 4, 7-14). Goyal argues that the Vākāṭakas had a feudal sociological set up without major urbanization. They were using cowries or inexpensive metal coins of the predecessors and contemporaries.

[74] The term Brāhmaṇism is as problematic as Hinduism. The former springs from reductionism; it denotes the ‘ism’ of one of the four varṇas that constituted the traditional varṇa-based Indian sociology; it leaves the other three undesignated. Whereas the latter term is anachronistic. Romila Thapar puts it better: ‘To use the general term Hinduism at this stage [early India] is historically something of an anachronism. The term “Hindu” was not in use in the early first millennium AD, and those who were supporters of what today we call ‘Hindu’ sects used their sectarian lables to identify their religion. Therefore, they identified themselves by the broader labels of Vaishnava and Shaiva or within these, by the narrower labels of Bhagavatas, Pashupatas and so on. The consciousness of a religious identity was that of the sect and not of an all-inclusive religion incorporating every sect.’ [ (Thapar 2003, 275)]

[75] See (Bakker 1997) and (Bakker 2004).

[76] The full verse is: ‘. . . . नत्यद्भुतपुण्यराशिः [|*] चक्रे भुव स्तूपविहारभूषां दानोदयैश्चार्थिजनप्रमोदम् || [२२||*]; . . . natyadbhutapuṇyarāsih [|*] chakre bhuva stūpavihārabhūṣāṁ dānodayeśćārthijanapramodam || [22||*]; . . . He, who has a very marvelous store of merit. . . . adorned the earth with stūpas and vihāras, and caused the joy of suppliants by conferring gifts (on them).’ [ (Mirashi 1963, 126, 129)]

[77] There was a brick monastery at Ajiṇṭhā itself, across the river. In 2012, the present author spotted the base of an ancient brick structure in the Ṭhāṇā village adjacent to Fardapur, a few km from the Ajiṇṭhā caves. This notice enables us to relate the Ṭhāṇā village with Cave 12’s inscription: ‘थानको देयधम्मम घनामदडस वणिजस सौववरको सौपा [सथो], Ṭhānako deyadhamam Ghanāmadaḍasa vaṇija[sa] sauvavarako saupā[satho],  The meritorious gift of a dwelling with cells (aparavaraka), and a hall (upāsaya), by the merchant Ghanāmadaḍa [of Ṭhāṇā]’ [ (Burgess and Indraji 1881, 68); (Burgess 1883, 116)]. At Ghaṭōtkacha, I have found fifth-century bricks scattered in the fields above the ghāṭs. There must have been a brick monastery there too, which is destroyed. There were brick structures at Ellora; some of them can be spotted even today. They are fast on the verge of destruction. Thus, there were many monasteries made of wood, stone, and brick, which are perished. It cannot be held that the rock-cut monasteries were the only type. Fa-hien and Xuanzang mention many monasteries in each region they visited.

[78] In 1880, James Burgess dated the Buddhist caves of Ellora to 350-550 CE (Fergusson and Burgess 1880, 368). In 1883, he dated them to 450-700 CE (Burgess 1883, 4). Yazdani dated them to seventh century CE: ‘The Buddhist shrines of Ellora excavated during the seventh century A.D. show clear influence of the Brahmanic architecture of South India…’ [ (Yazdani 1952, 2)] Geri H. Malandra closely follows Yazdani’s dating: ‘The site [Ellora] developed in three major phases, an early Brāhmaṇical phase (c. 550 to 600), a Buddhist phase (c. 600 to 730), and a last phase (c. 730 to 950), which comprised both Brāhmaṇical and Jain excavations sponsored by the Rāṣṭrakūṭa dynasty, the only known ancient patrons of Ellora.’ [ (Malandra 1993, 5)] However, Deepak Kannal has dated them to the sixth century CE (Kannal 1996, 130). Spink is close to Burgess, Yazdani, and Malandra in dating the Buddhist caves of Ellora to the seventh century CE (Spink, personal communicaiton).

[79] The eighth chapter of Danḍīn’s Daśakumāracarita is called Viśrutacarita. This chapter, like other works of Danḍīn is written in a fictional style. However, the plot of this fiction appears to be based somewhat on real historical events. The events were the gradual collapse of the Vākāṭaka dynasty. Danḍīn seems to have changed the names of the historical personas, and additionally introduced lots of dramatic and fictional elements to suite the objectives of a carita (ornate biography). Yet, to a discerning eye, there may be found relevant historical and geographical data, especially when studied with the archaeological evidence of Ajiṇṭhā and other historical sources relating to the Vākāṭakas. The historians in favour of this view are Collins (1907), Mirashi [ (1960), (1963)], (W. M. Spink 2005, 119-162, 169-178, 393-411), and (DeCaroli 1995). The topic is disputed. For an overview of the dispute, see (Singh 2009, 72-76).

[80] See (Singh 2015c).

[81] Suggested by (Spink, personal communication) and (Malandra 1993, 123).

[82] Yazdani observed that ‘The Chalukyas who succeeded the Vākāṭakas in the sovereignty of the Deccan during the first half of the sixth centry A.D. were perhaps not as favourable to the Buddhist religion as their predecessors, although the shrines which the Chalukyas built at their capital in Badami are largely copied from the Buddhist viharas.’ [ (Yazdani 1952, 2)]

[83] The word bahuśrutaḥ occurs in Buddhabhadra’s dedicatory inscription. We wonder whether it refers to the one among the eighteen nikāyas of the mainstream Buddhism. Chhabra seems to have left the word out from his translation of the verse: ‘यो बुद्धशासनगतिम् समबुद्ध्य जातो भिक्षुर्व्ययस्याभिनवेभिजनोपपन्नः [।*] बहुश्रुतः शीलविशुद्धचेता (लोकस्य मोक्षाय कृ)ताधिकारः ॥ [१६ ॥*], ‘Yo Buddhaśāsanagatiṁ samabuddhya jātō bhikṣurvya(rvva)yasyābhinavebhijanōpapanna[ḥ |*] bahuśrutaḥ ś(ī)laviśuddhachetā (lōkasya mokṣāya kṛi)tādhikāraḥ || (16 ||*), He who, born of a noble family, endowed with great learning, with his mind purified by righteous conduct (competent to lead the people on the path of liberation), having perfectly mastered the course of the Buddha’s teachings, became a monk in his early age.’ [ (Chhabra 1952, 116, 118, vs. 16)]

[84] See (Schlingloff 2013) and (Zin and Schlingloff 2007).

[85] For Mūlasarvāstivāda-Vinaya, Schlingloff cites German ed. (Schiefner 1876-1878) = Eng. Trans. (Schiefner 1882), Snk. ed. (Dutt 1947-50), Tib. ed. (Suzuki 1955-61), Eng. ed. (Gnoli 1977-78).

[86] For Āryaśūra’s Jātakamālā, Schlingloff cites inter alia (Kern 1891), (Speyer 1895), (Vaidya 1959), (Khoroche 1989).

[87] For Aśvaghoṣa’s Saundarānand, Schlingloff cites inter alia (Johnston 1928).

[88] ‘सिद्धम् देयधर्म्मो यम शाक्यभिक्षो म अपरशैल… नियस्य मातापितृ…, Siddhaṁ Deyadharmmō-yaṁ śākyabhikṣō m aparaśaila.i… nīyasya mātāpitṛi…’ [ (Cohen 2006, 331)]

[89] ‘शाक्यभिक्षो म अपरशैलनिकायस्य’ [ (Cohen 2006, 331)]

[90] ‘शाक्यभिक्षो[र]-म[हा]यान… [नियस्य मातापितृ]…, Śākyabhikshō[r*]-ma[hā]yāna [niyasya mātā-pitṛi]…’ [ (Chakravarti 1952, 112)]

[91] ‘विपश्वी सम्यकसम्बुद्ध चेतिक[…]रिकस्य [|*], vi(pa)śv(i) samya(k)saṁbu(ddhaḥ) cetika.rikasya.’ [ (Cohen 2006, 304)]

[92] For a detailed study of the so called ‘intrusive’ images at Ajiṇṭhā, see (W. M. Spink 2006, 116 ff.).

[93] Shastri (1997, 209-210).

[94] See Spink’s volumes (W. M. Spink 2005-2014). For some additional evidence relating to Caves 1, 2, 16, and 17, see (Singh 2012b); for Cave 26, see (Singh 2012a).

[95] For the Ajiṇṭhā Cave 16 inscription, vs. 17, see (Mirashi 1963, 108, 110)]. For the Ghaṭōtkacha cave inscription, vs. 13, see (Mirashi 1963, 117, 119). For the Ajiṇṭhā Cave 17 inscription, vs. 21, see (Mirashi 1963, 126, 129). Also, see (Burgess 1883, 128).

[96] Even Yazdani felt that the Vākāṭakas were supportive of the Buddhists; little did he know then that it was just Hariṣeṇa and not the entire Vākāṭaka dynasty that patronized the Buddhists. In Yazdani’s words: ‘The support given by the Vākāṭaka kings to the propogation of the Buddhist faith and art may have been secured through the religious zeal of Hastibhoja and his son Varāhadeva, the ministers of the last two kings of the dynasty, Devasena and Harishena, who apparently had embraced the Buddhist religion, and whose names are preserved in the inscriptions of Cave XVI at Ajanta and of the vihara at Ghaṭōtkaca in the vicinity (n. 3: V. V. Mirashi, Hydrabad Archaeological Series, No. 14, p. 2, and The Ghaṭōtkaca Inscription, p. 8-17).’ [ (Yazdani 1952, 1)]