IN THE PERIODsince this book was first published in 1961 there have been, not unexpectedly, fresh interpretations of the Mauryan period and of the activities of Aśoka in particular; new data is available as well. My own ideas on the subject have also undergone some change. I thought it appropriate to add an Afterword to this new edition of the book, rather than to attempt redrafting the book and incorporating the data and interpretations since 1961. There are therefore few changes in the earlier text except for some small corrections. In this Afterword I have surveyed the epigraphic and archaeological evidence, discovered and discussed in the last half century, some of which constitutes the more important fresh data. I have also referred to studies concerning this evidence and the treatment of other sources, such as the texts. I have concluded with a brief discussion of some recent views on Aśoka and Buddhism and on the Mauryan period generally.
In the listing of sources, the epigraphic data in the form of statements made by Aśoka has priority. Minor Rock Edicts (Minor REs) were discovered at Bahapur (New Delhi), Panguraria (Sehore Dt), Nittur (Bellary Dt), and Udegolam (Bellary Dt). These have been published and (p.274) discussed both individually and in collated studies.1 There is also a useful listing and locating of Aśokan inscriptions found upto the mid-80s.2 Major Rock Edicts (MREs) were found at Sannati (Gulbarga Dt) and at Kandahar (Afghanistan). Another edition of the Separate Edicts (SEs) also occurs at Sannati. A fragmentary inscription apparently in the style of the Pillar Edicts (PEs) series has been located at Amarāvatī and another is said to have come from Buner (near Peshawar) the last thought to be a fake. Aramaic inscriptions occur at Laghman. A claim was made that an Aśokan inscription was found at Ghuggus in the Chandrapur District of Maharashtra, but it remains untraceable.3 An attempt was made to read what were referred to as inter-linear inscriptions of medieval times in Sri Lanka, pertaining to the Mauryas, but these readings have been rejected.4 A complete coverage of all the Asokan epigraphs and their locations and a detailed annotated bibliography is now available in a valuable study.5
The length of the text of the Minor REs varies. In addition to the standard text there is, at some places concentrated in three districts of Karnataka, a longer text. It is also at sites in this area, as noticed earlier, that a kharoṣṭhī knowing scribe declares his presence. The sites in these three districts of Karnataka are close to earlier settlements, and to the Hutti and Kolar gold mines, gold reefs and areas known for semi-precious stones. The coinciding of areas geologically known to be gold-bearing and the distribution of Aśokan inscriptions is striking.6 (p.275)
The mines could have been under the control of the state, necessitating a nucleus of administration in the area to supervise the workings. Hence the references to the mahāmattas and the kumāras. In addition, local élites may have been the suppliers of resources and this would have required an administration largely limited to the senior levels. Reference is also made to the edict being dispatched throughout the king’s domain—sava-paṭhaviyaṁ.7
The reference to 256 nights occurs in most but not all the versions, the inference being that the standard text was issued for engraving at one point in time; nevertheless some changes were introduced in each version in the ordering of the text and the language. Sircar’s reading of the Bahapur version is that the 256 nights/days refer to the number of days since the relics were installed by the king.8 Narain’s alternative reading of the figure 256 as referring to the years after the death of the Buddha, in the Ahraura inscription, has not met with general acceptance.9
The additional part has some orders from the king. The rājūka was to be instructed by the mahāmattas and would in turn instruct the people. This included the raṭṭhikāni, possibly those with authority over a rāṣṭra, or local chiefs. The reference to instructing the elephant riders may have been to those associated with such chiefs. The Mauryan official would then have been the intermediary between the imperial authority and the local. The imperial attempt is not to change local society and bring it in line with mainstream Mauryan society, but rather to communicate with local society. The qualifier of putta in (p.276) Pāli, added to the name of a people, has been taken to mean that they belong to a tribe. Thus MRE II mentions the Kelalaputto and the Satiyaputto, and although the suffix putta is not used for the Coḍa and Pāṃdya, they are included in the general reference. (Interestingly Aśoka uses the Prākrit forms of the names and not the contemporary Tamil forms such as Cera-man and Adiya-man). In contrast, the Hellenistic kings are referred to as the yonarāja. The instruction consisted of the rudiments of dhamma: obedience to mother, father and teachers, mercy towards living creatures, speaking the truth and following the dhamma. Mention is made of the brāhmaṇas but not of the śramanas, who were presumably still rare in Karnataka.
The Panguraria inscription was located at a monastic site.10 A short inscription has been read as addressed to the kumāra Saṃva as a preface to the Minor RE but this reading has been questioned. Mention is made of the king named Piyadassi, which indicates a personal name. In the Nittur inscription he is called rājā asoka whereas in the Udegolam version he is rājā asoka devānampiya. The Panguraria inscription repeats the statement that this edict was to be inscribed wherever there are rocks and stone pillars.
Both the MREs at Sannati and Kandahar are significant but for different reasons. The stone slab of the Sannati inscription, was being used as a pīṭha pedestal for an image in a temple and was damaged by the cutting out of a section in the middle to hold the tenon at the base of the image. Both surfaces of the slab are inscribed.11 One surface has the partial texts of the MREs XII and XIV (Face A). The other surface has fragments of the SE I and the SE II (Face B), previously known only at the sites in Kaliṅga, and presumably replacing MRE XIII. The two groups of edicts may have came from different sources, since in the MREs la is used in place of ra whereas in the SEs ra is retained in rājā. According to Norman the positioning of the two (p.277) SEs reinforces the theory that SE I was engraved after SE II. That they have similarities and dissimilarities with the Dhauli and Jaugada versions would suggest that there was a master copy from which these versions derived.12 It had been assumed that since the SEs had so far occurred only at Kaliṅga they were a substitute for the MRE XIII in which the king expresses his remorse at the suffering caused by the campaign in Kaliṅga. But the occurrence at Sannati raises the question of why they should have been included at this site. Was there another campaign associated with this area which has not been mentioned in the sources? Was it an administrative error which led to its inclusion? Or were these edicts intended for administration in other parts of India as well, as Norman maintains, although they may have been for reading out and not for engraving. However if they were replacing any MRE then they must have been intended for engraving.
At Kandahar, apart from the Greek and Aramaic bilingual reported earlier (Kandahar I), a Greek version of part of the MRE XII and the opening section of MRE XIII, have also been found (Kandahar III). This is a fragment from a larger set of stone slabs. These are not translations but renderings in a condensed form, with some omissions.13
Fragments of what could be PEs have been recorded at Amarāvatī and Kandahar. The Amarāvatī inscription is on a stone which is not from Chunar (from where transportation would have been extraordinarily difficult), but from the closer Nallamalai range. The stone has traces of the characteristic Mauryan polish. The text, only partially preserved, does not conform to any of the known edicts although the palaeography and the language are akin to Aśokan brāhmī and Prākrit.14 (p.278)
Possibly it was intended as a new series or even another version of the existing edicts, meant for the peninsula, since this is the only one in that area. The site is close to the port of Dharanikoṭa which was doubtless in contact with other ports along the east coast and which also gave access to the Krishna valley and the Kolar goldfields. If the inscription was Aśokan then possibly the foundations of the stūpa could be dated to the Mauryan period, a date which is endorsed by excavations. However, it could also have been an edict of a successor of Aśoka, although Sircar’s attempted reading suggests sentiments similar to those of other edicts of Aśoka.15
At Kandahar, an inscription has been read as an Aramaic rendering of a part of PE VII (Kandahar II).16 The inscription is in Aramaic but with glosses to explain the terms used in Prākrit. An earlier reading of an Aramaic inscription from Lampaka, introduced the possibility of glosses.17 It would seem that kharoṣṭhī was not familiar in this area. The readings of these inscriptions have suggested parallels with a number of terms occurring in the MREs and PEs, which would in turn suggest that the Greek and Aramaic inscriptions were of a later date than the ones in Prākrit. Two Aramaic inscriptions from Laghman (Laghman I and II) refer to the name Priyadarśin as does the one from Taxila.18 The Aramaic inscriptions are concentrated along the Khyber and in Kandahar, whereas the Greek inscriptions remain in the vicinity of Kandahar. The Greek-speaking population seems to have preferred living in the cities (p.279) whereas the Aramaic-speaking people were to be found both in the cities and along the routes. The Greek presence in Kandahar is further confirmed by a Greek inscription on the base of a Hellenistic statue of this period.19 The earlier bilingual Graeco-Aramaic edict has also been examined in some detail and is said to be contemporary with the MREs.20 An up-to-date discussion of the Hellenistic background which incorporates this region, is available in a recent study.21
Recently a puzzling fragment of PE VI in brāhmī was said to have come from Buner near Takht-i-Bahi in the north-west—puzzling as it would be the sole inscription in brāhmī in this area.22 Its language agrees with other such edicts from the Ganges plain. Its genuineness is however not above doubt, especially as it was not found in situ. Taddei’s explanation is that there was a brāhmī-knowing scribe travelling with Aśoka in the north-west and he was asked to engrave the edicts. The inscription is on schist and not on Chunar sandstone. It would seem therefore that the main distribution of the PEs is still largely confined to the Ganges plain. However some attempts were being made to convey their ideas to the north-west. The locations were connected not only to the north-western routes but perhaps with central Asia as well, judging by the recently discovered Kuṣāṇa presence in Gilgit and Chitral.
The discovery of Greek and Aramaic inscriptions points not only to Mauryan control over parts of Afghanistan and the presence of Greek and Aramaic-Iranian speakers in the area, but also to Aśoka’s emphasis on maintaining a dialogue with people whom he regarded as important. It has been suggested that these were the areas described (p.280) as yonakamboja, where yona clearly referred to Hellenized Greeks and kamboja to the Iranians. The latter may have been a local community associated with Mazdaism.23 The significance of these inscriptions increases because they provide translations of various terms which have been controversial in the Prākrit versions. Humbach argues that contemporary Zoroastrian terms as well as ideas from Greek philosophy were used to render Buddhist concepts.24 It could also be said that the intention may have been to show the universality of some concepts. The word dhamma is translated as eusebeia meaning ‘piety, loyalty, reverence for the gods and for parents’. It has also been explained as that sentiment which the group or the individual entertained towards certain specified obligations such as respect for ancestors and parents and participation in civic sacrifices and offerings. Interestingly there is no reference to the Buddha, which might have been expected if the edicts were centrally concerned with propagating Buddhism among those unfamiliar with it. If the dhamma of Aśoka was to be translated as ‘law’ then as asserted by Schlumberger the Greek should have been nomos.25 The intention was to introduce dhamma to the local people and describe what was meant by it, while using their own idiom. The style is literary and the phraseology echoes Pythagorean and Platonic thought. The term eusebeia is used in a context very similar to the Aśokan in an inscription of Antiochus of the first century B.C. where he states that he considers eusebeia the safest and most pleasurable possession for all men and that he regards saintliness as his most faithful guardian. It would seem that he was familiar with the Greek inscriptions of Aśoka at Kandahar.26
The Greek translation of pāsaṃḍa is diatribe which at this time carried the meaning of a sect, but not necessarily a heretical sect, as was also the case in Prākrit. It refers more to philosophical schools (p.281) than to doctrinaire religious groups. The meaning of pāsaṃḍa was also to change in the latter direction in the post-Mauryan period. The absence of the conjoint bramaṇaśramaṇa reflects the king’s awareness that these two categories were absent in the land of the Yonas, to which he draws attention in MRE XIII. It is significant that the terms used for dhamma in Greek and Aramaic which are eusebeia and qsyr, have no doctrinal or sectarian meaning.27
Humbach prefers to call the Aramaic of these inscriptions, Armaeo-Iranian, as the Aramaic has an admixture of Iranian. Since these inscriptions generally belong to the period subsequent to Aśoka’s support for Buddhism, the earlier reading of the Taxila inscription should more correctly be, ‘Our lord Priyadarśin, the king’, which occurs as a formula. Dhamma in the Aramaic inscription, has been translated as ‘the conduct of the good’, or as Truth. Reference is also made to pious men being above judgement. This has been interpreted in the context of MRE IV, which these inscriptions seem to parallel, as referring to the officers of the law who have the power to judge.28 Again an alternative reading could be that it refers to the Zoroastrian notion of judgement after death. Since this idea was less familiar to the Greeks, it is not introduced in the Greek translation. Was Aśoka stating that those who observe the dhamma or the conduct of the good, would be regarded as pious and therefore exempt from judgement after death? This would show a degree of interest in and sensitivity to, the belief of those who were his subjects in this part of the empire, which is remarkable. (The same sensitivity is not however evident in his dialogue with his southern subjects, who were also at this time unfamiliar with Buddhism and with Prākrit. He may have been hampered in the use of a south Indian language because of the absence of a script specific to the language). That he was concerned with the after-life is evident from various versions of the Minor RE where he speaks of svarga, and interestingly not of nirvāṇa, as the goal of human existence. There is little reference in the Greek and Aramaic (p.282) inscriptions to officials or administration. This could be because the area was only recently and loosely under Mauryan control. Given this and the obvious presence of Greek speakers, the epigamia in the treaty between Seleucus and Candragupta would more likely have been a jus conubii, the freedom for Greeks and Indians to intermarry, rather than being restricted to a marriage alliance within the royal families.29 The discovery of the Aramaic inscriptions, suggesting Iranian connections, makes the governorship of Tuṣāspa in Saurashtra more plausible. It does raise the question however of how an Iranian would have fared in governing a non-Iranian population, unless one argues that Iranian settlements along the borderlands would already have had an admixture of Indian ways of life with which Iranians were familiar. Tuṣāspa may have been from a family sent by the Achaemenids to administer the deltaic area of their Indian territory.
An epigraphic confirmation of the existence of the Satiyaputtas comes from a Tamil-brāhmī inscription of the period just after Aśoka, at the site of Jambai in the South Arcot District. It records the donation of a cave shelter for monks and the donor refers to himself as Adiyan Neduman Anci, the Satiyaputta.30 He was also referred to as Adiya-man in the texts. It is significant that although the inscription is in Tamil, the equivalent name in Prākrit is also used. The poems of the Cankam collection would have been approximately contemporary with the edicts.
Attention has been drawn to the contemporary histories of the reign of Firuz Shah Tughluq and mention of his shifting and setting up Aśokan pillars from their original provenance to buildings which he had had built. The ones from Topra and Meerut at Delhi are well-recorded. The one said to have been brought from Kauśāmbi to Allahabad is again under discussion. But the histories mention two others: at Hissar (Punjab) and at Hissar (Haryana), whose identity as Aśokan pillars is close but not conclusive. They are now located in (p.283) mosques in the two towns. It has been suggested that the inscribed surface may have flaked off.31
The origin of the brāhmī script remains subject to periodic discussion. A detailed and careful study of Mauryan brāhmī supports the sub-continental use of the script although with some variations which are not consistent with regional characteristics.32 C.S. Upasak maintains that the script is neither borrowed nor entirely indigenous but evolved from both the functionality of the phonetic form of the language as developed by Sanskrit grammarians and contact with traders using a script. The innovation would have been pre-Mauryan as it was well-developed by the time of Aśoka. A survey of early Indian palaeography also supports the idea that the script was partially evolved from the structure of the language but also partially borrowed, probably from the Semitic.33 The semi-syllabic character of brāhmī was suited to the structure of the language, although there are problems in the placing of conjuncts. Another view is that the brāhmī script was invented in the early third century B.C. to facilitate Aśokan administration and that the idea and technique of writing came from the west with a borrowing from the Achaemenid model.34 If this was the case it is likely that the inscriptions would have been uniformly in Prākrit and brāhmī irrespective of local usage in language and script. The relationship between kharoṣṭhī and brāhmī has also to be considered. If kharoṣṭhī was also an invention to facilitate administration, then it could have been widely used elsewhere in the Mauryan empire for Prākrit edicts. This would have eliminated the need to evolve yet another script for administrative functions. Given the wide use of Aramaic in the Achaemenian empire and the fact that Gandhāra and much of the trans-Indus region were claimed as part of this empire, it is unlikely that a script was not known earlier, at least in the north-west. That (p.284) Aśokan brāhmī was well-developed would suggest a start in pre-Aśokan times, although its widespread use was Aśokan. A recent study of early Indian palaeography rejects Semitic links and revives an earlier theory of possible Greek influence on the earliest forms of brāhmī, the invention of which suggests that the script dates to the first Mauryan ruler.35 Opinion remains divided on the origin of brāhmī.
Pre-Mauryan royal inscriptions may not be discovered since pre-Mauryan politics may not have required these, but graffiti on potsherds, especially in conjunction with Northern Black Polished Ware, is a likely archaeological context for a script. Such graffiti at Amarāvatī has been dated to the early Mauryan period.36 Graffiti in brāhmī on local potsherds are also reported from Anuradhapura dating to levels of c. 450–350 B.C., thus earlier than at other archaeological sites. The reading suggests Prākrit names. Interestingly there are also sherds with graffiti symbols similar to those on megalithic B-and-R ware from peninsular India.37 These very early dates are so far restricted to this site but possibly other graffiti may provide evidence of a pre-Mauryan brāhmī. Similarly, Sircar dates the Piprahwa Relic Casket inscription to the third century B.C.,38 although such an early reading has been doubted. An inscriptional label on a casket would point to an established script. Inscribed potsherds suggest uses other than administration for it.
Nevertheless the late development of a script with a long lacunae from the Harappan period may have been because the pre-Mauryan litterati, the brāhmaṇas, relied on the oral tradition. The invention of a script would only have come from the need for a written tradition and probably drew on existing scripts in the neighbourhood. According to one view, even the Aśokan edicts despite being engraved, judging by the (p.285) speech breaks, suggest they were meant for recitation.39 The recording of the Buddha’s teaching eventually took the form of texts although the precise periods when various sections of the Canon came to be written remains uncertain. The questioning of a script in pre-Aśokan times comes largely from the statement by Megasthenes that the Indians did not know the art of writing. But this statement contradicts Nearchos.40 There has also been a substantial debate on whether the grammar of Pāṇini could have been composed in the absence of a script.
References to Indian traders in Iran and Mesopotamia during the sixth and fifth centuries B.C. establish the familiarity which some Indians had with Aramaic and the North Semitic script. A colony of Indians near Nippur is known, some of whom carried out military functions under their own chiefs, and an Indian woman is said to run a tavern in Kish.41 A contract of the fifth century concerning a female slave has an endorsement in an unknown script, which, it has been suggested might be an early form of brāhmī.42
Familiarity with written texts is evident from the edicts, where the term dhamma-lipi carries the notion of a written text and not just a brief proclamation. Orders to inscribe appropriate rock surfaces and pillars is again suggestive of an established experience with writing. If the script had been invented in the early third century for purposes of administration, and for propagating ideas on social behaviour and beliefs, then it is unlikely that the major effort would have been to engrave on stone, infinitely more difficult than writing on birch-bark or cloth. If the purpose was to facilitate administration then presumably much more would have been stated about the administrative functioning than is actually done. The SEs belong more to this category but these are only two, and the Schism Edict refers to the making of multiple copies for the saṃgha and for the officials. (p.286)
Some of the edicts are addressed to the ayyaputta/āryaputra, the kumāra and the mahāmattas, but their contents include far more than merely matters of governing. Rather than being personal statements written in a conversational style, the edicts would also have been drafted in a more official language. They are likely to have been dictated to a writer who in turn conveyed them to engravers in various parts of the Mauryan domain. Engravers of inscriptions are seldom highly literate so small mistakes and local variants occur in these inscriptions.
Literacy would in any case have been limited at this time to some brāhmaṇas, Buddhist monks entrusted with recording the teachings of the Buddha, merchants, officials and some other professions requiring a modicum of writing. The edicts therefore were not intended to be read by ordinary people but to be read out to such gatherings. The inscription would then be symbolic of a statement of power in an oral society, all the more so in the south where no concessions were made to the local language. It may well be asked whether the idea of reading out a text did not come from readings by Buddhist monks to the laity. That some elements of a Buddhist Canon existed is evident from the Bhabra edict.43 Recourse to writing would have helped in the centralization of state administration and communication with distant areas. Interestingly inscriptions in the Ganges plain do not occur in clusters in the same way as they do in the frontier areas in Karnataka and in the north-west. Where the edicts were in the vicinity of Buddhist monasteries, the monks would have read them out to the laity. Given the diversity of the empire, the fact that a body of texts were being propagated in every part would at least have encouraged the notion of a common ideology of persuasion.44
One of the fascinating aspects of the Aśokan pillars is that some of them came to be treated as a kind of historical palimpsest. What is thought to be the pillar from Kauśāmbi, now at Allahabad, carries a number of Aśokan inscriptions: the PEs I to VI, the Schism Edict and (p.287) the Queen’s Edict. It also carries the famous praśasti of Samudragupta as well as an inscription recording the visit of Rājā Birbal, presumably after Akbar had constructed the fort at the confluence. And there is an inscription by Jahangir giving his ancestry. In addition there are numerous short ‘scribblings’ as Cunningham called them, mentioning dates from the thirteenth century onwards.45 There must have been some perception of the pillar incorporating the past that it comes to be treated as an agency of legitimizing the present. If this legitimation had arisen out of its being regarded solely as an axis mundi or a liṅgam then it would have been venerated but would not have carried further inscriptions on its body.
Important epigraphic evidence comes from the Aramaic and Greek inscriptions which provide not only an insight into literacy, and the governing of the north-west, but also a further dimension to the discussion of the dhamma of Aśoka. The dhamma emerges as conceptually complex, enriched by his personal association with Buddhism and other belief systems of the time, and also as a policy reaching out to many dimensions of public life. Its significance as an ideology, both of the ruler and of his times, is thereby enhanced. The parallel concepts to the Prākrit inscriptions as well as the particular concepts of the Greek and the Aramaic, suggest meanings and nuances which would endorse the readings which have been made earlier in this book.
The evidence from archaeology is potentially a major source now on the Mauryan period. The archaeological activity in the last three decades has filled in some spaces largely with evidence of pre-Mauryan periods. For the Mauryan period excavations provide some data on settlements and the distribution of artefacts, although far more can be expected from future excavations.46 Information on urban centres (p.288) is relatively more accessible but of a limited nature.47 A more broad-based study placing urbanization in the context of the archaeology of early historic south Asia, provides not only a useful summary of the archaeological data but suggests the possibilities of the emergent picture from archaeological sources.48 Excavations pertaining to this period are almost entirely vertical and therefore useful for obtaining a stratigraphy of the site but of less help in reconstructing social and economic processes. There is therefore considerable need for more horizontal excavations.
The archaeological counterpart to the Mauryan period is included within the span of the Northern Black Polished Ware (NBPW) which occurs at an earlier date at some sites but is generally taken to have commenced by about 600–500 B.C. in the middle Ganges plain, diffusing and continuing until the second century B.C. Frequently associated with this pottery are other artefacts such as punch-marked coins, terra-cotta figures, beads of particular kinds and the increasing use of iron. These provide a similarity of items across a large area although their locations include other regionally identified artefacts as well. They have been taken as evidence of the Mauryan period in their later levels, especially at sites in the peninsula where the pre-Mauryan cultures continue. Since the archaeological span is prior to the dynastic time bracket, there can be some disjunction in co-relating archaeological data with the chronology of the dynasty. The evolution of urban centres which date to the pre-Mauryan period become in some cases more important in the Mauryan period. The hierarchy of sites within an area is difficult to ascertain as there has not been much in the way of systematic explorations with this question in mind.49
The NBPW is now given an Indian provenance, originating in the middle Ganges plain. Apart from other factors its association is with what is being termed the second urbanization, set in the context of (p.289) what is referred to by some as the Ganges civilization.50 Its technical quality has been attributed to the constitution of the clay and the method of firing.51 The NBPW reaches a high quality in the early Mauryan period although its use in the sixth century B.C. would be an acceptable date.52 Being a luxury ware it has a wide distribution although the associated ware is often more informative on local connections. The quality of the ware gives it precedence, but in terms of numbers there is generally a greater use of other wares such as the Black-and-Red, Black slipped ware, Red wares, thick grey wares and other more localized pottery. The distribution of the latter tends to be more regionally defined.
As part of the potter’s craft, terra-cotta figures of the elephant, horse, bull, ram, dog and nāga were popular. Models of carts, wagons and chariots occur in terra-cotta and in metal. Hand-modelled human figures are common at many places, especially representations of what might be mother-goddesses. The latter seem to occur in large numbers at many sites in the Ganges plain and could perhaps be identifiable with some religious intention.53 This provides an interesting commentary on what constituted popular religious belief and worship in this period. A distinction between pre-Mauryan and Mauryan terra-cottas is based on the larger number of human female figurines and their greater ornamentation as well as some element of a possible Hellenistic touch.54
Copper mirrors, dishes, ornaments of various kinds, and the usual arrowheads, antimony rods, nails and needles, occur at some sites. Iron objects suggest a more professional use. Weapons included caltrops, (p.290) and household objects ranging from nails to axes and adzes. Iron hoes occur in larger numbers at Megalithic sites in the peninsula. Crucibles from Ujjain Pd. II and smelting furnaces point to iron working on some scale. Mauryan period workings at mines near Udaipur in Rajasthan have provided some striking results.55 Silver and lead were the main products at Agucha and Dariba, and zinc was mined at Zawar (as it is to this day). Agucha has given C-14 dates going back to the third century B.C. and is in the vicinity of what is described as a Mauryan settlement. Dates for the mine at Dariba go back to the end of the second millennium, but the period for the most intensive work appears to be in the latter part of the first millennium. The timber revetment to the mine is thought to be perhaps the largest timber structure from antiquity. The evidence suggests a prodigious scale of lead/silver production in the Zawar area, which, it has been proposed, may well be tied to the demand for silver in the making of punch-marked coins. The technology of working the mines can be reconstructed from the remains, and points to sophisticated techniques.
A survey of ancient quarries in the vicinity of Chunar has been rewarding in terms of reconstructing the making of the pillars.56 Incomplete monoliths cylindrical in form, suggesting the proto-type raw material for the working of the Aśokan pillars, were found together with evidence that the quarries were used in later periods as well. The choice of Chunar is explained by its closeness to the river which facilitated the transportation of these monoliths. The cylinders have marks which might be mason’s marks or inscriptions. One carries a brief kharoṣṭhī inscription which might be of the Mauryan period. This, it is suggested, may point to quarry workers being from the north-west where there existed a familiarity with stone cutting and carving. But this also brings to mind the small presence of kharoṣṭhī (p.291) registered by the engraver of some of the Minor REs in Karnataka. Possibly it was the engravers who came from the north-west.
The western plain was essentially wheat-growing although some rice was found at Atranjikhera in the doāb.57 If this was locally cultivated rice, then it raises the possibility of the area being conducive to double-cropping. This in turn has a bearing on agriculture and has been used as evidence to suggest a fiscal crisis in the Mauryan economy requiring to be met with an enhanced revenue. Wet rice cultivation was concentrated in the middle and lower Ganges plain. It is thought to have been a staple in the peninsula together with millets, especially rāgi. The cultivation of cotton is indicated by the presence of cotton cloth at various sites.
Arrangements for irrigation were locally controlled as major irrigation works are sparse. Inscriptions mention some large scale irrigation works, a canal built by the Nandas in Orissa and the Mauryan dam on the Sudarśana lake in Saurashtra. The latter has yet to be conclusively identified,58 but the dimensions given and the frequency of repairs suggest that the dam irrigated an area that was large but not extensive. Excavations at Kumrahar also revealed a canal connecting the palace with the river, doubtless used for transporting goods.59 A canal lined with bricks and said to be of the Mauryan period was located at Besnagar from an early excavation. Small tanks and catchment areas are more common in the peninsula.
Mauryan expansion of trade and the facilitating of contacts between distant areas accelerated exchange particularly in small items such as glass, clay and shell beads, as also beads of agate, carnelian (including etched beads), jasper and lapis lazuli, ivory hairpins and combs, and ear-ornaments of bone. Silver and gold ornaments are more limited. It is likely that much of the silver was used in the making of punch-marked coins and silver was not widely available in India. Gold supplies (p.292) were mainly confined to Karnataka and possibly the auriferous soil of some parts of the extreme north. It is thought that the diamond mines of the peninsula were also worked.
The earliest fortified settlements appear to have been those at Ujjain, Kauśāmbi, Rajghat, Rajgir and Campa.60 The direction indicates an early route. Gradually other centres came to be fortified. Perhaps the most impressive were the wooden posts and nailed cross-beams from Pāṭaliputra. City sizes are difficult to determine even with relative precision in the absence of surveys. Literary sources often give exaggerated figures, more to emphasize the notion of an expansive settlement than to be accurate. City walls become essential and initially have a mud core with revetments of brick, rubble and stone and occasionally even wooden sleepers. Some have bastions, gateways with guard rooms or a moat. The wall was in part defence, in part flood control—where the town was on a river bank—and in part a demarcation between town and country, although the last was by no means definitive.
Burnt bricks of a relatively large size are used for dwelling complexes. Wedge-shaped bricks are used for circular structures. Bricks introduce a notion of permanency and an attempt to control the weathering of structures. Room sizes tend to be small. Houses are fitted with wells, some brick-lined and some ring wells—a construction of large terra-cotta rings, or with soakage jars open at both ends and placed one on top of the other. Ring wells become a characteristic feature at many sites. Some varieties of wells are linked to drains.
These characteristics of urban centres enable a recognition of those that were more important and those that were marginal. In the north-west Taxila/Bhir Mound and Charsada (the ancient Puṣkalāvatī), remain small but are important. Bhir Mound was not a planned city but did have differentiated streets, lanes, drains and other features pointing to a civic organization.61 Kandahar had earlier urban beginnings as a frontier town on the Indo-Iranian borderlands. The settlements at Ropar and Sugh in the Punjab were probably on the route to the (p.293) Ganga-Yamuna doāb. In Rajasthan, Bairāṭ was a major Buddhist site at which excavations have provided evidence of an improvement in living conditions at what are said to be the Mauryan levels, as at some other sites. The Aśokan connection is further indicated by fragments of what seem to be Chunar sandstone pillars and an umbrella of stone with a Mauryan polish.62 Excavations near Rupnath have provided evidence of a substantial improvement in living and the extensive use of artefacts in Mauryan levels.63
In western India, the port of Bṛghukaccha was associated with the manufacture of ivory and shell items and beads. Central Indian sites include the important centre of Ujjain as well as other sites such as Tripuri, Vidiśā/Besnagar, Maheshwar and Eran, some of which have a Chalcolithic history, but are in all cases pre-Mauryan settlements. Maheshwar was at the ford across the Narmada to Navdatoli and was on the route going south, the dakṣināpatha. The route may have passed through Panguraria and Sannati. There will now be an increased interest in the possibility of Mauryan artefacts being found in the vicinity of Sannati.
Moving eastwards, there are traces of copper mining in Singhbhum and also some evidence of iron working, but more detailed excavation is required before the date of the working of these mines can be guaged. In the lower Ganges valley the port of Tamluk/Tāmralipti played a major commercial role and together with Chandraketugarh had links with coastal Orissa, Andhra and further south, and eventually with Sri Lanka. Mahāsthān is currently under excavation. In Orissa the extensive site of Śiśupālgarh has been partially excavated but its possibilities have not been adequately investigated. Orissa and Andhra have been the subject of an excellent study of secondary state formation in which the process is seen from pre-Mauryan to post-Mauryan times.64 (p.294)
The north-western Deccan carries traces of contacts with the north at sites such as Nasik, Prakash, and Bahal. In the south-eastern Deccan and in Karnataka, Megalithic settlements both precede and coincide with Mauryan expansion. The Mauryan presence is not evident as an agency of destruction or of change because the Mauryas seem to have been more concerned with acquiring the resources of this region rather than restructuring the economy. Dharanikoṭa/Dhānyakaṭaka in Andhra is associated with the arrival of the NBPW after extensive use of Black-and-Red Ware of the Megalithic variety, and a quantity of silver punch-marked coins. A wharf in alignment with a navigational channel points to maritime links, probably with the Ganges delta, and it would also have controlled the route along the Krishna valley leading to the area of the gold mines. Excavations at Amarāvatī indicate a settlement in the Mauryan period or perhaps even a little earlier, which makes it possible that the stūpa was initially built at the time of Aśoka.65
Nevertheless, the important sites of Brahmagiri and Maski in Karnataka, both locations of Aśokan inscriptions, show the continued predominance of the Megalithic culture ways. The importance of the sites in this area was enhanced by the fertility of the Raichur doāb. Nittur has Megalithic stone circles in the vicinity of the inscription and Udegollam is an early historical site. There is considerable uncertainty about the identification of a few sherds of possible NBPW further south in the peninsula.
A comparison of the distribution of archaeological sites associated with the Mauryan period and the Aśokan inscriptions, indicates that there were three areas which might be called areas of isolation at this time, where urban centres and inscriptions are absent. These are the lower Indus plain, the eastern part of central India (which remained (p.295) an area of isolation for many centuries) and the far south.66 Elsewhere Mauryan control established routes with emerging centres of exchange, some of an urban nature. With the break-up of the Mauryan empire, the larger cities became the nuclei of new states. The importance of Pāṭaliputra is clearly indicated by its being larger in size than the other cities of the time.
The archaeological evidence reinforces the argument that there appears to have been an improvement in the standard of living during this period at many sites and particularly those which evolve, or had evolved, into urban centres. This enhancement doubtless drew not only from new technologies but also from forms of administration, communication and exchange, which were part of Mauryan policy. Those sites which have Aśokan inscriptions and have been excavated, point to the presence of artefacts associated with the period, except however for the sites in central Karnataka which curiously have not so far registered the presence of northern artefacts although there is a heavy cluster of inscriptions in the area. This perhaps says something about the nature of Mauryan settlement and control in these parts. What is noticeable, although not unexpected, is that the cultures immediately prior to, and often even contemporary with the Mauryan are diverse from region to region, both in identity and in degrees of historical change. This not only endorses the argument that Mauryan society was indeed a multi-cultural society, but also requires us to view the possible divergencies in the relations between the Mauryan administration and the local people of a region.
Punch Marked Coins
These coins were discussed in some detail at a seminar, the proceedings of which have been published.67 The coins appear in stratified levels at excavations often together with NBPW. The date of the NBPW has (p.296) been taken back to the seventh century at some places and the dating on the coins in this publication has also been sought to be made earlier, some taking it back to 800 B.C. General opinion however still adheres to the sixth century B.C. as the start of coinage. In any case they would have been in circulation prior to the Mauryas.68 The symbols on the coins are suggestive of an indigenous origin although the notion of coined metallic money was probably known from west Asian usage. Pre-Mauryan literary evidence, with the exception of Pāṇini, does not confirm the use of coins, but is familiar with using weights of gold and silver with a designated value. Punch marked coins would have increased in number with the greater availability of silver and this may have required close connection with west Asia and Afghanistan, apart from the more sporadic mining in the subcontinent.
In the study of punch-marked coins the work of D.D. Kosambi remains the most challenging. Kosambi’s papers on punch-marked coins, from various journals, were collected and published, thus bringing together his views on early Indian numismatics.69 This makes available his substantial analysis of these coins and even if one may not agree with it on every count, it still directs attention to issues which need to be studied. He treated coinage as the necessary tool of trade and argued that it originated with traders. At a later stage the state appropriated the system. Coinage gives a clue to trade relations provided that the identification of coins with those who issue them can be ascertained. Where he identifies the symbols with certain Mauryan rulers there can be considerable difference of opinion. He uses his own identification to assert a debasement in the metal and therefore argues for a fiscal crisis in the later Mauryan period.70
Another view questions Kosambi’s method of analysis and the early date of the coins.71 Cribb dates the coins to the fourth century B.C. (p.297) arguing that the association with the NBPW is chronologically uncertain. The date should be determined by dated coinage in the Chaman Hazuri hoard and by the reference to Alexander receiving talents of marked silver. An examination of the symbols suggests that there were local issues of varying weight which are rare in northern India and national issues with a standard weight found all over the subcontinent. The coins were developed in the north-west and were imitated in the Ganges towns. However the pre-Mauryan use of coins is suggested in other sources. The Nandas are said to have standardized weights and this is likely to have included coins.72 Coins were a familiar item to Pāṇini,73 which would date them to at least the fifth century.
Aśokan Pillars as Monuments
A major reassessment of the Aśokan pillars and their capitals was made by John Irwin in which he questioned a number of theories which have so far been taken for granted.74 Irwin maintains that the pillars which are called Aśokan are not all Aśokan and some are pre-Aśokan. This is attested by PE VII where the king states that the edict is to be inscribed even on existing pillars. Irwin attempts to demonstrate this by using the evidence from archaeology. The pillars, according to him, were originally cult objects, some being regarded as the axis mundi and often associated with funerary tumuli, and therefore some remain uninscribed. They formed a sacred triad together with the tree, and the stūpa. Aśokan pillars therefore do not mark the beginning of (p.298) monumental art but the culmination of an older tradition of setting up pillars, which was a religious tradition. With the exception of the Sārnāth pillar which shows a Perso-Hellenistic connection, the style of the other capitals is traceable to an earlier epoch when India was already familiar with the ancient Near East. The pillars were originally of wood and were placed directly into the earth. The pillar had associations with the sacrificial yūpa, and the dhvaja, and was symbolic of an established cult. The dhvaja was worshipped before battle because divinity was thought to reside in it.
Irwin discusses in detail the nine pillars around which there have been excavations. Of these, four were placed directly in the ground with no support in the fashion of the wooden pillars and these therefore could not remain upright and eventually fell. The other five were placed on a stone foundation and therefore survived for a longer time. Four of these were inscribed—Sārnāth, Topra, Rampurva (lion capital), Lauriyā Nandangarh—and the fifth is only a shaft at Gotihawa. The pillars have a bricked surround which was a pradakṣinā-patha. Irwin argues that the first group was earlier and the second group was later. The Vaiśāli pillar with a lion capital was also earlier as it seems to have sunk substantially into the earth suggesting that there was no foundation stone.
In discussing the chronology of the pillars his view is that the first edict issued was the Schism Edict, even though it would date to a later part of Aśoka’s reign if it is linked to the Third Council at Pāṭaliputra. His contention that the Allahabad-Kosam pillar has always been located at Prayāga is plausible in terms of the method of engraving the edicts, but there is little convincing explanation as to why the edict addressed to the mahāmattas of Kauśāmbi should be inscribed on a pillar at Prayāga, particularly as it is uncertain whether the Ganga-Yamuna confluence was sufficiently sacred at that time to merit a pillar. The major monastic site, similar to Sarnath and Sanchi, was at Kauśāmbi and not at Prayāga and the contents of the Schism Edict would be relevant to a major monastery. Interestingly there is a strong association in the Buddhist tradition of dissenting monks with Kauśāmbi.75 (p.299)
Discussing the capitals he argues that the lotus-and-palmette and the bead-and-reel designs are not Perso-Hellenistic but are ultimately derived from the Egyptian lotus which design came in various ways to Greece, Iran and India probably through Assyria, the Levant and Ionia in the mid-first millennium B.C. The links with the Near East have however been questioned.76 Irwin interprets the so-called ‘honey-suckle’ motif as a stylised lotus. The pillar for him is symbolic of that which separated heaven from earth in the creation myths. The elephant and the bull had Indian connotations, especially connected with fertility. The lion was heraldic and borrowed and to that extent represented new conceptions of royalty. The pillar then becomes the symbol of the divine role of ancient Indian kingship, a view not too distant from that of Coomaraswamy. It is not merely a Buddhist symbol but articulates the summation of all virtue. Aśoka’s imperial vision is said to have been influenced by the Achaemenid example.
The connections between India and Assyria and the Levant are plausible since they would also relate to the issuing of silver bent bar coins in India and the evolution of the brāhmī script. But such connections may not have been influential beyond these similarities. That some pillars are pre-Aśokan and others set up by Aśoka also seems a convincing argument and the technical investigation of the actual setting up of the pillars points to the kind of excavation and analysis which needs to be carried out elsewhere where there are Aśokan pillars. That some were pre-existing and that some were earlier than others even during the reign of Aśoka would not create problems. His statement that not all the pillars attributed to Aśoka were cut and polished at Chunar would have to be further investigated particularly after the recent work at Chunar. This would include a discussion of the Mauryan style sculpture in Chunar sandstone found near Kauśāmbi and dated to post-Mauryan levels, which Niharranjan Ray maintains, reinforces his argument that the Didarganj yakṣī is post-Mauryan.77 (p.300)
But some major questions remain unanswered. Why were some pillars inscribed and not others? What conditioned the choice? If these pillars had the cosmic meaning which is imputed to them, then why are they limited to the Ganges plain? The texts which Irwin quotes, such as the. Ṛgveda had a geographical context which was that of the north-west and the Punjab and the water-shed, an area which is devoid of pillars. The concentration in the Ganges plain is suggestive of a different tradition if we are to look for antecedents. The pillars attributed to Aśoka in Hsüan Tsang’s list are also largely in the Ganges plain. The inscriptions have no connection with the earlier cults of the Indradhvaja or such like and their contents tend to be practical concerns of administration, the welfare of the subjects, and social ethics as part of short discourses on dhamma, or alternatively are concerned with the functioning of the saṃgha. If Aśoka was resorting to traditional symbols in inscribing his edicts on pillars, the message he was conveying was not linked to the belief about the pillars. At most it could be said that in some cases Aśoka was drawing on the earlier associations with the pillars to legitimize his own message. This raises the question of whether he chose the pillars and pillar sites, which might be a partial explanation as to why they are confined to the Ganges plain, although of course the transportation of the newly cut pillars would have been a major consideration. There is little evidence of stone carving prior to the pillars, therefore the question of where the technique originated in India and how, remains as yet, unanswered. The function of pillars serving as imperial symbols, quite apart from their cultic or other significance, should not be ignored.
Certain other monuments and art remains pertaining to the Mauryan period have also been discussed. The Dharmarājika stūpa at Sarnath is now held to have an Aśokan nucleus.78 The Lomas Rishi cave has come in for detailed consideration.79 The list of Mauryan art (p.301) remains has sought to be enlarged but the dating of many items still remains controversial.80 Ring-stones and stone discs, associated mainly with urban centres have been described as Mauryan and on the basis of the reliefs carved on them, they are said to have possibly been votive objects in a religio-sexual context.81
The arthaśāstra of kauṭalya
A three-volume study by R.P. Kangle has now become the standard work on the text, including as it does a critical edition, translation into English, and an introductory study.82 Earlier English translations having been rather unsatisfactory, this is a substantial improvement. The notes which accompany the text are of considerable help in disentangling some of the more difficult passages. Kangle discusses the chronology of the text and is of the opinion that there is no convincing reason why this work should not be regarded as the work of Kauṭalya, who helped Candragupta to come to power in Magadha.83 He has elsewhere discussed more recent articles relating to various aspects of the Arthaśāstra.84 Among these Burrow argues for differentiating between Cānakya and Kauṭalya.85 An extensive bibliography on the Arthaśāstra is now available.86 A more recent translation has re-arranged the text to make it more accessible to the modern reader.87 But the justification for the re-arrangement is based on modern attitudes towards the administration of a state which takes the text outside its historical context. Why the text was arranged in the way it was originally, remains (p.302) unexplained. There has also been an analysis of the text with a focus on political economy and some aspects of Marxist historiography.88
A detailed consideration of the chronology of the text which is a departure from the older form of argument is the study of T. Trautmann.89 This includes the collation and analysis of various narratives concerning Cānakya/Kauṭalya and the differentiation between them in a variety of traditions. Turning to the text an attempt is made to arrange its various sections chronologically by using the statistical method with the aid of a computer. This involves listing the potential discriminators—the commonly used particles, such as evam, ca, tathā, va, and so on, the use of which in literary style helps identify and differentiate between the styles of various authors, if there is more than one. Sentence length and compound length were also used as discriminators. Trautmann’s analysis suggests that there are three distinct books within and prior to the compilation of the Arthaśāstra, Books II, III and VII, which appear to have been composed separately. The composition and authorship of the other Books is less certain. There would then have been several authors and it would be incorrect to refer to Kauṭalya as the author of the entire text. Arthaśāstra was a category of literature like Dharmaśāstra, and would therefore have had antecedent texts, and others compiled in the early centuries A.D.
Trautmann would date the compiling of the various texts into the Arthaśāstra at a date precluding the Mauryas and the Hunas and therefore of the early centuries A.D. possibly c. A.D. 250. Book II, entitled Adhyakṣa-pracāra, may have been the earliest. Books III and IV containing material akin to that of the Dharmasmṛtis probably pre-date Yajñavalkya. The elaborate doctrines of Book VII do not suggest a date prior to the Christian era. He further argues that although the existing Arthaśāstra may not be so old, the śāstra tradition which it contains probably pre-dates the Dharmasmṛti literature in view of (p.303) the discussion on the relationship between the two categories.90 The rājadharma section of Manu for example, could well have been drawn from an early Arthaśāstra. It may be suggested that this discussion has an added historiographical interest. Nationalist historiography maintained that the discovery of the Kauṭalya Arthaśāstra vindicated Indian civilization as having a rational and practical side as well. The inclusion of topics parallel to the Arthaśāstra in other texts such as Manu, or the Mahābhārata, indicates that this rational and practical side was more widespread than was earlier conceded.
Now that the Vedic corpus, the Mahābhārata, the Rāmāyaṇa, and many of the Dharmaśāstras are available in the form of data-bases for computers, the application of the statistical method is bound to increase. It does not supersede traditional scholarship but acts as a parallel investigation. Since the computer is largely an aid to the more mechanical side of the work, the discussion still draws upon references in various texts and a familiarity with the texts is all the more necessary to the reconstruction of the results of the computer analysis. Its potential also lies in the fact that the method can be further refined for more precise results. The introduction of this new technique of analysis in relation to the dating of the Arthaśāstra does require that those working on this particular subject should comprehend the technique and not set it aside as some do. This is a lacunae in a recent study which discusses the date of the Arthaśāstra.91 Trautmann’s method is dismissed because it is claimed that it cannot be verified in India, which is of course incorrect.92
That the Arthaśāstra was not a document describing the government associated with a single period is a widely held view. To the extent that it is a comment on how an efficient governmental (p.304) system should be organized, it would inevitably refer to more than one period. If the present compilation dates to the third century A.D. then items, artefacts and concepts of earlier times are likely to occur in the discussion within the text. The identification of these would stretch from the earliest association in time of the text—the Mauryan period—to later times. To accuse scholars of arbitrarily using the evidence when they date passages by such identifications, is meaningless. Texts in the early period were constantly revised and updated and the Arthaśāstra was no exception. Thus in many such texts some sections are early, some are late and some early sections carry interpolated passages of a later date. Interpolations can have many reasons and take many forms including the foisting of names into manuscripts. Interpolations can suggest in some cases, the point in time when additions were made. Therefore if it is said that the Arthaśāstra was edited over time, this is not an attempt to ‘explain away’ interpolations. The freezing of texts in their original form becomes more characteristic of the post-Gupta period and suggests changed attitudes to the oral tradition and to literacy. As long as the text is not being quoted as an authentic, descriptive narrative, but as a normative text, then the question of its date gets subordinated to the concession to interpolation, and therefore it can best be used in a comparative sense with other textual sources.
There is also the need to consider who the authors of a text may be and to whom the text is addressed. The statistical method shows that it is not the work of a single author but a compilation. If Book II is the earliest, then it fits in with the initial attempts to establish a monarchy based on efficient revenue collection and a degree of centralization through administration. There is no concept of empire in the text, but the notion of government is very different from that of the earlier janapadas which preceded the Nandas and the Mauryas. This part of the text suggests that the state is also beginning to enter activities such as control over forests, over mines and semi-precious stones, over coined metallic money, and so forth, which was a departure from the earlier practices. It is precisely in the area of the state’s control over economic resources that there is more agreement between Book Two of the Arthaśāstra and the Indica of Megasthenes. (p.305)
If the Arthaśāstra, the Aśokan inscriptions and the Buddhist texts maintain that the yonas have a different social system, it does not mean that the Arthaśāstra is later, but that the system existed at the time of Aśoka and was being commented upon by the other texts. The nature of the comment would indicate whether they were speaking of a contemporary situation in emphasizing the difference. The use of various languages such as Prākrit, Greek and Aramaic by Aśoka and not Sanskrit, was because he intended a particular purpose through his edicts and this intention would not have been met by the use of Sanskrit. When the intention changes then later kings use Sanskrit. Royal inscriptions, largely votive, continue using Prākrit for some time after Aśoka. Initially these are a small part of the many brief votive inscriptions made at this time by ordinary people.93 The Arthaśāstra is in Sanskrit because it obviously has a very different function and belongs to the śāstra tradition which was in any case distanced from the Buddhist moorings of many of the Aśokan edicts. Daṇḍin in the Daśakumāracarita states that Viṣṇugupta composed six thousand ślokas for the Maurya to learn daṇḍanīti.94 This passage has often been quoted with reference to the theory that the text may have been metrical in origin. What is more important for our purposes is the association with the Maurya, even at such a late date.
The Indica of Megasthenes
A fresh survey of references to India in Greek texts has recently been made.95 More specifically there have been a number of studies of the context in which Megasthenes was writing by various scholars of Greek and Classical history and these suggest that the text was not written in isolation from parallel texts at the time.96 Megasthenes was (p.306) viewing India through a Hellenistic perspective and Greek ethnographical writing. In two major articles, Zambrini made a detailed review of earlier scholarship on the subject and argued that there was a relation between the Indica and the Aegyptica of Hecataeus of Abdera and these have to be seen as polemical in the political tension between the Ptolemies and the Seleucids.97 Megasthenes was in part locating his Indian experience in a Greek view of the world. Bosworth’s study of both Arrian and the Indica leads him to maintain that Arrian is concerned essentially with the history of Alexander, therefore the Indica is a supplementary interest and has to be read alongside the major work.98 Arrian’s main interest is in the geographic and ethnographic details of the voyage of Nearchos. The Indica was propaganda in favour of Alexander but was not a polemic against Hecataeus who had made Sesostris the great world conqueror. He also suggests that Megasthenes may have been the representative of Sibyrtius who controlled Arachosia, rather than the ambassador of Seleucus and that his mention of visiting the court at Pāṭaliputra was probably not to repeated occasions of being at the court but to his single visit to Pāṭaliputra. Bosworth argues that it is more likely that Megasthenes visited India in c. 310, as the description of the Punjab and the Ganges plain seems earlier rather than later and prior to the loss of the Indus area to the Maurya. Candragupta, who is referred to as the king of the Prasii, is not said to be the only king ruling in India for there are other kings mentioned, as well as the autonomous cities. It should however be kept in mind that none of the authors who were supposedly quoting from Megasthenes were his contemporaries, therefore the quotations may well have been subjected to interpolations. Further, that Diodorus and Strabo are possibly closer to each other in what they say and that Arrian, perhaps because he is even later, not only paraphrases Megasthenes (p.307) more evidently, but also makes an occasional statement which is not found in the earlier two.
It is interesting that the historiographical locating of Megasthenes has become part of a more analytical discussion among specialists of Greek sources of ancient history, examining the intellectual assumptions of those writing for Hellenistic society. This will help to clarify some of the ambiguities in the quotations from Megasthenes. Independent of this discussion I have argued that in his description of Indian administration and society and the seven castes, Megasthenes was clearly writing in the context of Greek and Hellenistic experience and ideas, and I have tried to reassess the statements from Megasthenes on the divisions of Indian society and on some aspects of economic arrangements.99
Megasthenes has little to say on the practice and worship of Buddhism, except for the possible general reference to renouncers as one of the categories included among the philosophers and sophists. Yet the Buddhists were among the sects which endorsed renunciation but nevertheless at this time lived on the fringes of urban settlements from where they obtained alms. Megasthenes does however mention more than once, the Indian gods whom he identifies with Herakles and Dionysus, generally believed to be references to Kṛṣṇa and Śiva. An attempt to change this identification to Indra and a tribal deity has not met with much support.100
The Aśokāvadāna has received some attention. There is now an English translation of the first part of Przyluski’s study La Legende de l’Empereur Açoka, which is an analysis of the northern Buddhist traditions concerning Aśoka as described in the Aśokāvadāna.101 Subsequent to Przyluski the Aśokāvadāna has been edited with some reference to additional texts by S. Mukhopadhyaya and this edition has been used (p.308) as the basis of a new study and translation by Strong.102 The historicity of the legends from the avadānas has been discussed with reference to the Queen’s edict possibly endorsing a weakening of Aśoka’s control over the kingdom in his last years.103 It is said that the edict was not engraved at the order of Aśoka since it opens with the statement, devānampiyasa-vacanena, suggesting that it was issued by the queen herself. But a similar opening statement occurs in the SEs and should more reasonably be interpreted as officials recording the orders of Aśoka rather than his losing control. An unpublished manuscript containing the legend of Kunāla as part of the Aśokāvadāna-mālā has also been discussed.104 The complete text of the Lama Tāranātha’s history of Buddhism in India has been translated afresh into English.105
The claim by various scholars that the date of the death of the Buddha should be advanced by about eighty to a hundred years, would make some difference to the interpretation of the evidence from Buddhist texts.106 The changed date may suit events connected with the various Buddhist saṃghas as has been argued, but may not suit the chronological statements made about pre-Mauryan history, although it would not contradict the major historical processes in the Ganges plain associated with this period, such as state-formation and urbanization. Archaeological evidence supports, as we have seen, the start of urbanization at some sites in the sixth century B.C. with further changes in the subsequent century.
It has been argued that the sites associated with the life of the Buddha such as Lumbinī, Kapilavastu, Śrāvastī, Vaiśāli, Rājagṛha, Kauśāmbi (p.309) and Sārnāth/Rajghat, could not have been urban centres this early.107 Lumbinī was in any case not described as an urban centre in the narrative on the life of the Buddha. The settlement at Kapilavastu/Ganwaria has been dated to 800 B.C. with NBPW occurring in the time bracket 600–200 B.C. together with considerable evidence of structures.108 Kapilavastu is described as the central town of the Śākyas but is not included among the mahānagaras in Buddhist texts. At Śrāvasti there is an overlap between NBPW and PGW and a large variety of items suggesting that it was an exchange centre as early as Pd.I which is dated to the sixth century B.C.,109 but given the overlap could have had earlier beginnings as a settlement. The garh area of Vaiśāli was under occupation from at least 500 B.C., with an admixture of NBPW and BRW at the lowest level, although the structures are said to be later. Interestingly the NBPW at Ayodhya is being dated to the seventh century B.C. The extensive site of Rajgir, both the earlier fort and the later settlement have produced NBPW and at one spot the settlement has been dated to 500 B.C. or earlier. Kauśāmbi, even with the revised dates, was clearly an urban centre in the sixth century B.C. The settlement at Rajghat goes back to 800 B.C. with BRW, and urban characteristics are associated with the introduction of NBPW. In the absence of extensive horizontal excavations, the nature of urbanization for each site cannot be precisely guaged. On the basis of a comparative study of the sites of the Ganges plain it can be maintained that if the Buddha preached in the late sixth and early fifth century, some sites were urban centres and at others there was an initial change towards urbanization.
The evolution of urban centres requires a multiplicity of factors which, judging from the archaeological data of the Ganges plain, seem to have been present in the sixth century B.C. Proto-urban settlements would require the support of a large agricultural base, and the possibility (p.310) of double-cropping or of wet rice cultivation would have provided this. If the settlement was located on a route with potentialities for trade, as many of these early towns were, this would add to a tendency towards urbanization. In both these changes iron technology would have played a role even if it was not the major causative factor. Together with this, the emergence of kingdoms required a central settlement which would be the nucleus of political control and power. These are described in the texts as rājadhānis and as such would have attracted a larger population than other settlements. Population densities are evident in a comparison of pre-NBPW settlements with NBPW ones, where in one district which has been surveyed, the pattern is not only of an increase in the number of settlements but also larger numbers in most and in a few, a marked concentration of population. The latter are the recognisable urban centres.110 The larger number of settlements would have provided the support of agricultural and other resources, particularly as their spread is in areas with arable land. The bigger settlements are along the Ganges which doubtless acted as a route. Environmental factors also determined the size and pattern of the settlement.
The early Buddhist texts show an awareness of some towns being urban centres at an early date and others becoming so later.111 Kauśāmbi is of major importance both as a meeting point of routes and as the capital of janapada, apart from developing into a monastic centre.112 Rajgir is described as the old settlement of a hill fortress, Giribbaja, in the verse sections which are believed to be early, and the new settlement, Rājagaha, was a town at the foot of the hill built by Bimbisāra. In contrast, Pāṭaligāma is said to have been only a village in the lifetime of the Buddha and became the capital in the reign of Bimbisāra’s great-grandson.113 If the suggested later date is accepted then the main city providing the background to his teaching (p.311) ought to have been Pāṭaliputra, which is not the city referred to in the Buddhist texts. Rājagaha declined with the rise of Pāṭaliputra. Incidentally, the Mahābhārata describes it as a flourishing town and the capital of Jarāsandha, who is close in time to Bṛhadratha, which would make him earlier than Bimbisāra, although the date of this reference in the epic may be uncertain.114 Vaiśāli is said to have many caityas dating to before the Buddha. Śrāvastī was a centre of exchange and commercially important. The towns associated most often with the Buddha were Śrāvasti, Rājagṛha, Kauśāmbi and Vaiśāli, conforming to a limited area in the middle Ganges plain. Banaras and Taxila are more frequently mentioned in the Jātakas. In the typology of towns, even in the Pāli Canon, there is a differentiation between the incipient urban centres and the well-established.115 Kuśināra, although a capital of the Mallas is described as inferior and set in a forest. It carries the appellation of nagara as do the mahānagaras but is different, although it is not described as a gāma, as was Pāṭaligāma, thus pointing to the difference. The archaeological counterparts to these variants would also differ in this graded distinction of urban settlements.
The change to a mahānagara would have required mud walls and ramparts giving way to fortifications, therefore fortifications were not necessarily characteristic of the earliest period of urbanization. There is mention of the six mahānagaras or cities where the Buddha could have chosen to die, a list that is repeated more than once.116 Obviously by the time that the Canon was compiled these cities had become major. There would therefore have been a difference between incipient or even early urbanism and mature urbanism. A comparison of the chronology of sites on the basis of archaeological artefacts with literary references can be misleading. A study of the pattern of settlements in an area surrounding an urban centre is a better gauge of the degree of urbanization. (p.312)
With the juxtaposition of gaṇa-saṅghas and emergent kingdoms, the political change was evident. As rājadhānis, the early towns associated with this change were concentrations of power and recognized as such. As nagaras they introduced a different way of life. Even if they are not impressive in terms of their urban structures (and few cities were spectacular in this sense even in later times judging by the archaeological record) they nevertheless do represent a different complexity of governance, and of production, exchange and distribution, a complexity which becomes part of the historical context to the rise of Buddhism.
The archaeological correlation with the towns mentioned in the Pāli Canon is of less significance than the analysis of the Buddha’s teaching as recorded in the texts which in terms of its historical context is associated with the coming of monarchies and urban society. The Canon describes a condition of established monarchies and mature urbanism which may not be evident from archaeology for the period of 600 B.C. But this does not preclude the sixth century from being the period in which the urban change and the change to monarchy had been initiated. On the basis of the traditional reckoning the Buddha would have started his teaching in about 530 B.C. by which time the kingdoms had come into being. This would agree with Puranic chronology.117 To the extent that the Buddha’s teachings were receiving a hearing and gradually becoming popular, it can be assumed that this was in part related to these changes. The Middle Way was not just an alternative to Brahmanic ritual but also took cognizance of the newly emerging society, different from that visualized in Brahmanic ritual. Liberation from ritual, apart from its metaphysical concerns, was also a liberation from that which virtually destroyed wealth. This would have appealed to the kṣatriyas moving towards greater power through monarchical states, and who had prior to this begun to question the major sacrificial rituals as is evident from the early Upaniṣads,118 and to the seṭṭhis wishing to invest wealth, both of which were situations (p.313) evolving in the gradual change to state systems and to the introduction of commerce.
The compiling of the Canon is said to have occurred at the two Councils after the Buddha’s death. By this time the urban context was well-established and was even more so when the Canon was converted from oral memory to writing. There was a time difference between the three events: the death of the Buddha, the compiling of the Canon and the writing of the Canon, and these events cannot be collapsed into a single period. The Bairat-Bhabhra Edict refers to the teachings of the Buddha which were obviously familiar and recognizable in an oral tradition and which subsequently could well have been associated with texts such as the Mahāvagga, the Nikāyas and the Suttanipāta. For these teachings to have gained currency outside the middle Ganges plain may well have taken much more than a century, given the nature of communications at that time.
The Canon refers more frequently to the pre-Mauryan states such as Magadha, Kośala, and Vatsa, with their kings and their capital cities and to the Vṛjjis at Vaiśāli. If urbanization and state-formation evolved more or less at the same time, then these changes would have been initiated by the late sixth century. The historical context to the teaching of the Buddha is a situation of initial urbanization and a shift from chiefdoms to kingdoms, but prior to the emergence of large kingdoms or empire.119 In fact the association in the early Buddhist texts is not even with the Nandas. There is unlikely to be a radical postponement of the historical processes linked to this period if the date is changed. But the evolution of Buddhism has to be seen as proceeding in stages. Allowing the Buddha a lifetime of half a century of teaching would in itself have witnessed the change from an initial urbanism to a more established form. The settings associated with his teaching were more often those of early urbanism. Much of the teaching was in parks and groves on the outskirts of urban centres. (p.314)
It is argued that the figure 218 for the number of years between the mahāparinirvāna and the accession of Aśoka, is fictive since eighteen is a mythical number. If the figure from the Northern Tradition of 100 years is accepted then the mahāparinirvāna would date to 369/8, the period of Nanda rule. Some scholars basing themselves on the Northern tradition have suggested 386–4, which would of course again lead to 118 for the accession of Aśoka and coincide with the Nanda dynasty. Any date for the mahāparinirvāna subsequent to c. 400 B.C. would make the Buddha contemporary with either the Śiśunāgas or Nandas according to Puranic chronology. Even if a variant chronology based on a collation of the texts is considered, the rulers mentioned in such a chronology, such as Anuruddha, Nāgadasaka, Śiśunāga, do not figure in the early Pāli Canon and are mentioned briefly only in the Sri Lankan texts of a much later period, whereas the rulers prior to these are said to be contemporary with the Buddha.120 The association is with Bimbisāra and Ajātaśatru, who, even allowing for exaggerated regnal years, preceded the Nandas. The janapadas mentioned in narratives of the Buddha would have become part of the Nanda kingdom. Mention of the Nandas is late and marginal in the Buddhist texts, and this would be curious given that they would have been the pre-eminent political power during the lifetime of the Buddha, on the basis of this new reckoning.121 The period of the Śiśunāgas and Nandas is associated with events after the death of the Buddha.
In a series of papers, Eggermont has continued his discussion on Mauryan chronology which he includes in the time-bracket of 317–186 B.C.122 He maintains that the original date for the mahāparinirvāna was 368 and this was later changed to 486 and then adjusted to 483 to accommodate (p.315) Mahinda as the sixth ācārya and introduce the four-year interregnum prior to Aśoka’s accession, with a final change in the twelfth century A.D. to 544. He also questions the length of Aśoka’s reign of thirty-seven years according to the Mahāvaṃsa, arguing on the basis of expired and current years, that the correct length is twenty-seven years.123 He has tried to make the calculation coincide with some readings of the Purāṇas where Aśoka’s regnal years are given as twenty-seven. Eggermont largely accepts the list of the later Mauryas as given in this book, but disagrees with the idea that the empire after Aśoka was divided.
An attempt has been made by Guruge to reorganize the chronology of Aśoka’s life with a consecration dating to 265 B.C.124 The Aśokan inscriptions have been rearranged, in terms of rock slabs and stone pillars rather than content, an arrangement which is historically unacceptable. Aśoka’s conversion to Buddhism is dated to 262 B.C., in the fourth year of his reign which does not conform to what the king himself states in the Minor RE. The Minor REs are dated to 255–54 and a year later Aśoka is said to have issued the bilingual Graeco-Aramaic inscription at Kandahar and this is prior to the issuing of the MREs, which, given the linguistic analyses that have been made of the Aramaic inscriptions, do not permit this inscription to precede the MREs. Aśoka’s death is placed at 228 B.C. There are some curious readings of the texts of the inscriptions to suit the new chronology. For instance, in an attempt to give less chronological weightage to the reference to the five Greek kings, the verb it is said, can be read both in the past and present tense; Aśoka therefore knew that some of the five were not alive at the time of issuing the inscription. If this was so he would either have referred to their successors or not mentioned them at all. One edict is read as Aśoka setting himself the quantitative target of increasing dhamma by exactly 150 per cent, which is not what the edict says. The author adopts his own classification and numbering of the edicts, which is most confusing. (p.316)
Contemporaries of the Mauryan empire have received further attention. Among these the foremost are the Achaemenids.125 There is now a detailed study of the social and economic history of this period126 which provides both useful information as well as comparison with Mauryan studies. A recent discussion of the Seleucid state points to a more realistic assessment of Seleucid-Mauryan relations than previous works.127 Megasthenes’ Indica is seen as a legitimation of Seleucus’ non-conquest of India and describes an apparently well-organized empire whose people are freedom-loving and should be left alone. Megasthenes may have viewed some aspects of Mauryan life from a Seleucid perspective. Looking northwards, a collation of the legends relating to the founding of the kingdom of Khotan and some linked to Aśoka, have been analysed and suggest that the legends associating Aśoka were meant to promote the prestige of the country rather than reflect historical events.128 A recent translation and study of the Tibetan text, the Li-yul Gyi lo rgyus, establishes that Aśoka was believed to have visited Khotan, where in this version a son was born to him during the visit who was destined to rule Khotan, helped by the minister of Dharmaśoka.129
The reference to the five yona-rājās remains the bed-rock of ancient Indian chronology. Whatever fanciful turns may be given to the Sandracottos-Candragupta equation, the reign of Aśoka is pegged to the mid-third century B.C. What is of interest is whether he had separate links with each of the five or did he know them by association, since there were both hostilities and close marriage alliances between the five. These particular five kings were part of a political network and the list therefore is not a casual one. The exchange of (p.317) gifts at the level of the courts was a prelude to commercial exchange. The port of Sopārā is likely to have been important in maritime links with coastal Arabia.130
The names and titles taken by Aśoka have been subject to further examination. On the basis of Indo-European roots it is thought that devānampriya means ‘legitimate (child) of the gods’, although it degenerates in meaning to a ‘blockhead’ in later times.131 This is not the same as divine kingship for the gods are multiple and familiar to the humans and the legitimation is that of a special status conferred by the gods. Another argument maintains that Piyadassi was his personal name and Aśoka which occurs infrequently was a biruda bestowed on him by the Buddhist saṃgha.132 If this was so then obviously he would have used this name when addressing matters relating to the saṃgha, which is not the case. It is more likely that, as the legends maintain, he was called Aśoka and adopted the name of Piyadassi, more appropriate to a king, when he began to rule.
There has been some criticism of the argument put forward by Bhandarkar, Smith, Przyluski, and myself that Aśoka may not have been married to Devī.133 The earliest of the texts, the Dīpavaṃsa, makes no reference to a marriage and mentions samvāsa. The later text, the Mahāvaṃsa, says that he obtained the daughter of a merchant and refers to samvāsa. The Mahāvaṃsa ṭīkā, still later, refers to the consent of the parents. Samvāsa means to cohabit but according to Guruge it is used in the sense of co-residence. But if the result is the birth of two children then obviously it has other connotations. He quotes the Samantapāsādikā as stating that Aśoka took the daughter of Devaseṭṭhi, where the word aggahesi also means to seize or to capture. It is frequently forgotten that such unions were common in royal families (p.318) and did not carry moral censure. However the sons of these unions were often debarred from succession. A. Chattopadhyaya is right in her suggestion that Samghamittā could have married a nephew of Aśoka if it had been a cross-cousin marriage.
Aśoka and Buddhism
The virtual dichotomy between those who give priority to the edicts as a source for the activities of Aśoka and those who would accept the versions in the Buddhist sources, is made apparent in Guruge’s biography where not only are the Buddhist texts privileged but among them the Sri Lankan Buddhist sources are given more credence.134 It is tedious to go through what is described as the bias of Indian historians and various mistakes in readings. These on investigation more often turn out to be those of Guruge’s misreading of the primary and secondary sources. Since most of his criticisms are inconsequential I am discussing them in a lengthy footnote.135 He provides us with (p.319) his curriculum vitae of devānampiya piyadassi Aśoka, third emperor of the Mauryan dynasty’ which is substantially an endorsement of the narrative of the Sri Lankan texts. Aśoka was the instrument of the establishment of Buddhism in Sri Lanka and has therefore to be presented as the greatest patron of Buddhism.136
As in all sectarian literature, Buddhist texts have also to be viewed as emanating from a sectarian perspective. This is not to negate the value of this perspective but to be aware of the purposes of the authors, an awareness which is required even for using non-sectarian texts. Such an approach may well be uncomfortable for those who are adherents of the particular sects but is inevitable in any historical analysis. It has been pointed out repeatedly that the narrative of the life of Aśoka differs in the two traditions, the Northern Buddhist (p.320) tradition and the southern Pāli Sri Lankan tradition. Not only is the projection of the individual not identical, but in the first the focus is more on him and his acts and in the second there is the added dimension of the Buddhist mission to Sri Lanka. The edicts provide a historically more reliable account of Aśoka’s relation with Buddhism and the nature of Buddhism in the third century B.C. At this time the Buddha was pre-eminent and the historical Buddha was more visible perhaps than the divine Buddha of later times.137 The law of dharma as discovered by him placed some limitations on the supernatural powers that could be claimed for the Buddha, limitations which faded out in subsequent periods.
A new reading of the Schism Edict questions the view that Aśoka tried to unite the saṃgha and was a follower of the Theravāda.138 The argument is that the Schism Edict does not deal with doctrinal differences, but with monastic practices and relates to the authority of the monastery. A similar reading maintains that this edict does not prove the historicity of the legend of the Third Council.139 However the severity of the punishment—expulsion, would suggest that the matter concerned more than just monastic practice and discipline and may well have included doctrinal matters, although the king may not have been intending to intervene directly in the decisions of the major monasteries. Dissensions are referred to in the Pāli Canon, some associated specifically with the monastery at Kauśāmbi.
It has been argued that there were two Aśokas: one was the historical figure from the inscriptions and the other was the legend, and the legendary Aśoka was the model for Buddhist writers.140 Whether he (p.321) was faced with the contradiction reflected in the model king as the cakravartin and the model Buddha as the bodhisattva as suggested, would depend on when these concepts became current and possibly both were post-Aśokan. The exercise of power may be viewed as incompatible with the ultimate Buddhist ethic yet royalty is beneficial for king and subjects. The temporal power of kingship had to be reconciled with the spiritual power of the saṃgha. Aśoka was attempting to resolve this contradiction. However, the distinction between the two Aśokas requires that the historical person be viewed as separate from the legend and assessed from the perspective of his thoughts and actions as depicted in his edicts, and in the context of his being at the helm of an empire. Even this study tends to use the legend as an entry point to the man, although it does so to a much more controlled degree than many others. The distinction between the two is pertinent to a historical view since the texts recording the legend are many centuries subsequent to the reign of Aśoka. The legend therefore already has a historiographical function. If the notion of the cakravartin is taken more literally then there remains the question of whether the concept had a political connotation during the time of Aśoka and of which he was aware, or whether it developed later, partly perhaps using him as a model.141
A more wide-ranging analysis of the role of the legend of Aśoka in Buddhist tradition and in the interaction of religion with the socio-political order, begins with an assessment of the historical Aśoka and then sets the legend in the context of various polities of south and South-east Asia.142 This study has helped to shift the focus from primarily antiquarian concerns to a debate on the broader issue of the interrelations between the state and the saṃgha. The dichotomy between renunciation and conquest as projected in the institution of the saṃgha and the rājā is not an absolute dichotomy since the two are interdependent. One may argue that if the saṃgha helps imperial policy by being supportive, political authority also assists in aiding the saṃgha financially and in other ways. This interdependence is heavily (p.322) underlined in the legends of the northern Buddhist tradition regarding the last years of Aśoka and the activities of his successor. State polity in the Buddhist tradition is opposed to the kṣatriya vidyā as defined in the Arthaśāstra, but the opposition is not as sharp as suggested by Tambiah. To the historian, the Buddhist model of kingship remains a partial and particular view and comes to be treated as such.143
Tambiah rightly points out that the dhamma was an efficacious ideology of pacification, political stability and security. It may also be said that it could have softened the thrust of Mauryan control especially in the peninsula. It is interesting that there is a considerable presence of Buddhism in the peninsula after the Mauryan rule, as also some visibility of Jainism in the south. The exhortation to the king to protect the saṃgha is a complex matter for it is the prosperity and property of the saṃgha which also requires protection and which in return guarantees the legitimacy of the protector, a relationship by no means unique to Aśoka and Buddhism, evident as it is in many situations of royal patronage to a religious order. Religions indigenous to India have imprinted the image of the ascetic on Indian thought. This in the past has been described as a life-negating principle, but as Tambiah shows, and other discussions would support it, the ascetic ideal far from being life-negating is in fact viewed as an alternative source of power, especially among orders of renouncers.144 The ascetic/saṃnyāsin or renouncer/bhikṣu, is seen as being outside and above political concerns, although often the ascetic becomes the focus of movements in opposition to established authority. He deliberately contravenes social mores in order to terminate social obligations, nevertheless he is not relegated to low caste status as would happen with others contravening social norms, but is accepted as being outside the framework of caste society—a notion which is hinted at in Megasthenes’ description of the philosophers. (p.323)
Aśoka’s relationship with Buddhism evokes therefore a range of interactions for that period, which have often been by-passed by historians and Tambiah’s study underlines the need for looking afresh at the texts. That the legend was important to the Buddhist tradition has long been accepted and it has its own historiographical evolution in various Buddhist texts. Nevertheless, the historical Aśoka has to be viewed from a different standpoint. His ideas on dhamma borrow from the current debate but are set within an imperial framework. An attempt at uniformity at the ideational level is emphasized in the rock and pillar inscriptions. Dhamma as he defines it, was an ethical principle with an appeal to the broadest social spectrum. Moving away at one level from the usual hegemony of imperial systems, he was nevertheless endorsing a process equally important to imperial needs, namely, acculturation. This he sought to encourage through a policy of persuasive assimilation in which conforming to the ethical ideals of dhamma was encouraged. The cultural norm implicit in this was that of the Ganges plain and although some concessions are made in the use of language, the message comes from a particular source.
If language lies at the boundary of the person wishing to communicate and those receiving the communication, then the social process of communicating becomes significant. The edicts are not to be examined only for the degree to which they reflect Buddhist ideas, because they also reflect the king’s relations with his subjects and this varies in different parts of the empire. The universalistic ethic of the new sects was different from the varna framework of brahmanical norms. Renunciatory orders tended to cut across caste and clan and weaken these identities and where the orders were loyal to the state they provided a network of support across the empire. The search for uniformity was at one level the personal vision of an individual but its articulation related to his historical role.
The Mauryan Empire
A couple of monographs provide an overview of this period. The earlier one is a useful summary of the main events and some of the discussion (p.324) around these.145 The later monograph discusses society and economy in some detail and therefore provides a different focus.146 The centrality of the economy as an aspect of state control is emphasized as also the importance of iron technology to the growth of urbanization.147 The production of iron is recorded but that it was the sole catalytic agent in the creation of cities is now regarded as debatable. The working of the south Bihar mines which would be an obvious source of iron remains enigmatic and the extensive use of iron in the Megalithic sites of the peninsula largely precedes urban centres and the arrival of the Mauryas. Urbanization would have been accelerated by multiple factors apart from a change in technology and these factors included the establishing of administrative and commercial centres given the commercial expansion under the Mauryas. Much of the useful discussion in this monograph tends to get deflected by a wide use of sources some even of the pre-Mauryan period and others post-Mauryan. Admittedly the chronology of sources remains uncertain but perhaps a greater discrimination involving a comparison of different categories of evidence is now required.
In the discussion of the extension of agriculture the point that there was a multiplicity of forms in the working and ownership of land which would support multiple levels of development, is well made. That the description of Megasthenes seems to have carried the imprint of forms familiar from the Hellenistic kingdoms is now acknowledged and the practice in these areas may be responsible for some contradictions in the quotations from his account.148 Diodorus writes of the absence of private ownership in land with cultivators probably paying a rent to the king in addition to one-fourth of the produce. Strabo refers to a chora basilike or royal lands and states to (p.325) the contrary that the cultivators are given a wage as well as keeping one-fourth of the produce. Clearly there is a discrepancy here. Arrian speaks of the cultivators paying a tax on the land which they cultivate and makes no mention of royal ownership. Bongard-Levin draws attention to the differentiation between rent and tax.
The question of whether śūdrakarṣakaprāya in the Arthaśāstra II.1.1. should be translated as ‘śūdras and cultivators’ or as ‘śūdra agriculturists’149 has a bearing on the nature of agrarian arrangements as well as on the status of the śūdra where, as a cultivator it would have been better than as a labourer. It also relates to whether the major part of agricultural activity was that of slaves and śūdra labourers under state control as has been argued,150 or whether there was a greater diversity of economic arrangements in which this was one among other systems. The sītādhyakṣa, supervising the cultivation of sīta lands, is required to use dāsakarmakaras and those providing labour in lieu of fines. The former were to be paid the abysmal sum of one and a quarter paṇa per month and given food.151 But in the chapter on organizing settlements in rural areas, apart from the śūdra agriculturalists to be settled in villages, reference is also made to various other forms of tenancy. Mention is made of people from other professions who receive tax-free land. Furthermore some categories of tax-payers were to be allotted arable land for life. Assistance of various kinds was to be made available to cultivators, so as to enhance the income of the treasury, and such assistance was to be forthcoming especially to those who took the initiative to clear waste land and set up water systems for irrigation. Various categories of arrangements relating to the cultivation of land even on the sīta land are advised. Had there been an overwhelmingly dāsakarmakara or śūdra labour based cultivation by the state, throughout its territory and providing a sustained income, then the empire would probably have survived for a longer period. Such cultivation may have been unlikely in many areas such (p.326) as the peninsula where the Megalithic settlements remain unchanged during the Mauryan period. But some state controlled agriculture would have been carried out in the Ganges plain.
Bongard-Levin, in agreement with many other historians, does not subscribe to the view that there was a slave mode of production in the Mauryan period although he agrees that some slave labour is evident in both agriculture and craft production.152 Slaves used in domestic work predominated over slavery for production. It could be argued that the use of the conjoint phrase, dāsakarmakara, referring to slaves and hired labour, makes it difficult to assess the amount of slave labour.153 Megasthenes comments on the absence of slavery which I have tried to explain as an absence not of slavery in any form, but specifically, the predominant Greek form of the doulos which was the use of slave labour in production.154 The dāsakarmakara were paid a minuscule amount but the fact of payment places this category in a status different from the doulos. The Arthaśāstra makes a distinction among the dāsās and refers to three categories: one that is permanently a dāsa; one who is kept as a pledge, the ahitaka; and the ārya who was temporarily enslaved but regained freedom after a stipulated time or a payment.
Associated with the theme of labour, the definition of viṣṭi frequently interpreted as forced labour has been questioned on the basis that in the Arthaśāstra it meant labour employed by the state or labour in lieu of taxes and not forced labour. It became forced labour in the post-Mauryan period.155 The degree to which labour was forced (as was the case in a system of corvée where it was additional to the normal taxes), or alternatively was one among a list of regular taxes and therefore contracted for, probably varied according to context, as the references are not consistent in its description. Most of the references in the Arthaśāstra are of the latter kind, although at one point the king is advised to protect those cultivators who may be oppressed (p.327) by daṇḍa-viṣṭikara, suggesting perhaps forced labour.156 Megasthenes makes no reference to a labour tax for cultivators, and artisans are said to pay a tax/phoros to the state and render prescribed services/leitourgai. The meaning of the latter term was not forced labour but a service performed at one’s own expense for a superior or for the state.157 State oppression is in any case not limited to categories of taxation and can occur in other forms as well.
In addition to multiple forms of land working, Bongard-Levin also emphasizes the existence of the village community, said to emerge either from clans or from settlers moving to vacant lands, where the association of villagers enjoyed equal rights as owners of plots although as a self-governing unit it included other groups in the population. This argument seems to contradict that of multiple forms of land working and ownership. Information on the agrarian and fiscal economy of the Mauryan period and later, has also been gathered in another publication which is a useful addition even if it carries little theoretical discussion of the categories to which its refers.158
The concept of the ‘tribe’ remains opaque in most discussions of this period, in part because its modern usage covers such diverse situations that it does not allow for precision. For the Mauryan period it would seem to include in various secondary works, the forest dwellers, members of the gaṇa-saṅghas or those under the jurisdiction of self-governing cities and more generally what are referred to as ‘tribal peoples’. One characteristic is that they are treated as peripheral groups and some are drawn into the mainstream whilst others remain outside. The occurrence of either one or the other would doubtless have depended on the degree of control which Mauryan administration wished to exercise in the area. This appears to become a pattern for state systems in the post-Mauryan periods and the relationship between the state and its ‘tribal’ peoples is a recurring theme, where the attitude of the former depends on the nature of resources (p.328) controlled by the latter and the degree to which such resources are sought to be appropriated by the state. Even among the oligarchies, there is a difference between those of the Ganges plain which were absorbed into the imperial system and those of Rajasthan, Punjab and the watershed area which retained much of their identity in the post-Mauryan period. The reasons for the difference would have to do with the nature and importance of Mauryan interest in their habitat. The process of absorption was not simple as these groups would have required not only a change in political form but also a conversion to caste society. The preconditions to this would have differed in the oligarchies from the monarchies of the pre-Mauryan period.
A juxtaposition of Brahmanical and Buddhist sources does suggest a rather different view of caste in each. Whereas the former differentiate largely between varṇas, the latter seem to assess the hierarchy and ranking more in terms of jātis although mention of varṇas is also made. This has led to the debate on whether Megasthenes’ description of the seven groups which constitute Indian society refer to varṇa or to jāti. A detailed discussion of the text relating to this and a return to investigating the possibilities of meanings in the Greek of that time, suggests that the reference was to the system of jātis rather than varṇas, although there is a garbled inclusion of categories and rules derived from other sources as well.159 Both Diodorus and Strabo use the word meros for the general divisions of Indian society and genos for the categories of birth, marriage and occupation. Arrian uses only genos and this may be because he tends to be paraphrasing Megasthenes to a greater extent than the earlier two.
A notable departure from the administration of the pre-Mauryan period lies in the description of how towns were administered. Allowing for a degree of exaggeration in the meticulousness of the Arthaśāstra in this matter, what does come through is the consciousness of urban living posing new and different problems from those associated with rural life. This is reflected not only in urban architecture, in the various offices located in the capital, in the markets of the city, but also in the precautions which are discussed to ensure the welfare of the urbanites. (p.329) Among these is the concern for hygiene and medical aid, doubtless made necessary by the heavy concentration of population at a large city such as Pāṭaliputra.160
The emphasis on administration was necessary because of various factors: the extent of the territory under the control of a single state which was unprecedented, the need to rationalize taxation where the number of taxes still remain less than in later times but had to cover both agricultural and commercial activities, and the recognition of services not just as dues but sometimes even in lieu of taxes. Given the extent of the empire there could not be a directly controlled administration uniform throughout the territory. The inscriptions make mention of various categories of administrative divisions: as for instance, janapada, pradeśa, deśa, and āhāra and the precise connotation of each remains problematic, particularly as these may be premised on varying considerations. Representation of Mauryan authority was accordingly ranked but doubtless drew both from those recruited at the capital as well as local potential in the provincial areas. In a section of the Arthaśāstra which could well be later, there is a listing of salaries.161 These need not be taken as literal but point to an important aspect of the evaluation of various levels of administration. Where the clerk gets a salary of 500 paṇas, the minister gets 48,000 paṇas, a ratio of 1:96. The upper levels of the bureaucracy tend to be well paid and the difference with the lower levels is striking. The questions which remain unanswered are whether this trend was set by Mauryan administration and it may well have been, and whether salaries were paid in cash, which may not have been the norm. Such salaries would require the regular and systematic collection of revenue, an activity which characterizes state functioning. High office would then be conducive to the accumulating of wealth and ranks of kinsmen may well have closed in to make this more effective.
The question of the degree of centralization and the efficiency of communication as an aspect of administration has been analysed in (p.330) terms of the nature of communication in the Mauryan empire.162 Both Megasthenes and the Aśokan edicts speak of the construction of roads. Yet the Arthaśāstra does not mention any special service concerned with the fast delivery of messages. It has been estimated that a courier leaving Patna would have taken thirty days to reach Kandahar and much more to Karnataka. During the monsoon it would have been even longer. The deployment of troops, as for example those sent when Taxila rebelled for the first time, could not be accomplished in a hurry. Therefore troops would have to be posted permanently at strategic points and even important decisions would have to have been taken at the local level by the local administration particularly in the areas distant from Pāṭaliputra. This would imply that the local representative of the king was invested with power.
Fussman makes the further point that the inscriptions in kharoṣṭhī and in Greek and Aramaic would imply that the local officials were from the area and were not officers posted from Pāṭaliputra. Mauryan administration was not homogeneous in its recruitment or practice. The empire included a variety of tribes as well as erstwhile kingdoms and Arrian refers to autonomous cities. The edicts were not uniform because they were not directly transmitted by Aśoka but were sent to the provincial officials who passed them down to lower levels in the hierarchy. Fussman maintains that there was a central, absolute power dependent on the personal activity of the sovereign who relied on the army and on efficient officers; regional administration was of a non-systematic kind and there was greater liberty the further away it was from the centre. The centrifugal factors were the freedom of the high officials of the provincial administration, the continued existence of their powers derived from pre-Mauryan rulers and difficulties in communication. These factors explain the disappearance of the Mauryan empire.
Admittedly there has been in the past an overly emphatic description of the Mauryan state as rigidly centralized and controlled by a powerful bureaucracy and monarch. The recent trend of questioning (p.331) centralized systems for early empires takes the Mauryan empire into consideration as well. However not all elements of centralization in the process can be discarded for there is also the fact of differing patterns in different areas. The Mauryan state is best viewed from the perspective of two foci: one is the processes of state formation and the creation of the early state which occurred in the Ganges plain prior to the Mauryan period and which can be seen as occurring in some other areas even into the post-Mauryan period:/and the other is to observe the Mauryan system by providing a definition of empire.
The emergence of the state is a gradual process and not all areas which get demarcated as states go through an identical process. The early state focuses on a large redistributive system involving tribute and booty, and revenue from various sources including taxes and labour. Legitimacy which becomes an important component of the state requires consensus, coercive power and frequently an ideology which invokes the supernatural.163 The mature state stretches these activities to a different level where legitimation, bureaucratization and control over economic resources become major items co-related with the size and the spatial distribution of the population. Together with these, social divisions, some of which take on class functions, become apparent.
With the evolution of an empire, the typology of the state changes and the structure of the state becomes more complex. It was previously argued that this complexity lay in the extension of centralized control, but in effect, the complexity lies in variations in control which have to be manifest in the system. I have attempted to examine the Mauryan state from this perspective.164 The definition of an empire includes not only extensive conquest with a monopoly of force, but also territorial control which may not be of a uniform nature and domination over peoples regarded as culturally alien or at least different. Territorial (p.332) conquest involves a frequency of wars since wars have an economic purpose as well: they are the agency for obtaining tribute, booty and loot as also prisoners-of-war to be used as slaves or labourers, and can result in the exploration of new kinds of resources. Thus if enhancing revenue is the motive then fertile agricultural land and trade routes are likely to be the prize of campaigns. Adjustments in administration become an on-going process if there are continual conquests. Thus roads have to be built in conquered areas and officers posted to these or at least local persons recruited to office. The expenses for this have to be offset by the tribute and booty. Therefore at some point an optimum territory would be the solution for any state and whatever is conquered in addition may be either deliberately or because of weakness, more loosely controlled.
The administration of Mauryan territory and economy may be seen as a relationship between three categories of control: between the metropolitan state, the core areas and the peripheral regions. The metropolitan state would be the nucleus of the empire with a history of having evolved from an early state to a mature state. In the case of the Mauryas this would be Magadha which was an area of primary state formation and which had in the pre-Mauryan period become a hegemonic state in the Ganges plain. With the enlargement of Mauryan territory it may have brought into its ambit those parts of the Ganges plain with which it had interacted earlier. Resources and revenues would come to the metropolitan state and be distributed from here through channels of public expenditure, such as administration and the maintenance of an army. This area would be under a high-powered, centralized, bureaucratic control, being the nucleus of the empire, and may have approximated the pattern of government described in the Arthaśāstra. The text therefore may be said to relate not to an imperial system, but possibly to the functioning of a metropolitan state within such a system. This might also explain why Aśoka takes the title of rājā-māgadhe, seemingly simple but effectively powerful. The metropolitan state is juxtaposed with other categories of territories.
The core areas would be those which prior to their conquest were states, but had failed to develop into major stares, such as Gandhāra and Āvanti, or else were areas of incipient state formation such as (p.333) Kaliṅga or Āndhra.165 These areas were not part of the metropolitan state, but were nevertheless governed by a hierarchy of officials. References to the different levels of the hierarchy as is apparent from the edicts suggests that the senior levels were familiar to both provincial and metropolitan administration. At the upper levels the administration was imitative of the metropolitan state but at local levels it would be more distanced. Core areas drew revenue from agriculture which required a certain investment and development on the part of the metropolitan state, even though much of this may have been local enterprise. The extension of sīta lands would in part be in such areas. Those who were deported after the Kaliṅga campaign are likely to have been settled on such lands.166 These would be areas where there would be some restructuring of the economy with the introduction of multiple tenures ranging from the cultivation of sīta lands to privately owned lands. Such restructuring is suggested by the fact that the only two major irrigation works—the dam on the Sudarśana lake and the Nanda canal—are both in core areas. Varieties of privately controlled irrigation mechanisms were prevalent elsewhere. Core areas were also those which had a commercial focus with centres of exchange such as Taxila, Ujjain, Dharanikoṭa, Suvarṇagiri and were therefore on trade routes. Hopefully horizontal excavations in the future will provide more evidence on such centres. Trade routes acted as corridors connecting core areas and possibly the expansion of a kingdom such as that of Magadha was partly conditioned by the wish to control such areas, which is suggested by the points of maximum control in the subcontinent under the Mauryas. With the disintegration of the empire, the core areas often evolved into the mature states of later times.
The third category was peripheral regions where there had not been state systems previously and which had therefore been areas of isolation to begin with. They were located in the interstices of rich agricultural belts and trade routes. Their resources were frequently (p.334) those which could be mined, such as iron or gold, or could be obtained directly, such as precious and semi-precious stones, timber and elephants, all much valued according to the Arthaśāstra. These areas therefore did not require bringing large tracts of land under cultivation or a major restructuring of the economy. Even the hierarchy of officials was probably more evident only at the senior levels where they worked in a supervisory capacity. Their prime function would have been accessing the natural resources and this the Mauryas do not seem to have regarded as necessitating a new imperial enterprise. Senior administrators in such areas may well have come from the north, since the language of administration in Karnataka for instance was not the local language but that of the metropolitan state. The large number of Aśokan inscriptions in this area are a contrast to the scant archaeological presence of artefacts associated with the Mauryas. Was this the result of enthusiasm on the part of the local Mauryan officials in having the edicts engraved at many places? That there was some variation in the contents of what was engraved seems apparent from the opening sentences of the Minor REs in Karnataka and the intrusion of some kharoṣṭhī.167 It would seem that the existing channels of resource mobilization were tapped, but little attempt was made to change them and introduce systems from the metropolitan state. These were the regions inhabited by the forest peoples mentioned in the edicts and local tribes who had their own hierarchy of functioning under their own chiefs. Such areas would have been frequently associated with what I have elsewhere called, lineage-based societies.168 The conversion of forest land and waste land to agricultural land would here be a more marginal and gradual activity as compared to tapping the existing resources. The administration would be relatively liberated from the control of the metropolitan state so long as the revenues and the resources reached the centre.
Needless to say, core and peripheral areas were not permanently one or the other. Core areas did often develop into independent kingdoms but could occasionally lapse into regions of lesser importance. (p.335) Peripheral areas more frequently, could develop into core areas of new kingdoms, given an investment in resources and the participation of the inhabitants in this investment. The latter was significant and initiated a cultural interaction as well. This it would seem, was well-understood by Aśoka and partially relates to his need to inscribe edicts which carry within them aspects of this interaction. One of the historical processes of change would have been the mutation of peripheral areas into core areas.
The distribution of population was not uniform as is evident from even an impressionistic view of the nature of archaeological sites in the different parts of the Indian subcontinent. The heaviest concentration was in the Ganges plain and around the hub of the core areas. A more limited density occurs in areas which were exceptionally fertile, such as the Raichur doāb. The peripheral regions were more sparsely populated with fewer concentrations.169 An attempt was made to estimate the population of the Mauryan empire but it was based on what seem to be exaggerated numbers for the Mauryan army and resulted therefore in the figure of 181 million, which is excessive for that period.170 Other methods of computing population using archaeological data suggest a much lower figure.171
In a reassessment of the decline of the empire, the policies of Aśoka could be viewed more dispassionately, and a further investigation made into possible administrative and economic problems. Aśoka’s dhamma may well have assisted in acculturation without the usual upheavals associated with imperial systems. Given the availability of resources and the manner in which the imperial system exploited these, the cause for upheaval may have been more limited. However the coming apart of the empire in so short a time would be suggestive either of oppressive features or inefficiency or both. The Northern Buddhist tradition insisting on the revolt at Taxila may indicate that the attempts at acculturation were resisted in some places. The existence of a multiplicity of cultural perceptions and norms would also have tended (p.336) to encourage local alignments and where the larger unity may have been distant. The legend about Aśoka building a vast number of stūpas, if taken literally would indicate a heavy drain on imperial finances. But the archaeological evidence suggests only a few and such as were hardly likely to seriously dent the Mauryan income. A comparative study with other ancient empires does point to the relatively small number of what might be called imperial monuments. Pāṭaliputra, built by Candragupta Maurya, was certainly in the tradition of an imperial capital. Curiously Megasthenes does not mention other cities with such monumental architecture.
Again if Greek and Latin authors are to be believed, the size of the standing army would have drawn away a substantial amount of imperial revenue. In the absence of conscription it would have been necessary to maintain a large regular force. Yet the figures quoted for the army do seem exaggerated and were intended to explain why Alexander did not campaign further into India. The Mauryas, having consolidated their strength in the Ganges plain and the north-west and controlling as they did the routes going south, had little need for further conquest.172 Whereas successful campaigns do bring in an income, nevertheless the maintenance of a large army quickly terminates the income. The forsaking of conquests by Aśoka was doubtless motivated by his reaction to the horrors of the Kaliṅga campaign, but possibly he was also nudged by the financially negative effect of such campaigns.
That the administrative system may not have been able to hold together the demands of empire has been stated often enough and has recently been discussed in some detail as we have seep. A properly efficient system for such a large territory requires a considerable revenue. It may be more appropriate to comment on this. If the senior bureaucracy received the emoluments that are suggested then these would have been a major item of expenditure. The junior levels of the bureaucracy, probably locally recruited would not have involved the same financial outlay. The question then is whether there was a sufficient tapping of resources and collection of revenue to generate (p.337) the required wealth to maintain the army, the administration and the overall imperial system. The empire was short-lived and this may have been because it was primarily concerned with extracting revenue from existing resources and possibly not sufficiently with creating new revenue bases. To maintain the hegemony of the imperial system, it was not enough just to integrate other areas. A substantial reorganizing of the economy of these areas would also be required. Significantly, when the empire breaks up, the differentiations observed within the imperial system come into their own. Magadha continued as the nucleus of large kingdoms incorporating the Ganges plain even though the successor dynasties were unrelated to the Mauryas. The consolidation in the north-west was linked to the politics of states once regarded as Hellenistic but now more mixed. Successor states elsewhere in Orissa, Andhra and the western Deccan were virtually impelled by the break-up of the Mauryan state. The pattern of the erstwhile gaṇa-saṅghas is equally interesting for those in the Terai and north Bihar were assimilated into the monarchical system of the Ganges plain, whereas those in Punjab and Rajasthan reverted to their oligarchic forms. It is not therefore as if the Mauryas were succeeded by the Suṅgas and the system continued. There is a distinctive pattern in the way in which the Mauryan state comes apart and this pattern is a guide to post-Mauryan history.
(1) D.C. Sircar, Asokan Studies, Calcutta, 1979. P.K. Andersen, Studies in the Minor Rock Edicts of Aśoka, I, Critical Edition, Freiburg, 1990. M.A. Mehendale, ‘North-western and Western Influence on the Versions of Aśoka’s Minor Rock Edict’, BDCRI, 17, pp. 81–97. K.R. Norman, Collected Papers, vols I–IV, London, 1990–93, analyse many edicts.
(2) F.R. Allchin and K.R. Norman, ‘Guide to the Aśokan Inscriptions’, South Asian Studies, 1985, 1, 43–50.
(3) A.M. Sastri, ‘Fifty Years of Epigraphical Studies in India: A Brief Survey’, Puratattva, 1994–95, 25, p. 27. Incidentally, mention is also made of a second post-Aśokan reference to Aśoka in a fifth century inscription in Sanskrit at a rock shelter at Narasinghagarh in the Rajgarh District of Madhya Pradesh. Ibid., p. 28.
(4) S. Paranavitana, The Greeks and the Mauryas, Colombo, 1971.
(5) H. Falk, Aśokan Sites and Artifacts, Mainz am Rheim, 2006.
(6) R. Nanda, The Early History of Gold in India, Delhi, 1992, has a map of the gold bearing sites. Map. 1, p. 2. F.R. Allchin, ‘Upon the Antiquity and Methods of Gold Mining in Ancient India’, in JESHO, 1962, 5, pp. 195–211.
(8) D.C. Sircar, ‘New Delhi Inscription of Aśoka’, Ep. Ind., 1970, 38, 1, pp. 1–4. M.C. Joshi and B.M. Pande, ‘A Newly Discovered Inscription of Aśoka at Bahaptir’, JRAS, 1967, pp. 96–8. K.R. Norman, ‘Notes on the Bahapur Version of Aśoka’s Minor Rock Edict’, JRAS, 1971, pp. 41–3.
(9) A.K. Narain, ‘A New Version of the Minor Rock Edict I of Aśoka’, Bharati, 1961–2, 5, 1, pp. 1–5; D.C. Sircar, ‘The Ahraura Inscription of Aśoka’, Ep. Ind., 1965, 36, 5, pp. 239–48. K.R. Norman, ‘Notes on the Ahraura Version of Aśoka First Minor Rock Edict’, IIJ, 1983, 26, pp. 277–92. S. Sankaranarayanam, ‘Ahraura Inscription of Aśoka’, IHQ, 1961, 37, 4, pp. 217–24. J. Filliout, ‘Studies in Aśokan Inscriptions’, Calcutta 1967.
(10) D.C. Sircar, ‘Panguraria Inscription of Aśoka’, Ep. Ind., 1981, 39, 1, 1–8.
(11) K.R. Norman, ‘Aśokan Inscriptions from Sannati’, South Asian Studies, 1991, 7, pp. 101–10. K.V. Ramesh, ‘The Aśokan Inscriptions at Sannati’, Indian Historical Review, 1987–8, XIV, 1–2, pp. 36–42. J. Howell, ‘Note on the Society’s Excavations at Sannati, Gulharga District, Karnataka’, South Asian Studies, 1989, 5, pp. 59–62. I.K. Sharma and J. Varaprasada Rao, Early Brāhmī Inscriptions from Sannati, Delhi, 1993.
(12) M.A. Mehendale in, ‘What was the Place of Issue of the Dhauli and Jaugada Separate Edicts’, BDCRI, 1953, 17 and in JOIB, 1, pp. 240–4 had suggested that linguistic forms in these edicts, ‘show affinity with the north-western language of the Aśokan inscriptions, thus pointing towards that area as the most probable source of the issue of these two separate edicts’.
(13) D. Schlumberger, et al., ‘Une Bilingue Gréco-Araméene d’Asoka’, JA, 1958, 1–48. E. Benveniste, ‘Edits d’Asoka en traduction grecque’, Journal Asiatique, 1964, pp. 137–57. K.R. Norman, ‘Notes on the Greek Version of Aśoka’s Twelfth and Thirteenth Rock Edicts’, JRAS, 1972, 2, pp. 111–18. D. Schlumberger and E. Benveniste, ‘A New Greek Inscription of Aśoka at Kandahar’, Ep. Ind., 1967, 37, 5, pp. 193–200.
(14) D.C. Sircar, Aśokan Studies, p. 121.
(15) D.C. Sircar, ‘Fragmentary Pillar Inscription from Amaravati’, Ep. Ind., 1963, xxxv, part 1, pp. 40–4. I.K. Sarma, Studies in Early Buddhist Monuments and Brābmī Inscriptions in Andbra Pradesh, Nagpur, 1988.
(16) E. Benveniste et al., ‘Une inscription indo-araméenne d’Asoka provenant de Kandahar (Afghanistan)’, Journal Asiatique, 1966, 254, Fas 3–4, pp. 437–70. See also Allchin and Norman, op. cit. S. Shaked, ‘Notes on the New Aśoka Inscription from Kandahar’, JRAS, 1969, 118–22.
(17) W.B. Henning, ‘The Aramaic Inscription of Aśoka Found at Lampaka’, BSOAS, 1949, 13.1, pp. 80–8.
(18) A. Dupont Sommer, ‘Une nouvelle inscription araméene d’Asoka trouve dans la vallée du Laghman (Afghanistan)’, in Academie des Inscriptions et Belles-Lettres, 1970, pp. 158–73; G. Ito, op. cit., 1979. G. Ito, ‘Asokan Inscriptions, Laghman I and II’, Studia Iranica, 1979, 8, pp. 175–84. B.N. Mukherjee, Studies in Aramaic Edicts of Aśoka, Calcutta, 1980.
(19) P.M Fraser, ‘The Son of Aristonax at Kandahar’, Afghan Studies, 1979 (1980), 2, pp. 9–21.
(20) G. Pugliese-Carratelli and G. Garbini, A Bilingual Graeco-Aramaic Edict by Aśoka, Rome, 1964. See also J. Filliozat, ‘Graeco-Aramaic Inscription of Aśoka near Kandahar’, Ep. Ind., 1961–2, 34, 1–8.
(21) K. Kartunnen, ‘India and the Hellenistic World’, Studia Orientalia, 1997, 83, 264–271.
(22) K.R. Norman, ‘A Newly Found Fragment of an Asokan Inscription’, South Asian Studies, 1988, 4, pp. 99–102. M. Taddei, ‘Nuovo Iscrizione di Asoka del Nordvest’, in G. Gnoli and L. Lanciorti (eds), Orientalia Iosephi Tucci Memoriae Dicata, Rome, 1988. B.N. Mukherjee, ‘A Brāhmī Inscription of Aśoka from the North-West’, IHR, 1989–90, 16, 1–2, 130–1.
(23) F. Scialpi, ‘The Ethics of Aśoka and the Religious Inspiration of the Achaemenids’, East and West, 1984, 34, 1–3, pp. 55–74.
(24) H. Humbach, ‘The Aramaic Aśokan Inscription from Taxila’, Journal of Central Asia, 1978, I, 2, pp. 87–98.
(26) F.R. Adrados, ‘Aśoka’s Inscriptions and Persian, Greek and Latin Epigraphy’, in S.D. Joshi (ed.), Amṛtadhārā, Delhi, 1984, pp. 1–15.
(27) A.L. Basham, ‘The Rise of Buddhism in Historical Context’, Asian Studies, 1966, 4, p. 405.
(29) U. Scerrato, ‘A Bi-lingual Greco-Aramaic Edict by Aśoka’, Series orientale Roma, 1964, xxix, pp. 1–27.
(30) I. Mahadevan, ‘Recent Discoveries of Jaina Cave Inscriptions in Tamil Nadu’, in Rishabha-saurabha, Delhi, 1994, pp. 123–7.
(31) W.H. Siddiqi, ‘A Study of Aśokan Pillars: Re-erected by Firuz Shah Tughluq’ (mimeographed), 1976. B.Ch. Chhabra, ‘Aśokan Pillar at Hissar, Punjab’, Vishveshvarananda Indological Journal, 1964, p. 3.
(32) C.S. Upasak, The History and Palaeography of Mauryan Brahmi Script, Nalanda, 1960.
(33) A.H. Dani, Indian Palaeography, Oxford, 1963, pp. 29–30.
(34) S.R. Goyal, Brahmi—an Invention of the Early Mauryan Period, pp. 16–18.
(35) H. Falk, Schrift in Alten Indien: Ein Forschungsbericht mit Anmerkungen, Tubingen, 1993. See also, R. Salomon, JAOS, 1995, 115, 271–80.
(36) Indian Archaeology—A Review, 1973–4, p. 4.
(37) S.U. Deraniyagala, ‘Radio-carbon Dating of Early Brāhmī Script in Sri Lanka’, Ancient Ceylon, 1990, 11, pp. 149–68. The Prehistory of Sri Lanka, II, Colombo, 1992, p. 739 ff. F.R. Allchin, The Archaeology of Early Historic South Asia, Cambridge, 1995.
(38) D.C. Sircar, Select Inscriptions …, Calcutta, 1965, p. 81.
(39) K.L. Janert. ‘About the Scribes and their Achievements in Aśoka’s India’, German Scholars in India, vol. I, pp. 141–5.
(40) L. Gopal, ‘Early Greek Writers on Writing in India’, PIHC, 1976, pp. 544–52.
(41) N.A. Dandamaev and V.G. Lukonin, The Culture and Social Institutions of Ancient Iran, Cambridge, 1989.
(42) G.V. Brobinsky, ‘A Line of Brāhmī(?) Script in a Babylonian Contract Tablet’, JAOS, 1936, 56, pp. 86–8.
(43) A.L. Basham, ‘Aśoka and Buddhism: a Re-examination’, The Journal of the International Association of Buddhist Studies, 1982, 5, 1, pp. 131–43.
(44) R. Thapar, ‘Literacy and Communication: Some Thoughts on the Inscriptions of Aśoka’ Cultural Pasts, 439–52.
(45) A. Cunningham, Inscriptions of Aśoka, CII.I. London, 1879.
(46) A. Ghosh (ed.), An Encyclopaedia of Indian Archaeology, Delhi, 1989, has a number of useful entries and one hopes that it will be up-dated soon. Most entries include references to reports of excavations.
(47) D.K. Chakrabarti, The Archaeology of Ancient Indian Cities, Delhi, 1995. A. Ghosh, The City in Early Historical India, Simla, 1973.
(48) F.R. Allchin, The Archaeology of Early Historic South Asia, Cambridge, 1995.
(49) G. Erdosy, Urbanisation in Early Historic India, Oxford, 1988, is an attempt at such a study relating to the evolution of Kauśāmbi and its umland.
(50) T.N. Roy, The Ganges Civilisation, New Delhi, 1983.
(51) K.T.M. Hegde, ‘Scientific Basis of the Technology of Three Ancient Indian Ceramic Industries’, in B.M. Pande and B.D. Chattopadhyaya (eds). History and Archaeology, Delhi, 1987, pp. 359–62.
(53) S.C. Ray, ‘Evolution of Hindu-Buddhist Art in the Gangetic Valley: A Study on the basis of Stratigraphic Evidence of Archaeological Excavations’, in Investigating Indian Art, Berlin, 1987, pp. 273–90.
(54) G.A. Poster, From Indian Earth—4000 Years of Terracotta Art, New York, 1986. D. Desai, ‘Terracotta Dancing-girls’, Marg, 1985, 37, pp. 72–4.
(55) P.T. Craddock et al., ‘The Production of Lead, Silver and Zinc in Early India’, in A. Hauptman, E. Pernicka and G.A. Wagner (eds), Old World Archaeology, Bergbau Museum, 1989, pp. 51–69.
(56) V. Jayaswal, ‘Evidence for Resource Exploration for Art Activities at Chunar’, in R. Varma (ed.), Art and Archaeology of the Vindhyan Region, Rewa, pp. 133–44. P.C. Pant and V. Jayaswal, ‘Ancient Stone Quarries of Chunar: An Appraisal’, Pragadhara, 1990–91, no. 1, pp. 49–52. V. Jayaswal, From Stone Quarry to Sculpturing Workshop, Delhi, 1998.
(57) K.A. Chaudhuri, Ancient Agriculture and Forestry in Northern India, Bombay, 1977. R. Thapar, From Lineage to State, p. 29.
(58) R.N. Mehta, ‘Sudarśana Lake’, JOI(B), 1968, 18, pp. 20–8, is an attempt at locating the site and seems probable but not certain.
(59) Indian Archaeology—A Review, 1954–5, p. 19.
(61) A recent overview of the history and archaeology of the sites at Taxila is available in A.H. Dani, Taxila, UNESCO, 1986.
(62) Indian Arhaeology—A Review, 1962–3, p. 31. D. Mitra, Buddhist Monuments, Calcutta, 1971, pp. 42ff.
(63) R.K. Sharma and S.N. Mishra, Excavations at Kakrehta (Rupnath), Delhi, 1992.
(64) S. Seneviratne, ‘Kalinga and Andhra: the Process of Secondary State Formation in Early India’, in H.J.M. Claessen and P. Skalnik (eds), The Study of the State, The Hague, 1981. Social Base of Early Buddhism in South-East India and Sri Lanka, Ph.D. Thesis, JNU, 1985.
(65) I.K. Sharma, ‘Early Sculptures and Epigraphs from South East India: New Evidence from Amarāvatī’, in F.M. Asher and G.S. Ghai (eds), Indian Epigraphy, New Delhi, 1985, pp. 15–23.
(66) F.R. Allchin, ‘Patterns of City Formation in Early Historic South Asia’, South Asian Studies, 1990, 6, pp. 163–73.
(67) A.K. Narain and L. Gopal (eds), Chronology of the Punch-Marked Coins, Varanasi, 1966.
(68) S.C. Ray, Stratigraphic Evidence of Coins in Indian Excavations and some allied issues, Numismatic Notes and Monographs, 1–14, Varanasi, 1959. See also M.K. Dhavalikar, ‘The Beginnings of Coinage in India’, World Archaeology, 1975, 6, pp. 330–8.
(69) D.D. Kosambi, Indian Numismatics, New Delhi, 1981.
(71) J. Cribb, ‘Daring India’s Earliest Coins’, in South Asian Archaeology, 1983, Naples, 1985, pp. 535–51.
(72) Pāṇini II.4.21.
(73) V.S. Agarwala, India as Known to Panini, Varanasi, 1963, pp. 259–74, 474.
(74) John Irwin, ‘Asokan Pillars: a Re-assessment of the Evidence’, The Burlington Magazine, 1973, cxv, pp. 706–20; loc. cit., 1974, cxvI, pp. 712–27; 1975, cxvII, pp. 631–43; 1976, cxvIII, pp. 734–53. ‘The Prayaga-bull Pillar: Another pre-Aśokan Monument’, in H. Härtal (ed.). South Asian Archaeology, 1979, vol. II, Berlin, 1981, pp. 313–40. ‘Origins of the Pre-Aśokan Pillar Cult at Prayaga (Allahabad)’, JRAS, 1983, 3, ‘True Chronology of the Aśokan Pillar’, Artibus Asiae, 1983, XLIV, 4, pp. 247–65. ‘Buddhism and the Cosmic Pillar’, in G. Gnoli and L. Lanciotti (eds), op. cit. Some of Irwin’s views have in turn been questioned by S.P. Gupta, The Roots of Indian Art, Delhi, 1980, but the questioning has not furthered the debate in any essential sense.
(75) Vinaya I.337–57; Kosambi Jātaka, no. 428.
(77) Niharranjan Ray, ‘Maurya and Sunga Art’, in Indian Studies Past and Present, 1964–5, VI, pp. 53ff. J.C. Harle, in The Art and Architecture of the Indian Subcontinent, Harmondsworth, 1986, p. 31 supports a date of the first century A.D.
(78) D. Mitra, Buddhist Monuments, Calcutta, 1971, p. 24.
(79) J.C. Huntington, ‘The Lomas Rishi: Another Look’, Archives of Asian Art, 1974–5, 28, pp. 34–56.
(82) R.P. Kangle, The Kauṭilīya Arthaśāstra, Bombay, 1965.
(84) R.P. Kangle, ‘Some Recent Work on the Kauṭilya Arthaśāstra’, JAS, Bombay, 1968–9, 43–4, pp. 227–38.
(85) T. Burrow, ‘Cānakya and Kauṭilya’, ABORI, 1968, 47–9, pp. 13–17.
(86) L. Sternbach, Bibliography of Kauṭilīya Arthaśāstra, Hoshiarpur, 1973.
(87) L.N. Rangarajan (trans.), Arthaśāstra, New Delhi, 1992.
(88) Eva Ritschl and Maria Schetelich, Studien zum Kauṭilīya Arthaśāstra, Berlin, 1973.
(89) T.R. Trautmann, Kauṭilya and the Arthaśāstra, Leiden, 1971.
(90) G.N. Dwivedi, The Age of Kauṭilya, Agra, 1966; J.D.M. Derrett, ‘A Newly Discovered Contact between Arthaśāstra and Dharmaśāstra: the Role of Bharuci’, ZDMG, 1965, 115, pp. 134ff.
(91) S.R. Goyal, Kauṭilya and Megasthenes, Meerut, 1985.
(92) The application of the statistical method to the Harappa script for example, a far more complicated procedure than the linguistic analysis of a text, was carried out entirely in India by I. Mahadevan and published by the ASI, The Indus Script, Delhi, 1977.
(93) R. Thapar, ‘Patronage and Community’, in B. Stoller Miller (ed.). The Powers of Art, Delhi, 1992, pp. 19–34.
(94) P. Peterson (ed.), The Daśakumāracarita of Daṇḍin, II, Bombay, 1891, p. 52.
(95) K. Karttunen, India in Early Greek Literature, Helsinki, 1989.
(96) O. Murray, ‘Herodotus and Hellenistic Culture’, Classical Quarterly, 1972, 22, pp. 200–13.
(97) A. Zambrini, ‘Gli Indika di Megasthene’, Annali della Scuola Norm. Sup. di Pisa, 1982, 3, 12, 1, pp. 71–149 and 1985, 3, 15, 3, pp. 781–853.
(98) A.B. Bosworth, A Historical Commentary on Arrian’s History of Alexander, II, Oxford, 1995. ‘The Historical Setting of Megasthenes’ Indica’, Classical Philology, 1996, 91, pp. 113–27. See also, ‘Aristotle, India and the Alexander Historians’, TOPOI, 1993, 3/2, pp. 407–24.
(99) R. Thapar, The Mauryas Revisited, Calcutta, 1988, pp. 32ff.
(100) A. Dahlquist, Megasthenes and Indian Religion, Uppsala, 1962.
(101) D.K. Biswas (trans.), The Legend of Emperor Aśoka, Calcutt, 1967.
(102) S. Mukhopadhyaya, Aśokāvadāna, Delhi, 1963. J.S. Strong, The Legend of King Aśoka, New Jersey, 1983. See also, The Legend and Cult of Upagupta, N.J., 1992. L.W.de Jong ‘Notes on the Text of the Aśoka Legend’, in G. Pollet (ed.), India and the Ancient World, Louvain, 1987, pp 103–13.
(103) G. M. Bongard-Levin, ‘The Historicity of the Ancient Indian Avadānas: A Legend about Aśoka’s Deposition and the Queen’s Edict’, Indian Studies Past and Present, Calcutta, 1971. pp. 123–41.
(104) G.M. Bongard-Levin and O.F. Volkova, ‘The Kunāla Legend and an Unpublished Aśokāvadānamālā Manuscript’, Indian Studies Past and Present, Calcutta, 1965.
(105) Alaka Chattopadhyaya and the Lama Chimpa, Tāranātha’s History of Buddhism in India, Simla, 1980.
(106) H. Bechert (ed.), The Dating of the Historical Buddha, I and II, Gottingen, 1991.
(107) H. Härtel, ‘Archaeological Research on Ancient Buddhist Sites’, in Bechert, op. cit., pp. 61–89. H. Kulke, ‘Some Considerations on the Significance of Buddha’s Date for the History of Northern India’, ibid., pp. 100–7.
(109) A. Ghosh, An Encyclopaedia of Indian Archaeology, II, p. 420.
(110) Makkhan Lal, ‘Summary of Four Seasons of Exploration in Kanpur District, Uttar Pradesh’, Man and Environment, 1984, VIII, pp. 61–80.
(111) K.T.S. Sarao, Urban Centres and Urbanisation as Reflected in the Pali Vinaya and Sutta Pitakas, Delhi, 1990, pp. 213 ff.
(112) Vinaya Piṭaka, I.337–57. Kosambi Jātaka, no. 428.
(113) Vinaya Piṭaka, I.226–230. Dīgha Nikāya II. 86ff. F.E. Pargiter, The Purana Texts of the Dynasties of the Kali Age, London, 1913, p. 22.
(114) Sabhāparvan, 19.1ff.
(115) K.T. Sarao, Urban Centres and Urbanisation as Reflected in the Pali Vinaya and Sutta-pitakas, Delhi, 1990.
(116) Dīgha Nikāya II. 146 and 169.
(117) F.E. Pargiter, The Purana Text of the Dynasties of the Kali Age, London, 1913.
(118) R. Thapa, ‘Sacrifice, Surplus and the Soul’, History of Religion, 1994, pp. 305–24.
(119) R. Thapar, ‘Ethics, Religion and Social Protest in the First Millennium B.C. in Northern India’, In Ancient Indian Social History: Some Interpretation, New Delhi, 1978, pp. 40–62.
(120) H.C. Raychaudhuri, Political History of Ancient India, Calcutta, 1972 (7th edn.), p. 201.
(121) F. E. Pargiter, The Purana Texts of the Dynasties of the Kali Age, Delhi, 1975 (rpt.), pp. 21ff.
(122) P.H.L. Eggermont, ‘New Notes on Aśoka and his Successors’, Persica, 1965–6, II, pp. 27–70; 1969, IV, pp. 77–102; 1970–1, V, pp. 69–102; 1979, VIII, pp. 55–93.
(123) This confusion over expired and current years, and mistakes of much more serious kinds occur in H. Alahakoon, The Later Mauryas, Delhi, 1980. See my review article in The Sri Lanka Journal of the Humanities; 1981, VII, 1–2, pp. 153–65.
(124) A. Guruge, Asoka, a Definitive Biography, Colombo, 1994.
(125) I. Gershevitch (ed.), Cambridge History of Iran, vol. 2, Cambridge, 1985. S. Chattopadhyaya, The Achaemenids and India, Delhi, 1974 (rpt.).
(126) M.A. Dandamaev and V.G. Lulconin, The Culture and Social Institutions of Ancient Iran, Cambridge, 1989.
(127) S. Sherwin White and A. Kuhrt, From Samarkand to Sardis, London, 1993.
(128) G. Yamazaki, ‘The Legend of the Foundation of Khotan’, Toyo Bunko, 1990, 48, pp. 55–80.
(129) R.E. Emmerick, Tibetan Texts Concerning Khotan, Oxford, 1967, pp. 15ff. Guruge’s contention that there is no mention of Aśoka visiting Khotan (p. 389) is incorrect.
(130) R. Thapar, ‘Epigraphic Evidence and some Indo-Hellenistic Contacts during the Mauryan Period’, in S.K. Maity and U. Thakur, Indological Studies, New Delhi, 1987, pp. 15–19.
(131) M. Hara, ‘A Note on the Sanskrit Phrase, Devānampriya’, Indian Linguistics, 1969, 30, pp. 13–26.
(132) M.C. Joshi and J.C. Joshi, ‘A Study in the Names of Aśoka’, JOI, 1968, XVII, 4, pp. 415–24.
(133) A. Chattopadhyay, ‘Vidiśā Devī’, JOI, 1970, xx, 2, pp. 115–20.
(134) No wonder that President Premadasa of Sri Lanka writes in his Preface to the book that ‘From Dr Guruge’s study the Pali sources of Sri Lanka emerge as being nearest to the truth’.
(135) Guruge is not above contradicting himself: at one point Aśoka is a Buddhist when he builds the Hell (107) but a few pages later is being converted through this agency (119); I am faulted for saying that Aśoka declared Sampadi the heir-apparent and the text quoted states that he was the Yuvarāja (275). My reference to Daśaratha is not to Przyluski but to the Viṣṇu Purāṇa (352). The list of later Mauryas is from the last section of the Aśokāvadāna as I have stated and not from the A-yu-wang-ching as Guruge maintains. Guruge confuses the abbot Yaśas with the minister and the latter obstructing Aśoka is referred to in the Divyāvadāna, XXVIII, 382ff. Guruge needs to read Przyluski more carefully. If he does so he will discover that I am not presenting a confusion of texts, but that Przyluski refers to a large number and my references in turn are to the specific text which he quotes and not just to his book. Fa-hsien states unambiguously that Gandhāra was where Dharmavivardhana, the son of Aśoka ruled: J. Legge, p. 31 (380). Guruge contends that Aśoka was never viceroy at Taxila, a view based on an erroneous understanding of the statement in the sources, for the Aśokāvadāna for instance, states that Bindusāra recalled Susīma and appointed Aśoka to govern Taxila. If this is unacceptable then it could equally well be argued that the Sri Lankan Pali sources invented the viceroyalty at Ujjain in order to give status to the initial missionaries to Sri Lanka. HsÜan Tsang maintains that Aśoka was viceroy at Taxila: Watters, I, 241. Khallata is not a name in this instance but a qualifier meaning bald-headed: Divyāvadāna, XXVI, 372–3. My statement that Aśoka was in Taxila prior to Ujjain docs not mean that he went from Taxila to Ujjain but that he was in Taxila at an earlier period (379). There have been further discoveries of edicts in the last ten years but none of these date to the last years of Aśoka’s reign (271). Jalauka is linked to Gandhāra because the two areas in contemporary sources are often a compound term: Gandhāra Jātaka 406; Milindapanba 327–8 (331). To argue that the Aśoka described in the Rājataraṅginī was not the Mauryan king but a petty ruler of Kashmir, would be rejected by most historians (333). The story of the half āmalaka in the northern tradition cannot be rejected only because it is not confirmed in the southern Buddhist texts, for the latter have their own prejudices. Guruge states that the stories of Aśoka’s conversion in the Divyāvadāna and in Hsüan Tsang are dissimilar whereas I have pointed out a similarity. On this Watters states that Hsūan Tsang probably condensed the story from the Divyāvadāna and the Tsa-a-han-ching as they agree closely in all the main incidents and differ in some particulars: II, 89 (381) Guruge states that I have misconstrued Eggermont’s arguments in my discussion of Mauryan chronology (403) but curiously Eggermont in a more recent paper referred to above, agrees with me!
(136) Op. cit., pp. 452ff. The purpose of the book is made even more explicit not only in the statement that it was undertaken at the request of President Premadasa (519) but in the fuller explanation elsewhere: ‘The historical tradition of Sri Lankan (sic) had commenced with the founding of the Sinhala nation by Northern Indian Aryans who had migrated into this island in several waves of immigration. This nation founded in a new country, separated by a strong and rapidly expanding block of Dravidians from their fellow Aryans in the north of the subcontinent, had apparently felt the need to assert its cultural identity’. He goes on to say that a fully developed historical sense among Sinhala Buddhists brought about a national Sangha.
(137) A. Bareau, ‘The Place of the Buddha Gautama in the Buddhist Religion during the Reign of Aśoka’, in S. Balasooriya et al. (eds), Buddhist Studies in Honour of Walpola Rahula, London, 1980, pp. 1–9.
(138) H. Bechert, ‘The Importance of Aśoka’s so-called Schism Edict’, in L.A. Herns et al. (eds), Indological and Buddhist Studies, Canberra, 1982. See also N.A. Jayawickrama, ‘A Reference to the Third Council in Aśoka’s Edict’, University of Ceylon Review, 1959, 17, pp. 61–72.
(139) G. Yamazaki, ‘The Lists of the Patriarchs in the Northern and Southern Legends’, in H. Bechert, The Dating of the Historical Buddha, p. 320.
(140) R. Lingat, Royautés Bouddhiques, part I, Asoka et la Fonction Royale, Paris, 1989.
(141) A.L. Basham, ‘Aśoka and Buddhism …’, is of the view that it developed later else he might have referred to himself as such in the inscriptions.
(142) S.J. Tambiah, World Conqueror and World Renouncer, Cambridge, 1976.
(143) E. Lamotte, History of Indian Buddhism, pp. 233–4. R. Thapar, ‘Religion and the Social Order’ in Contributions to Indian Sociology, 1987, 21, 1. pp. 157–65.
(144) R. Thapar, ‘Renunciation: the Making of a Counter Culture?’, in Ancient Indian Social History: some Interpretations, New Delhi, 1978, pp. 63–104. ‘Householders and Renouncers in the Brahmanical and Buddhist Traditions’, in T.N. Madan (ed.), Way of Life, Delhi, 1982, pp. 273–98.
(145) S. Chattopadhyaya, Bimbisāra to Aśoka, Calcutta, 1977.
(146) G.M. Bongard-Levin, Mauryan India, New Delhi, 1985.
(147) This was initially argued by D.D. Kosambi and questioned by A. Ghosh, op. cit. and supported by R.S. Sharma, ‘Iron and Urbanisation in the Ganga Basin’, IHR, 1978, I, pp. 1ff and V.K. Thakur, Urbanisation in Ancient India, Delhi, 1981, p. 65, and further questioned by N.R. Ray, ‘Technology and Social Change in Early Indian History’, Puratattva, 1975–6, 8, pp.132–8.
(151) Arthaśāstra, II.24.28.
(152) Cf. D. Chanana, Slavery in Ancient India, Delhi, 1960.
(153) R. Thapar, From Lineage to State, New Delhi, 1984, p. 83.
(154) R. Thapar, The Mauryas Revisited, pp. 32ff.
(155) G.K. Rai, Involuntary Labour in Ancient India, Allahabad, 1981.
(156) Arthaśāstra, II.1.37.
(157) R.Thapar, The Mauryas Revisited, pp. 32ff.
(158) N.N. Kher, Agrarian and Fiscal Economy in the Mauryan and post-Mauryan Age, Delhi, 1973. See also B.C. Sen, Economics in Kautilya, Calcutta, 1967.
(159) R. Thapar, The Mauryas Revisited, pp. 32ff.
(160) D.V. Subba Reddy, Glimpses of Health and Medicine in the Mauryan Empire, Hyderabad, 1966.
(161) Arthaśāstra, V.3.
(162) G. Fussman, ‘Central and Provincial Administration in Ancient India: The Problem of the Mauryan Empire’, Indian Historical Review, 1987–8, XIV, pp. 43–72. ‘Quelques Problemes Asokéens’, Journal Asiatique, 1974, pp. 376ff.
(163) H. Claessen and P. Skalnik (eds), The Early State, The Hague, 1978. H. Claessen, ‘The Internal Dynamics of the Early State’, Current Anthropology, 1984, 25, 4, pp. 365–79.
(164) R. Thapar, ‘The State as Empire’, in H. Claessen and P. Skalnik (eds), The Study of the State, The Hague, 1981, pp. 409–26. The Mauryas Revisited, Calcutta, pp. 1–31.
(166) D.D. Kosambi, An Introduction to the Study of Indian History, Bombay, 1956, pp. 185 and 218.
(168) R. Thapar, From Lineage to State.
(170) J.M. Datta, ‘Population of India in about 320 B.C.’, Man in India, 1962, 42, 4.
(171) Makkhan Lal, Archaeology of Population, Varanasi, 1984, pp. 40ff.
(172) R. Thapar, ‘The Role of the Army in the Exercise of Power in Early India’, in A. Chaniotis and P. Ducrey (eds) Army and Power in the Ancient World, Stuttgart, 2002.