Ancient Kosala And Mmagadha

Author(s): Dharmanand Kosambi
Publisher: D D Kosambi
Catalog link: https://jainqq.org/explore/269365/1

Page #1
ANCIENT KOSALA AND MAGADHA By D. D. KOSAMBI The coins of The Paila Hoard (which are grouped in chronological order in another paper) belong in all probability to ancient Kosala, and to a period ending in the 5th century B. C. The ultimate problem is to identify the groups with kings whose names survive in the scanty and conflicting literary records, for the marks on the coins are heraldic, not epigraphs. No manual of heraldry is known that covers these, nor have any inscriptions been discovered for the locality and period. Most of the archaeological work has concentrated upon the Buddhist construction that covers the best-known sites, without troubling to discover what lies below. There exists no connected document which could be glorified with the title of a chronicle, annals, or history. For all these reasons, it is not only necessary to set down our hypotheses, but to outline the historical background. This sketch will be found to differ from those of books most commonly accepted as authoritative. The relative chronology of our coin-groups remains independent of the historical reconstruction. The essential features have usually been missed altogether, in the traditional manner of foreign historians when faced with the comparatively meagre, unreliable, "historical" data that can be gleaned from Indian documents. The principal characteristic of the times is a great social ferment which ended with the stabilization of a new society. The underlying historical processes 1 Whenever no direct reference is given to save space), the ultimate source will be found quoted and discussed in G.P. Malalasekara's Dictionary of Pali Names (2 vol., London 1938); CAI=Cambridge Ancient History of India, vol. I, Ancient India, ed. E. J. Rapson, Cambridge 1922. I have generally neglected Jain sources, of which a review can be found in J.C. Jain's thesis Life in Ancient India (Bombay 1947). Though CAI gives a clear, charming and sympathetic but at the same time superficial picture of Indian society, it brings Buddha's death down to 483 B.C. whereas tradition puts it 60 years earlier and this tradition is accepted by Vincent Smith in the Oxford History of India (2nd, ed. S.M. Edwards, Oxford 1922). For primary sources, I have used the Pali Text Society's editions for DN=DighaNikaya, MN-Majjhima- Nikaya, Fat=Jataka, DhA=Dhammapada Atthakatha. Citations are by sutta and story number. Jataka translations by J. Dutoit (in German) and DhA by Burlingame (Harvard Oriental Series 28-30) were most useful in finding references quickly. But for analysis of Buddhist sources, I can only acknowledge my great debt to my father, who first pointed out the economic basis of ahimsa before 1912 ; his Bhagavan Buddha (in Marathi, Nagpur 1940) still contains some of the best published critique of the sources, and the second chapter of the first volume gives the best discussion I know of the 16 traditional kingdoms of the Anguttara-Nikaya, showing that most of them had ceased to have any independent existence or political importance by the time of the Buddha. For brahmin sources, the most used is F. E. Pargiter's DKA="The purana text of the dynasties of the Kali age"; the same author's name is used in citing his rather optimistic rationalization of very late, dubious, and not yet critically edited sources : "Ancient Indian Historical Tradition". A refers to the Arthasastra, for which I usually follow T. Ganapati Sastri's text and commentary. Asoka's edicts in Hultzsch's edition (Ep. Ind.I.) are referred to as R. for the rock-cut and P. for the pillar texts, by number.

Page #2
ANCIENT KOSALA AND MAGADHA 181 manifest themselves in two different aspects, religious and imperial. The first led to the spread of Buddhism over the entire sub-continent; the second also culminated with Asoka, in the Magadhan conquest of India. 1. FREE ARYAN TRIBES. These two aspects of the superstructure cannot be treated as curious accidents unconnected except by the chance conversion of Asoka to Buddhism. The canon compares the Buddha again and again to a cakravartin. The two great non-violent religions, with their basic idea of ahimsa, arise in UP at the same time as the beginning of Magadhan expansion. The Buddha and Mahavira are not only contemporaries but close in geographical situation. Their religions differ by very little--in spite of bitter theological controversies and polemics--in philosophy or in the organization of their monastic orders Both these teachers made their own contributions as the culmination of various systems proposed by a long line of ascetic predecessors. Finally, both are ksatriyas of one particular type, namely men belonging to clans that are proud of their Aryana descent without following the highly brahminized vedic ritual. The clans are free oligarchies where the power vests in the general assembly of all members. This has left its mark upon the Buddhist monastic order, which is called by the same name as that for such a tribe, samgha. The initiate becomes a "son of the Sakyans', and must behave nobly. The tribe of the Licchavis, in which Mahavira was born, undoubtedly preserved this assembly, the sabha. The constitution of the Buddhist Samgha clearly preserves the old tribal constitution as adapted to monastic purposes. We know that the voting was by sense of the meeting, not a counted majority ; the decisions were carried out by the elders of the various families; the king was one of these senior oligarchs, usually elected in rotation but without any of the absolute power that characterizes, in theory, the later Indian monarch. There is a great difference between the tribal oligarch rajan and the absolute maharaja. Buddha shows his high admiration for the Licchavis again and again; they differed from his own tribe, the Sakyans, only in their greater military strength, acknowledging no overlord, and having greater power over the surrounding population ; the Sakyans still put their hands to the plough (Jat, Nidana). . The tribes could and did produce such great ascetic leaders for two main reasons. First, that they had not been penetrated by Brahminism though of ! I take Aryan as denoting at this period some special tribal organizations and their derivatives, linked together by affinities of language, ritual, a common aggressive mode of patriarchal life; there is no question of racial purity, though original racial differences as well as later pride in purity of lineage must be admitted. In Pali ayyo is used as synonymous for free man as against a dasa, acquiring gradually the same value as 'mister'. J.J. Meyer in translating A 3.13 does not know this meaning of Arya-prana, "living as a free man". 3. This would be proved by the story of Bhaddiya, Vinaya Pitaka, Cullavagga 7.

Page #3
182 D. D. KOSAMBI ancient "Aryan" lineage, and secondly that the free democratic tribes were themselves breaking up: The first point is a natural consequence of the fact that the ancient vedic religion and ritual placed enormous importance upon the sacrifice which had become progressively so complicated as to require the services of professional priests, and was of use only to powerful and aggressive kings. The tribal householders performed their own simpler ritual. Brahmins as such received no respect from the Sakyans (DN 3). The Buddhist samgha is firmly against all ritual (samkhara). The point is also proved by the low positions that later brahminical works assign to such tribesmen, namely that of mixed castes. For example, the Manusmoti 10.22 'derives the Mallas and the Licchavist from mixed ksatriya and vratya descent, the vratya being defined two stanzas earlier (and in 2-39) as he who had not gone through the brahminical initiation ceremony to which he was entitled--which meant that he belonged by birth to the Aryans. We know independently that the Licchavis, also called Vajjis (=vrajins, nomads) can be equated to these vratyas. The Atharvaveda xv makes a desperate but unsuccessful attempt to capture the vratya ceremonial. Finally, the Licchavis were certainly not of low birth, for a marriage alliance with them was sought by powerful kings of dubious lineage as a method of ennobling themselves. The best known case is that of the Gupta kings; the termination--gupta is to be used only by vaisyas according to orthodox rules. But the founder Candragupta I married the Licchavi princess Kumaradevi, struck coins in their joint names and with their joint images, though the Bhagavad-gita shudders at the idea of miscegenation. The son and successor Samudragupta carefully proclaims himself son of a daughter of the Licchavis. Yet the actual royalty cannot be derived from the Licchavis, who ceased to be of any political or military. importance about eight hundred years earlier. The alliance is clearly patent of nobility, for the Puranas in mentioning the first Guptas as local rulers in the Gangetic basin (DKA 53, 73) call them, along with others, 'niggards in graciousness, untruthful, very irascible, and unrighteous', though we get an entirely different picture from the great Sanskrit literature of the Gupta period at its best. This ascent however, began long before the Guptas, and at a time when the Licchavis could be dangerous enemies. Bimbisara, king of Magadha and special friend of the Buddha, had a Licchavi girl named Cellana among his queens. 4 The reading is nicchivi or nicchavi, a known variant for Licchavi as seen from Malalasekara subLicchavi. 5 She is taken as mother of his son, murderer, and successor Ajatasatru, who is called Vedehiputto (DN 2), son of Vedehi or of the Videha princess. But Videha is not the home of the Licchavis, who were based upon Vaisali, the modern Basarh and Raja Bisal Ka Garh near Muzaffarpur. The Satapatha Brahmana i. 4. 1. 17 makes both Videhas and Kosalas descendants of Videgha Mathava. Bimbisara's chief queen is called Kosaladevi. The atthakathas explain vedehiputta as "son of a learned woman" which, though unconvincing, shows that the accepted Buddhist tradition makes Ajatasatru's mother a Kosalan. Cellana, whether she bore Ajatasatru or not, would still count as his mother, along with all the other wives of king Bimbisara, according to the older rule.

Page #4
ANCIENT KOSALA AND MAGADHA 183 That the Malla Bandhula and his nephew Karayana were in Pasenadi's service (DHA. 4. 3) though the Mallians did not acknowledge any suzerain, indicates, as surely as the fact of many such tribesmen becoming monks, that the tribal life and system no longer gave satisfaction to its ablest members. Later feudal developments are foreshadowed by the absolute monarch granting whole villages to his priests or officials. Payasi, called a rajanya (DN 23) holds a village by grant of Pasenadi, not by force of his own arms nor conquest by his tribe ; of course there is no question of the consent of the villagers. ained as they wharata grudgingly Muh. 13.35 V The "Aryanization" of some aboriginals like the Matangas while others of the same tribe remained as they were and sank to the status of a very low caste is demonstrable. The Mahabharata grudgingly recognizes such wild tribes as ksatriyas, degraded by the wrath of brahmins (Mbh. 13.35 Vulgate). In particular Mbh. 12. 297 (Vulgate) mentions "The ksatriyas called Atirathas... Svapakas, Pukkasas, Nisadas ... Candalas" as sprung from the original four castes by intermixture. These are clearly aborigines whose survivors in developed localities retain their primitive ways and rites, hence are abhorred. For example, the Candalas, whose inclusion in the above list of fallen ksatriyas is notable, lived in their own special villages apart from all others, had their own language ; (Jat. 498) though brahmin girls washed out their eyes when defiled by sight of a Candala, some of the cleverer ones could manage to secure the necessary education at a distant center like Taxila, and ultimately pass themselves off as members of a higher caste, even brahmins ! Perhaps the best case is that of the Koliyas, against whom their neighbouring Sakyans fight by methods forbidden in traditional Aryan warfare, e.g. poisoning the water, reminiscent of some American tactics against the Apaches. Yet the Koliyans claimed and received a share of the Buddha's ashes ; verses at the end of DN 16 identify them as the Nagas of Ramapura, clearly pre-Aryans? on the frontier of the developed region. Later, the Mahavastu (355) assigns them respectable ancestors, the sage Kola of Benares and Sakyan mothers. Then we find Naga kings in the south (DKA 49, 53) while aboriginal Nagas still exist in the forests of Assam. This is what is meant by progressive Aryanization. It is one of several processes going on simultaneously, along with decay of the older Aryan tribes, rise of autocracies, of new religions, and greater volume of trade. The process continues in the same marginal regions, to the present day in a different guise, namely a primitive aspect of the class struggle. The 6 The Candalas and heretical ascetics are to dwell beyond the cemetery and burning grounds, outside a city, according to A 2.4. The low status of Candala and the often synonymous svapaka seems to derive from their cating dogs, which to the Iranian branch of the Aryans at least would make them as unclean as cannibals. ? Mahavagga 1.63 shows that Nagas could not join the order, and this exclusion has left its mark on the Buddhist initiation ceremony in the form of a question, "are you a human being ?", meaning thereby that the novice should not be a naga.

Page #5
184 D. D. KOSAMBI Census Of India, 1921, vol. V, pt. 1 p. 347-8 gives lists of castes that were then trying to advance themselves groupwise in the social scale, while those already so advanced denied their claims. Many of the 'castes' in the list are occupational, developed from tribes ; remnants of tribal assemblies appear now as caste sabhas (p. 346). But registration as a higher caste by the census has an economic basis which does not appear there, being more clearly seen in the standard method : the hiring of brahmin priests by the 'caste' or by a section thereof as soon as a certain amount of money has been acquired. This generally means revision of ritual, often rewriting of the tradition by assimilation to some legend in the puranas or Mahabharata, and changes in dress. The Census volume (chap. xi) has discused (without realizing the implications !) Europeans, Anglo-Indians, Christians also under caste (p. 368), showing a parallel process : individuals of the latter two trying to upgrade themselves by one stage, just as in the USA individuals having a negro ancestor try very often to conceal the fact if physiognomy permits. This is clearly a class phenomenon, for social and ultimately cconomic advantage, not one of what is strictly speaking immutable caste. Hence it is obvious that the origins of caste also lie, at a more primitive level, in the means and relations of production; that the institution survives only as long as it is not incompatible with the form of property. There is no question but that the higher castes were and are, on the whole, economically better off; where they are not, or have not greater opportunity, we find an absence of caste tension while no attempts are made at 'passing over'. Similarly for the 'Aryanization' in the early period, which was obviously due to the superiority of 'Aryans' in war and in the struggle against nature; this meant that not only their tools and weapons, but also their forms of property, ritual, social institutions, and (more slowly) the language (cf Census p. 296) were adopted ; it seems clear that the brahmin caste played a very similar role in this older process also. 2. ABSOLUTE MONARCHS. The most powerful king of Buddha's day was Pasenadi, king of Kosala whose capital at Savatthi (skt. sravasti) is now represented by the ruins of the double village Saheth-Maheth on the frontier of Bahraich and Gonda districts. This king also desired a marriage alliance with a tribe, the Sakyans. The reason given is his veneration for the most famous Sakyan, the Buddha. But actually, the respect the Sakyans enjoyed apart from their connection with the Buddha was the main cause. The demand for a maiden who would be Pasenadi's chief queen was, surprisingly enough at first sight, highly embarrassing to the Sakyans. Two different views are reported of them on this occasion : "We live under the suzerainty of the king of Kosala. If we do not give him one of our daughters, great enmity will arise thereby; but if we do give him one, the traditions of our clan will be destroyed thereby. What is to be done?" (Jat 465). "The king is an enemy of ours. Therefore, if we refuse

Page #6
ANCIENT KOSALA AND MAGADHA 185 to give him what he demands, he will destroy us. Moreover, he is not of cqual birth with ourselves. What is to be done?" (DHA. 4.3). Both of these are substantially correct. Pasenadi's suzerainty over the Sakyans is elsewhere attested (D:N 27). The low birth of Pasenadi is confirmed by his marrying Mallika, flower vendor and daughter of a flower-gardener, (Jat 415) who became his chief queen. The Lalita Vistara 3, in reviewing objections to all the royal families in which the Buddha could have taken birth reports kausalakulam matangacyuty-upapannam, na mitr-pitr-suddham hinadhimuktikam na kuloditam ... The interesting phrase is matanga-cyuty-upapannam, descended from the Matangas. The word, according to Chinese and Tibetan translators, means a low and savage tribe, apparently our present Manga ; matanga can also mean an elephant, (like naga, which again means cobra, noble, and demon), presumably a totem. But that Pasenadi was low-born is clear. The Sakyan solution of the problem itself proves this. They cheat him by offering the beautiful Vasabha-Khattiya, daughter of Mahanama Sakya, by a slave woman Nagamunda; the mother's name combines the names of two well-known savage tribes. Pasenadi is deceived successfully about the girl's birth, and makes her his chief queen for a while. The insult is later discovered but forgiven by Pasenadi on the Buddha's intercession. The son of the union, however, succeeds in wiping out the Sakyans as a measure of revenge. The royal houses of Kosala and Magadha had intermarried. Bimbisara and Pasenadi had each taken the other's sister in marriage ; after a certain amount of fighting with his own nephew, Pasenadi cemented the truce by giving his daughter Vajira to Ajatasatru. Incidentally, this extremely close relationship would not be admissible under current brahminical rules. It follows that both royal houses were on the same low level of descent. The puranas call these Magadhan king, ksatrabandhavah (DKA 22); the malevolence and spite of the termination -bandhu is brought out by the Chandogya Upanisad 5, 3.5, where a discomfited brahmin calls the more learned ksatriya rajanyabandhu. For that matter, Magadhan brahmins who took part in vratya rituals and were otherwise lax are called brahmabandhu. Now Prasenajit is assigned high lineage by just one source, among the Iksvakus in the Puranas (DKA 11); this can only be by confusion with another Prasenajit of Kosala before the Kali age (Pargiter 145, 150). The purana list is obviously garbled in its attempts to report all Iksvakus, whence just before Prasenajit we find Sakya, Suddhodana, Siddhartha, Rahula. The reason is that the Sakyans claimed descent from Iksvaku as did the older and perhaps mythical Prasenajit. The puranic list ends with Sumitra, the last of the Iksvakus according to an ancient traditional sloka (DKA 12)-in spite of the much later southern Iksvakus (DKA 49). But this Sumitra, in the more accurate Buddhist records, was the last king of Videha, with capital at Mithila (Lalita Vistara 3), and

Page #7
186 D. D. KOSAMBI had no successor, the line having ended with him. Pasenadi never claims to be an Iksvaku like the Buddha, but only a ksatriya and a Kosalan like him as well as of the same age (MN 89). The point is of considerable importance, for the genealogy shows that the kings were of a type foreign to the previous traditional ruler, the elective or hereditary tribal king who was accompanied in battle by all the able-bodied men of the tribe. There the king had to be a member of the tribe, whether by birth or adoption, and his powers were limited by the assembly. His army was the armed tribe as a whole. But such are not the kings of Magadha or Kosala as we see them. The Kosalans no longer function as a tribe, if indeed a Kosala tribe still existed at the time. The king rules over subjects, the vast majority of whom do not belong to his tribe, nor perhaps to any tribe at all. The Magadhans were also a tribe, but the same is true of the Magadhan kings from Bimbisara ; Magadhan later means trader (Manusmoti 10.47), so that either the first traders in the region where the Manusmoti was written came from Magadha, or the remnants of the tribe had developed into a guild ; the two are not mutually exclusive. The Arthasastra Magadhan appears both as bard (A 3.6, 7;5.7; 10.3) and a mixed caste. Malla survives only in the meaning of wrestler, like Pehlawan in Persian. One main reason for this collapse of old institutions was the development of private as against tribal property, following conquest over aboriginal populations and the development of the tribe into an oligarchy. The process was not confined to the Kosala and Magadha tribes. Thus we find the Sakyans and certainly the Licchavis in the oligarchic stage, living by the labour of others. The Licchavis are not only individually rich, but collect tolls, have spread into Videha and Kosala where, with the Mallas, they made a confederacy of 18 clans which was the strongest power to oppose the kingdom of Magadha. Their expansion is not pioneering8 in a virgin wilderness but conquest of settled territory which had its own rulers before them. Keeping the Licchavi tribe pure and exclusive meant conquest and exploitation, the development of a tribal democracy into a tribal oligarchy. It therefore meant slackening of tribal bonds. Both would naturally lead the more sensitive members like the Buddha and Mahavira to turn to the contemplative life. It was by completing this decline and dissolution of the tribal constitution through his intriguing minister Vassakara that Ajatasatru succeeded in conquering the Licchavis. The intrigue is supposed to have taken only three years, but the decay must have begun much earlier, to be accelerated by the gain in personal wealth among the Licchavis. Viewed on a larger scale, this was the inevitable consummation of Aryan conquest, of the development of helotage as the sudra caste, exploited $ The word nigama, sometimes translated as 'suburb', seems rather to mean tribal settlement in a new territory, an act of pioneering perhaps in the ancient Greek or Roman sense of 'colony'. The Sakyan niganna where Pasenadi is supposed to have had his last meeting with the Buddha was at a distance of three yojanas from the nearest city, so that the meaning of suburb cannot apply here,

Page #8
ANCIENT KOSALA AND MAGADHA 187 by an armed body of warriors which had become the ksatriya caste and begun to exploit other Aryans too, starting with the vaisyas. There are two concomitants of this new monarchical system. The internal is characterized by a standing army and permanent officials without tribal bonds, owing allegiance to the king alone ; and of course the kings being of uncertain birth, neither supported nor controlled by tribal custom, show a remarkable tendency to usurpation. The external development is conquest, in particular the great efforts made to suppress what was left of the free tribes. The king has now a special office, that of army commander sena pati. Bimbisara is called seniya in Jain records and offers command of his army to the Buddha. Pasenadi has the Malla Bandhula as minister and perhaps army commander. Pasenadi's son Vidudabha had the title sena pati in his father's lifetime. Pasenadi had Bandhula killed by treachery, suspecting him of ambition beyond his station. The fear was not unjustified because later Viludabha usurped the throne, being invested with the regalia by Bandhula's nephew, the minister Digha-karayana ('long' Karayana; possibly the Dirgha Carayana of A 5.4), when Pasenadi had gone on a final visit to the Buddha (MN 89 atth, DhA 4.3, Jat. 465). Pasenadi himself had been crowned just after completing his studies by his own father (DhA 4.3) king Mahakosala. Bimbisara was placed upon the throne at the age of fifteen by his father (Mahavamso 2. 28). But abdication did not save Bimbisara, who was imprisoned and finally murdered by his son Ajatasatru. The pattern is thus well-established. The state as a mechanism of violence unconnected with and therefore hostile to tribal power has come into being, and its control could be acquired by violence. Purity of descent or legitimacy are of minor importance. One of Bimbisara's queens was the beautiful courtezan Padumavati miraculously brought from Avanti to be the nagarasobhini of Rajagpha. Mahapadma Nanda, the last great king before the Mauryans, is reported to have been the son of king Nandin by a sudra woman (DKA 25). 3. CONFLICT WITH THE SAMGHAS. One remarkable external consequence of this new state is the extermination of free samgha organizations. The puranas are impressed by Mahapadma's complete annihilation of all ksatriyas, like a second Parasurama (DKA 25); but the context makes it quite clear that he destroyed only those ksatriyas who lived in tribal units. The names are given : Iksvakus, Pancalas, Kaseyas, Haihayas, Kalingas, Asmakas, Kuravas, Mithilas, Surasenas, Vitihotras. Of these, the royal Iksvakus were already extinct, as also the kings The sreni in A 7.11 and elsewhere seems to mean an association of settlers, who also bore arms; and could be dangerous if provoked beyond endurance, Thus seniya may mean leader of a sreni; other senivas are known, such as the naked ascetic of MN 57, who lived like a dog.

Page #9
188 D. D. KOSAMBI A Kuru king is mentioned in the Ratthapala sutta (MN 82; Kuru territory in DhA 2.1.4-5), but again the context shows that we have to do with a petty tribal king. The tradition seems to connect an extermination of tribal ksatriyas with the ending of certain ancient tribal king-lists. Mahapadma's action is founded upon good precedent, and the whole process emphatically set out for Indian kings by the Arthaldstra, book XI, (sections 160-161), which is dedicated to handling samghas. "The acquisition of a samgha is better than that of an army (danda) or an ally. Samghas are invincible by others because of their unity (or organization samhatatcat.) Those favourably disposed should be conciliated by gifts; those opposed, destroyed by sowing dissension. In Kamboja, Surastra the ksatriya groups live by husbandry and weapons. The Licchavis, Vrajikas, Mallakas, Kukuras, Kurus, Panicalas and the like. live only by title of kings." The book goes on to give details of the political technique whereby the samgha organization may be weakened. The importance of this section seems to have been missed by our historians. It shows first of all the existence-whether actual or traditional at the time of writing does not matter-of certain samghas, whose names we have met elsewhere. It also states explicitly (I follow the text and commentary of T. Ganapati Sastri, not R. Shamasastry's translation) that such samghas were generally too strong to be destroyed by military action. This is fairly well confirmed by Alexander's campaign in the Punjab, where the stiffest opposition came from such tribal organizations; the Purus still existed, for king Poros had a nephew also called Poros, whence both should be Pauravas in the old tradition whereby the king is called by the tribal name. We also hear of the Malloi, identified with the later Malava though the lack of the extra syllable should make them a branch of the Mallas; and the Oxydrakai, presumably ksudrakas. The Arthasastra does not overestimate the formidable power of such samghas in resisting direct assault. So a technique of promoting internal discord is carefully set out, being exactly the same as that practised by Ajatasatru and his minister Vassakarale against the Licchavis. Ajatasatru completed his conquest by force of arms. The Licchavis survived for nearly a thousand years thereafter, down to the Gupta period but not as a samgha with independent military power, obeying none but their own assembly. Again this is not an isolated example. Vidadabha attacks the Sakyans, without first sowing discord, supposedly to wipe out the insult of his own birth. But though the action is represented as a revengeful massacre, Sakyans did survive (DhA 4.3); kings from places as distant as Ceylon are said to have married Sakyan brides (CAI 607). Clearly, they did not survive as a selfgoverning tribe; the Kosalan king, though their nominal overlord, must have feared their tribal existence and semi-independence. The parallel here is a bit too close, for the Jain sources have Ajatasatru's grandfather, the Lic 10 I have shown in the ABORI 37, 1952, pp. 53-60 that the Sanskritization of the name as Varsahara cannot be correct. One should take it as rasyakara, he who wins over, a nickname after his great feat.

Page #10
ANCIENT KOSALA AND MAGADHA 189 chavi (Haihaya) Cedaga, drowning himself as does Vidudabha's grandfather Mahanama. Thus the kings ruined by direct attack those tribes whose institutions still retained a dangerous measure of democracy combined with military power. This put the finishing touches upon the process of decay and dissolution which we have already pointed out as affecting all tribes of the sort in U.P. at least. 4. CONFLICT BETWEEN KINGDOMS. Aggression did not stop with destruction of the tribes, but extended also to other kingdoms. Anga had already been absorbed by its western neighbour Magadha, apparently at the time of Bimbisara. Kasi fell to Kosala scme generations before the Buddha, for the name is joint, Kasi-Kosala, as is Anga-Magadha. There are many scattered references to fighting between Kasi and Kosala long before Pasenadi (Jat 51, 100, 156, 336, 355, 428, 371, 532). The really interesting conflict, however, is the struggle between Kosala and Magadha, which was inevitable, granted the tendency of the times to powerful, central, absolute monarchies. Kosala is by far the greater kingdom, extending from the lower Nepal to the Ganges. But, as we shall see, Magadha is in many ways strategically better placed. During Pasenadi's lifetime, the conflict is inconclusive (Jat 239, 289, 492; DhA 15.3), victories being reported on both sides, with a marriage alliance at the end. The cause is a village near Kasi, part of Ajatasatru's mother's dowry, which Pasenadi wished to rescind. The fact that Magadha conquered ultimately is uncontested, for before the time of the Mauryans, Kosala is already part of the Magadhan cmpire. For the Paila hoard of punch-marked coins, it is of the utmost importance to determine just when Kosala faded out as an independent kingdom. This could not have been later than Mahapadma Nanda, and I suggest that it was far earlier, either at the time of Ajatasatru or immediately after. In the first place, we hear of no ruler of Kosala after Viludabha. The centre of gravity shifted to Magadha, where the first council after the Buddha's death is held at Rajaglha; but Kosala was more important during the Buddha's lifetime. The Buddha himself spent at Savatthi, no less than 25 "rains" during the period of his ministry, far more than those spent in all other residences put together; 871 suttas of the four Nikayas are supposed to be pronounced at Savatthi. There were many converts at Savatthi; the most famous lay follower being the trader. Anathapindika (Sudatta), whose gift of the Jetavana (purchased at a fantastically high price: a very rare case of land being bought outright for cash) is one of the great themes of Buddhist art. King Pasenadi was also a patron, though preoccupied. Vidudabha's massacre of the Sakyans did not mean persecution of the samgha, any more than Ajatasatru's killing Bimbisara and smashing the Licchavis prevent the Jains claiming him as one of themselves. Thus the shift to Magadha was part of a general trend, not confined to Buddhism. Finally, Ajatasatru had a claim upon the Kosalan throne which

Page #11
190 D. D. KOSAMBI could not have been pressed long after him: he was nephew as well as sonin-law of Pasenadi. It was to him that the aged Pasenadi, abandoned by all except a servant woman, fled for help against Viludabha, only to die outside the walls of Rajagpha; Ajatasatru performed the obsequies. His ministers are supposed to have dissuaded him from attacking Vidudabha, but the principal reason for abstinence is undoubtedly the Licchavi-Malla campaign. Without a secure rear, an attack against Kosala was unthinkable. On the other hand, Viludabha was engaged at the same time in crushing the Sakyans, so that he could not have attacked Magadha; soon after the Sakyan massacre, he was swept away with a great portion of his army, by a sudden freshet (DhA 4.3), the camp having been pitched in the dry bed of the river Rapti (Aciravati). Thus towards the end of Ajatasatru's reign Kosala had neither king nor much of an army; that we hear of no successor to Viludabha is therefore peculiarly significant. In view of the 'legal' title, there would be no opposition to any Magadhan attempt upon the throne of Kosala. atasatru bishing his son certainly of the stoc Pradyota had takena King's daughter Vis The puranas (DKA 21) state that Sisunaga, supposed founder of the dynasty to which Bimbisara and Ajatasatru belonged, defeated the Pradyotas, ruled in Girivraja (older Rajagpha), establishing his son as ruler in Kasi. This is impossible without a conquest of Kosala, which certainly could not have taken place four generations before Ajatasatru. As a matter of fact the Pradyotas rule Avanti, supposedly with capital at Ujjain. This is much too far for any real fighting with Magadha at so early a priod. We have two stories of conflict with the Pradyotas, the first being with the Vatsa or Vamsa prince Udena (skt. Udayana), king of Kosambi (skt. Kausambi). This ends with an elopement of Udena with the Pradyota King's daughter Vasuladatta (DhA 2. 1. 4), after the Pradyota had taken him prisoner. But that queen then drops out of the stories which concern an Udena who was contemporary of the Buddha, being husband of the beautiful and proud brahmin girl Magandiya (DhA 2.1.5) first offered by her father to, and rejected by, the Buddha. This story has a historical appearance, for the gotra is found in brahmin lists as Maganda or Maganda (but not the Makandika of the Divyavadana); the queen did her best to avenge the insult offered by the great Sakyan. Therefore, the other Udena romance must (if historical) be of some king not the contemporary of the Buddha. My work on the Taxila hoard shows the possibility of some Buddhist records being correct; they mention a Susunaga (Mahavamso 4.5-6), fifth after Ajatasatru, who was placed upon the throne by the people. This would be the proper time for an invasion by and repulse of the Pradyotas, and the line of Magadhan kings would end as Saisunagas. The Upanisads (Bs. Up. 2.1; Kaus. Up. 4.1) mention an Ajatasatru, king of Kasi, and it is quite possible that towards the end of his reign our Ajatasatru did acquire Kasi, perhaps the whole of Kosala, just as he had retained his claim to the village in the Kasi territory by fighting with Pasenadi. Ajata

Page #12
ANCIENT KOSALA AND MAGADHA 191 satru cannot be contemporary of Janaka; CAI 122 misunderstands the king's wish to emulate the already legendary Janaka. The Mallas were not under Pasenadi. They exercised, as did the Licchavis, powers of life and death over all in their territory, though otherwise such powers were the prerogative of a maharaja, not of lesser oligarchs like the Sakyans. A Malla-Licchavi confederacy, as reported by Jain sources (Jain 382) is highly plausible (though the Licchavis having an absolute chief at Vesali who was a Haihaya, is not). But many of the 18 confederate clans belong to Kasi-Kosalan territory, which means the absence of any real monarchy in Kosala. In destroying the federated tribes, Ajatasatru must also have invaded what had been Kosalan territory or at least sphere of influence. We hear of Ajatasatru fortifying Rajaglha a few years after the Buddha's death (MN 108) against an expected invasion by the Pradyota king, though such an attack would have been quite difficult; we hear nothing more about the actual fighting, if any took place. The fortifications however exist, being in' all probability the cyclopean walls of Rajgir. There is nothing against Ajatasatru having turned his attention westwards after settling the Licchavi problem; this would make it still more likely that Kosala was annexed to Magadha either by him or by his immediate successor. 5. ECONOMIC BASIS OF CONTEMPORARY RELIGION AND STATE. What was the mainspring of this expansion ? What is the change in society that brings to the surface two great and similar religions, not to speak of similar parallel sects, such as the Ajivakas? The answer seems to be fairly simple : the increase of population and of trade. The very style of Pali texts is vastly different from those of the brahmins. The unmistakable character of brahmin writing is not its insipidity but its rustic stamp, though brahminism is characterized by its extreme, facile adaptability in practice combined with rigid and apparently immutable theory, which leads to a blinking of the facts, to a neglect of reality paralleled only by other priest-ridden societies and literatures. The real brahmin of the Gshya-sutras is firmly based upon the village, despising and avoiding the town (CAI 237); the theory goes with him unchanged even when he has moved over to the town and forgotten the very appearance of the tutorial grove or forest; when he has learned to indulge in and profit by almost any trade. But in the early Buddhist literature the prominent class of lay followers is that of the traders and 'treasurers', people who bought and sold; also of those who followed the professions for a living. In addition, people of any caste could change their means of livelihood at will (CAI 203-4), in strong contrast to the smotis. Buddhism and Jainism preach ahimsa successfully because it was an economic necessity. The killing of animals in yajila sacrifice had become an intolerable

Page #13
192 D. D. KOSAMBI burden upon the subjects whose cattle were taken away without compensation; the fruits of the yajna were success in battle, and the constant warfare implied thereby meant heavy losses to the traders, and general distress. The whole basis had to be denied, and non-killing, which is what ahimsa means literally, cuts at the roots of both sacrifice and war. Primitive religion differs from civilized religion in one main respect, not matter how much of the ritual is carried over. The earlier type concentrates upon the cult and its ritual alone; the correct performance of certain rites. bring certain fruit. In the commoner cases, this yield is to be seen in the present world, in the way of better crops, avoidance of disease and of the deadly spirits that might cause it; perhaps success in ventures. But a very important portion deals with the problem of continuity, of what happens to a person after death. Something of the individual is supposed to survive, though it might be in changed form and of impalpable substance as in the dream world. Here comes the essential difference between the two types of religion: the later substitutes, to a considerable extent, good deeds and the good life for correct ritual. The change is due to the existence of a form of society far more heterogeneous than the original simple tribe-community. It is a discovery of the first magnitude that religion can be used to ease the difficulties of early society, to make the common life of diverse elements of society easier, The Buddhist sermons make clear the duties of a householder; its prime virtues such as truthfulness are social virtues. There is no Buddhist ritual whatever. The basic theory of transmigration may have been totemic in origin, the dead man reverting to the original totem animal that had first been the food of the tribe, hence its very substance; and ultimately become taboo (another basis for ahimsa) except in ritual feasts by one of those dialectic inversions that mark crucial changes in society. But observe that a man's action, karma weighs his soul, and automatically fits him for the body he will occupy in his next birth. Buddhism claims to point the way to the negation of this otherwise unending cycle of rebirth, but the action of Karma is a trade, barter, or wage-payment transaction. It differs as greatly as possible from the earlier ritualistic conception of an after-life, and from the later developments which substitute a short cut by means of bhakti, faith in a personal god. Ritual is a matter of individual choice for the layman, and often there are polemics against ritual as well as against brahminical pretensions (Jat 545). Finally, the new proponent of the religion bases himself upon the goodwill of society, living only by its alms; he does not support the king in sacrifice nor does he receive whole villages in fee as does the brahmin priest. In its initial stages, therefore, Buddhism succeeds because the karma doctrine, the social rules deduced from it, and the much cheaper mode of life of the almsman all appeal, economically as well as philosophically,

Page #14
ANCIENT KOSALA AND MAGADHA 193 to the society in which trade had gained a new importance, and which consists of many diverse types of tribal or local units in close contact. Exactly the same needs demand an absolute universal monarchy, the tyranny of one rather than the tyranny of many ; along with this, freedom from robbers that infested the great forest which still covered much of the land between settlements, freedom from irregular and excessive tolls charged by petty rulers or tribal oligarchs, and the opening of new trade routes. This can be proved very easily; at the same time it will be seen that Magadha had to be the center of the new kingdom, whether it conquered the others or was conquered by one of the others. The first land clearing in this region ran along the Himalayan foothills, as beginning of the transition from pastoralismi to agriculture. The course of development is quite clear. The Gangetic plain is alluvial, and in its original state, must have been densely forested, swampy in many places, and certainly devoid of heavy settlement till the age of metals, specifically the age of iron, had set in. This is one of the reasons for dismissing Pargiter's theory of ancient Indian historical development, outward from the Gangetic plain. The Nile, the Mesopotamian rivers, and the Indus permitted the first urban civilizations to develop because of the surrounding desert. Heavy land clearing was unnecessary. Along the Danube, the loess corridor allowed neolithic man and his successors to farm without fighting heavy forest. The rainfall and warm climate of the Gangetic valley would make riparian settlements impossible for neolithic man. Even at the time of the Buddha all movement ceased during the rains, which shows that the roads were mere cart tracks, rivers crossed by ferry or ford, and the land mostly jungle dotted with settlements. Under present conditions, the summer with its heat and dusty winds would be a far worse season for the wandering almsman in U. P. than the monsoon. When Pasenadi began a campaign during the rains (Jat. 176, 226) he not only lost to the borderland rebels, but also scandalized the Buddha by this infraction of what had become a scriptural rule. Alexander's success against Poros seems in part due to this sacrilege. 11 - The first river settlements, presumably at Allahabad (prayaga), perhaps also Benares (kast) and the like must almost certainly have been founded by refugees from the Punjab, driven eastwards by the Aryan invaders ; undoubtedly, their enemies must have followed in their wake. When Hastinapura was destroyed by flood, Nicaksu shifted his capital to Kosambi (DKA 4-5). But the major early settlements all lie along the northern foothills. The process of development is set forth in the Satapatha Brahmana i. 4. 1. 14 ff; the land was burned over, swampy places dried up, and cultivation could then 11 However, A 9.1 seems to imply the possibility of a campaign in the rains, so that the old rules, were abandoned in practice, by the time of the Mauryans, perhaps following the Greek example, or because the increasingly great distances made it necessary. 10

Page #15
194 D. D. KOSAMBI begin. The extreme limit of this went as far as Vidcha. The trade route then swung down to Patna, and across the river to Rajgir. This is the route from Savatthi to Rajagrha followed by traders like Anathapindika, and which the Buddha had begun to traverse in the opposite direction from Rajagrha when death overtook him near the Malla settlement at Kusinara. The road would not be difficult to trace, as it passed through Vesali, Setavya, Kapilavastu, Kusinara, Pava, nara, Pava, Bhoganagara. Another went from Savatthi south to Saketa and Kosambi, thence presumably to the Avanti kingdom. Both of these are good targets for systematic archaeology. The unit distance is the yojana, approximately nine miles (though the Arthasastra yojana was about 5 miles), the distance after which carts were outspanned. On special roads, there seem to have been royal stages for horse-chariots, the rathavinita from which MN 24 derives its name, and which must have been about the same as the yojana, for six yojanas separated Savatthi from Saketa, while king Pasenadi completed his journey between those two capitals in the seventh stage. The stages and outspanning places should be traceable even now. The western route from Savatthi led to Takkasila 147 yojanas away, and presumably allowed trade with the Delhi-Mathura region on the way. Taxila was a great center of learning, as is seen from the tradition that Pasenadi, the Malla Bandhula who was later his minister, and the Licchavi Mahali who afterwards took. service with Bimbisara had studied there together in their youth (DhA 4-3). The much earlier Kosalan prince Chatta fled to Taxila after Brahmadatta of Benares stormed Savatthi and took his father captive; he studied the three vedas there, and returned to find buried treasure by aid of which he recovered his kingdom (Jat. 336). Brahmin tradition supports this, for eastern brahmins travel regularly to the distant northwest to learn their main business, the firesacrifice (Br. Up. 3. 4. 1, 3.7.1). The punch-marked coins found in great profusion at Taxila belong to Magadhan kings for the greater part, even before the Mauryans. Rajagrha is an exception to this, being on the opposite bank of the river, though still in the foothills, this time of the Vindhyas. I suggest that its original importance derived from the minerals; in particular iron, which is found as easily smelted surface deposits in Dharwar outcrops, of which the hills about Rajgir consist. More would be available by trade from Chota Nagpur, where there still exist iron-working aboriginal tribes. From the Nepal hills, copper would be the likelier metal, but even this is more easily available in Bihar. No metal is to be found naturally in the immensely fertile alluvial portion of the Gangetic plain, the portion which is the most densely settled today and which had to be cleared with the greatest labourlabour which would have been impossible without iron. But given iron tools, the rapid opening up of U.P. proper was inevitable. According to J. A. Dunn, Memoirs of the Geological Survey of India, LXIX, part 1,1937 (reference by courtesy of Prof. K. V. Kelkar), Singhbhum district has very

Page #16
ANCIENT KOSALA AND MAGADHA 195 ancient workings of copper mines (p. 54-5), though the date is uncertain; the working technique was most efficient. At the time under discussion, the great demand and logical outlet for metals would be towards the fertile but heavily forested plain of the Ganges, with its great and increasing population. Of course, India's greatest deposits of iron as well as of copper occur in Singhbhum and Dhalbhum. The location of Rajgir is due also to its straddling this trade-route for metals, north-west to the U.P. The distance from the richest sources of ore is less than two hundred miles; Gaya is nearer, but would be more difficult to settle without cheap metal tools, whereas the hills about Rajgir make for less dense forest. The later port for this region was Tamluk, and place names beginning with tama mark ancient though later forgotten sites of copper deposits. The absence of pumps seems. to have stopped mining at water level, and may have been one of the causes of later Magadhan decline. With heavier trade which would follow the rapidly increasing population near the river, the Ganges becomes the easiest major route for mass transport, as compared to the old, slow, expensive, and unsafe, land routes. Thus Kosalan expansion towards Kasi is the natural search for a port, an outlet. Rajagrha was similarly attracted towards Patna, which inevitably became the major trade center, the greatest port, and so the capital. The last step, of shifting the capital, was taken by king Udayi (DKA 22), son according to Buddhist (DN 2) and grandson by puranic sources of king Ajatasatru. But we are very fortunately placed in being able to discover the first steps also. These were taken during Ajatasatru's reign by the ubiquitous mahamatya Vassakara, in building a fortress at Patna. The place was already a depot where parcels of merchandise were opened for barter (putabhedanam). But the Licchavis and Ajatasatru both levied toll upon the traders, who were ruined thereby; the Licchavi tolls seem to have been more irregular, and less justified, or at least with less protection attached thereto. So in the last year of the Buddha's life, we hear (DN 16 and atthakathas) that the stockade built by the traders was being magnified, and the city properly founded as a base of operations against the Licchavis. The most profitable trade then was that down the Himalayan watershed, and consisted to a great extent of perfumes and cosmetics, which are of the utmost importance in a hot climate where bathing is a necessity, refrigerants a discovery for the future. Of course, it is clear that a strongly fortified city at Patna would reduce the importance of Benares as a great port, and would at the same time enable a blockade of the great river to be set up. That is undoubtedly the reason why AjataSatru refused to give up his foothold in the Kasi territory, though as usual only personal motives are alleged for such economically necessary actions. Trade is primarily by caravans sent by the leading merchant of one center to that of another distant center, and the caravans bring back local

Page #17
196 D. D. KOSAMBI goods in exchange (Jat. 90). Nevertheless, the new society needs and uses coinage, which cannot be older in India than the 7th century B.C.; that is about the period at which it appears elsewhere in the world. In Buddhist works, in any case, the kahapana (skt. karsapana), meaning thereby the silver coin at the earlier stages and probably a bronze coin by the time of the commentators, is standard measure of value. Commodity production as a whole is still at a very low level, but certain luxuries (such as fine cloth from Benares) and certain necessities like salt meant trade over a long range. Had metals been more plentiful, and commodity production at a higher level, chattel slavery would have been far more general. But we find the chattel slave rare, usually some sort of an accomplished servant, and all slaves are treated better than the paid servants (DhA 5.1 gloss). The Arthasastra forbids the sale even of a sudra living as a free man (aryaprana), and the sudra was the great source of expropriable labour. Even brahminical works insist upon avoidance of the bride-price (Manusmoti 3.51), for it might seem purchase of the woman, thus degrading her to the status of a slave. I suggest that the absence of slavery and the ultimate victory of brahminism are both connected with the low density of commodity production, which made trade in human beings unnecessary, and country life as easy as that in the cities. 6. CHRONOLOGICAL POSSIBILITIES While the outline seems simple enough, its details are not, and it is necessary to point out the virtual impossibility of constructing a detailed chronology from conflicting sources. As far as kings are concerned, there is no need to prove the statement, which will be found true by anyone who takes the trouble of comparing our records. But even for uncontested Buddhist sources, the dates are not only doubtful but the time sequence is often impossible. Look, for example, at the story of Pasenadi, Vidudabha and Ajatasatru. The last is supposed to have become king eight years before the death of the Buddha. Pasenadi, being of the same age as the Teacher, must then have been in his 72nd year. After that he fights several battles with Ajatasatru; during flights from one of which he sees the sixteen year old Mallika and makes her his chief queen forthwith. But she seems to have alternated with Vasabha-khattiya as chief queen, and survived Pasenadi according to the Anguttara-nikaya. Vidudabha visits the Sakyans during his father's lifetime, being himself then sixteen years old, overhears a servant woman talk of him as "that slave-girl's son", learns of the deceit, and vows to wash his throne with Sakyan blood. This vow he fulfills at the very earliest opportunity. But the usurpation takes place in Pasenadi's (and the Buddha's) 80th year. After that, Viludabha makes three attempts (which I should take as three separate annual campaigns) upon the Sakyans, turning away each time because of the Buddha's quiet intercession. The fourth time, the Buddha recalls that the evil karma of the Sakyans in poisoning

Page #18
ANCIENT KOSALA AND MAGADHA 197 the river against their Koliyan neighbours has come to fruition like a maturing debt, and must be paid, so does not intervene. But the Buddha himself died in his eightieth year, while travelling slowly on foot from Rajaglha on the great trade route. He did not get as far north as Kapilavastu, dying at Kusinara on the way. Moreover, he passed the rains near the Licchavi headquarters, which would give no time for a meeting with Pasenadi, and makes the triple saving of the Sakyans a pure miracle. Even the commentators are not clear about Buddha's meeting with Sariputta in the Mahaparinibbana-sutta, for Sariputta was supposed to have died before this period. Some of these details may undoubtedly be clarified by discarding the miraculous and fabulous element-not always simple with Buddhist or any other ancient Indian works--but it seems to me that the precise chronology we need cannot be worked out by these methods. 7. MeGASTHENESE AND THE ARTHASASTRA The account of 12 Megasthenes, properly evaluated, enables us to date the highest development of one particular and hitherto unrecognized form of production in India. Seven distinct classes (meros, Diodorus and Strabo; genea, Arrian) of people are reported which amount to castes, seeing that custom and law forbid intermarriage, change of profession, or transition between groups. That Megasthenes actually said this is not to be doubted, from the accord among the quotations (Diodorus 2.40-41=M 38-41; Strabo 15.1.39-48=M 83-86; Arrian 10-12=M 214-8). The confusion arises from modern scholars' attempt to reconcile the seven classes with the traditional four Indian castes. The classes are: 1) brahmins and ascetics; 2) husbandmen; 3) herdsmen-hunters; 4) artizans and retail merchants; 5) fighting men; 6) overseers who report to the king or the magistrates of free cities, 7) the great councillors who determine policy, officer the armed forces, and regulate all affairs of state. It is denied by modern scholars that these seven major castes could have existed, because the sudra is absent, the husbandman and merchant should both be members of the vaisya caste, and there could not possibly have been enough overseers nor courtcillors to form two separate castes by themselves. Of course, the Arthasastra sticks to the theoretical four castes in its preamble, and is simultaneously condemned as a later document, making no mention of Pataliputra, or Candragupta, or any other king, city, or empire; nor showing acquaintance with imperial affairs such as must have concerned the ministers of Candragupta Maurya. 1. Text fragments collected by E.A. Schwanbeck, Bonn 1846 ; for Diodorus Siculus and Arrian text and Latin translation by Carl Muller, Paris (Firmin Didot) 1878. The English translation by J. W. McCrindle, Calcutta 1926 (reprinted from the Indian Antiquary 1876-77) is cited as M with page number. The two Kautaliya-Studien of E. Breloer (Bonn 192-6, 1928) are useless.

Page #19
198 D. D. KOSAMBI There is no doubt about the first class, that of brahmins and sramanas who have been included in one category. This does not detract from the value of Megasthenes's testimony, for the two are grouped together in that order, by one compound word, in Asoka's edicts. They clearly enjoyed equally high position as holy men, the classification into a single caste being permissible because the sramanas were celibate in any case. We are also told that becoming a 'philosopher' was the only change of caste allowed to any of the others (M 86, 218). Caste 5 of warriors, second greatest in numerical strength, is clearly the ksatriyas. They pass the time of peace in idleness and sport, so well are they paid by the king; all their equipment comes from the royal arsenal, the elephants and horses from the king's stables, the care of these in war or peace being the task of others. The implication of this has been missed, for it means that with negligible exceptions, all able-bodied male ksatriyas had been organized into the king's paid standing army. For support, we find in the Arthasastra that the hereditary army, entirely of highly trained ksatriyas, is the best (A 6.1); the author himself says in emphatic contradiction of preceding theorists, that a compact standing army is better than disbanded forces (A 7.9); in 9.2 we find auxiliaries mentioned, but the order of importance is always the hereditary standing army maula, temporarily hired bhrtaka recruits (both noted together by Megasthenes as the 5th caste), guild-soldiers sreni (perhaps remnants of tribal or city fyrds), troops of an ally, and wild tribes. Megasthenes's third caste, herdsmen, settle neither in towns nor cities, but live as pastoral nomads and hunters in tents (cf. A 2.29); this is interesting as a survival of the traditional ancient mode of Aryan life, inevitably degraded in the presence of civilization. The main difficulty arises from equating the second caste, husbandmen, to the vaisyas. It is true that the Greek georgos, Latin agricola does mean husbandman, exactly as did the original, traditional vaisya: 'settler'. But what are the characteristics of this class of cultivators according to the Greek envoy? First, that they form by far the greater part of the population, producing the surplus upon which society rests; for none of the other six classes produce any food at all, except the herdsman in some trifling measure. They don't own the land, for a 25% tribute (M84) plus rent (M 39-40) is to be paid to the king or the free cities. They never enter the cities for any purpose (M 84), but live all the time in the villages with their families (M 39). They are not furnished with arms (M 216) and are 'exempted' from military duties (M 39, 83, 216), so that they may be seen ploughing calmly within sight of contending armies, which never harm them (M 84, 216) nor devastate the land (M 39). This idyllic life, however, meant that the cultivators were totally disarmed; for even the wild beasts and birds that damage their crops have to be killed by the third caste, (M 40, 84 "who alone are allowed to hunt, and to keep cattle"), not by the second. Moreover, they have no property at all, for the land does not belong to them nor have they any cattle (which

Page #20
ANCIENT KOSALA AND MAGADHA 199 belong to the "herdsmen"), and they cannot follow any other profession; they can not enter the cities for any purpose, whereas the vaisya could and did, being a leading citizen in the free city. So we have in reality a description of the sudra. Once again this is confirmed by the Arthasastra (42.1): The king should establish new settlements or repopulate deserted ones by inviting foreign immigrants or deporting settlers from his own overpopulated centers. Each village is to be made up of between a hundred and five hundred families of sudra cultivators (sudra-karsaka). Administrative centers are to be set up for groups of 10, 200, 400, 800 villages. A few priests actually serving in the locality are to be granted tax-free lands. Superintendents, functionaries, accountants, doctors, veterinaries, messengers and such officials may be endowed with land which the individual holds for life, cannot alienate by sale or mortgage; however, officials in general have a regular stipend, as do the watchmen and specially hired labourers (42.24). If the holder failed to cultivate it properly, the land would be reassigned by the state to someone else except perhaps when the holder had just brought waste land or wilderness under the plough. No ascetic other than one who has adopted a forest retreat for his own individual salvation is to be admitted to the villages; this means that no one proselyting for a monastic order could gain entrance. No place of communal entertainment, nor resting place may be built in the village; professional actors, dancers, singers, musicians, raconteurs, bards are also completely excluded. "For from the helplessness of the villages and exclusive preoccupation of the men with land comes wealth" of all sorts. This wealth is not for the cultivator, who can hardly use it, but for the state treasury. We are very far here from the idyllic life depicted by so many who wrote of Indian villages from modern observations made at a safe distance and from the property-owner's point of view. The 'idiocy of village life' was deliberately intensified as a method of exploitation. There are two reasons for misinterpretation of the sudrakarsaka-georgos as a vaisya. One is the projection of a supposedly immutable village system from 18th and 19th century India into antiquity. The other is that the Greeks were deeply impressed by the fabulous nature of a country which produced incredible beasts like the elephant, plants like cotton and sugarcane, sweet tasting rock crystals (sugar!), deciduous trees that never shed their leaves, tremendous rivers, and an extraordinary soil yielding two bumper harvests every year. On the same level (from the Greek point of view) as the fantastic, mythical, gold-digging ants was the attested truthful and law-abiding nature of the Indian in the absence of written law-codes or written contracts. Arrian 12 (M 217) says explicitly, "but indeed no Indian is accused of lying." This could never be said of the Graeculus esuriens. Most striking of all was the fact that the Indian economy flourished, so incomprehensibly to the Greek mind, without chattel slavery. "Of several remarkable customs existing among the Indians there is one prescribed by their ancient philosophers which one may

Page #21
200 D. D. KOSAMBI. regard as truly admirable. For the law ordains that no one among them shall, under any circumstances, be a slave, but that, enjoying freedom, they shall respect the equal right to it which all possess; for those (they thought) who have learned neither to domineer over nor to cringe to others will attain the life best adapted for all vicissitudes of lot. For it is but fair and reasonable to institute laws which bind all equally, but allow property to be unevenly distributed" (Diodorus Siculus ii. 39 end, M 38; my italics). The trouble here is not with Megasthenes, but with the sentimental and idealizing embroidery of Diodorus, proved by the utopian island which he reports as visited by one Iamboulos (ii. 55-60), at the end of the very same book; which may be a traveller's tale based upon the name of the island of Socotra, derived from sukha-dhara, the land of bliss. There is no such concept of equality in what survives of Megasthenes nor in any Indian source. The case is made much worse by misleading translation, which I have italicized. The last sentence in C. Muller's Latin conveys exactly the opposite sense, namely that it would be the act of a simpleton to promulgate equal laws for all with unequal status and opportunity (between slave and free). The Greek euethes and exousias anomalous are surely better represented by 'stultum' and 'inaequalitatem facultatum', than by the English phrases above. The Greeks could not recognize slavery in India. "All the Indians are free, and not one of them a slave. The Lakedemonians and Indians here so far agree. The Lakedemonians, however, held the Helots as slaves and these Helots do servile labour; but the Indians do not even use aliens as slaves, and much less a countryman of their own" (Arrian, M 210-213). This is not to be doubted. The Greek doulos originally meant a slave by birth, to be distinguished from andrapodon13 the captive (taken in war or kidnapped) sold into slavery. Such war-prisoners were a legitimate and most valuable part of the booty in Greek and Roman campaigns. If the Indians had any such type of slavery, Megasthenes would certainly have known of it; he was the ambassador of Seleukos, who had just lost a war against Candragupta. We have already examined the tradition in India as regards slavery. The Arthasastra forbids the sale of any one, even a sudra, who lives as a free man and does not belong to some (obviously unimportant) recognized category of slaves. Inasmuch as the Greeks saw a parallel between Sparta and India, we have only to consider what Lakedemonian helotage had been originally, before approximating to slavery in a powerful slave-holding and slave-trading environment. "The Lakedemonians held the Helots as state-slaves in a way, having assigned to them certain settlements to live in and special services to perform" (Strabo, 8. 5.4; trans. Loeb Classics, H. L. Jones, 4.135). The the ans an ch less a cohe Greek dorthe captive to ? andre steek of their cians 13 The Arthasastra has a corresponding word, dripada=biped, grouped with the catuspada (Greek tetrapodon) as two-legged and four-legged animals for sale (-4 3.15). There is no doubt that some human beings were sold, as the context shows this to refer to humans, not poultry. But the main consideration is that they were unimportant as a source of labour, and A 3.13 shows mainly contract labourers whose rights are carefully protected, though they are dasas.

Page #22
ANCIENT KOSALA AND MAGADHA 201 Mauryan sudra is a helot, as had been his predecessor and equivalent, the dasa in vedic tribal economy; dasa is indiscriminately equated to sudra, servant, and slave. The sudra caste made slavery unnecessary, and yet enabled the forested Gangetic plain to be brought under cultivation. Thus Megasthenes and the Arthasastra corroborate and supplement each other in a remarkable, unexpected way. The latter work has no feature to give it scale; it can, on a first superficial reading, easily be mistaken for a manual of administration for some princeling who rules no more than a county (janapada). The extraordinary situation revealed to us by the Arthasastra and Megasthenes is that crown property is overwhelmingly the main basis of production, and its profitable administration the chief preoccupation of the state, which replaces private enterprise to an extent never seen before or after in India. Even prostitution, like mining, and wines, is directly under the exchequer with a special ministry (4 2.27). There are no feudal landlords worth notice, nor any nagaratres the like Anathapindika who is richer than the princes whose lands are crossed by his caravans. If Roman and Greek historical writers deal primarily with events of national and international importance, it is only because the exploitation of occupied territory was the natural prerogative of certain classes, not taken over directly by their instrument, the state. With Kautalya's everything is state-regulated to the last detail: cities and markets; road construction and transport; rivers, canals, irrigation; care and convoying of foreigners; land measurement, division, taxation; registration of births and deaths (even now unsatisfactory in India!); levy of sales taxes and tolls of every sort; public supervised sales of manufactured articles, as well as supervision of artizans. (Strabo 15.1.50-52, M 86-88). Both sources agree here, the Indian giving detail, the Greek the immense scale upon which the operations were conducted. Given the caste structure of Indian society, there is absolutely no reason to doubt that the multitude of lower state functionaries would develop into a separate caste, as also the higher ministers, at that period. This is what Megasthenes reports, though the sastras don't trouble to recognize the transient phenomenon, which disappeared with the over-centralized, government-entrepreneur empire, just as the Arthasastra was forgotten except 14 Dionysus, in Megasthenes, does not indicate the Indian Siva as so many have assumed without consideration, but obviously the conqueror Indra of the Rgveda, who is so often invited to fill himself with Soma. The three-peaked mountain (M 162) seems to be a development of Trikakud in the Trisiras Tvastra saga. Spatembas would be, in the most reasonable Sanskrit equivalence, Asvatthaman, one of the 7 immortals in later legend though his role in the Mahabharata is not of first rank. The Greek envoy, therefore, gives a better report of Indian tradition than his translators. 15 Ganapati Sastri, in his learned preface, insists that the gotra is Kautalya. from Kutala, the change to Kautilya being indicative of the reputation for Machiavellian crookedness ascribed to the author. However, the gotra Kutala, Kautalya, Canaka or Canakya is not found in any surviving list (unless one sees some connection with the Canakaraj-niti in the Tibetan Canon), whereas Kautilya survives in several such lists, and families actually belonging to that gotra exist, as for example among the Madhyandina brahmins of Maharastra. Nevertheless Kautalya is the uniform reading of the 12th century Patan MSS in Mani Jinavijayaji's printed but unpublished edition. 11

Page #23
202 D.D. KOSAMBII . by name. The existence of free cities mentioned by Megasthenes appears to be supported by some Mauryan punch-marked coins that bear the emperor's personal mark, with homo-signs or some group-mark in place of the usual symbols of sovereignty. It must be remembered that such cities would have a tribal origin, and a guild structure. The other fact to be kept in mind is that both the Arthasastra and what remains of Megasthenes are concerned with the major features, not with every single minor form of production ; India is a country of long formal survivals even after substance and basis are obselete. The essential is that the city, free or not, is an administrative center, no longer a center of tribal assembly and warfare. There is no class structure between the primary producer and the coercive mechanism (the State); the mechanism of violence is controlled by the center, namely the king and his bureaucracy. This system must have reached full bloom with the Nandas (about 400 B.C.), whose proverbial wealth and prosperity are attested by the great variety and accurate minting of their silver punch-marked coins. Its main effect was upon the Gangetic and Indus valleys, and the regions about a few cities elsewhere, the rest Asoka's empire being undeveloped, with great variety of survivals. The inevitable expansion carried with it the seeds of decay. In the absence of slavery, with the sudra deported from the cities, and tight control of manufactured goods and its prices from raw material to the finished article, civic production would certainly not suffice for the whole countryside from Afghanistan to Bengal and the Himalayas to Mysore. Nevertheless, more land is steadily being brought under cultivation by means of the tools made in the cities. The immense distances and poor transport would also make central administration most cumbrous, accounting progressively more complicated, official peculations increasingly harder (A 2.7-9) to check. Certainly, the Mauryan silver coinage as a whole, particularly after Candragupta, shows far greater proportion of copper with much rougher weight adjustment than before-symptoms of tremendously increased demand for currency, logical corollary of the opening up of the whole peninsula for the first time. The minute reverse marks of the pre-Mauryan coinage also disappear, to be replaced by a single issue-mark, presumably at the mint, which again confirms rigid state control, to the exclusion of former important merchant guilds. A natural step would be that the artizan, or at least his technique, first moved to the countryside, thus bringing about the full development of self-sufficient rural units. The state machinery would have to be dispersed in consequence. Increasing the number of stipendiary superintendents would be impossible, which means the necessity of creating new classes of owners, new intermediate relations of production between state and producer. That is, the ksatriya has also to move to the countryside with small forces scattered over village territory. The brahmin is as important in making the people submissive; he has already a foothold as priest and medicine-man, with tax

Page #24
ANCIENT KOSALA AND MAGADHA -203 free lands; the rusticity of the Grhya-sutras is now turned to good account, becoming incurable, if indeed it does not first come into existence at this stage. That my argument is not purely a priori may be seen by comparison of the king's duties in the forgotten (because the productive forms were obsolete) Arthasastra with the later but much more pretentious and sanctified Manusmrti (Ms. 7.111-143; 7.195-196; et. al.). Taxes have fallen below the Arthasastra level because they are mainly levied upon property owners; the sudra though free still has no property, nor can release by any master free him from servitude (Ms. 8.414) to the three Aryan castes. The main revenue of the Arthasastra was the tribute on grain from state lands sita, which has disappeared altogether; the minor taxable forms called rastra by the Arthasastra are now synonymous with 'country'. This means that state production no longer exists. Nothing is said about transport, irrigation, or supervision. The villages are grouped in much smaller units, with a chief over every village and unit whose compensation is ownership of a certain amount of land. Local garrisons appear (Ms. 7.114-15) whereas the Arthasastra wanted only frontiers to be strongly held. The artizan and the sadra labourer of the Manusmrti give the state one day's free labour per month, the corvee in lieu of taxes (Ms. 7.138). Worst of all, the immunity of land and villager that struck Megasthenes has disappeared. The peasant is no longer allowed to witness the ruin of empires while concentrating upon some miserable patch of land, for the royal invader is enjoined to ravage the countryside thoroughly whenever the enemy cannot be directly attacked (Ms. 7. 195-6). At this stage, the Mauryan central administration is unnecessary. The Gupta capital Ujjain was never the world's greatest city, as Patna had been in its prime, though the center of as large and even more prosperous an empire. Patna itself dwindled to a village before the time of Hiuen Tsang. While total production increased, commodity production went down in density if not volume. This was the difference between the land seen by the Greeks and the apparently changeless countryside visited by later foreign travellers. The victory of Manusmrti brahminism means also the absence of common law, each caste, guild, profession and locality being entitled to its own customary legal traditions (Ms. 8.41). The Arthasastra recognized the need for a legal structure which transcended these barriers, though caste-privilege was, to a certain extent, allowed in its jus gentium, being a primitive form of class privilage. There are two important points of contrast. The Manusmrti administrator is appointed by the king but not paid a salary; he lives off the country. The village chief has as his perquisites the food, drink, firewood which would be supplied to the king (if in residence). The head of ten villages enjoys the income of a family holding (kulam, Ms. 7.119); the chief of twenty gets five times as much, that of a hundred the revenues of an entire village, and of a thousand the revenues of a whole town. This is clearly the beginning of feudalism, and it may be worth noting that G. Buhler in his translation of the

Page #25
204. D. D. KOSAMBI smoti (Sacred Books of The East, xxv, preface, pp. cvi f.) believed the work to have been completed between the 2nd century B.C. and the 2nd century A.D. Secondly, the merchant is to be taxed (Ms. 7.127) with due regard to the prices of purchase and sale, depreciation and expenses of the trip, length of the road to the market and expenses for protection against thieves etc. on the way. This shows a consideration entirely foreign to the Arthasastra, where the merchant was in fact a competitor of the royal monopolies, and treated as a natural criminal of sorts. Cash transactions are so low that royal menials are paid only 1 to 6 copper panas a day, a drona of grain monthly, and clothes every six months (Ms. 7.126). According to A 2.24, the sitadhyaksa (of a janapada) is to cultivate all the best (most-ploughed) land under his own direct supervision, with free (karmakara) labour or forced labour including the royal dasas and the prisoners who work out their fine. What is left over, however, allows private enterprise to enter; the half share cultivators (ardhasitikas) may be allowed to work such surplus land, or the quarter or fifth sharers, the difference between the two being that the latter have nothing except their own bodily effort to put into the land. It follows that even in the sita those who owned oxen and implements of their own had a certain interest. The essential is that this did not develop into permanent tenure or private property during the time of the Arthasastra; when it did by the period of the Manusmrti, the sita would cease to exist, becoming indistinguishable from the rastra. 8. TERMINATION OF THE ARTHASASTRA SYSTEM The foregoing section expresses a definite opinion, that the essential features of the Arthasastra do characterize the government described by Megasthenes ; while it does not follow that the whole of the document must be authentic, the work is not entirely theoretical, it has a sound basis in reality, and the bulk thereof must belong to the time of Candragupta Maurya, though reporting upon older theory and tradition. That the book may contain later additions or revision is not denied; still less, that the author was a brahmin pedant. But the book is unusually consistent as compared to the Mahabharata or the Manusmoti. The brahmin and ksatriya are allowed legitimate heirs (savarna) from women of one caste lower each. Differences between the four-caste theory (A 1.3) and practice appear with the vaidehaka, synonymous with trader throughout the book, but in A 3.7 the offspring of a ksatriya woman by a vaisya father; in A1.6, a Vaideha king Karala is held up as bad example. This corresponds to the historical development of a tribe into a guild and then a caste. The mixture of castes is frowned upon, as in all brahminical works, but it is really the fault of the king, who won't go to heaven if he tolerates it. For other transgressions, the Arthasastra relies upon a complete system of standardized cash fines, the separate items filling nine columns in the index to Shamasastry's translation.

Page #26
ANCIENT KOSALA AND MAGADHA 205 The Arthasastra is not to be compared to Bismarck's memoirs, where a chancellor tells what actually happened, while trying to justify his own policy. We have no support here for the Mudraraksasa Grey Eminence tradition, except possibly in the final colophon-probably a later insertion. The book is not meant like that of Castiglione for a gentleman-courtier's training in superficials. Machiavelli's Principe draws different conclusions (in comparably rugged prose) from the career of princes like Cesare Borgia, presumably in the hope that one such would some day arise to unite Italy. If the 14th section of the Arthasastra is genuine, with its extraordinarily futile magic formulae, the treatise cannot possibly be meant for the vulgar gaze. The select few who would read would necessarily be other ministers like Canakya, who are here made acquainted with traditional principles and theories, (not history), brought up to date and fitted to the reality as the author saw it, with considerable instruction as to the working of administrative departments. Though the principles and methods apply to states of any size, the declared purpose is aggrandizement, and the system could not long remain static, for growth or collapse are the only two effective alternatives it could admit as proved by history. The janapada, which most of us mistake for the whole country at a first reading of the Arthasastra, is really the natural unit of administration. Originally, it means the territory of a tribe, with some city or group of small towns as headquarters, hence the paurajanapada of 2:1 contrasted to the new settlements. The Gangetic janapadas were widely separated by dense forests, which yielded to the highly profitable sita cultivation of the Arthasastra king, as distinct from the rastra that had been settled by tribal or private enterprise. The first kings naturally develop from the tribal chiefs of a janapada ; preservation of the janapada unit would enable local custom to be allowed for, as in A3.7, end. This would also account for the 'free cities' of the Greek observer, as the paurajanapadas, the newer settlements being ruthlessly taxed and controlled. The imperial superstructure covers, without destroying or standardization, an immense variety including the wild tribes (atavikas) which had not yet been Aryanized ; on the other hand, though the correct definition of sita and rastra appear in A 2. 15, the differences tended to be obscured, as seen in A 2.6, presumably by the royal power encroaching upon the prerogatives of older settlements, as well as penetration of state settlements or wilderness by individual pioneers. Nevertheless it is only an indigent king (A 2.1) who eats up the old settlers! No grasp of the Arthasastra is possible without understanding the fundamental difference between the few older free cities and the tremendous royal settlements peopled mostly by unprivileged sudras. The pragmatic note is heavily stressed throughout. Practical considerations are ever the prime ones : artha eva pradhanah says Kautalya in his own

Page #27
206 D. D. KOSAMBI. . name in A1:3 ; spiritual and sensual considerations follow from the practical. Similarly, in A 8:1 he again contradicts his preceptors : the army has its roots in the treasury; without a treasury, the army goes over to another, or kills its own master ; sarvabhiyogakaras ca koso dharma-kama-hetuh, the treasury is the first cause of religion and enjoyment. The writer has also grasped what few Indians after him realized : that real wealth depends upon command of production, particularly of raw materials, cereals, and heavy industry; the first is stated explicitly in A1.4. The last is made still clearer in A 212: akara-prabhavah kosah, the treasury depends upon mining, the army on the treasury, and he who has both wins the rich earth. The mine is the womb of war materials : khanih sangramopakarananam yonih. No brahmin, however well-read, could merely dream all this up for himself, for the true brahmin mind runs smoothly in quite different grooves, as for example the ritual exorcisms of A 4.3 against national calamities. The absolute control of metals would be easy for pre-Mauryan Magadha, which covered the natural outlets for India's richest deposits of iron, copper, and other metals in Bihar. By the mining industry, the author does not mean only precious metals and gems, but emphatically all sorts of tool-making metals and alloys, loha=kalayasa, tamra, vitta, kamsya, sisa, trapu, vaikyntaka, arakuta (A 2. 17). In A 2.12, minute directions are given for the state director to mine and smelt these ores, and to regulate their trade. In A 2.21, the common metals and grain are among forbidden commodities, as are arms, military gear of all sorts, gems, and other royal monopolies. It then seems a contradiction to read in A 2.22 the toll rates upon grain and metals, but this is explained by the line jatibhumisu ca panyanam avikrayah. No goods is to be sold by the private trader where it originates. In other words, the merchant had to add value to the commodity by transport to the ultimate market; incidentally paying tolls on the way. At the end, he would find not only his profit regulated, but also his wares sold by a state-organized sale such as Strabo reports from Megasthenes (M 87). This has been laughed off as ridiculous and impossible, but the Arthasastra does confirm it once again. Not only are sales of local goods state monopoly (A2.16) but A 4.2 directs the chief of the.city market to separate new from second-hand wares (of which the ownership must be proved before sale) and when there is a glut, to gather all the merchandise in one place for a regulated sale, during which time no other goods may be sold. Transport, particularly land transport, was difficult, and only here was private enterprise encouraged in any way. : Could all this administrative routine have been the fantasy of some idle theorist? What of technical terms like aksapatala for exchequer ? Each state granary and storehouse must have a standard raingauge (A 2.5) for land is classified and its yield estimated according to rainfall (A 2. 24). This is the only instance I know of so practical a step being mentioned in any Indian

Page #28
ANCIENT KOSALA AND MAGADHA 207 text; Plato's Republic could do nicely with religious or philosophical axioms, but such technical details imply working methods in actual use. When the treasury is in great difficulties (A 5.2) and a capital levy has to be made, the merchants in precious articles should pay as much as 50 per cent, cloth and lower metals 40 per cent, grain traders and wagon-train organizers 30 per cent. and so on. In later Indian history, this would simply be done by force, as often as it pleased the absolute monarch. But Kautalya warns that this levy is not to be made twice, and that it should be on a voluntary basis as it were, by the chief collector of revenues begging from the people. He suggests seeding the populace with pseudo-volunteers who come forward enthusiastically to contribute apparently large amounts. Yet there is no question of civil rights, fair play, or from each according to his means-let the goldsmiths have all their property confiscated, the richest travelling merchants be quietly murdered as if by a robber or enemy, and the loot piled into the state treasury. It would be inconceivable for a brahmin theorist to recommend as does the Arthasastra, confiscation of temple treasures (by the corresponding adhyaksa, naturally !), setting up of new cults, faking of sudden miracles, all to collect money from the credulous for the king, not for the brahmin. Here we have support from Patanjali who reports : Mauryair hiranyarthibhir arcah prakalpitah (on Pan. 5. 3. 99), that the Mauryans had established cults for the sake of money. That the Mauryan treasury had actually been in difficulties again and again is shown by progressive debasement of the coinage, which in the later Taxila hoard (dated to about the reign of Asoka by a mint-condition coin of Diodotos) contains from 60 to 75 per cent copper, as against the 25 per cent recommended by the Arthasastra (A2.12). Even when Kautalya speaks against Bharadvaja (A 5.6) to the effect that the minister shouldn't make himself king if the ruler dies suddenly and a good chance offers, he is being practical on moral grounds. Bharadvaja's advice was followed when the Sunga army commander Pusyamitra ended the Mauryan dynasty; again, when nasla thain: the Kanvayana brahmin minister Vasudeva succeeded the last Sunga. It may be a mere coincidence that one or the Sunga and Kanvayana are both Bharadvaja gotras. After all, what is the main purpose of this treatise on political economy? The conquest and guarding of the whole earth16 by one king. The work begins with those words, thrthivyam labhe palane ca. From the seventh book onwards, this purpose becomes the main aim, and the thirteenth deals with the most difficult task of straightforward aggression, the reduction of fortified places and cities. But even though no treachery down to murder and poisoning is too foul to get rid of the opposing king, the countryside is to be preserved from all damage, as Megasthenes noted to the astonishment and joy of writers 18. The world outside India doesn't count for imperial purposes : desa prthivi, tasyam himaval samudrantaram udicinam yojana-sahasra-pariminam tiryak cakravarti-ksetram (A 9.1).

Page #29
208 D. D. KOSAMBI like Diodorus ; cutting off forage and perhaps damaging some crops is a desperate measure for the most extreme cases (A 13.4). It is towards building up the resources for conquest that the whole cumbrous mechanism of the state is directed unscrupulously. An accurate permanent census, Domesday Book account, and register is to be kept at all times of everything and everyone as was noted by Megasthenes : boundaries, wealth taxed and untaxed; males, females, children, servants in each household and their work, if any; the purpose and duration of each voyage undertaken by anyone (A 2.35,36) both in city and countryside. To this end, there are not only registrars (gopa) for every ten to forty houses in the city, five to ten villages in the country, but a system of passes at city gates and a whole army of spies in disguise. These last are distinct from the informers (sucaka) who discover embezzlement (A 2.8). Spying is the only check upon officials, says Kautalya (A 2:9). Apart from embezzlement, the official who neglects any chance of making revenue for the king, by inattention to the reports of underlings and spies, is to be fined, though collecting more than the amount due is frowned upon. Whoever diminishes the revenue eats up the king's wealth (A 2.9), which he must make good, with a fine if his intentions were evil ; whoever doubles (or more) the king's revenue eats up the land, but is let off with a warning if the whole amount be deposited into the treasury. There is no question of refund, in spite of book-keeping and regular receipts. Similarly for all the special taxes on merchants based on the theory that every one of them cheats on all possible occasions in spite of regulations, checking of balances and measures, and other state control including spies with every caravan (A 2.21); but the taxes go to the king. Even in selling a plot of land, or a holding-which may be bought by relatives, neighbours, creditors in that order, and only in their absence by anyone else (A 3.9), forty other people are invited to be present and to bid up the price; the excess over the predetermined price goes to the treasury along with the sales tax ! This system would become intolerably oppressive except when in a state of expansion, and even then expansion into paying territory. A complete scale of payment for every conceivable state servant is set down in A 5.3. The highest is 48,000 panas (per year) for the great ministers, royal priest, dowager queen-mother, chief queen, and crown-prince. It descends to 60 for the lowest categories, of stablemen, camp-servants, and drudges for heavy unskilled labour, visti.17 17 There is the usual ambiguity, caused here by the ending of the long compound :vistibandhaka Shamasastry takes this as one unit, the mean the press-gang foreman, but then the pay of 60 panas a year would be disproportionately low, in comparison to the rest. Ganapati Sastri's interpretation of visti and bandhaka as two separate categories is proved by two definitions of visti : A 10. 4 gives the camp-visti, whose work is to clear the camp, roads, wells, fords, etc., carry fodder, take the oifensive and defensive weapons to and from the battlefield, and carry away the wounded work which would be difficult to entrust to unpaid forced labour, which is Sham usastry's idea of visti. In A 2.15, we have the visti for the great royal storehouses, consisting of porters, the weightsmen, and so on. Here the category ends-disa-karmakaravargas ca vistih, which is generally taken to include both slaves (bondsmen) and ordinary labourers. It follows that even slaves were paid for such heavy labour; presumably, the money would count, as in A 3.13, towards their ransom and eventual freedom. si

Page #30
ANCIENT KOSALA AND MAGADHA 209 Dunn (loc. cit. p. 3, p. 54) writing in 1936 when wages in Singhbhum had fallen to a miserable 7 annas a day still felt that the obvious and very thorough working of the same mineral deposits in very ancient times could only have been by "an inexhaustible supply of cheap labour-slave labour." Unfortunately for this piece of imperial British yearning, the saila-khanaka was paid from 500 to 1000 panas ; the scarcity of metals, great demand, state monopoly, possible payment to the head of a whole family of workers, might be the explanation ; but the Arthasastra miner was free. As for the level of prices, Megasthenes reported that the soldiers and their families lived very well ; the Arthasastra pay for a soldier is only 500 panas, though more for veterans, with graded cash bounties on the field of battle (A 10.3). The spy in disguise gets 1000, the ordinary spy 500, intermediary 250 ; spying must have been very well worth the while. There is even a system of pensions for the dependents of those who die in service, and of gratuities. It is incredible that this scheme of cash payments, unparalleled in a country where feudal tenure is not yet forgotten, should be purely imaginary or designed for a mere janapada king. A characteristic suggestion at the end of A 5.3 is that spies disguised as merchants and sutlers should sell to the soldiers in camp at double rates, thus clearing off the king's goods as well as recovering the money paid out in salaries. The best proof of reality is the advice that villages should not be given away, that only a poor king should give land, raw materials, or cattle rather than cash. Contrast Pasenadi's giving whole villages18 to his priests ; the Arthasastra gift is merely the trifling brahmadeya lands (A3. 10) and groves (A2.1; M 99) though the author is himself a brahmin. The ruler is to spend only a quarter of his revenues for all such cash payments; when we recall the statement of Megasthenes that a single camp of Candragupta held 400,000 men (M 68), the total revenue in cash must have been enormous though the budget is nowhere given. Yet the Mauryans changed none of the pre-existing local forms, as in the rastra; a conquered king (and therefore his officials) are to be retained in office (A 7.16). Thus the chief benefit of conquest would come not from direct tribute but rather from promotion of trade, and settling of new lands; the latter is given as one of the main objects of policy in A 7. 11-12, with the advice, strange when coming from a brahmin, that the land predominantly settled by men of the lowest caste is best. That the southern trade routes were richer than the Himalayan in all goods except wool, hides, and horses, is explicitly noted (A 7.12), against the older tradition. But how far could this go when the territory of a thousand yojanas from the Himalayas to the seas, the cakravarti-ksetram (A 9.1) had actually been 18 This may tally with the comparatively poor Kosalan coinage, and with the Lalita Vistara's otherwise incomprehensible remark that the Kosala royal family was na caparimita-ratna-nidhi-samutthitam. In any case the Magadhans rapidly became by far the richer state, by holding the metals, the trade routes, and the river. 12

Page #31
210 D. D. KOSAMBI brought under a single ruler ? The new lands did not compare in fertility to the alluvial plains of the U. P. and the Punjab ; and settling them would not only be difficult, but far less paying in proportion to the expenses of administration. The strong rule of one man, which had seemed so desirable with the ancient warring principalities and tribal areas of three centuries earlier would now be oppressive beyond measure to the merchant, who is treated as a super-criminal in the Arthasastra, with virtually no civil rights ; yet the merchant would naturally tend to gather in a progressively greater share of the currency. The Arthasastra is remarkably deficient in one respect, namely detailed knowledge of ascetic sects--though spies are very frequently disguised as holy men, a custom observed down to the Gosains of the Peshwas. The solitary mention of Buddhists is in A 3.20 : Sakya, Ajivakas, and such other monks as recruit from the lowest castes may not be fed at a feast to the gods or the manes, on pain of a 100-pana fine. The passage is taken by J. J. Meyer and others as a commentary on the Yajnavalkya smoti 2. 235-7, (sudrapravrajitanam ca daive pitrye ca bhojakah. . . satadandabhak) which gives no details of sects ; nevertheless, the smoti as we now have it is later than the Arthasastra, for the pana there (1.365) is of copper, as against the silver pana of Kautilya, which dates it to a much later period of debasement. Now the term sakya for Buddhists is early, for even the Harivanga qualifies it: sakya-buddhopajivinah. Ajivaka denotes a follower of Makkhali Gosala, who resembled the Jains in their nakedness, and the Buddhists in many of their opinions, being hated by both; they were a firmly rooted Magadhan sect, though they too spread to the far south and survived very late.19 I suggest that this passage dates the Arthasastra (in essence) to the later years of Candragupta's reign. A pre-Mauryan date is highly unlikely, because of the reference to Saurastra, debased coins, cults established for gathering money, etc. Asoka's conversion to Buddhism, respect for all sects, gift of three caves in the Barabar hills to the Ajivakas (followed by three more by Dasaratha in the Nagarjuni hills) exclude the possibility of so late a date.20 In the Arthasastra the Jains are not mentioned 19 In this connection, see A.L. Basham's authoritative and comprehensive work : "History and doctrines of the Ajivikas", London (Luzac) 1951. 20 The Arthasastra is supposed to be dated to the end of the 3rd century BC or later by its reference to Chinese silk : A 2.11, cinapatta, cinabhumija. The name China extends to the whole country from the principality of Ts'in after the final victory of Ts'in over Chu and the coronation of Shih Huang Ti as first emperor about 223 BC. This argument, however, is completely irrelevant for cina in A 2.11 denotes the source from which India actually derived the silk at that time ; nothing is said about 'the whole of China'. Any modern history of China (cf. W. Eberhard A History of China, trans, E. W. Dickes, London 1950; chap. iii-v) will show the great importance of Ts'in from the 8th century onwards. Not only did it straddle the trade route to the west and India, but its rise to supreme power is due to its having ended feudalism before the rest of China, which in turn is duc precisely to the local development of a strong merchant class. A merchant became chief minister under the first emperor, scandalising the feudal nobility. For us, the point is proved by A 2.11 itself, where occurs the only other mention of China in the whole book : cinasi, a fur imported from Balkh ; this shows both the trade route and the paramount importance of Ts'in in trade.

Page #32
ANCIENT KOSALA AND MAGADHA 211 (though they occur as Nigankhas in R 13) simply because their shift to the south must have taken place before this, (presumably in the face of local Magadhan competition) not later than the great migration under Bhadrabahu; incidentally, this implies that the south had been penetrated by Magadhan traders before the armies marched there, so that the classical routine of imperialism : traders, missionaries, armies, is well established even in so long a past. Then there is the Jain tradition of Candragupta himself (Tiloyapannatti 71) having been converted, to die in the faith. So, this passage is impossible for the later Mauryans, say after Bindusara; we may recall that the Asokan district officer rajjuka is not mentioned in the Arthasastra, which has, however, a rajju tax. It is even further out of place for the Kusanas, the Satavahanas, and the Guptas, all of whom preserved and increased the great Buddhist foundations. Even the Sungas, supposedly hostile to the religion, dedicated structures at Sanci. On the other hand, for a real persecutor like Sasanka, such a ban would be much too mild, for the advice would then be to exterminate the heretics. Kautalya knows nothing of monasteries like Nalanda, saying merely that (A 3.16) ascetics should live together without mutual contact, in places set apart ; earlier comers are to make room for the later, on pain of ejection. Any ascetic is punished in full for serious crimes, fined for lesser offences; if unpropertied he may pay the fine by specially painful prayers or fasts for the king, which implies merely that he was not to work them off as a prisoner or slave. Asoka's conversion is not only a cause but even more a symptom of decline of the Arthasastra21 system, which was inadequate to maintain a stable empire. After conversion of the emperor, it would no longer be possible to keep the monks out of the villages, and indeed Asoka found a far greater use for the monks than spying: they were an integral part of his new system of administration, by conversion of the people to gentler ways, by new relations between king and subject which must correspond quite well to new relations of production. The final move towards feudalism of a priest-ridden type is obvious only from the Manusmrti, but an immediate change is also to be discerned in Asoka's edicts. Though Asoka had suppressed two revolts at Taxila as viceroy (Divyavadana 19 ), the bloody Kalinga campaign was the final turning point (R. 13). Thereafter the army was obsolete; he is proud of using it for spectacles and religious processions only, not for war (R. 4). There should be no further killing of animals (R. 1), though criminals are still to be executed --with a special grace of three days after the sentence (P. 4). The king makes a new departure in administrative routine by travelling constantly, and visit 21 Asoka suppressed the samaja by edict (R. 1). This appears in A1.21 as some sort of saturnalia ; in A 2.21, it meant freedom to drink for three days. In A 5.2, it provides occasion for some royal spy in disguise as a merchant to allow himself to be 'robbed', so as to swell the treasury with a real merchant's goods; at the end of A 10.1 the samija is forbidden altogether in the army. This is one of the few direct contacts between the Arthasastra and Asoka. The taxes remitted in the Rummindei edict occur in the rastra list.

Page #33
212 D. D. KOSAMBI ing all classes of his people, particularly the brahmins and ascetics; by making gifts directly to the aged and needy, instructing and examining his subjects in morality (R. 8). He honours all sects (R. 12), but means thereby that the essential principles of all should be followed. This is more than a fashion with him, for new officials, the Dhamma-mahamatas now appear (R. 5), who are occupied with the morality of the whole population, the welfare of all prisoners, adjustment of schisms and controversies-all matters that didn't concern the Arthasastra. The church, therefore, is an adjunct of the state, a method of reducing the expensive mechanism of violence. Moreover, all royal officials are to take their own 5-year tours of inspection (R. 3), which again implies lesser number of spies. The king does worry about administration, for he expressly says that he will hear reports at all hours however late, and in all places, no matter how private (R. 6); that he should thus "discharge his debt to all living creatures," is a new and inspiring ideal of kingship. Whether he changed the former attitude to the traders is not explicitly stated; he recommends minimum expenditure, and collection of as little goods as possible (R. 3 : apavyayata apabhamdata sadhu). Nevertheless, it is clear that the trader finds his relations with the state entirely changed. Though Ujjain and Taxila have their own viceroys (Dhauli 1), the greatest monuments of his and succeeding reigns are not there, but on trade route junctions such as at Sanci, or Karli in the Deccan. The trader comes forward to contribute bis share voluntarily, alongside that of the king, to the Sangha ; the caves at Karli, at Kanheri, and elsewhere are often cooperative enterprises in which each contributes what he can of his own free will. This would not be possible unless the trader had benefited, and in addition found some reason to like the state-as the Arthasastra state could not be liked. Certainly, the king's opening up new territory by dhammavijaya, planting useful and medicinal trees, building rest-houses, digging wells, and repairing roads far beyond his own frontiers was effective in spreading his influence where his arms had not reached. This would be of double benefit to the trader, and we have noted that Buddhism, Jainism, and such Magadhan religion are specially attractive to the traders; even though the king still reveres the brahmins, as does the trader, they now have a far better meeting ground in religion which was absent in brahmin ritual with its exaggeration of caste differences. Asoka's conversion served as a measure of democracy, bringing the king nearer to the people, making him accessible to public opinion without the expensive, dubious intermediacy of spies. In the earlier stages, the religion of ascetic celibate almsmen was more economic than brahminism with its costly antisocial multiplicity of sacrifices and endless fees for everything. Thus, except among the Yonas, where indeed men were not attached to any (Indian) sect or religion (R. 13), the Asokan policy had immense success. Even then invaders like Menander and the Kusanas found it expedient to be converts, so that the success was really greater than Asoka claimed. His empire fell apart because it was no longer economic nor even necessary. The country

Page #34
ANCIENT KOSALA AND MAGADHA 213 had been united, and settled in a new path of development for which a strong centralized mechanism of violence was not essential. It is a reasonable guess that the end of the empire was caused, or marked, by local officials turning into feudal landholders, in what had been the seta. On the other hand, Buddhism spread to other lands as an adjunct to unity and auxiliary of the state mechanism in Ceylon, Burma, China ; in Tibet, and much later Mongolia, it replaces the state mechanism altogether. This is not the original Buddhism of 5th century Magadha, but its continuity of function in new forms * adapted to new types of society is striking. Conquest to the natural fron tier, poor transport, vast distances made it inevitable that the absolute central administration should be replaced by a dispersed feudal structure, both in India and China. In India, however, Buddhism staved off the feudal period till the change from the sita to the rastra was completed ; in China the new religion was welcomed even by the warring feudal landowners, to promote internal peace. In both places, it performed the minor economic task of returning to circulation-in the construction of pious works--the wealth accumulated by prince, landlord, and merchant. Espionage is necessarily directed against enemies of the state, whether they be foes from without or internal class-enemies that threaten the state which is itself the manifestation of some particular class. The universal espionage of the Arthasastra state proves that its essence, the king, had virtually no friends; it was not at that period the tool of any important class though warrior and trader profited during expansion; the old tribal basis for kingship had vanished. The Asokan change found such friends for the king and for his state. It also found new forms of expression in architecture and sculpture, for which the technique came from woodworking, the themes from popular legends, and the taste perhaps from Asoka's holiday spectacles. The court poet (or for that matter any professional secular poet) is not mentioned in the Arthasastra, nor apparently in Pali literature; the change made it possible for him to come into existence. Just as the new Magadhan religions had been developed by members of a class in decay, ksatriyas of the older free tribes, the new literature at its brightest (in the Gupta period) would develop from brahmin myths, in the brahmin language, by brahmin writers. This means that the older priesthood also had to undergo a change when there arose a new type of kingship, based upon private ownership rather than state enterprise and monopoly.