D D kosambi and the study of early indian Coins

D D KOSAMBI: THE MAN AND HIS WORKEconomic & Political Weekly EPW july 26, 200897D D Kosambi and the Study of Early Indian CoinsB D ChattopadhyayaThis article sets out to explain what drove D D Kosambi to take up the study of early Indian coins. Kosambi’s research in numismatics beginning in the 1940s marked a radical departure in the field from the practices and interests in the previous 100 years. There can be questions about how historians, including Kosambi, may have used coins as markers of socio-economic change. But Kosambi’s use of numismatics was such that historians can no longer ignore numismatic evidence for societal history. It was inevitable that some of the major findings of Kosambi would not stand the test of time. But the need to revisit his findings only demonstrates the value of his pioneering perspectives: we need to think differently and prepare a new agenda for examining material from the past – if necessary by turning current perspectives upside down.DD Kosambi’s explorations into the study of early Indian coins, which seem to relate only marginally to his other-wise extensive work on Indian history, society and cul-ture, represent his early attempts to gain familiarity and experi-ment with a substantial body of early Indian material remains. In explaining the beginnings of his experimental work with earlyIndian coins Kosambi mentions that he took up two initial problems “to teach myself statistics”: one was an analysis of the examination marks of first year college students, and the second – “a more fruitful problem” – was a quantitative analysis of early punchmarked coins. Kosambi’s initial finding was that the apparently crude bits of “shroff-marked” silver were coins as carefully weighed as modern machine-minted rupees. The effect of circulation on any metal currency is obviously to decrease the average weight inpro-portion to the time and to increase the variation in weight. The theory of this “homogeneous random process” is well known, but its applica-tions need careful work on whole groups of coins [Kosambi 1986]. This statement about how he initially became interested in the study of early Indian punch-marked coins separates Kosambi from those who are drawn to numismatic studies deriving from Indological or antiquarian interests. This also does not mean that he was in any way unaware of the historiography of such studies, but the distinction may explain how his concerns and the way he sought answers to his queries was a complete departure from the tradition of numismatic studies, as it had been evolving for over a 100 years before his time. This short paper is an attempt to under-stand his concerns: What it cannot attempt is to explain the tech-nique and logic of his numismatic statistics. It should be left to statisticians to explain the method and applicability of his statis-tics, a task to which they have not addressed themselves so far.Numismatic MethodologyTo appreciate properly where Kosambi took a different path calls for a brief statement on numismatic methodology in India in gen-eral. The beginning of colonial interest in Indian past, and the institutionalisation of this interest in different forms, had led to a steady accumulation of a wide variety of remains from the past: Manuscripts, objects of art, inscriptions, coins, stone tools and so on. The initial task was to classify them in chronological or some other order, and edit them, where necessary, for publication. The procedures and methodology of coin study were no different. Coins were collected (although many found were melted, hap-hazardly distributed or auctioned even by officials, or found their ways into the oblivion of private collections) as stray or accidental finds sometimes in the form of either small or large hoards, and sometimes during archaeological excavations, though generally not in large quantity. B D Chattopadhyaya (bdchattopadhyaya@gmail.com) retired from the Centre for Historical Studies, Jawaharlal Nehru University, New Delhi.

D D KOSAMBI: THE MAN AND HIS WORKjuly 26, 2008 EPW Economic & Political Weekly98Arranged chronologically into different groups or series, mostly on the basis of “legends” or writings on them, the descrip-tive methodology of coins, listed group-wise or variants within the group, picked on mainly the following variables: The weight of individual pieces, metal, description of the obverse and reverse symbols, including readings of legends, where they appear, and a few other details. When coins were uninscribed, as the punch-marked coins studied by Kosambi were, meticulous documenta-tion of their symbols, weights and variations would form the basis of classification. The documentation, as Kosambi’s trenchant criticisms running through his essays repeatedly point out, was not always impeccable, but the above would roughly characterise numismatic studies, from early standard works of A Cunningham (1891) or E J Rapson (1898) to the preparation of systematic museum coin catalogues by V A Smith (1906), E J Rapson (1908), John Allan (1936) and others and of detailed coin hoard studies by Indian scholars like A S Altekar (1927), P L Gupta (1963) and others. General problems deriving from descriptive studies centred around metrology or weight standards and their different de-nominations, comparing textual references with weights of ac-tual specimens, the relationship between areas of distribution and political control, and the tricky question of the origin of coin-age in India. Texts, considered a major source for coinage study, came in for a detailed analysis for providing vital clues in these matters. Considerations on the economic significance of coins – when discussions on the theme were undertaken – tended to ac-cept the simplistic generalisation that a substantial volume of coins or currency of high-value coins in gold and silver would suggest economic and commercial prosperity for the country; a deviation from it would suggest economic decline. An additional theme which received considerable attention from numismatists was the significance of symbols which figured on coins; in particular, the various combinations in which symbols appeared on India’s earliest coins – the uninscribed punch-marked coins – offered a puzzle evoking varieties of solution.1These were not themes which initially motivated Kosambi to study the ancient metallic specimens. As it has been pointed out in the beginning, it was the possibility of using the coins as sam-ples in his experimental statistical work that drew him to them. It is true that some of the queries arising out of his work were those of coinage experts too, but he would not, even in a somewhat changed scenario of numismatic study in terms of historical ar-chaeology, handle the problems the way they did. The second aspect in which his work differed from that of coin specialists was in the choice of his samples. A numismatist exam-ines all varieties of coins from the point of view of variables cho-sen by him: a single coin, a highly corroded specimen of which the present weight may be at considerable variance with its origi-nal weight, or a coin series of which only a few specimens are available. What Kosambi, on the other hand, required was a sub-stantial body of specimens, and, further: The coins must have been cut with sufficient accuracy at the begin-ning so that their initial variation is not much greater than the changes caused by circulation. This excludes copper, pewter, and even billon coins of the ancient period…Again, the circulation must be regular enough to have the proper effect, which excludes gold coins in general, almost always hoarded with the minimum handling, but liable also to be clipped or, in India, rubbed on the touch stone. Finally, the groups must have sufficiently large members with comparable history, i e, should be members of the same hoard [Kosambi 1956: 164]. All these preconditions stated in clear terms much later must have been in Kosambi’s mind when he set out to make a statistical study of early coins; they explain why he chose specimens from fairly large-sized hoards and why he took up only the “punch-marked” series of coins, issued mostly in silver with satisfactory conformity to an ascertainable weight standard and found in a fairly good state of preservation. His basic work on coins was published between 1940 and 1952; the later publications or publi-cations referring to his own generalisations on coins were based on his findings during this period. Since, given all the precondi-tions set for his work, he chose to work only on punch-marked coins, found only in hoards, it may be in order to make a few com-ments on the nature of these coins. Possibly closely associated with earlier crude silver pieces, also sometimes found together in hoards, punch-marked coins derived their name from the way they were prepared: they were sheets of silver, cut to a standard required weight (corresponding to the present ‘rak-tika’ or ‘rati’ standard) and were hammered or punched with a regular group (usually five) of symbols on one side. The other side generally remained blank, but where tiny marks appeared on the reverse, their number, as Kosambi attempted to bring out their significance in relation to their circulation history, could rangefromone to 20 or more. As punch-marked coins repre-sented, after they came into circulation, the most extensively manufactured and used coins in all parts of the country in early times, there were significant variations in the use of symbols and weights, and being in circulation for a long span of time (and therefore called ‘purana’ or old), along with other series more local in character, they went through certain significant stages of evolution.Preliminary StudyHowever, all this knowledge emerged only gradually and would not have been the central concern of Kosambi’s work. What con-cerned him, and the methodology of his approach, are both present even in his preliminary study based on two hoards of coins, both found at the Bhir mound in Taxila (near Islamabad in Pakistan) and published in detail by E H C Walsh [Memoirs 1939], with several inaccuracies and inconsistencies uncovered by Kosambi. For the larger hoard of heterogeneous series of coins, assigned to a date earlier than the other, Kosambi selected 1,059 punch-marked coins, divided into square-shaped, and round-shaped groups, and arranged them further into 10 groups, this time the grouping being done on the basis of the number of reverse marks on them. It is true, Kosambi admitted, that the number of reverse of marks in some cases could be as many as 16, but on the few specimens that the symbols appear, it was difficult to count them properly and, in any case, these specimens would not have made any significant difference to his statistical generalisation. Perhaps the simple arrangement can

D D KOSAMBI: THE MAN AND HIS WORKEconomic & Political Weekly EPW july 26, 200899be best represented by reproducing a sample table that he prepared [Kosambi 1984:2]: From this Kosambi immediately came to highlight what ap-peared to emerge as the most important point: “there is a regular drop in average weight with increase in the number of reverse marks”. That the reverse marks – irregular as they may appear – were of a major significance in the study of the circulation of punch-marked coins was the first point that Kosambi established through his simple method, buttressed by statistical tests. The correlation between weight loss and increase in the number of reverse marks, however, by itself did not explain either the pres-ence of reverse marks or variation in their number. So, even at this initial stage of his work he offered what he considered the only plausible hypothesis: “The only hypothesis that can account for our results is that the reverse marks are checking marks stamped on by contemporary regulations or controllers of currency, at regular intervals” (emphasis in original). He advanced also some firm suggestions regarding the significance of the regular group of five symbols figuring on the obverse, with of course con-siderable variations in them: in doing so he admitted the tenta-tiveness of what he was suggesting: “…I review the usual discus-sion of the symbols on the coins, and add my own pennyworth to the existing welter of conjecture”, and further (ibid: 15), “This is conjecture, not statistics, but after all a working hypothesis can always be produced, to be modified by newer evidence.” Refer-ence to his reconstruction on the basis of this hypothesis would be briefly made again later.Weight VariationThe calculation of weight variation between coins followed from an initial appreciation of the fact that no two coins, even though freshly minted and with modern technology, are of identical weight. As in several other cases, in initial variation as the start-ing point, Kosambi got more than 200 freshly minted coins tested at Bombay mint, and comparing the scale of variation with evi-dence from the older Taxila hoard (ibid: 66), came to the conclu-sion that “the ancients did a pretty good job of this coinage” in a normal period, the variation being wide in a period of abnormal minting. Another example of comparison with current coins was when a sample of 3,000 rupee coins was taken out of circulation to test how “coinage weight” would decrease with increasing length of circulation. It was found that it did “with about the same regularity” as with the square group of coins of the older Taxila hoard.To reiterate briefly as a layman reader, Kosambi’s analysis involved various stages. One was identifying different series present in a hoard deposit and trying to demarcate them in terms of their distinguishing features and of the chronology of minting. For example, the older Taxila hoard consisted of six groups, apart from two coins of Alexander, one of his half-brother Philip Arrhidaios, and one Daric of pre-Alexandrian Persian Achaemenid empire, the cumulative evidence suggesting that the hoard may have been buried around 320 BC. The second stage of course would be accu-rate recording of the weights of all coins; re-weighing coins dis-covered years back may be an impossibility unless they are all in an organised and accessible collection. The third would be select-ing the coins for carrying out the particular tests since tests were not undertaken on all groups present in the hoard, although found together and circulating alongside the selected group, they too would provide some idea of the nature of circulation.All these point to Kosambi’s insistence on making the initial documentation as comprehensive as possible and recording ob-servations on relevant aspects of the find. For example, simulta-neously with concrete details of how coins can be cleaned,1 one finds him also ridiculing the idea that some naive suggestions could be instantly made by numismatists on the condition of the coins immediately upon discovery. Centuries of burial in damp soil draw up the copper alloy of silver coins to the surface “leaving spongy silver behind”. Instead of understanding this process of “decuprification”, numismatists were making “the technically impossible supposition that molten copper had been poured on to the silver coin to bring up the weight” [Kosambi 1956: 165]. Equally impeccable needed to be the recording of stratigraphic and spatial context:Something could be done with a chart of findspots, but not in the ac-cepted dilettantish manner. If the findspots are accurately marked with groups, and the numbers counted instead of just the occurrence of single coin of the type, we would make better conjectures…… For this, however, will be needed not only better grouping of information but also far more information from new excavations and more thorough-going surface collections……..it would have been of value to know the stratification of the coins of the older Taxila hoard (Kosambi 1984, ‘Scientific Numismatics’, pp 43-44).These observations proceeded from Kosambi’s focus on two Taxila hoards which remained the constant reference points of his writings, and it is from this technical perspective that he examined afresh the reports on other finds of punch-marked coins for comparative analysis and commented on the particular inadequacies in their documentation. He also physically examined the coins of Bodenayakkanur hoard of Tamil Nadu. The reports reassessed by him were of Paila hoard of Uttar Pradesh (ibid, Essay No 10), Purnea hoard of north Bihar, east Khandesh hoard of Maharashtra (ibid, Essay No 7) and so on. The re-examinations helped him further to attribute the coins to specific regions, identify their special attributes and comment on the chronology of their circulation. For example, the Paila hoard was found in what could be considered the ancient ‘janapada’ of Kosala. Two features of the coins are distinct: the obverse symbols number 4 and the average weight of the coins would be two-thirds of that of the coins of other Taxila hoard. Since Kosala was incorporated within an expanding Magadhan kingdom by the 5th century BC, x----o 1 2 345Square n 224128132856446 m 53.26 52.93 52.74 52.47 52.53 52.17Round n 58342928 2510 m 58.35 52.84 52.75 51.90 52.29 51.67 6 7 8 9 10Square n 212510 9 8 m 52.03 51.67 51.40 51.47 51.01Round n 13 8 9 3 3 m51.8252.2351.2350.1051.20…….n is the number of coins with the number x of reverse marks given at the column head, and m the average weight in grains.

I

III II

D D KOSAMBI: THE MAN AND HIS WORKEconomic & Political Weekly EPW july 26, 2008101Persian siglos, on punch-marked coins of the Taxila hoard, is attrib-uted to their having circulated in the north-western region, despite the coins themselves having originated in eastern India. Further, the simultaneous currency of several series is envis-aged: The heavy “bent bar” coins of the Taxila area, not found outside the zone, the Magadhan punchmarked coins which con-stituted the major group in the Taxila hoard, the lighter-weight series of Kosala of the Paila hoard and so on. PM coins were suc-ceeded by other series. However, the continuity of the coins as exchange-credible in later centuries is suggested on the basis of both hoard study and modern parallels where very old pieces have been found acceptable in a “mix” of metallic currency. In the case of thePM coins the longevity of the series in post-Kosambi archaeological study is confirmed not only by the evidence of their being manufactured locally in the early Christian era but also by stray casers of coins with legend, manufactured by the casting technique again sometime in the early Christian centuries.Clues from HistoryClues connected with history from coins are found throughout Ko-sambi’s writings on punchmarked coins. Two concrete cases may be cited. One was offered by the contrast between two Taxila hoards. The coins of the later hoard contrasted with the earlier not simply in terms of chronology: both stratigraphy and the presence of a coin of Diodotus I, Indo-Greek ruler of the middle of the 3rd century BC as well as symbols assignable to the Mauryas, dated the coins to the Maurya period. What was significant was the nature of the Mauryan coins themselves. The initial variation between individual pieces was large, and the coins were adulterated in the sense of being heavily alloyed. This was a puzzle in the context of the largest empire in Indian antiquity, and in resolving this puzzle and in order to find a parallel Kosambi examined a large number of coins in circulation in 1940-41 when too currency was debased and “the legal remedy had been abandoned” (ibid 154-55). The expla-nation then was that like the British empire during the second world war, the Mauryans too experienced a heavy shortage of cur-rency. This was indicative of various pressures on the huge empire, leading ultimately to the decline of its structure. Allied to this was the destruction of the local economy of Taxila. While the pre-Mauryan hoard suggests greater uniformity and regularity in circulation and exchange; the absorption of the region into a system of rigid bureaucratic control under the Mauryas spelt its doom. The other historical deduction related to a much later period, i e, the period after the middle of the first millenniumAD. Kosambi pointed to the contrast – the contrast being a theme of much empirical work by later scholars – between the abundance of coins in circulation up to the Gupta period and its paucity afterwards. This contrast is accounted for by a major change in economy, acquiring more and more the character of a feudal economy [Kosambi 1956: 168-69]. “The self-contained village was hereafter the norm of production. Taxes had to be collected in kind, for there was not enough to allow their conversion into cash.”The considerable amount of research following that of Kosambi on pre-Mauryan, Mauryan and later periods may or may not invalidate Kosambi’s specific deduction that he dared to make in the barren field of numismatics, single-handedly and with as

D D KOSAMBI: THE MAN AND HIS WORKjuly 26, 2008 EPW Economic & Political Weekly102much care for empirical details as he devoted to other fields of his academic pursuit.3Kosambi’s off-cited emphatic statement: “…every hoard of coins bears the signature of its society” (ibid: 174) reiterates a point made in his other writings too of the urgency of looking at coins not sim-ply as objects of numismatists’ or collectors’ curiosity but in the context of when they were produced and the way they circulated. How important coins are in distinguishing between stages in history is a point used by him in his critical review [Kosambi 2005] of a volume edited by R C Majumdar in the History and Culture of the Indian People Series, which, even when it describes the beginning of Imperial history and of the historical period does not mention at all the beginning and the nature of metallic currency in India. There may be debate about the ways in which historians, including Kosambi himself, have used coins as markers of socio-economic change, but there is hardly any historian any more to ignore numis-matic evidence altogether for societal history. Numismatic studies have changed considerably over the years from amateurish (including among the professionals) handling of coins and initial explorations into quantitative techniques, al-though it would be impossible to establish any direct connection between Kosambi’s writings and the new numismatic historio-graphy. One can, however, in retrospect only recall what Kosambi wrote, as a plea to future numismatists, in 1941-42: As dated hoards are rare enough, and yet provide the only method of studying our punch-marked coins, I suggest that our numismatists and treasure trove officers pay more attention to numbers and weight, before and after cleaning. This does not mean that hereafter an ar-chaeologist must also know statistics; an acquaintance with the ele-ments of arithmetic and proof-reading will do [Kosambi 1984: 77]. Greater precision in documentation, and emphasis on full documentation and not in piecemeal, as a first step toward mean-ingful analysis are needs felt by all numismatists now without dis-tinction, although coins continue to disappear before documenta-tion can begin.What is inevitable with progress in knowledge, some of the major assertions of Kosambi present in his different writings, are proving to be replaceable. It seems today quite surprising that despite his knowledgeable accent on the growing importance of archaeology and archaeological stratigraphy, Kosambi did not consider it worthwhile to modify his chronological estimate about the emergence of metallic currency in India from the early 1960s onward. After all, the first systematic statement of the archaeo-logical position of a mean date somewhere between 6th and 4th centuriesBC had already been published in the 1950s [Ray 1959] and then regularly. Kosambi, for reasons not known, did not comment on the implications of this statement in his later notes. His other major finding, namely, that there is a positive correla-tion between increasing weight loss and increase in the number of reverse marks with progress in circulation, may require some modification. A fresh examination of coins, in the collection of the British Museum, apparently of the same series but not of the same hoard, is claimed to have shown that the correlation was not always positive [Susmita Basu Majumdar nd]. Even granting that the marks were put by bankers or shroffs and that they suggested some kind of monitoring, the possibility that a mark, already present on a coin, may have been recognised and made putting another mark redundant for the time, cannot be altogether ignored. A similar possibility would exist for the presence of a substantial number of marks. Thus there would be every likelihood that the number of reverse marks would not be exactly proportional to the period of circulation of the coin or duration of a coin during the process of its circulation, perhaps suggesting an impossible task that one should try and work out the circulation history of each coin in a group.Kosambi’s other major point, which concerned his calculation of the rate of absorption of a certain number of coins during circulation, may call for a different strategy of calculation in a different context. For example, in a different historical situation – in the Tomara- Cahamana period in early medieval Delhi – it has been assumed, on the basis of a hoard study, that “Bull and Horseman” type coins were issued continuously from c AD 1120 to 1193; and, further, that the rate of production was constant. Given these conditions and the high velocity of circulation due to prevailing circumstances, the rate of absorption too was quite high, the “half-life” or survival rate of coins being approximately 20 years [Deyell 1990: 172-74]. Suggestions of new possibilities or the new estimates only point to the enduring value of the pioneering work; they demon-strate that we need to turn to Kosambi not for the finality of his pronouncements but for pointing out that we need to think dif-ferently and prepare a new agenda for examining material from the past, be they coins or relics of a different nature, if necessary by turning our perspective upside down. Notes1 In his essay on Bodenayakkanur hoard Kosambi wrote about cleaning the majority of its coins: “The process consisted of soaking overnight (or longer if necessary) in a 10 per cent solution of for-mic acid, washing in pure water, and scrubbing carefully with a soft toothbrush. The museum chemist’s [at Madras] cleaning was more thorough than mine, and he coated the cleaned specimen with celluloid varnish..”, Kosambi 1984, p 124.2 Kosambi (1984), Essay No 4 in particular: ‘On the Study and Metrology of Silver Punchmarked Coins’, pp 14-80.3 Kosambi (1984); for a summing up see Kosambi (1956).ReferencesAllan, John (1936):Catalogue of the Coins of Ancient India(in the British Museum), London.Altekar, A S (1927): Catalogue of Gold Coins in the Bay-ana Hoard,Bombay.Cunningham, A (1891):Coins of Ancient India from the Earliest Times to Down to the Seventh Century, London.Deyell, John S (1990): Living without Silver: The Mon-etary History of Early Medieval North India, Ox-ford University Press, New Delhi.Gupta, P L (1963): The Amaravati Hoard of Silver Punchmarked Coins,Hyderabad.Kosambi, D D (1956): An Introduction to the Study of Indian History, Bombay. – (1984): Indian Numismatics (Introduction by B D Chattopadhyaya), Orient Longman, Hyderabad, reprinted 1991. – (1986): ‘Steps in Science’ inScience, Society and Peace, collection of papers brought out by Academy of Political and Social Studies, Pune, 1986.– (2005): ‘What Constitutes Indian History’ in Combined Methods in Indology and Other Writings, compiled, edited and introduced by Brajadulal Chattopadhyaya, Oxford Paperbacks.Majumdar, Susmita Basu (nd): ‘Kosambi and Scientific Numismatics’, mimeo. Memoirs of the Archaeological Survey of India (1939): No 59, Calcutta.Rapson, E J (1898): Indian Coins, Strassburg. – (1908): Catalogue of the Coins of the Andhra Dynasty, the Western Kshatrapas, the Traikutaka Dynasty and the Bodhi Dynasty,London.Ray, S C (1959): Stratigraphic Evidence of Coins in Indian Excavations and Allied Issues,Varanasi.Smith, V A (1906): Catalogue of the Coins in the Indian Museum, Calcutta, including the Cabinet of the Asiatic Society of Bengal, Vol I, Oxford.