The Date of Kanishka since 1960 (Indian Historical Review, 2017, 44(1), 1-41)

The Date of Kanishka since 1960 The paper which follows is the final text submitted to the Indian Historical Review in late 2016. The final paper was accepted and will appear in Indian Historical Review 44(1) (pages 1-41 according to the proofs I have received). SAGE publications do not permit the type-set version of the paper to be shared electronically but do allow the submitted text to be made available from my personal website (for which purpose I use Academia.edu and ResearchGate), which is one of the two open access models I am happy to publish under1. As such you should always make clear in a reference that you depended on this copy, and not on the one published in IHR2, I suggest something along the lines of: B a e , ‘. The Date of Kanishka si e I dia Histo i al ‘e ie , : 1-41 autho s te t, do loaded [ e site add ess], [date do loaded]. And make sure your library subscribes to the original journal, the IHR editorial team put in a lot of hard work making sure bibliography and references lined up and generally doing a patient job editing the final version (the bibliography is a lot less tidy here, you will just have to put up with that). The paper covers only one aspect of the controversy around the Date of Kanishka, essentially new evidence and arguments developed post-1960. It does not, as touched on in the introduction concern itself with the public reception. It is worth eade s ea i g i i d that hile the e is i suffi ie t e ide e to de o st ate the date is AD 127 I personally think it is very likely such evidence will eventually be found. However there is enough evidence to make 78 AD or a date in the third century impossible3. The paper was written from that perspective, and also in part to demonstrate why the weight of varied evidence (not any particular piece) requires that conclusion. 1 The other is an embargo period of up to two years after which the typeset version can be used. As a personal hobby horse I consider systems where authors pay journals to be utterly reprehensible though they have attracted a substantial amount of advocacy. 2 Especially important as I do not consider the text sacrosanct and may well edit it, possibly without flaggi g a diffe e e i e sio i fa t I e al ead o e ted a few grammatical errors and typos). 3 By which I mean so unlikely as to require serious distortions of existing data or special pleading to make either of those options intellectually defensible. The Date of Kanishka since 1960 Robert Bracey British Museum This paper is dedicated to Awadh K Narain (1925-2013), Robert Göbl (1919-1997), Evgeny Zeymal (1932-1998) , B N Mukherjee (1932-2013) and that generation of scholars which made a solution to this problem possible. Abstract The 1960 London Conference on the Date of Kanishka involved many leading scholars of Central and South Asian studies and had a profound impact on the field. This paper examines the historiography of the central problem posed at the conference: in what year did the era of Kanishka commence? It traces the advances in evidence that led to the solution of AD 127 between 2000 and 2010. The complexity of this process is often omitted in historiographical accounts, which opens the final solution to criticism and also fails to address why the field polarised after 1960 and found it so hard to reconcile new evidence. The paper suggests that eventual solution was a result of the cumulative effect of new data. It also shows that the field as a whole arrived at a solution long before it arrived at a consensus. This suggests that the failure of new evidence to bring about a solution more quickly is a major challenge to South and Central Asian studies in the future. Keywords Kanishka, Kushan, Chronology, Numismatics, Epigraphy, Wisdom of Crowds. The 1960 London Conference In 1968 the proceedings of a conference held on Kushan chronology at the School of Oriental and African Studies in London in 1960 was published. This publication marked a critical moment in the historiography of Kushan studies. More than a century of scholarship up to 1960 had come to the firm conclusion that establishing a fixed date for the accession of the Kushan king Kanishka would resolve most of the chronological problems which plagued the early centuries AD in Northern India. Attempts to tackle this problem up to 1960 are well documented1. In essence two groups of evidence, inscriptions in an unknown era and coins issued by Kushan emperors, defined the problem. The inscription with the earliest date was found at Kosam, near Allahabad, on the pedestal of a Buddha image and reads [ma]harajasya Kan[i]shkas[y]a savat[s]are 2 he 2 di 8, referring to the reign of Kanishka and to the 8th day of the 2nd month of winter in the year 2 with a partially abbreviated dating formula (he for * I am grateful to Joe Cribb for his comments on this paper and to Ysa Frahse for reading the text. The paper was p epa ed hile o ki g o the E‘C s e g g a t Be o d Bou da ies . For the historiography presee C i ‘edis o e i g the Kusha s su se ue t de ates up to the o fe e e a e tou hed o Whitehead The fi st o fe e e o the date of Ka ishka a d o e ed in more detail by Puri, Kusana Bibliography, with Russian contributions are summarised by Stavisky et.al Soviet Central Asian Archaeology and the Kushan Problem. Subsequent material and non-English contributions are covered by Fuss a , G Co t i utio s des sa a ts f a çais… ; Chronique des études kouchanes (1975and Chronique des études kouchanes (1978. 1 hemata, winter, and di for divase, day) that is common for the city of Mathura2. Other inscriptions, both in Brahmi and also from Gandhara in Kharoshthi, name Kanishka with dates up to 23. From the year 28 inscriptions naming Huvishka, and from the year 64 by Vasudeva, were also known. Amongst the sequence of Kushan coins there was a clear succession for th ee ule s hose a es i Ba t ia ΚΑΝΙϷΚΙ Ka ishka , ΟΟ ϷΚΟ Hu ishka , a d ΒΑ ΟΔ Ο (Vasudeva) clearly coincided with the order in the inscriptions. Clearly there was an era and it must have commenced early in the reign Kanishka. The era was often assumed to begin in the fi st ea of Ka ishka s reign though no evidence demonstrates that. Scholars believed that if they found that date it would provide a firm chronological framework for the early centuries AD. The historian A K Narain, then doing research at SOAS, is to be credited with realising how many fields the problem touched upon. He invited a large number of scholars from many countries and disciplines to attend a seminar in 1960. Papers were circulated in advance allowing the discussion to probe deeply into the problem. The resulting conference papers, published eight years later have served as a state of the art for the field since. This made the conclusion, summarised by the chairman art-historian A L Basham, one which should hardly have been disputed subsequently: Unfortunately the conference ended in an agreement to differ, for it was clear at the last meeting that the opinions of those present were fairly equally divided between AD 78 and a date some fifty years later. The great labour and learning which were put into the papers written for the conference, and into the discussions at the conference itself, have not solved the problem, which is bedevilled by evidence, none of it a solutel o lusi e, poi ti g i diffe e t di e tio s. 3 As Basham correctly summarised the problem was not a matter of method or scholarly e t e h e t, it e ai s a ope o e a d e e ide e alo e a fi all a s e it 4. For the next forty years three solutions dominated, a date of 78 AD, a date in the first half of the second century between AD 100 and AD 144, and a date in the early to mid-third century. Partisans of one solution or another would at various points summarise the existing evidence to support their position5. More objective accounts would return to the same conclusion voiced by Basham that no piece of evidence was decisive and the problem was essentially unsolvable6. However amongst specialists the frame of the argument was gradually shifting. Fewer scholars entering the field favoured AD 78 or a third century date, though no clear candidate in the first half of the second century appeared until 2001. At least in part this is because what Basham hoped for, new evidence, was discovered. Incrementally new evidence added weight to a date around AD 128. Gos a i, Kosa I s iptio of the eig of Ka ishka: Yea ; Shrava, S. Kushan Dated Inscriptions, no.10 (henceforth Sh#). I agree with Goswani that the image and inscription are almost certainly Mathuran products transported to Kosam, as in all likelihood are all of the images erected by the Nun Buddhamitra and the monk Bala in the middle Ganges region. 2 3 Basham, Papers on the Date of Kaniska, p.x. Basham, Papers on the Date of Kaniska, p.xii 5 Mitterwallner, Kuṣā ̣ a Coi s a d Sulptu e; Göbl, Donum Burns; Pu i, The Kusha s ; Cribb, 'The Early Kushan Kings: New Evidence for Chronology'; Soper, Recent Studies Involving the Date of Kanishka ; Bivar A Cu e t Positio o “o e Ce t al a d “outh Asia Ch o ologies ; Daffi à, Il Di attito “ulla Data Di Ka iṣka 6 Al a , I do-Parthian and Early Kushan Chronology: The Numismatic Evidence 4 That new evidence can be divided into two main types, numismatic and epigraphic. There are now several synchronisms between eras of India and Central Asia which were unavailable to the conference, and evidence of at least one era unknown in 1968. The coinage is far better understood today than it was in 1960, and thus relative chronologies much better established. This is not just true for Kushan coinage, but also for related dynasties such as the Kushanshahs, Kidarites, Guptas, Indo-Parthians, and Western Ksatraps. Numismatists have also established links between Kushan coins and these other dynasties which are more informative than the tenuous links between Kushan and Roman coinage discussed in 1960. The numismatic advances have fed into and complemented epigraphic discoveries. New inscriptions have been published, including those in Bactrian which were almost entirely unknown in 1960. Though less important than accumulation of numismatic and epigraphic data other sources have played a part, particularly the translation of the Yavanajataka, while some new archaeological and scientific dating techniques have been applied. Lastly there has been a shift in methodology. In 1960 scholars looked for a key piece of evidence hoping to fi l esta lish Ka ishka s a essio in order to resolve the chronology of the whole period. Today the date of Ka ishka s e a is identified with AD 127 or very close to that date but scholars have also come to realise that the Date of Kanishka is symbolic of, rather than central to, a more general problem. By focusing on the range of chronological issues from the first to fourth centuries the uncertainty of each event or reign has been gradually narrowed until that whole context today provides the solutio fo Ka ishka s e a rather than the other way around. This paper will summarise these new strands of evidence and how they have shifted the understanding of Kushan chronology since 1960, focusing first on the numismatic, then the epigraphic, and finally the other sources. I. The Numismatic Evidence 1984 Munzpragung der Kusanreiches The Viennese numismatist Robert Göbl was invited to the 1960 conference to present on the relationship between Kushan and Roman coins. Though originally a Romanist his interests had ranged increasingly eastwards and resulted in a paper in 1957 on the Kushans. It was this that brought him in contact with the Iranian Archaeologist Roman Ghirshman who arranged his invitation to the 1960 London event7. Robert Göbl would go on to publish a corpus of Kushan coins 1984 under the title System und Chronologie der Münzpgrägung des Kuša ei hes (henceforth MK). Robert Göbl believed very strongly that coinage existed within a political and economic system, as he puts it for the sake of control the coins must have been issued according to certain laws or customs. These we are able to rediscover, at least in part, with the help of a s ste ati e a i atio 8. This led him naturally to two new processes which would significantly advance Kushan studies, and at least one which caused considerable confusion. The first of these methods was the die study. As the dies which struck ancient coins are handmade they can be distinguished from each other9. If enough examples are examined then much about the technical details of production can be established. However, many more Al a , Gö l, ‘o e t . Göbl, Nu is ati E ide e ‘elati g to the Date of Ka iṣka . 9 Ea l appli atio s i luded ‘ose field, The D asti A ts of the Kusha s; fo Coi age of Wi a Kadphises . 7 8 u e t issues see B a e , The coins must be examined than there are dies and this led Göbl to the need for a systematic corpus. Previous publications of Kushan coins had illustrated, or listed, a representative sample of Kushan coins or all of the material in a single collection. Göbl began to collect every example of a Kushan coin which he encountered – those in public and private collections and those seen in auction catalogues. He recorded these on small file cards still available in Vienna today. Today such a systematic technique, and the intention that it would serve a die study, is a normal part of numismatics, but in the 1960s it was innovative. His third position concerned how to date coins relative to each other on the basis of typology, this will be examined below. MK is a monumental book featuring 155 larger than A3 plates illustrating Kushan and Kushano-Sasanian coins at 1:1 scale. In the gold coinage, where-ever possible, Göbl illustrates an example of every combination of obverse and reverse die of which he was aware, and lists additional examples in the text. Presenting data in this way does more than catalogue what is available, it actually creates a new source of evidence for the study of problems. The data demonstrated the order of issue of coins subsequent to the rule of the Kushan Emperor Vasudeva which would be important in interpreting the sequence of the inscriptions. Particularly this demonstrated the existence of a king, Vasishka, who post-dates Vasudeva. The use of die studies has also allowed an increasingly precise attribution of particular events usually disruptions in the history connected to the dating problem. At the 1960 conference a number of papers focused on the Sasanian conquest of Bactria. The evidence presented was a passage in the early Arab author Tabari, attributing the conquest to Ardashir, and an inscription of the Sasanian king Shapur10. This event must have taken place in the reign of Vasudeva (known from the inscriptions) or his successor, Kanishka II11, long known from coins, some found in Bactria12. The difficulty was establishing the synchronism, when in the reign of Ardashir or Shapur did the invasion take place and when did this happen in the reign of Vasudeva or Kanishka II. Die studies have resolved at least part of this problem. Though Göbl did not recognise the significance the data he gathered contained enough information to establish when the event took place in the Kushan succession. Much of the details of this were resolved by Joe Cribb, curator at the British Museum, in the 1980s and 1990s. 10 The relevant passages of Tabari are quoted from the 1879 edition on pages 391-3 of Basham, The Date of Kanishka. This has ee t a slated o e e e tl The Histo of al-Ṭa a ī C.E Bos o th, se . - pp. The he e t a k f o “a ād to Iṣtạ kh , a d the e to “ijistā , the e to Ju jā a d the e to A a shah , Ma , Balkh, a d Kh ā az , as fa as the fa thest f o tie s of Khu āsā , afte hi h he etu ed to Ma . He killed a la ge u e of people, a d despat hed thei heads to the fi e te ples of A āhīdh. The he etu ed from Marw to Fā s a d took up his ua te s at Jū . E o s f o the ki gs of the Kūshā , of Ṭū ā , a d of Mak ā , a e to hi offe i g thei su issio . F e has p epa ed a t a slatio of the i s iptio of “hapu si e the o fe e e, ‘. N. Frye The History of Ancient Iran. Handbuch der Altertumswissenschaft ; 3. Abt., T. 7 (1984), the opening of which reads " and all of the mountain chain of Pareshwar, Media, Gurgan, Merv, Herat and all of Aparshahr, Kerman, Seistan, Turan, Makuran, Paradene, Hindustan [ = “i d], the Kusha shah up to Pesha a , . As the i s iptio relates the defeat of the Roman Emperor Valerian it must indicate the situation after AD 260, though it fails to mention any Kushanshah. 11 Homonymous kings have caused considerable confusion in Kushan studies. Throughout this paper, to avoid confusion, the modern terminology which applies Kanishka II or Vasudeva II to the second Kushan Emperor of that name is used unless specifically stated otherwise. Until the correct sequencing of the inscriptions was understood many scholars incorrectly placed inscriptions referring to Kanishka III between Kanishka I and Kanishka II. This meant that Kanishka II was often referred to as Kanishka III or Kanishka II/III in publications, especially from the 1970s to early 1990s. 12 Cu i gha , Late I do-“ thia s correctly identifies the issue of later coins in the name of Kanishka though his exact attributions are often incorrect and Vasishka is not identified. At a macro-level there are two mints13 producing gold coins for Vasudeva and only one mint for Kanishka II. One mint of Vasudeva continued to make coins in the name of Vasudeva and this long series of coins eventually became the coinage of the Kushanshah, the Sasanian backed kings of the former Kushan territory in Bactria. The other mint changed from making coins in the name of Vasudeva to those in the name of Kanishka II. Du i g Vasude a s eig ew control marks were introduced at both mints twice during the reign, the first time a trident above the altar, the second time a nandipada symbol in the right field. In both cases the number of obverse dies employed between these events at the first mint is about twice that at the second. As these indicate broadly the relative production it is possible to count the number of dies between the introduction of the nandipada and the first coins of Kanishka II, at the subsidiary mint, and then find the corresponding point at the first mint. That point seems to correspond with the introduction of very atypical control marks consisting of small dots in various parts of the design. This indicates that the loss of the mint, and thus presumably Bactria, to the Sasanians corresponds very closely to the end of the reign of Vasudeva. This clarifies one side of the synchronisms, when in the sequence of Kushan kings the event happens. Despite these advances Göbl s a ou t had e little immediate impact. The numismatic evidence would lead to the gradual acceptance of a second Kushan dating sequence (see below) but much else that could be derived from the evidence was ignored. In part at least this was about presentation. MK was written in German, and though a translation was prepared by Warren Schwartz in April 1985 this was never published and exists today only as photocopies of a typescript. Thus the plates without the text were frequently employed and these can be confusing because Göbl attempted to combine both a type and a die catalogue in the same presentation and these have very different organisational principles. The book was expensive and its print run was small so they are largely restricted to specialist institutions. The other reason for the difficulty is that Göbl rushed the later sections of MK. The coins are presented comprehensively but the interpretation was not complete. This was not corrected until the late 1990s when he prepared a catalogue of the collection at the Bernisches Historisches Museum in Bern, Switzerland. The museum had received a gift of the private collection of Craig Burns which was particularly strong in terms of the late Kushan/KushanoSasanian period. Burns had apparently fallen in love with the romantic setting of the museum14 and been particularly grateful to the then curator, Kapossy. The Bern publication gave Göbl the opportunity to revisit the question of the numismatic sequence. Unfortunately, though the presentation is more complete (covering pre-Wima Kadphises) and straightforward than in MK the publication was still in German and thus inaccessible to most Kushan scholars. Table 1: Göbl’s Reconstruction of the numismatic sequence Kujula Kadphises Soter Megas Mostly attributed today to Wima Takto Vima Kadphises Kanishka I Huvishka Vasudeva I These a e Go l s Mu zstatte A t pes to A) and Munzstatte B (types 511 to 536) of Vasudeva, slightly confused as he divides B between two homonymous kings. For the coins of Kanishka II Gobl recognises there is only a production at Munzstatte B (types 538-557). 14 Göbl, Donum Burns. 13 Kanishka II Vasishka Vasudeva II Maio es do us And contemporary early KushanoSasanian imitations which Göbl labelled Xodēšah Including coins Göbl labels as Vāskušā a Göbl s te fo the fi al Kusha ki gs Mahi, Shaka, and Kipunadha, and the usu pe Maś a. Table 1 shows the reconstruction of the coins offered by Göbl. By the mid 1980s this arrangement had reached at least specialists15. However, the nature of the scholarship prevented it reaching more general works so the most recent general history of ancient India still gives an erroneous order similar to that understood in 196016. The Third Century Dating Confusion At the 1960 conference one contributor had sought to place Kanishka in the third century AD. R C Majumdar was the general editor of the History and Culture of the Indian People but not a Kushan specialist and his grounds for the dating (AD 248-9) were flimsy. Other authors, including the Russian numismatist Zeymal17 (1974) have considered that the problems of chronology could be swept aside by a radical rethinking such as moving the date to the third century. Robert Göbl pressed his case harder and was generally more influential than Zeymal18. In 1963, while researching a comprehensive work on the Hun kings, Göbl was shown a jewellery piece at the British Museum combining types from the coins of Constantine and Huvishka. Known as the Romano-Kushan Medallion he published it in a short note the following year. Göbl would claim repeatedly that the medallion was not the basis for his dating of Kanishka to 232 AD. The medallion i hai of e ide e as e el a sort of adde du a d othi g o e were his final words on the matter (1999: 164). Disinterested readers will, as all of his contemporaries did, have difficulty believing this statement. After his initial publication he returns to the medallion for two further articles19 and spent a disproportionate amount of time on it in his final publication. There is an irony that Göbl often criticised others in strident terms for failures of method while perpetuating two very serious ones in his own work. The first was a conscious error that he held too rigorous a view on the problem. Göbl believed absolutely that only typological connection, links based on the borrowing of motifs, could resolve the issue of dating. Typological connections are easy to see if you look for them. Göbl managed to see them between Kushan coins and second century Roman coins in 1960 when he presented to the 15 Czuma, Kushan Sculpture: Images from Early India; Mitte all e , P o le s of Ch o olog i the A t of the Kusa a Pe iod e t fu the tha Gö l a d e og ised that the B ah i i s iptio s of the Maio es do us e e the names of Kushan Emperors. 16 Singh, A History of Ancient and Early Medieval India. Reconstructions differed in 1960 but most authors followed a similar scheme in which there were eight kings finishing with Kanishka II. Later kings were omitted, Wima Kadphises was assumed to be the son of Kujula, and Vasishka was placed between Kanishka and Huvishka, while the Ara inscription was taken as Kanishka I or as second Kanishka thus creating the problems of nomenclature discussed in note 11. 17 Zeymal, Na hal a a data Ka ishki ; D ev ie o et Tadžikista . 18 The o l u e t ad o ate of su h a positio is Ni holaus “ hi del, see fo e a ple “ hi del A dashi Kusha shah a d Vasude a the Kusha … 19 Gö l, ‘ ) ei eue Te i i fü ei )e t ales Datu … ; The ‘o a -Kusha ia Medaillo … conference and then see them again between the same Kushan coins and fourth century Byzantine coins after he had examined the medallion. In this, the medallion was a trap. It was copied from both a coin of Constantine (AD 272-337) and a coin of Huvishka. If typological connection was to be applied rigorously it had to work for this object as well as others and it bound Göbl into a third century date for Kanishka. In fact copying demonstrates only that a copy post-dates the original, a terminus post quem. In the absence of other evidence such borrowings might make contemporaneity more likely than not, but once the connections begin to run counter to other evidence it is important to remember is only a terminus post quem. Though rigour in its general sense is useful, in its formal methodological sense, as a singular application of method, is a liability to the historian who must suit their methods to the evidence – not force the evidence to their method. The less palatable, but understandable within the time and place, failure on the part of Göbl is one of orientalism. Göbl constructed the Kushans with the constraints of his own understanding of the oriental; passive, receptive, unchanging. So in order to originate a coinage they had to be recipients of Roman influence including Roman designs, Roman pattern books, and even Roman mint workers. It was impossible for him to contemplate the possibility that Kushan coins were made largely for local reasons using local prototypes by local artists, or that a design element could be copied by a Roman engraver from a Kushan coin20. Numismatic Synchronisms In 1960 coins were used to synchronise the sequence of Kushan kings with contemporary dynasties, in particular by assuming that Kushan coins copied their designs from Roman prototypes. This was unsatisfactory as there was no way to establish how long a copy postdated its prototype by, and there was considerable subjectivity in identifying a borrowing and deciding in which direction it had occurred. As discussed above it led Göbl into serious error. Numismatic studies since 1960 were able to offer better synchronisms between the Kushan coinage and other dynasties. Typology played a part in this but hoards, metrology, and overstrikes were also brought to bear on the problem. Finds of Kushan coins in conjunction with other dynasties were known before 1960 of which the recovery of Roman coins including an Aureus of Sabina (AD 128) with Kushan gold (including a coin of Huvishka) in the Buddhist stupa of Ahin Posh, had received the greatest attention. There have subsequently been associated finds. In the burial mounds at TillyaTepe in Afghanistan, excavated in 1978, there is both an obol of the heraus-sanab type (Kujula Kadphises) and an aureus of the Roman Emperor Tiberius (AD 16-21) which Zeymal21 thought told against an early date for Kanishka. At Khairabad-tepe (Azarpay, 1970: 256) a coin of Nero was found in the same levels as Wima Kadphises, Kanishka and Huvishka coins, and coins of Vasudeva in the same levels as coins attributed to the Sasanian emperor Hormizd II (AD 302-9). However, after 1960 numismatic effort focused more on establishing links between the Kushan coinage and its near contemporaries than with Rome. Specifically in the Gangetic valley Gupta and in Central Asia Zeymal22 began to collate finds both new and historic. In the 20 [Subsequent note, Apr 2017] This remains an issue. I have read several terrible accounts that have managed to pass muster for peer review in the last few months. The idea that ideas must consistently originate in the west and pass to the east, whether that is coin design or dramatic literature is still pervasive in academia. 21 )e al, Till a-Tepe ithi the o te t of the Kusha Ch o olog , p. 22 The publication by Zeymal D ev ie o et Tadžikista has been subsequently expanded by Dovudi, Monetie Kladi Tadjikista … a d Gorin, 'Klad Med Mo et… . Gupta s o t i utio (1957) is prior to the 1960 conference 1980s and 90s Cribb published groups of coins which had appeared in the trade23. Though these lacked provenance they were probably hoards from Pakistan and Afghanistan. Subsequent to 1960 the count of hoards which mixed Kushan and Gupta gold increased to five. Three of these hoards24 include Samudra Gupta (#22, #24, #26) while two run as late as Kumara Gupta (#27, #28), and the Kushan coins with two exceptions (one doubtful) are all post-Vasudeva, mostly from Vasishka to Shaka25. Numismatists have invested some time in attempting to establish which of the late Kushan coins acted as prototypes for the early Gupta designs of Samudra26. An answer has only been achievable subsequent to the proper classification of the Kushan types in 1984. What is clear is that the Gupta mints adopted a variety of elements from different coins which complicates the question of prototypes but probably the most interesting are control marks which appear in the top right of the reverse above the throne on which the goddess is seated. Some of these are unique to Gupta coins, but some such as the Brahmi cha a te tha a e dated very precisely to the end of the reign of Kanishka II and the early part of the reign of Vasishka27. These borrowings place Samudra subsequently in the reigns of Vasudeva II to Kipunadha and probably in the reign of Shaka, as suggested by the Allahabad pillar. The Allahabad pillar inscription was first published in 1834 but the section reading daivaput a-shāhi-shāhā ushāhi-śaka- u u ̣ ḍaiḥ was not understood until relatively recently28. Though the inscription was noted several times in 1960 no-one raised this section. When Bhandarkar, Chhabra, and Gai revised the Corpus Inscriptionum Indicarum of the Early Gupta kings they dedicated a substantial discussion to the reading of this line29 and it can be seen they came close to an understanding though there are still a number of errors. They recognised that the first part is a Kushan title devaputra shao shaonano (as it appears in Bactrian on the coins) and that the last part murunda is a title used by the dynasties of the northwest. Unfortunately misled by earlier readings they continue to read this as a list, and the term śaka as a people, the Scythians30. The realisation that the Allahabad pillar inscription in fact referred to a single, particular, Kushan emperor required the recognition of the long sequence of coins labelled Shaka in Brahmi. This was only possible after Robert Göbl s pu li atio of MK a d u fo tu atel ot possi le fo hi as his h o olog e ui ed Vasude a II to e “a ud a s o te po a . Joe Cribb made the connection between the coins and the inscription in 199931. Metrology and was concerned principally with the eastern extent of Kushan rule, it led to more comprehensive surveys, for example by Singh 'Numismatic Evidence of Kushana-Murunda Rule…', a d the pu li atio of a hi al date, ota l Srivistava, Coin Hoards from Uttar Pradesh. This material is summarised and collated in Bracey, The Geography of the Kushans. 23 C i , Ga dha a Hoa ds of Kusha o-“asa ia a d Late Kusha Coppe , a d C i , I itatio s of Kusha Coppe Coi s . 24 Nu e s fo hoa ds a e take f o B a e Gold Coi Hoa ds . 25 The coins in the name of Samudra in #25, a trade parcel, which contained Kipunadha types are in fact Kidarite imitations of Gupta types. 26 The opinions are summarised in Raven, Gupta Gold Coins with a Garuda-Banner, p. 25-28 27 The other control mark is three dots, usually employed in the reign of Vasudeva II. Joe Cribb has shared these observations with colleagues and at many presentations but they remain unpublished. 28 C i , The Ea l Kusha Ki gs , pp. . 29 Bhandarkar, Inscriptions of the Early Gupta Kings, pp.26-29. 30 D C Sircar had recognised that the titles referred to a Kushan king as early as 1951 but without the numismatic identification of Shaka was unable to identify the king. 31 Cribb 'The Early Kushan Kings: New Evidence for Chronology' p.187. It was apparent after Robert Göbl s pu li atio that the diminution in weight of Kushan copper coins is mirrored in contemporary issues by other dynasties. The weight standard employed by a ruler is likely to indicate their approximate date relative to the Kushan rulers who issued coins on similar standards. The same is broadly true of the gold coins where the gold content is adulterated over time32 and other contemporary gold coinages seem to follow that reduction. The post Vasudeva Kushan copper issues diminish in weight in a series of steps. This, and the attribution of copper coinages to particular kings has been explored in detail by Khan33. In the time of Vasishka and Kanishka III the weight of the Kushan copper denomination has fallen to about 5g. It fell again to about 3.5g under Vasudeva II, and subsequently under Shaka and Kipunadha. The coinage of the Indian Kotas begins at about the 5g standard and falls to a standard closer to Vasudeva II and then Shaka. Following the conquest of Bactria by the Sasanians another series of copper coins is issued. These copy their obverse design from the copper coins of Kanishka II but use the reverse type (the god Wesho standing before a bull) used on coins of Vasudeva I. In broad outlines (see Khan, 2010 for details) the earliest feature a variety of control marks and weigh about the same as Kanishka II coins. They are soon replaced by coins which repeat the same mark, a delta or triangle, and conform roughly to the standard of Vasishka, and then as they become much cruder they drop to a standard close to that of Vasudeva II. In fact (Rajgor, 1991) such coin types continue to drop in weight suggesting the series outlives both the Kushanshah state that issued them and the Kushan state they were imitated from. A significant number of hoards are now known from Mathura to Taxila which contain Kushan copper down to the reign of Vasishka (and sometimes Kanishka III) as well as imitation coins of the same weight standards, but not coins naming Kushanshah rulers. Four relevant hoards are published in Khan (2010: table 4a). Hoards A, B, and C show the same pattern of Kushan and imitation coinages down to weight standards of about 5g. Hoard D has no Kushan coins, its imitations are of a lower weight standard (3.5-4g) and it contains Kushanshah issues of Hormizd I and Peroz II. Table 2: Weight Standards of Kushan Copper and relation to other coinages Aprox. Weight Kushan Emperor Coinage on similar Standard standards 16-17g Wima Kadphise to tet ad a h Huvishka 8-11g Huvishka (later Yaudheya coinage reign) 8-9g Vasudeva (early) 7-8g Vasudeva (late) 5-6g Kanishka II Kushanshah Early Kota coins imitations with various symbols 5g Vasishka Kushanshah Kanishka III imitations with delta 3-4g Vasudeva II Crude imitations Kushanshah coins (Hormizd & Peroz) C i & Odd De ase e t a d se ue e of late Kusha a gold oi s ; B a e & Odd The a al sis of Kusha pe iod gold oi s “pe ifi G a it . 33 Kha , Coppe oi s of Vasude a a d su esso s f o Ta ila . 32 2-3g Shaka 1.5-2.5g Kipunadha Shapur II (AD309379) Table 2 summarises these connections. Though this has refined the relationship between Kushan coinage and other dynasties it lacks the precision to help resolve the problem which attracted most attention in 1960, the first year of the Kanishka era. These links have been supplemented by other evidence, including overstrikes. Overstrikes Interest in overstrikes began to grow in the 1970s34. An over-strike occurs when a coin that is already in circulation is struck using new dies. The new design should obliterate the original but in some cases it can leave traces which are detectable. At the least an over-strike demonstrates minting of the over-type cannot have finished before minting of the undertype had begun. If many overstrikes are found that may indicate the types are contemporary, which is certain if there are mutual over-strikes. In the early part of the Kushan series it was established that Kujula Kadphises over-struck the coins of Gondophares35. Gondophares coins were also over-struck by Zeionises whose coins as noted above circulated with those of Kujula in Kashmir36. The sote egas oi age hi h followed these issues of Kujula were over-struck on Gondophares successor, Sasan37 and were over-st u k i la ge u e s o e of “asa s su esso s, Pako es38. The same types were further counter-marked by the Western Ksatrap ruler Damazada in large numbers39. Cribb40 also highlighted over-strikes of Kujula Kadphises coins by the Kings of Khotan. While the Indo-Parthian and Western Ksatrap kings such as Gondophares, Sasan, Pakores, and Damazada, are all dated relative to Kushan kings the Khotanese kings could be dated independently, through Chinese sources. The Khotanese kings with the name Gurga overstrike Kujula issues while their successors Inabasa/Xiumoba (c.AD60- , […]doga(sa)/Guangde (cAD66-AD86), and Panadosan/Fangquan (?-AD132) used designs which appeared to borrow elements from Kujula s issues i Kash i . This would require Kujula s eig to o e e efo e AD . There are no known over-strikes in the period from Wima Kadphises to Vasudeva, though some Kushan coins are counter-marked in Choresmia, which was known at the time of the 1960 conference41. This is largely because an overstrike is only practical if the coins of the under-type and over-type are of the same weight and composition. Otherwise the coins need to be modified in order to re-use them. As Kushan coins dropped in weight and other dynasties begin to issue copper coins of similar size in the third century overstrikes occur again. Some of the last Indo-Parthian coins, Te h i all the fi st ele a t pu li atio is “i o etta, A essa o the so-called Indo-G eek oi age published a few years before the 1960 conference but though there were several numismatic contributions none mentioned overstrikes. 35 Wide a , U e su f appe de Go dopha es… . 36 Senior, A Catalogue of Indo-Scythian Coins, p.150, type 213. 37 Sims-Williams & Cribb, A e Ba t ia i s iptio … ; cf. Alram, I do-Parthian and Early Kushan Chronolog , p.34 38 Mitchiner, Indo-Greek and Indo-Scythian Coinage, #1103, Sims-Williams & Cribb, A e Ba t ia i s iptio … , p.119; Senior, A Catalogue of Indo-Scythian Coins, type 269, Alram, I do-Pa thia a d Ea l Kusha Ch o olog p.43. 39 Senior, A Catalogue of Indo-Scythian Coins, vol.4 p.27. 40 C i , The “i o-Kha oshti Coi s of Khota …pa t a d The “i o-Kha oshti Coi s of Khota …pa t . 41 Tolsto , Dated do u e ts f o Top ak-kala pala e… ; 34 imitations of Pakores, are over-struck on imitations of Vasudeva42 but the most extensive over-striking is of Kushanshah coins by Vasudeva II, particularly his seated types on those of Hormizd43. Table 3: The coinage of the Kushan kings Kushan Emperor Coins circulate with Kujula Kadphises Coins circulate with IndoParthian rulers “ote Megas Wima Kadphises Kanishka I Huvishka Circulate with late Mitra kings of Ayodhya Vasudeva I44 Kanishka II Vasishka Vasudeva II Mahi Shaka Kipunadha Coins overstruck →Go dopha es ←Gu ga ule s of Khotan →“asa ←Da azada ←Pa o es → ou te a ked i Choresmia Coins circulate with earliest Kushanshah Circulate with Kota coins I itatio Pa o es → imitation Vasudeva →Pi uz I →Ho izd I ←Ho izd I Kanishka III Coins circulate with Samudraguptas Circulate with Kidarite In conclusion there were two major outcomes of numismatic research between 1960 and 2000. The first was an accurate sequencing of the coins, which had only previously been available in rough outline and only for the period from Kujula to Vasudeva I. This critically identified and placed in context the coins of Vasishka and Shaka. The second was a series of relationships between Kushan coinage and the coinage of other dynasties, particularly the Indo-Parthians and the Kushanshahs. There are good reasons these advances had very little impact. It took a long time for the details to be worked out, so nothing resembling table 3 was published in English until 199245. And though these numismatic advances are a cornerstone of resolving Kushan chronology they do not provide any answer in themselves. The chronologies are relative, not absolute, with the lengths of individual reigns unknown and the linked dynasties having equally uncertain chronological frameworks. In 1960 all that was in known with confidence was the length of the reigns of Kanishka I, Huvishka, Vasudeva I, and the Indo-Parthian Gondophares. Everything else would require new discoveries in epigraphy. Epigraphy 42 Senior, A Catalogue of Indo-Scythian Coins, hoard no.51. Pa ti ula l Ho izd t pes ith the a e Meze , see C i , Ga dha a Hoa ds of Kusha o-Sasanian and late Kusha Coppe , so e o e st ikes of Vasude a II o Pi uz I a e also k o , see p. ; so e Kha , Coppe oi s of Vasude a… p. -61, 44 Cribb (pers. Comm.) confirms there are unpublished overstrikes of Kushan coins by the Kotas. 45 The table in Errington & Cribb, Crossroads of Asia omits Mahi from the Kushan list, as well as the relationship with Sasan, Pakores, Damazada, and the Kotas. 43 The second Kushan era (the identification of Vasishka) An established sequence for the coins would lead ultimately to a solution for problems with the known inscriptions. Dated inscriptions belonging to the Kushan period are relatively common and it was well established before the 1960 conference that they were dated in more than one era (see below). The largest group, mostly from the city of Mathura were assumed to be in the era of Kanishka from the year 2 to the year 98. However two scholars, Johanna E van Lohuizen de Leeuw and John Rosenfield46, had proposed that even these dates might actually represent more than one reckoning. Lohuizen argued that the hundreds had been omitted from the inscription as an abbreviation, and that some of the lower dated inscriptions should be exactly 100 years later, while Rosenfield argued that a second era began about 100 years after Kanishka I. Both were art historians and both rested much weight on their ability to date sculpture by the palaeography of its inscriptions (Lohuizen) or its style (Rosenfield). Neither took the important precaution of establishing first a reference sequence whose dates were known independently and it is therefore unsurprising that their work was not widely accepted, contained many contradictions, and erroneously attributed a number of inscriptions. In 1960 even these authors accepted a sequence of inscriptions that ran Kanishka-VasishkaHuvishka-Kanishka-Vasudeva-Kanishka, often with complex explanations of joint rule. The subsequent numismatic developments demonstrated a clear sequence Kanishka-HuvishkaVasudeva-Kanishka-Vasishka-Kanishka. The first to connect this in print was the epigrapher Damsteegt47 who based his analysis on Robert Göbl s al ead pu lished o k a d an unpublished dissertation examining the palaeog aphi a gu e ts fo Lohuize s dropped hundreds theory by Plaeschke. Damsteegt unfortunately followed Göbl in employing a third century date for Kanishka which confused matters. Czuma48 followed Damsteegt, and Lohuizen49 recognised that the numismatic evidence supported her arguments in the last article she published. By the late 1980s the d opped hu d eds positio as suffi ie t idel a epted that a de ate e isted et ee the sho t ea s a d lo g ea s h o ology for stylistic developments in the art of Mathura championed by those who agreed or disagreed with the hypothesis50. Stylistic and Palaeographic Analysis Some art historians and epigraphers51, have tended to favour an early date for Kanishka, such as AD 78, in part because of the dates bearing on a number of art historical or linguistic questions. If the date of Kanishka was AD 78 then material at Mathura would be fifty years earlier than it actually is. This would push back the early uses of Sanskrit at Mathura and the early images of the Buddha. Stretching back to the art-historian Ananda Coomaraswamy in the inter-war years there has been a current of academic opinion that would prefer the first anthropomorphic forms of the Buddha to be made in an Indian, ie Mathuran, rather than a Ga dha a ith its G eek o e to es o kshop52. There is also an understandable desire to 46 Van Leeuw, The Scythian Period; Rosenfield, The Dynastic Art of the Kushans Epigraphical Hybrid Sansktrit, pp.9-12. 48Czuma, Kushan Sculpture: Images from Early India. 49Lohuize The “e o d Ce tu of the Ka ishka E a . 50Willia s, The Case of the O itted Hu d eds… . 51Salomon, Indian Epigraphy, pp.183. 52Cooma as a The O igi of the Buddha I age fi st ad o ated the idea, ost e e tl it has ee ha pio ed at le gth Lohuize , Ne E ide e ith ‘ega d to the O igi of the Buddha I age a d y Sharma, 47Damsteegt, make Mathura, which is clearly a centre for the production and export of art and inscriptions, a centre of innovation. Unfortunately these epigraphic and art historical problems have too often been made to have a bearing on the date of Kanishka. A quotation from Czuma will illustrate this: The ost o pelli g e ide e of all, ho e e , is offe ed the isual ate ial, su h as that o ie i the u e t e hi itio . The logi al p og essio of st le f o ”uńga to ea l Kuṣā ̣ a a t do u e ts a u i te upted o ti uit of st le. “hould o e suggest a late date fo Ka iṣka, su h as AD / , it ould lea e the fi st e tu AD fai l a e, si e the ea liest Kuṣā ̣ a artworks would date only from the first half of the second century AD. 53 While accepting that art history is heavily impacted by the resolution of these problems no seriously objective scholar would grant it more than a very minor role in resolving it. Palaeographic and stylistic dating requires first the establishment of a reference sequence of e a ples. That is possi le toda e ause the se o d se ue e s a solute date a d elatio to the coins is well established. In the eighties and nineties the order of the reference se ue e as depe de t o hethe ou a epted a lo g t o se ue es o sho t (one sequence) chronology, and which inscriptions you assigned to which eras. Thus any argument for the date of Kanishka from the art-historical evidence, such as that by Czuma, incorporated a chronology in its assumptions and was entirely circular. The other difficulty is that individual hands could have long working lives and so even when forms change rapidly and a good reference sequence is available it is rarely possible to achieve a precision greater than ±40 years. These methods therefore lack the precision to contribute usefully on the date of the Kanishka. However, as Damsteegt realised such methods are essential in attributing the bulk of dated inscriptions and sculptures which lack historical information to a particular era. The problem in reference to inscriptions has been recently discussed54 but a re-examination of the sculpture remains to be undertaken. New Inscriptions The period after 1960 saw a substantial increase in the body of epigraphic material that could be brought to bear on the problem of Kushan chronology. Not only were new inscriptions discovered but it is important to remember that the conference had no systematic publication of the available data while three such surveys55 were published subsequently. These new inscriptions emphasised the overlap between Vasishka and other Kushan kings in date. In 1960 Vasishka was known from four inscriptions in the years 2256, 24, 26, 28, all in Brahmi (Sh#58, 59, 61, 62), and the Ara inscription of the year 41 (CKI#158) which mentions his son Kanishka III. While the earliest Huvishka inscription then known was the pillar of the year 28 (Sh#64) dated unusually in the Macedonian month Gorpias. The Kamra inscription (CKI#230) was found in the early 1970s and editions were prepared by both Mukherjee and Dobbins. The inscription named Vasishka, and appeared to Buddhist Art: The Mathura School, particularly p.156. Notably both Lohuizen and Sharma favoured AD 78 and Sharma went as far as to reject a second century a d the d opped hu d eds theo . 53Czuma, Kushan Sculpture: Images from Early India, p.40. 54B a e , Ka kali Tila a d Kusha Ch o olog . 55Janert, Hei i h Lüde s; Mathu ā I s iptio s; Fuss a , Do u e ts pig aphi ues Kou ha s a d Do u e ts pig aphi ues kou ha s II ; “h a a, Kushan Dated Inscriptions. 56The year 22 inscription reads vasuksha ̣ asya.but it is not possible to offer a plausible alternative to identifying this with Vasishka as the date would not fit Vasudeva II. commemorate the birth of his son, Kanishka. However, both Mukherjee and Dobbins read the date incorrectly and it was not until Bailey that the correct reading of year 30 was made57. The Amitabha inscription (Sh#65) was found at Govindnagar on the Western outskirts of Mathura in 1977. The inscription drew much attention for its possible Mahayana content58 but its date is more relevant to our discussion. The dating formula reads sa 20 [?] va 2 di 20 6. The character after 20 is undoubtedly not 8 though it has been published several times as such. Though a 5 is possible, and was read from the rubbing, the photos published by Gregory Schopen suggest the character is the same as that in the day, a 6. Whether 25 or 26 this creates at least four years of overlap in the reigns of Vasishka and Huvishka59. Though these inscriptions had little impact on the debate amongst those already involved they made it increasingly difficult for new scholars to accept that the Kushan inscriptions were a single sequence. The resulting arrangement of dates which was clear from the association of coins and inscriptions is shown in table 4. Table 4 The two Kushan sequences Years Known King 2-23 Kanishka 26-60 Huvishka 64-98 Vasudeva 15 Kanishka II 22-30 Vasishka 41 Kanishka III Göbl (1984: 54) recognised the coincidence between his coin study and the inscriptions and accepted a second century (which he preferred to a second era as proposed by Rosenfield). Despite the adoption of this by many scholars, the idea of two sequences, one represented by the inscriptions from 2 to 98 featuring Kanishka, Huvishka, and Vasudeva, and the second by inscriptions from 22 to 41 naming Vasishka and Kanishka III, has been strongly resisted by many specialists. This has been partly because of entrenched positions but also because advocates of d opped hu d eds o si ila positio s have overstated the case. So far, no-one has been able to separate the inscriptions of Kanishka II from those of Kanishka I60, nor is it possible to establish if two centuries or two eras are involved. The Kanishka era was just one of a number of related dating systems in use in the Northwest in the early centuries AD. As all were intertwined advances in understanding the other eras impacted on understanding of the date of Kanishka. The Azes Era 57Mukhe jee, Ka a I s iptio of Vajheshka vāsishka ; Dobbins, The Kamra Kharoṣṭhī Inscription of Vāsiṣka ; Baile , The Dati g i the Ka a Kha osthi I s iptio 58 “ hope The I s iptio o the Kuṣān Image of Amitā ha… . 59The year 23 inscription of Kanishka from Sonkh (MI#136) also overlaps with the reign of Vasishka but as the s i e o itted ost of Ka ishka s ame and the year 22 of Vasishka (see note 51) has an odd rendering of his name this overlap was relatively easy to dismiss. In previous publications I have taken the Amitabha Buddha inscription to read 25 based on the rubbing but based on the photographs in Schopen The I s iptio o the Kuṣān Image of Amitā ha… figs 2a, 2b I think the reading 26 is certain. 60 Lohuize s a d ‘ose field s e o s i is-dating Vasishka and Kanishka III have already been mentioned. Puri (1994), who also tries to attribute a year 14 to Kanishka II misdates the Surkh-Khotal inscription attributing it to Kanishka III. The chronology of Northwest India and Central Asia actually rests on a number of different eras, most of which have suffered from the same problems as the date of Kanishka – and the solutions to which have gradually impacted upon the solution of the Kanishka era. In 1960 it was very clear that multiple eras must have been used in Northwest India and Pakistan. As well as the inscriptions dated in years from 2 to 98 which name Kanishka and his successors there were inscriptions up to 399. Some of these name Indo-Parthian kings, notably Go dopha es, as ell as Ka ishka s p ede esso s i ludi g Kujula Kadphises. The bulk of these inscriptions are in Kharoshthi and can be consulted in the corpus of the Early Buddhist Manuscript Project (http://gandhari.org/a_inscriptions.php). Today there are 101 dated Kharoshthi inscriptions known of which the majority must be in pre-Kanishka eras. There are far fewer examples in Brahmi but the inscriptions of Sodasa and Rajavula, and the inscription of the year 299 were well known in 1960. Even after the conference there was very little agreement on how many eras were represented or what the inception points of any of these eras was. Various suggestions had ee ade fo hat as ofte te ed a old “aka e a . These ould e o t asted ith the most popular solution, by John Marshall, of an era related to the Indo-Scythian king Azes. Marshal (1914: 973) could not establish exactly when this Azes era commenced but was able to establish an approximate solution. The Takht-i-Bahi inscription (CKI #53) is dated in the year 103 and also in the year 26 of the reign of Gondophares. Assuming this was the king whose coins were issued in the middle of the second century61 Marshall could establish the era commenced in the mid-first century BC, and based on a term in several other inscriptions he argued it was named after the king Azes. Finding a specific date convenient Marshall attributed it to the closest Indian era known at the time, the Vikrama, though he was not confident about the link: "... the identity of the era of Azes and the Vikrama era can hardly be regarded as fully established, and to my mind, it is quite possible that the era of Azes will be found to have commenced a few years earlier or later than 58BC"62 The solution of the Azes era made an incremental step forward in 1979 when Bivar published the Indravarman casket63. This inscription very clearly referred to the era as that of the IndoScythian king Azes. Though this article added no new evidence for an equation between the Azes a d Vik a a e a the de o st atio that Ma shall s a e had ee o e t see ed to give his specific date more weight with subsequent scholars. In fact an equation had always looked unlikely. The earliest known certain Vikrama era inscriptions were a group from Rajasthan i the ea s to hi h used the te k ita in their dating formulas, something known later in inscriptions confidently placed in the Vikrama era. This was a great distance from the region in which the Azes era was used and left a substantial gap in both time and space that could not be explained. The Yona Era That gap between the Vikrama inscriptions and those of the Kushan empire is not immediately apparent because there are a number of Northwestern inscriptions dated later than the year 282 and a Mathura Brahmi inscription dated in the year 299 giving the titles of 61 The lone dissenting voice on this point is that of Senior, A Catalogue of Indo-Scythian Coins, p.4, who believes the inscription refers to a later Indo-Parthian king. 62 Ma shall, The Date of Ka ishka , p. . 63 Bi a , The Azes E a a d the I d a a a Casket a d fo the su sta tial additio al alue of the askets i s iptio see “alo o & “ hope The I d a a a A a a Casket I s iptio ‘e o side ed… . an un-named Kushan king. However, it was clear in any close study of the inscriptions that the Azes era alone could not explain all of these inscriptions and some, probably all, of those with high dates had to be inscribed in another unknown reckoning with an even earlier inception point. Two inscriptions are known of Azes era years 122 and 136 (CKI # 59 & 60) which mention an un-named Kushan king, and an inscription of the year 284/7 names Wima Kadphises (CKI# . CKI # s date as disputed with some scholars reading one hundred rather than two hundred. A date of 287 would put Wi a s eig i the thi d e tu if the i s iptio were ascribed to the Azes era, too late for almost any chronology. Similarly the Bactrian inscription from Dasht-i-Nawur named Wima Takto and gave the year 279. It created the same problem of forcing the date of these pre-Kanishka kings into the third century AD if read as Azes dates. It was impossible that both were wrong and so they demonstrated clearly that yet another sequence had to exist with a much earlier inception point. The problem of this other era was solved with the discovery and reading of CKI #405, a reliquary inscription with a triple dating on it64. It gave the year in the reign of the Apracaraja king Vijayamitra (year 27), as well as specifying the Azes year 73 and the year 201 of what it called yoṇaṇa vaṣaye the year of the Greeks. This was a very important synchronism, as it confirmed the use of two eras, the Azes and the Yona, and that the Yona preceded the Azes by 128 years. It meant, for the question of Kushan chronology, that the pre-Kanishka inscriptions naming Kushan kings could now be expressed in a single reckoning. This made a dating of AD 78 for Kanishka even more difficult as it would require the Azes era to commence before 81 BC, which would in turn push the Indo-Parthian dynasty much earlier than could be reasonably sustained. Table 5 summarises the inscriptions naming Kushan kings65. Table 5 Kushan kings in the Yona and Azes eras Inscription Kushan King CKI #59 Un-named Kushan King CKI #60 Un-named Kushan King Dasht-i-Nawur Wima Takto CKI #62 Wima Kadphises Mathura 299 Un-named Kushan Inscription King Yona era [250] Azes era 122 [264] 136 279 287? 299 [151] [159] [171] This had clearly advanced the chronological understanding, however it had not solved everything. The Patika copper plate date of year 78 has been read maharayasa mahaṃtasa Mogasa which is usually assumed to refer to Maues, who issued coins in the early first century BC. This might indicate an additional era present in the inscriptions and arguments have been made for eras relating to both Gondophares66 and the Kings of Odi67. The kings of Odi pose a particular problem because in the Senavama scroll (CKI#249) Kujula Kadphises is mentioned. The Senevarma scroll is the longest extant Kharoshthi inscription, a gold sheet found in a gilded miniature stupa, dedicated the Odi ki g “e a a a i the fou tee th ea – – of o , The I do-G eek E a of / BC . first tabulation of this correspondence is Cribb, The G eek Ki gdo 66C i , The Ea l Kusha Ki gs… . 67“alo o , Th ee Kha oshti ‘eli ua I s iptio s , p. -9. 64Salo 65The of Ba t ia, Its Coi age a d its ollapse . the Lo d “e ̣ a a a, lasti g a thousa d ea s, o the eighth – 8th – da of the o th ” ā a ̣ a 68 it was first published by Bailey69. Amongst the various beneficiaries of the merit of the donatio a e listed “adaṣka ̣ a, so of the g eat ki g, hief ki g of ki gs Kujula Kadphises, so of the gods 70. Sadashkana is otherwise unknown, and the year 14 of Senavarma, either a regnal year or potentially another era, cannot be fixed. These discoveries have not resolved a central problem of how to assign inscriptions to the correct era. It is worth making clear how complex this problem remains. An inscription can be marked in various ways, the term yona marks inscriptions in Greek eras, variations on the name ayasa mark the Azes era, while the names of certain Kushan kings mark the era of Kanishka (Huvishka or Vasudeva) or the second sequence of dates (Vasishka or Kanishka III). However, when an inscription is not marked in this way, which is most inscriptions, its attribution is ambiguous. The presence of low and high dates had initially given a clue that two eras were involved in the Yona and Azes sequences but the possession of a low or high date is not sufficient to allow an attribution. In 1988 an inscription was discovered at Mathura dated ava a aj asa 71 ṣoḍ aśotta e va ṣaśate , the yona era 116. So this is a low dated inscription but clearly marked as in the yona rather than the Azes reckoning. The discovery of this inscription put in doubt the attribution of the bulk of the low dated inscriptions which had been assumed to be Azes era but had no variation on ayasa in their dating formula. When the triple dated inscription was first found most scholars assumed the Yona era commenced in 185 BC assuming that the Azes era was equivalent to the Vikrama era of 57 BC, which as discussed above is doubtful. Cribb72 dissented on this point and attempted to link the Yona era to the inauguration of the Bactrian king Eucratides. Cribb s discussion of Bactrian Greek chronology rested on four hoards of coins, published in 1965, 1973, 1981, and 2000, all subsequent to the 1960 conference. He assigned the inaugural date as 174 BC, which would mean that the year 299 inscription from Mathura was dated in AD 125, which Cribb believed was two years before the first year of the Kanishka era. This led Cribb to suggest that the Kanishka era was not a new era but a continuation of the Greek era with the centuries omitted. Bactrian Inscriptions Bactrian was known as a language in 1960 but from rather meagre evidence. The coins of the Kusha E pe o s afte the fi st ea s of Ka ishka s eig , a d those of thei su esso s employ it, but the only lengthy text available at the time was the main inscription from Surkh Khotal. Since then monumental royal inscriptions have been published from Dasht-i-Nawur (1974), Dilberjin (in 1976), Airtam (in 1981) and Rabatak73. The significance of Dasht-i-Nawur has already been touched upon and Rabatak will be treated in a few moments. Bactrian studies has also been expanded by the publication of seals, graffiti, and recently an extensive corpus of documents of the post-Kushan period, which contain dates that provide further context to the chronology. Bau s, Catalog a d ‘e ised Te ts a d T a slatio s of Ga dha a ‘eli ua I s iptio s. , p.232. Baile , A Kha oṣṭ ī I s iptio of “eṇavarma, King of Oḍi. . 70 Bau s, Catalog a d ‘e ised Te ts a d T a slatio s of Ga dha a ‘eli ua I s iptio s. , p.232. 71 Quintanilla, History of Early Stone Sculpture at Mathura, p.254-5. 72 C i , The G eek Ki gdo of Ba t ia, Its Coi age a d its ollapse . 73 For general introduction see Sims-Willia s Ba t ia La guage a d fo a o e ie of the i s iptio s SimsWillia s, Ba t ia Histo i al I s iptio s of the Kusha Pe iod . 68 69 The Rabatak inscription was discovered in 1993 in northern Afghanistan. It was initially published by Sims-Williams and Cribb in 1996 and provided the second complete inscription from a royal sanctuary after Surkh Khotal. Its importance for chronology was the genealogy it gave of the Kushan kings, including a name new to scholarship, Wima Takto. The subsequent revisions of the text criticisms and commentaries74 have not altered its relevance for chronology. This king corresponds to the bulk of the a o ous oi age k o as sote 75 76 megas after the Greek titles on it . MacDowall had suggested subsequently to the 1960 conference that the soter megas coins had to represent an additional ruler between Kujula and Wima Kadphises, now a name could be attached to that. This king must also be the Wima whose sculpture appears at the shrine of Mat and named in the Dasht-i-Nawur inscription77. Whether Wima Takto is the same individual as Ku ula Kataphśa s so Sadaṣka ̣ o78 mentioned in the Senavarma scroll or not is unclear. It is possible Wima Takto is a throne name or that Sadashkano died before ascending the throne but our sources do not give enough information for us to speculate. Combined, the Bactrian inscriptions confirmed the existence of a three kings before Kanishka and gave an additional date, 279, for one of these79. They were supplemented in the 1990s by various letters in the Bactrian language. The Bactrian Letters In 1991 Nicholas Sims-Williams was first shown a document in the Bactrian language written on leather. Within a few years it was clear this was just the first example of a substantial archive of material found somewhere in Afghanistan. As well as Bactrian language material the letters also included Arabic documents which we will return to in a moment. As these letters are administrative or financial documents they are frequently dated. There was little doubt when these were first published that the era was the same as that used in a group of inscriptions in the Tochi valley in Pakistan. The Tochi valley inscriptions were first published in 1964 and subsequently discussed by Humbach in 1966-7. At his request his original contribution to the 1960 conference was withdrawn and replaced by a short note on the Surkh Khotal inscription and the inscriptions from the Tochi valley. At this site inscriptions in Arabic and Sanskrit are juxtaposed on one rock and inscriptions in Bactrian and Sanskrit on another. Humbach took the juxtaposed inscriptions to be contemporary and therefore that the dates were comparable, the first pair being hijra 243 (AD 857) and samvat 32, the second a Bactrian chṣono 635 and a samvat 38. This samvat is undoubtedly the Laukika era and the equation gave a date of AD 232 for the commencement of the chṣono era of the Bactrian letters, which Sims-Williams modified slightly to AD 233 in his initial discussions. There have been attempts to use this date to fix the o uest of Ba t ia the “asa ia E pi e, a d thus to date the sta t of Ka ishka s eig Sims-Willia s, N. Fu the otes o the Ba t ia i s iptio of ‘a atak ; “i s–Willia s, N. The Ba t ia inscriptio of ‘a atak: a e eadi g . 75 The position may be more nuanced as will be argued in a forthcoming article by Cribb but that the sequence of coins Kujula - Soter Megas - Wima Kadphises – Kanishka could be mapped onto the sequence of kings in the inscription in broad terms is beyond doubt. 76 Ma Do all, “ote Megas, the Ki g of Ki gs . 77 Most contrary views rest on errors either in reading the original inscriptions, or in the most developed case by Os u d Bopea a h hi o a se ies of oi s hi h a e ode fo ge ies, see B a e , The oi age of Wi a Kadphises . 78 Bailey, A Kha oṣṭrī I s iptio of “eṇavarma, King of Oḍi , p.23. 79 Othe dates appea , the ea i the u fi ished i s iptio a iousl ead see Bivar, The Ka ishka Dati g f o “u kh Khotal , and the year 31 in the reign of Huvishka, neither of which have any bearing on the solution to this problem. 74 fi i g the e d of Vasude a s. It is in a sense the other half to the synchronism provided by the numismatic demonstration that the Bactrian mint of Vasudeva was lost at the end of his reign. The earliest of the dated Bactrian letters is for the year 110 (doc.A), so belongs to the Kushanshah period which post-dates Kushan rule in Bactria, and the latest is in the year 549, after the Arab conquest. The bulk of the late documents appear to originate from a single family archive, the family of Bek, and particularly concerned his son Mir. They are of interest because the same family members are mentioned both in documents written in Bactrian and dated in the Bactrian era (doc.X year 527; doc.Y year 549) and also in various tax receipts written in Arabic and dated in the hijra. Equations between these have been attempted, and other synchronisms offered based on the references to Hephthalites or Sasanian emperors in earlier documents. De Blois has argued from these synchronisms that the inception date must be earlier than AD 233, and has offered a Sasanian era of AD 223 instead. At least one scholar, Nicholas Schindel, would prefer to see the documents as dated in the Kushan era, a d hile that see s u likel it should e ad itted De Blois dis issal ests o e o his preference for his own solution than any actual evidence80. While interesting these synchronisms remain uncertain and they are not the most important effect of the Bactrian letters on the date of Kanishka question. For much of the period up to and including the 1960 conference there was a strong reluctance to suggest more eras than were strictly necessary to explain the known dated inscriptions. This was a very sound methodological position, usually referred to as O a s ‘azor, and it was frequently cited to suppo t a e uatio et ee Ka ishka s e a a d the ”aka e a of AD. Dates et ee AD 100 and 144 required the invention of a new era to support them, whereas 78 AD synchronized all of the inscriptions of North India. However, as we have already seen by the 1990s it was clear there must be two sequences of dates in the Kushan inscriptions, and at least two in the pre-Kanishka inscriptions. The Tochi valley and Bactrian letters added two more (the Laukika and the chṣono81). Very few of these can be matched to known eras or synchronized and the possibility of others e ist. A gu e ts f o O a s ‘azo had lea l failed and with it the last line of argument for AD 78. The evidence of the Tochi valley inscriptions showed that by the eighth century the Laukika era was in use in the region once ruled by the Kushans. In 2006 Melzer published an inscription which added an additional relevant inscription. The inscription was dated in the year 68 and makes reference to a series of known Hunnic rulers: Khingila, Toromana, Sadaviha, Javukha, and Mehama. Apart from being an important document in itself it has also 80 [Update April 2017] This section managed to reach the final stages of publication criminally under-served by references. The letters themselves are obviously published in Khan, G. (2007) Studies in the Khalili Collection Vol V: Arabic Documents. I e see se e al efe e es to pu li atio s ith de Blois h pothesis - de Blois, F. (2006) Du ou eau su la h o ologie Ba t ie e post-hellé isti ue: l ère de 223ap. J.C. , A adémie des Inscriptions et Belles-Lettres Comptes Rendus, vol.2, 2006, p.991-997; de Blois, F. (2008) The Arabic inscriptions from the Tochi valley. [Lecture]. Presented at: Bactrian chronology workshop, Cambridge. I have not read the latter. The summary and counter-argument in Schindel, The E a of the Ba t ia Do u e ts , hi h thankfully did manage to make the bibliography refers to the former. While this remains a messy and disputed detail it is important to remember that it does not really matter. It was important what the date of Kanishka was when it might shift the first images of the Buddha or early use of Sanskrit at Mathura by 50 to 100 years, that impacted other historical conclusions. The +/- 5 years being argued over whether the Bactrian documents use a Kushan or a Sasanian reckoning do not have the same impact. 81 Chṣono is simply the Bactrian equivalent of samvatsare and could be used in the dating formula of any inscription, I use it here as a convenient shorthand for the era in the Bactrian documents. been used to suggest the use of the Laukika era as early as the fifth century (AD 492/3). It could be a new reckoning, and a regnal year is just plausible. A regnal year is a less likely for the Spina inscription published by Khan in 2001, the inscription dated in the year 78 and making reference to a king transcribed as Ya+ Mīhusa thut a by Khan. The year is too high to be a regnal date, Khan suggests that it might be dated to the second Kushan sequence or the era of the Bactrian documents but the palaeography suggests that is too early. An alternative interpretation would e to att i ute it to the sa e e ko i g as Melze s i s iptio o to the Laukika82. The interest of the Laukika era lay in the way in which the century was omitted when writing it. We shall return to this era below but this Al-Biruni s eleventh century account of the reckoning: Co o people i I dia date the ea s of a e te iu , hi h the all samvatsara. If a centennium is finished, they drop it, and simply begin to date by a new one. This era is called lokakâla, i.e. the era of the nation at large. But of this era people give such totally different accounts that I have no means of making out the truth. In a similar manner they also differ among themselves regarding the beginning of the ea 83 Other Evidence At the 1960 conference radiocarbon dates had been offered as evidence in connection with the Toprak Kala site. Unfortunately these predated the widespread recalibration of such dating techniques and so have little value. Subsequent to the conference and as the papers went to press an additional set were presented by Dani84. These came from a building assumed to have been destroyed in the reign of Kanishka at Charsada. Again they came before the large scale calibration. For examples the first measurement read at the time as 147 BC to AD 93 would in fact be 105 BC to AD 130 using the IntCal04 calibration and 55 BC to AD 235 using ShCal0485. As is clear the margins of error are very large indeed. More recently86 other radiocarbon results have been offered as evidence. These were based on finds of Buddhist scrolls taken from Buddhist sites and offered on the international market. Like the Bactrian letters these lack context but they seem to have originated in Pakistan or Afghanistan and belong to the early Kushan period. The team working on them took radiocarbon samples from scrolls in the Schoyen collection and assumed that two sampled scrolls were prepared at the same time to combine the dates and claim a reduced uncertainty (AD 130-250). They then compared these dates to that on a pot associated with the scrolls, year 12. The team presumed the pot was dated in the era of Kanishka, thus claiming to eliminate the possibility of AD78. Unfortunately the reasoning here is dubious as there is no real way to know which era year 12 refers to and there are a great many uncertainties in the relationship between when the preparation of the birch bark began (the point measured by radiocarbon dating) and the date engraved on the pot. Probably the most interesting aspect of the evidence presented by Allon et al. is that it sho s ho pe siste t the ea l dati g fo Ka ishka s eig AD is. Kha , B āh ī i s iptio of the ea f o “pi a… . O B ie , The Ancient Chronology of Thar, p.20 84 Da i, ‘adio-Ca o E ide e of the Dati g of Ka ishka . 85 The use of ShCal04 may seem odd, as it is a Southern Hemisphere calibration, but the original calibrations are based on tree-ring on data from Europe and North America and until similar calibrations are possible for North India and Central Asia there is no way of knowing which would be more appropriate for Kushan finds. 86 Allo et.al, ‘adio a o Dati g… . 82 83 Though by the 2000s the evidence overwhelming demonstrated the date could not be as early as the first century, in an article published in 2006 there were enough proponents of that solution that Allon et al. phrased their article as negative evidence against AD 78. Due to its wide margins of error radiocarbon dating, while useful for establishing the broad period of a text is unlikely to be applicable to problems such as the date of Kanishka. The only scientific method likely to yield results would be dendrochronology, a method for establishing the date at which trees are felled, which can be precise to a single year. Though scholars have speculated on applying this method to the problem of Kushan chronology87 it has no practical application at present. Firstly, the basic sequences necessary for the dating have not been prepared fo “outh Asia, though su se ue tl to Bi a s spe ulatio s a chronology in Nepal back to the fourteenth century88 has been established. Secondly, wooden artefacts are rare with no wooden inscriptions surviving from the Kushan period, though wooden statues have recently been recovered at Mes Aynak89 and Bivar was able to recover one sample for analysis. Literary sources have had less impact subsequent to the 1960 conference as they were very well covered up to that point. In a moment we shall turn to the one text that was translated subsequently to the conference. Amongst those literary sources examined at the conference several have been subsequently re-examined. The Periplus Mari Erythraei, a sailors guide to the Indian ocean, was revisited by MacDowall90 who analysed its implications in detail and it has subsequently been translated freshly by Casson91. The main change in understanding has been a shift in the chronology of Arabia which adds little as the text is clearly a composite of information from different periods. Similarly an examination by John Brough of the Niya documents does not offer much new. Brough was interested in the titles used by the kings of the Kingdom of Shan-shan, which lies on the Southern side of the Tarim basin to the east of Khotan. Local documents are dated in the reigns of kings, and in the seventeenth year of Amgoka the titles change. Those before are inspired by the Kushan form maharaja rajatiraja devaputra, and those after Brough elates to a Chi ese o upatio f o AD . The aga ies a e too g eat to e e se B ough s reasoning and use the synchronism to date the Kushan king but it is important because he called into question in the process another synchronism. It is generally assumed that people designate as Yuezhi are Kushans, but he argues that it is at least possible they are from Shanshan. In particular he wonders about the correspondence between the embassy to China of AD a d the egi i g of the Ni a do u e ts se ue e of Ki gs. Though B ough s analysis was criticised and his suggestion of extended Kushan rule in the Tarim Basin largely rejected it remains an important detail in the chronologies92. The Chinese sources are open to considerable interpretation. This is in part the ambiguity inherent in the classical Chinese texts themselves and is well illustrated by a subsequent a gu e t i that Kujula s o uest of Kao-fu could be dated before AD 54 by Daffinà93. Bivar & Bridge T ee ‘i gs fo Ka ishka Nash, A aheologi al T ee ‘i g Dati g at the Mille iu , p. . 89Engel, New Excavations in Afghanistan. 90Ma Do all, The efe e es to the Kusha as i the Pe iplus… . 91Casson, The Periplus Mari Erythraei. 92 B ough, Co e t on Third-Century Shan-“ha … a d “upple e ta Notes o Thi d-Century Shan-“ha fo iti is s see Loe e, Chi ese ‘elatio s ith Ce t al Asia, . 93 Daffi à, The Ha Shu His Yu Chuan Re-translated , pp. -22 87 88 a d The various texts and their assessment have played an important role throughout the debate on the date of Kanishka. Several have been translated or edited more recently than )u he s contribution to the conference but none of these have revealed any new material relating to chronology94. The Yavanajataka of Sphujidhvaja Pingree did not attend the 1960 conference though he published within a few years what would become one of the most important synchronisms yet identified. In verses 14 and 15 of the 79th chapter of an Indian astronomical text written probably in the third century AD is a method of converting between different dating systems: Whe ea s of the ”akas ha e elapsed, that is the t uth ie., fou datio of the calculation of time. At dawn on Sunday begin that year and the yuga of the Sun. Take the u e of ea s that ha e passed of the Koṣā ̣ as, add , a d su t a t from this su the ti e of the ”akas i.e., the ea i the ”aka e a ; the e ai de is the number of years i the uga hi h ha e elapsed. 95 In hindsight it seems odd that this was not read as indicating a Kushan era commencing 149 years from the Saka era, ie in AD 227. Pingree does not help matters with a roundabout calculation of AD 61 and in fact with a general presentation that is not easy on the reader96. In 1978 of course this calculation could only have supported the advocates of a third century date such as Göbl, and though it was read by Kushan specialists (Bivar and Mukherjee at least) it made no impact until the matter was taken back up by Falk. Falk la ed Pi g ee s a gled t a slatio of the passage a d offe ed a e o e: The elapsed years of the Kuṣâ ̣ as i o i atio ith ha ge i to the ti e of the ”akas. “u t a ti g f o this ”aka ti e [plus ] the elapsed yuga, i.e. 165 years) (produces) the elapsed years of the second yuga. 97 Falk took a third century date to be widely unacceptable a d that Lohuize s theo of a d opped hu d eds as ge e all a epted a t-histo ia s 98. From this he concluded this was clear proof that the era of Kanishka commenced in AD 127. Falk was over-stating his case. There was certainly no reason to prefer Lohuize s issi g e tu ies solutio to ‘ose field s se o d e a and even the existence of the two sequences was far from widely accepted, as we have seen. As was pointed out in discussions at the time the Yavanajataka was actually good evidence that Lohuizen was wrong. If the centuries were missing because of abbreviation, and this is very plausible because the inscription dates already include other abbreviations, it made no sense for an astronomical text to drop the century and thus render the calculation useless within a few decades. 94 Hulsewe, China in Central Asia retranslates chapters 61 and 96 of the former Han, while Hill, Through the Jade Gate to Rome, translates the account of the Western Regions in the later Han history, while the relevant extracts have been re-e a i ed ost e e tl Thie Yuezhi et Kouchans Pièges et Da ge s des “ou es Chi oises . 95 Pingree, The Yava ajātaka of Sphujidhvaja, II, 187. 96 Pingree, The Yava ajātaka of Sphujidhvaja, II, a d o his ge e al st le see ‘o he ‘eview: The Ya a ajataka of “phujidh aja Da id Pi g ee . 97 Falk, The yuga of Sphujiddhvaja and the era of the Kuṣâṇas 98 Falk, The yuga of Sphujiddhvaja and the era of the Kuṣâṇas p.121. Falk revisited the problem99 and tried to argue that this was not a case of the centuries being abbreviated but of them being deliberately omitted in a cyclical dating system. He also linked the reckoning to two inscriptions in the Gupta era which were dual dated. The first had the years 112 and 5, and the second the years 121 and 15, in which the higher dates are clearly the Gupta era. These inscriptions were all new since the 1960 conference, the first of them having been published by Iyer in 1973. Falk used the dual dating to show that year 1 of the other reckoning was equivalent to AD 426. He then argued that reckonings separated by approximate multiples of 100 should be considered part of the same system of reckoning. Falk s app oa h as iti ised100 as several arguments in favour of cyclical eras were erroneous and all of the dual dated inscriptions had dates low enough they could (as Iyer had) be attributed to the reigns of kings rather than eras. Bracey also introduced a thought e pe i e t to de o st ate the p o le ith Falk s easo i g. The laukika e a hi h e et ith a o e is also sepa ated f o Falk s date of AD a ultiple of a out o e hu d ed years (depending on shifts in the calendrical system its year 1 can coincide with the year 23 or the year of o o e a e tu ies . If Falk s a gu e t as that the AD 227 Kusha e a and the AD 426 of the dual-dated Gupta inscriptions must be part of the same reckoning as the Kanishka era because they are separated by gaps of about 100 years it follows the Kanishka era must also be equivalent to the laukika era. Independently Cribb had made a similar argument to connect the Yona era to the Kushan so that Yona 301 = Kanishka 1 = AD 127, as discussed above. Falk has subsequently adopted some of these arguments and attempted to introduce a further synchronism based on intercalary years to date the Azes era101. The solution is tempting and would make the Kanishka era one small part of the longest continuously used reckoning anywhere in the world. However, all of this is suggestive rather than conclusive. What do we know? In 1960 Bivar concluded there was insufficient evidence to resolve the problem of when Ka ishka s e a ega . Such a conclusion would no longer be true and it is probably opportune to list what is now known with reasonable certainty:    Between the first century BC and fourth century AD at least five sequences of dates are known in large numbers of inscriptions; the Yona, Azes, Kanishka, the second Kushan sequence, and the Bactrian (chṣono) era. In the same region there are many dating sequences which are poorly attested; the Maues, the reckonings of Sodasa/Rajavula; the kings of Odi; the dual-dated Gupta inscriptions; and the Hunnic (Laukika?). These and other sequences (such as the Toprak Kala documents) have an unclear relationship with the five better attested sequences. The relationships between these sequences is still unclear except that the Azes commences 128 years after the Yona; and the second Kushan sequence very close to one hundred years after the Kanishka. Falk The Ka iṣka E a i Gupta ‘e o ds ; Falk A ie t I dia E as: A O e ie B a e , A ote o the “ae ulu a d the Kusha E a the ta le p ese ted of o o da es o tains a number of labelling errors, notably confusing the Tochi Valley inscriptions with the Bactrian documents and Arab material. 101 Falk & Be et, Ma edo ia I te ala Mo ths a d the E a of Azes . 99 100    None of these sequences is equivalent to the Saka era of AD 78 widespread in Western India and the Gangetic valley at the same time and it seems unlikely any are equivalent to the Vikrama era first known from inscriptions in the third century AD The second Kushan sequence almost certainly commences in AD 227. The margin of error for the other sequences is relatively small, the Yona commences in the early second century BC; the Azes in the late first century BC; the Kanishka in the early second century AD; the Bactrian (chṣono) in the early third century AD. The details of all of the synchronisms and relative chronologies that can be reasonably firmly established are summarised in appendix 3. This has essentially solved the problem that was presented in 1960 and if the date of Kanishka is not AD 127 (which seems most likely) it is at least close enough to exclude most of the previously considered alternatives102. This solution has however raised a range of other problems. For historians there is the problem of refining the relationships between these sequences, exactly how many eras are present and what are the precise inception points. As in 1960 this problem probably requires more evidence. For epigraphers the major problem is attribution – how can you establish which of these reckonings an inscription corresponds with. For art historians the greatest task is posed. The city of Mathura is a major centre of sculpture, its works are important in the development of Buddhist, Jain, and Hindu iconographies in the first to third century and many of them are dated. There is essentially no study of that school of art which does not rest on a chronology now known to be flawed. There is a compelling need to revisit the stylistic development of Mathuran art with the firmer basis provided by these chronological advances. A Solution The conclusion will assume that a Kanishka era commencing in AD 127 with a subsequent century or second era in AD 227 is the correct solution. This will help to explore why the field polarised before 2001 rather than moving towards a consensus as new evidence accumulated. The current author does not think the evidence specifically for AD 127 is particularly strong, though it is the most plausible of the possible solutions. What is certain is that there are at least two sequences of dates in the inscriptions and the Kanishka era must have fallen in the first half of the second century AD. No individual piece of evidence demonstrates this. For example, Pi g ee s pu li atio of the Yavanjataka had no impact on the field though we know proponents of both early and late dates for Kanishka had read it103. E e ith Falk s e-presentation it is certain that the source does not refer to what we call the Kanishka era (but probably to the second sequence or chṣono). So it is possible, but not probable, for this single piece of evidence to allow even for a date as early as AD 78 or as late as AD 227. The same is true of all of the evidence, the most plausible interpretation of each piece is a date in the early second century but other options are possible. The problem is that it is necessary to make that case for all of the evidence (the Yona/Azes inscriptions, the Sasanian conquest, Indo-Parthian overstrikes, relationship with the Guptas, and so on) to arrive at a date earlier than AD 100 or later than AD 150, or even a 102 AD 128 is certainly possible as are other movements of a few years such as AD 132, but even a few decades such as AD 115 or AD 144 would be difficult to accommodate. 103 Mukhe jee, The Ya a ajātaka of “phujidh aja, the ”akakāla a d the Ka ishka E a espo ds di e tl to the pu li atio . “ope , ‘e e t studies i ol i g the date of Ka iṣka p. also e tio s it a d Bi a i a e ie shows he is familiar with the publication. date more a few decades from AD 127. And to consistently reject a probable interpretation in favour of a less likely but possible one is a form of special pleading. Conclusion – Incremental Advances In hindsight, the weight of new evidence published since 1960 which bears on the problem of Kushan chronology, and the conclusions it obliges us to draw are obvious. It is important to remember that those involved did not have that perspective. Each new piece of evidence added only incrementally and so were dismissed or incorporated by advocates of particular solutions immediately. Singularly these pieces of evidence lacked sufficient weight to overturn positions that had been forged in consideration of the full range of evidence in 1960. Kushan specialists should be wary of writing too simplistic a historiography on the date of Kanishka. It did not remain a vexed issue without advances until Falk s suggestio of AD (though too often that is how it is presented). The generation of scholars who were young in 1960 spent the last half century seeking out and debating new sources of evidence. It is that evidence that makes it possible today to firmly reject any equation of kanishka = saka = AD 78, and state that the era commenced within a few decades of AD 127. At the same time the scholars who had participated in 1960 were too embedded in the debate to be able to see the cumulative effect of the evidence they had gathered. Enough time would have to elapse for a new generation to become involved in the problem. As in the famous remark by Max Planck: A e s ie tifi t uth does ot t iu ph o i i g its oppo e ts a d aki g the see the light, but rather because its opponents eventually die, and a new generation g o s up that is fa ilia ith it. The date of Kanishka problem is one that should engage historians of early India today because it shows how the discipline has not been able to establish reliable methods of navigating scholarly disputes on such issues. In the period from 1960 to 2010 the discussion was dominated by polemical summaries and attempts to find one critical piece of evidence which would resolve the problem. Neither of these proved to be particularly fruitful routes to a solution. At the 1960 conference twenty-three of the participants expressed a firm opinion either in papers or in the subsequent discussion (Basham, Eggermont, Gershevich, Jaini, Lohuizen, Shama, Sircar, Staviskiy, Tolstov, Warder for AD 78; Thapar AD 78-112; Zurcher AD 80-120; Narain AD 100; Rosenfield AD 110-115; Conze AD 120; Allchin AD 115-143; Bivar AD 128-9; MacDowall AD 128-144; Dani, Göbl, Gupta Pulleybank AD 144; Majumdar AD248-9). The largest body of opinion, ten participants, were further from the correct answer (49 years) than any participant except Majumdar. In fact the mean date suggested, AD 110, is closer than the majority of the participants. That the aggregate result of the conference should have been better than the majority of its participants is not a surprise. Though in hindsight we know there was insufficient evidence to answer the question it is a well-known phenomenon that aggregate results from diverse groups can often outperform individuals or even the data itself. Sometimes known as the isdo of o ds o e of its earliest applications was in the tracking of the asteroid Ceres by the mathematician Carl Friedrich Gauss, who aggregated the diverse measurements of astronomers to achieve a closer result than that of the individuals. Various authors reject the idea that a consensus of opinion should carry any weight in debates. This position is given forcefully by Gerard Fussman quoting Robert Göbl: …as ‘.Göbl rightly points out scientific truth is not decided by a majority nor is it attained a o p o ise o ea et ee t o e t e e positio s. 104 No doubt Fussman was reacting to scholars such as Narain who described his own position as a compromise, or Walton-Dobbins who advocated the use of AD. 100 as the mean year in the controversy over the date of Kaniska 105. Walton Dobbins is not of course using the term mean correctly, as he has simply picked a round date between two of the possible solutions (at the time AD 78 and AD 128) arbitrarily. Fuss a s iti is does ot alte the i te esti g point that the field as a whole, measured crudely by taking the mean, is more accurate that its individual members. After the 1960 conference opinions actually polarised, with four of the scholars in appendix 2 advocating dates in the third century, but the mean value of all the scholars being AD 128. Where-as 5 of the 23 scholars had outperformed the mean in 1960 of the 20 scholars in appendix 2, who participated in the discussion post-1960 a d efo e Falk s i te e tio the ast ajo it performed worse than the average. The wisdom of crowds is a useful way to look in hindsight at how the field as a whole performed but it cannot solve problems. If only those scholars who changed their opinion or were new to the debate are considered the mean rises to AD 141 and it becomes clear the generation after 1960 were in part reacting against the majority position of AD 78, the apparent wisdom of the crowd is an expression of the polarised nature of the field. The correct answer was ultimately established by evidence not analysis of scholars. This discrepancy between scholarly consensus, the aggregate of the field, and the actual answers that are being sought is one that should vex students of Indian historiography. Though the date of Kanishka is now, mostly, a settled matter106 the way in which the field failed to arrive at a solution for so long is troubling. Many similar issues await solution, some of which no doubt require further evidence but many of which may require better methodologies. The wisdom of crowds analogy highlights the troubling phenomena of more evidence leading to less accurate scholarly analysis. The evidence was undoubtedly sufficient for a complete solution by the late 1980s but no consensus was reached until the mid-2000s. Even when it was reached historiographic accounts focused on a single recent advance, leaving the apparent consensus vulnerable to dissenters, such as Mukherjee, Fussman or Schindel. The importance of the Date of Kanishka debate is helping us to understand how scholarly debate can block as well as facilitate progress. There are clearly lessons to be learnt. Fuss a , A decisive stage in the study of Kusha a oi age . Bi a , ‘e ie of the Date of Ka ishka ; Walto -Do i s Ga dha a A t f o “t atified E a atio s 106 Though there remain advocates for other positions, Schindel for the third century and Fussman for AD 78, there have been no new entrants familiar with the evidence who have sought a solution far from AD 127. 104 105 Appendix 1 Time-line of significant events and publications post-1960 1960 Second Conference on the Date of Kanishka subsequently published in 1968. 1961 Posthumous publicatio of Lüde s Mathu ā I s iptio s 1964 Tochi valley inscriptions first published. 1968 Publication of the 1960 conference by Basham. Ja e t. Conference on the Date of Kanishka in Dushanbe. 1972 First overstrikes of Kushan coins published. 1974 First publication of Dasht-i-Nawur inscription. 1973 Kamra Inscription first published by B N Mukherjee. 1978 Pingree publishes a translation of the Yavanajataka of Sphujiddvaja. Excavation of the Tillya-Tepe burials. 1979 Pu li atio of the I d a a a asket i the Azes e a y A D H Bivar. 1980 Publication of the scroll of Senavarma, King of Odi. 1982 Bailey first reads the date on the Kamra inscription correctly. 1984 Publication by Robert Göbl of Munzpragung der Kusanreiches. 1984-5 Joe Cribb publishes connection between Kujula coins and the Sino-Kharoshthi coins of Khotan. 1991 First Bactrian letter seen by Nicholas Sims-Williams. 1993 Joe Cribb attributes Heraios-Sanab coins to Kujula Kadphises. 1993 Discovery of the Rabatak inscription, first published in 1996. Publication of the collection at the Bern Museum, Switzerland. 2001 New reading of the Yavanajataka of Sphujiddvaja by Harry Falk. 2005 Publication of the triple dated reliquary by Richard Salamon. 2006 Publication by Melzer of Hun inscription of the year 68. 2009 Publication by Falk and Bennet of a synchronism based on inter-calary months. Appendix 2 Contributors This is a partial list of those who contributed to the debate on the Date of Kanishka after the 1960 conference. Arthur Llewellyn Basham a poet and novelist completed a PhD on Indian philosophy in 1952 and went on to hold the chair in Indian History at SOAS. He was instrumental in organising, chairing, and ultimately ensuring the publication of the papers from the 1960 conference. Basham favoured a date of AD 78107. Adrian David Hugh Bivar (1926-)108, was appointed as lecturer in Central Asian Archaeology at the School of Oriental and African Studies in 1960, he attended but did not present at the 1960 conference and went on to make a number of contributions on the epigraphic problems related to the date of Kanishka. He favoured a date of AD128109, or slightly early, 120-128. Joe Cribb joined the Department of Numismatics at the British Museum in 1970 where he initially worked on Chinese numismatics, and was head of department from 2003. His work on Kushan coins began in the late 1970s at the British Museum. For more than thirty years he was the curator responsible for the Kushan coins at the British Museum, the o ld s most significant collection of the material and has made major contributions to their study (Burnett, 2011). Cribb began work on Kushan studies after the 1960 conference and held to a second century date though he advocated different dates on different occasions. In his last major summary of the problem before 2000 he favoured a date of AD107-120 but accepted AD127 later and has since advocated the equation Yona 301 = Kanishka 1. Stanislaw Czuma, was the Curator of Indian and Southeast Asian Art at the Cleveland from the early 1970s. His only significant foray into Kushan studies was a major exhibition on the art of the period in which he argued for a date of AD 78 for Kanishka. Theo Damsteegt (1948-) completed an important PhD, later published, on the epigraphical hybrid Sanskrit used in Mathuran inscriptions in 1978. He went on to work on Hindi narrative at the Kern Institute in Leiden, but had argued for a late date of AD200. Gerard Fussman (1940-) Professor at the College de France and prolific researcher whose principal contribution was the collecting and editing of Kushan inscritpions. Fussman has been one of the most vocal advocates for AD78. Robert Göbl (1919-1997) began studying in Vienna after the second world war where he was the first chair of the Department of Ancient Numismatics and Pre-Islamic History of Central Asia. He had a profound impact on numismatics generally of which his voluminous work on the Kushans and associated dynasties was just a part. He attended the 1960 conference where he favoured a second century date but soon adopted, and was one of the strongest advocates of a third century date, particularly AD 232110. E . ikipedia.o g/ iki/A thu _Lle ell _Basha [ o sulted De ]; Thapa , ‘ A L Basha Economic and Political Weekly Vol.21 No.9: 381 108 Ma Do all, D W Be ell, A “ “hah azi A “h. A D H Bi a Bulleti of the Asia I stitute: I a ia “tudies i Honour of A D H Bivar Vol.7: 1-8 109 Dochesne-Gullemin, J et.al. (1979) Acta Iranica Vol I: Bibliographies de 134 Savants: 56-59 110 Alram, M (2001) www.iranicaonline.org/articles/gbl- [consulted Dec 2014] 107 Helmut Humbach (1921- ) an Iranian linguist who worked at the Universites of Saarbrucken and Mainz. He attended the 1960 conference but withdrew his original paper in favour of a short comments on the Surkh Khotal inscription of the year 31. His major contributions included the study of the Tochi inscriptions which he dated to an era of AD 232 and thus derived an era of AS 134 for Kanishka111. Johanna Engelberta van Lohuizen-de Leeuw (1919-1983) spent most of her career at the U i e sit of A ste da . “he de eloped he theo of the d opped hu d eds i he PhD thesis which was published in 1949 and would go on to defend it and expand the attributions of inscriptions to particular eras many times after 1960. She participated in the conference on the date of Kanishka and continued to favour a date of AD78 throughout her career.112 David W MacDowall subsequently to attending has never committed firmly to a date continuing to favour AD 115 to 144, but later inclined towards Harmatta s solutio of AD 134 Gritti von Mitterwallner (-2012) was Professor of Indology at Munich from 1978 to 1991. She published a volume trying to reconcile the chronological problems of Mathuran sculpture and Kushan coins as a preliminary to her own interests in later numismatic material in which she advocated AD 144 as the date of Kanishka. B N Mukherjee113 (1932-2013) was a dominant figure in Kushan studies and had attended the 1960 conference having been a student of A L Basham at SOAS, his research encompassed all parts of the field but particularly numismatics and epigraphy. He was throughout his career a staunch advocate of AD78. A K Narain114 (1925-2013) was the energetic force behind the original conference and remained interested in problems of Kushan studies throughout his long career, he continued to favour a compromise proposal of AD 100. Baij Nath Puri has been head of the Department of Ancient Indian History and Culture at Lucknow University and president of the Indian History Congress, he did not participate in 1960 but in a summary of the problem in 1999 favoured a date of AD 142 John Rosenfield (1924-2013) taught at Harvard and went on to be curator at the Fogg Art Museum, his principal contribution was the Dynastic Art of the Kushans in which he argued for a date of AD 110-115. Richard Salamon a noted epigrapher currently works at the University of Washington where he is head of the early Buddhist manuscript project, he accepted AD 78 until the publications by Harry Falk and now advocates a date of AD 127. R C Sharma was trained in Museology in France and went on to be director of the National Museum and the Bharat Kala Bhavan. He is particularly noted for his publications on the sculpture of Mathura and favoured a date of AD 78. 111 Acta Iranica vol.I 1979 p.232ff “ i i asa , D M I Me o ia : Joha a E. a Lohuize -de Leeuw (1919A ti us Asiae Vol.46 No.1/2: 149-153; 113 Chak a a ti, ‘. ‘e e e i g a E t a-La ge “ hola : B.N. Mukhe jee I dia Histo i al ‘e ie , , pp.151-154 114 http://en.wikipedia.org/wiki/Awadh_Kishore_Narain, consulted Mar 2015 112 Satya Shrava (1924-) published an important compilation of epigraphic material as well as a work on Kushan numismatics. Like many epigraphists he rejected the idea of two dating sequences and favoured AD 78. Nikolaus Schindel a numismatist working at the Austrian Academy of Sciences whose specialism is Kushan Sasanian and Sasanian coins is the sole remaining advocate for a third century date, which he places in AD 227, su se ue t to Ha Falk s pu li atio o the Yavanjataka. Nicholas Sims-Williams born in 1949 studied was a student of Ilya Gershevitch, amongst other works he made a huge contribution to the study of the Sogdian language. In the 1990s he was entrusted by the art collector David Khalili with the publication of the Bactrian documents and has gone on to become the foremost expert in their decipherment. For a full list of his publications see Sundermann et al. 2009. Evgeny V. Zejmal115 (1932-1998), head of the Oriental Department of the Hermitage Museum completed a PhD on the coins of the Kushan dynasty and went on to publish a number of important volumes on finds of the coins in Central Asia. He was the principal advocate in Russian for a third century date for Kanishka, AD 278, a position he adopted at the 1968 conference in Dushanbe. Nikitin, A. Shkoda, V. Alram, M. ‘In Memoriam Evgeny Vladislavovich Zeymal’ Bulletin of the Asia Institute Vol.11, 1997, pp.1-2. 115 Appendix 3 Synchronisms Eras Yona Era 116 [137] 201 [78]103 [250] [264] 122 136 [267] 139 279 284/7 [151] [156/9] 318 359 384 399?119 Earliest Yona date Earliest Azes date CKI459 Azes Era 9 73 [206]- [231] 303 Notes Reigns Azes I Vijayamitra (Apracaraja) Gondophares117 Inscriptions of a U - a ed Kushan King Kujula Kadphises CKI405, triple dated inscription allows conversion between Azes and Yona eras116 Later than Artabanus III (AD 27/8) Reign commences ← AD Latest Azes e a date118 Kanishka Era 2 – 23 26 – 60 64 – 98 Wima Takto Wima Kadphises Kanishka I Various, mostly private, donative inscriptions Second Kushan Sequence Huvishka Vasudeva Kanishka II 15121 Era of Bactrian letters Other synchronisms 20 to 30 Vasishka 41 Kanishka III Vasudeva II Earliest dated Bactrian letter Mahi Shaka AD 230 Embassy of Po tia120 Yavanajataka shows year 1 = Saka Era 149 = AD 227 Ardashir 226-241 credited by Tabari as conquering Bactria Shapur II 241-72 claims to rule Bactria Contemporary of Hormizd I Kushanshah Contemporary of Samudra Gupta (c.335-375) The thi d date, , is p esu a l Vija a it a s eig . Tha Takhti-i-Bahi inscription is dated in the era of Gondophares and also in the 26th year of his reign. 118 CKI#563, though there are inscriptions with higher dates thought to be dated in the Azes era this is the latest known inscription that uses the term. 119 CKI178, 111, 116, 124, 133, the Skarah Dehri inscription is not a standard dating formula with numerals and so remains somewhat doubtful. 120 The Memoirs of the Three Kingdoms, compiled in the third century records an embassy on 26 January AD230. It is not clear that Po-t iao is Vasude a ut this is a plausi le Chi ese t a slite atio of the Ba t ia ΒΑ ΟΔ Ο (Zurcher, 1968: 371). 121 Though various attributions of inscriptions to the period of Kanishka II have been made, the only inscription on which there is general agreement are those associated with the Jain Nun Vasula which do not mention the king. 116 117 110 Contemporary of Shapur II (AD 309379) Gupta Era 112 121 260 295 Gupta period Mathura Sequence122 5 15 68 478 525 549 Year 1 of this sequence = AD426 Letters mentioning Hephthalite tax and inscription mentioning Hunnic rulers First mention of family of Bek First mention of Arab poll tax Last Bactrian letter Cyclical era year 1 of each century corresponding to 24 or 25 of common era century. Laukika Era 32 38 Kipunadha Kumara Gupta Tochi valley inscriptions Falk, The Ka ishka E a i Gupta ‘e o ds also wants to attribute to a reckoning commencing in AD 326 two inscriptions with the years 61 and 70 from Mathura but while compelling this attribution is uncertain. 122 References Inscription numbers preceded CKI are references to the Corpus of Kharoshthi Inscription maintained at the Early Buddhist Manuscripts project http://gandhari.org/. Sh# refers to inscriptions in Shrava, Kushan Dated Inscriptions, MI# to inscriptions in Janert, Heinrich Lüders; Mathura Inscriptions. Allo , M. “alo o , ‘. Ja o se , G. )oppi U. ‘adio a o Dati g of Kha oṣtḥ ī f o the “ hø e a d “e io Ma us ipt Colle tio s i Je s B aa ig, ed, Buddhist Manuscripts Vol.III, Oslo, 2006, pp.279-292. Al a , M. I do-Pa thia a d Ea l Kusha Ch o olog : The Nu is ati E ide e i M. Alram and D.E. Klimburg-Salter, eds, Essays on the pre-Islamic History of the Indo-Iranian Borderlands, Vienna, 1999, pp.19- 48. Al a , M. Gö l, ‘o e t Encylopaedica Iranica, http://www.iranicaonline.org/articles/gbl-, 2001, updated Feb 2012, consulted Oct 2014. Aza pa , G. The Kusha Co fe e e i Dusha 257. e Archaeology vol.23 no.3, 1970, pp.254- Bailey, H. W. A Kharoṣṭ ī I s iptio of “eṇavarma, King of Oḍi The Journal of the Royal Asiatic Society of Great Britain and Ireland, 1980, no.1 p.21–9. Baile , H.W. The Dati g i the Ka pp.79. a Kha osthi I s iptio i Afgha “tudies ol. -4, 1982, Basha , A.L. A New Study of the Saka-Kusa a Pe iod Bulletin of the School of Oriental and African Studies vol.XV, no.1, 1953, pp.80-97 Basham, A.L. Papers on the Date of Kaniska, Leiden, 1968. Bau s, “. Catalog a d ‘e ised Te ts a d T a slatio s of Ga dha a ‘eli ua I s iptio s. n David Jongeward, Elizabeth Errington, Richard Salomon and Stefan Baums, Gandharan Buddhist Reliquaries. Gandharan Studies, Volume 1, Seattle, 2012 pp. 200–51. Bhandarkar, D.R. Corpus Inscription Indicarum Vol.III: Inscriptions of the Early Gupta Kings, New Delhi, 1981 Bi a , A.D.H. The Ka iṣka Dati g f o “u kh Kotal Bulletin of the School of Oriental and African Studies, vol.26, no.3, 1963, pp.498-502 [JSTOR: 611561] Bi a , A.D.H ‘e ie of the Date of Ka ishka Journal of the Royal Asiatic Society of Great Britain and Ireland, No. 2, 1971, pp. 192-193 Bivar, A.D.H. The Kuṣāṇa T ili gual Bulletin of the School of Oriental and African Studies vol. 39 no. 2, 1976, pp.333-340 Bi a , A.D.H. The Azes E a a d the I d a a Archaeology, Berlin 1979, pp.369-76. a Casket He e t Hä tel, ed, South Asian Bi a , A.D.H. ‘e ie : I dia a d the Helle isti Wo ld Klaus Ka ttu e Bulletin of the School of Oriental and African Studies vol. 62 no. 2, 1999, pp.366-367 Bi a , A D H A Cu e t Positio o “o e Ce t al a d “outh Asia Ch o ologies Bulletin of the Asia Institute Vol.14 pp.69-75 Bi a , A.D.H. B idge, M.C. T ee ‘i gs fo Ka ishka South Asian Studies vol.5 no.1, 1989, pp.149-56 B a e , ‘. A ote o the “ae ulu Dec 2014. a d the Kusha E a , www.kushan.org, 2005, consulted B a e , ‘. Odd , W.A. The a al sis of Kusha pe iod gold oi s Gandharan Studies vol.4, 2010, pp.31-8 “pe ifi G a it B a e , ‘. Ka kali Tila a d Kusha Ch o olog i “. Bha da e a d “. Ga g, eds, Felicitas: Essays in Numismatics, Epigraphy & History in Honour of Joe Cribb, Mumbai, 2011, pp.65-79 B a e , ‘. The Coi age of Wi a Kadphises i Ga dhā a “tudies, ol. , : -74 Bracey, R. The Geography of the Kushans, http://www.kushan.org/sources/coin/, consulted Mar 2015. B a e , ‘ Gold Coi Hoa ds http://www.kushan.org/sources/coin/goldhoards.htm, consulted Jun 2015. Brough, J. Co e ts o Thi d-Century Shan-“ha a d the Histo of Buddhis the School of Oriental and African Studies, vol.28 No.3, 1965, pp.582-612 i Bulletin of B ough, J. “upple e ta Notes o Thi d-Century Shan-“ha , Bulletin of the School of Oriental and African Studies, Vol. 33, No. 1, In Honour of Sir Harold Bailey, 1970, pp. 39-45. Bu ett, A. Joe C i : a Eulo-joe- i “. Bha da e a d “. Ga g, eds, Felicitas: Essays in Numismatics, Epigraphy & History in Honour of Joe Cribb, Mumbai, 2011, pp.25-34 Casson, L. The Periplus Maris Erythraei, Princeton, 1989. Coo e as a , A.K. The O igi of the Buddha I age , The Art Bulletin, Vol.9, No.4, June 1927, pp.287-329. Cunningham, A. Coins of the Kushanas or Great Yue-t hi Numismatic Chronicle, 1892, pp.4082 98-159. C i , J. Ga dha a Hoa ds of Kusha o-“asa ia a d Late Kusha Coppe Coin Hoards vol.VI, 1981, pp.93-10. Ci , J. I itatio s of Kusha Coppe Coi s Coin Hoards vol.VI, 1981, pp.301-305. C i , J. The Sino-Kharosthi Coins of Khotan: Their Attribution and Relevance to Kushan Ch o olog : PA‘T Numismatic Chronicle vol. 144, 1984, pp. 128-152. C i , J. The “i o-Kharosthi Coins of Khotan: Their Attribution and Relevance to Kushan Ch o olog : PA‘T Numismatic Chronicle vol. 145, 1985, pp. 136-149. C i , J. Nu is ati E ide e fo Kusha o-“asa ia Ch o olog Studia Iranica vol.19 pt.2, 1990, pp.151-193. C i , J. The He aus oi s: thei att i utio to the Kusha ki g Kujula Kapdhises, .AD 80, in M.Price, A.Burnett, R.Bland, eds, Essays in Honour of Robert Carson and Kenneth Jenkins, London, 1993, pp.107-134. - C i , J. A e Ba t ia i s iptio of Ka ishka the G eat, pt. : The ‘a atak I s iptio , Its Historical Implication and Numismatic Context Silk Road Art and Archaeology vol.4, 1996, pp.98–142. Cribb, J. 'The Early Kushan Kings: New Evidence for Chronology' in M. Alram and D.E. Klimburg-Salter, eds, Coins, Art, and Chronology: Essays in the Pre-Islamic History of the IndoIranian Borderlands, Vienna, 1999. pp.177-206. C i , J. The G eek Ki gdo of Ba t ia, Its Coi age a d its ollapse i Os u d Bopea a h hi and Marie-F a çoise Boussa Afgha ista A ie Ca efou e t e l Est et ; Ouest, Tu hout, 2005, pp.207-226. C i , J. ‘edis o e i g the Kusha s i Eliza eth E i gto a d Vesta “a khosh Cu tis, eds, From Persepolis to the Punjab: Exploring ancient Iran, Afghanistan and Pakistan, London, 2007, pp.179-210. C i , J. Odd , W.A. De ase e t a d se ue e of late Kusha a old oi s i A.K. Jha and S. Garg, eds, Ex Moneta, Essays on Numismatics, History and Archaeology in Honour of Dr David W MacDowall, Nashik, 1998, pp.253-268. Cu i gha , A Late I do-“ thia s , Numismatic Chronicle, vol.xiii, 1893, pp.93-128, [subsequently reprinted as a monograph in 1895]. Czuma, S.J. Kushan Sculpture: Images from Early India, Cleveland, 1985. Daffi à, P Il Di attito “ulla Data Di Ka iṣka A p oposito di u li o e e te Oriente Moderno, Anno 52, Nr. 3/4 (Marzo-Aprile 1972), pp. 225-232. Daffinà, P. The Ha “hu His Yu Chua ‘e-t a slated T ou g Pao, Vol. pp.309-339 , o. / , , Damsteegt, Th. Epigraphical Hybrid Sanskrit: Its rise, spread, characteristics and relationships to Buddhist Hybrid Sanskrit: Its rise, spread, characteristics and relationships to Buddhist Hybrid Sanskrit, Leiden, 1978. Da i, A.H. ‘adio-Carbon Evidence of the Dati g of Ka ishka i i A.L:Basha , ed, Pape s o the Date of Ka iṣka, Leide , , pp.436-7. Do i s, K. W. Ga dha a A t f o “t atified E a atio s East and West, Vol. 23, No. 3/4 (September-December 1973), pp. 279-294 Dobbins, K. W. The Kamra Kharoṣṭhī Inscription of Vāsiṣka East and West Vol. 25, No. 1/2 (March-June 1975), pp. 105-109 Dovudi, D. Monetie Kladi Tadjikistana (1980-2008) Dushanbe, 2009. Engel, N. New excavations in Afghanistan: Mes Aynak, 2011 www.nationalmuseum.af/images/stories/PDF/mes%20aynak.pdf [consulted Dec 2014]. Errington, E. Cribb, J., eds, The crossroads of Asia – transformation in image and symbol, in the art of ancient Afghanistan and Pakistan, Cambridge, 1992. Falk, H. The yuga of Sphujiddhvaja and the era of the Kuṣâṇas Silk Road Art and Archaeology vol.7, 2001, pp.121–36. Falk, H. The Ka iṣka E a i Gupta ‘e o ds Silk Road Art and Archaeology vol.10, 2004, pp.167-76. Falk, H. A ie t I dia E as: A O e ie 46. Bulletin of the Asia Institute vol.21, 2007, pp.131- Falk, H. The Pious Do atio s of Wells i Ga dha a i E.J.‘. Me isse a d A. Ba e ji, Prajñadhara: Essays in Asian Art History Epigraphy and Culture in Honour of Gouriwar Bhattachary, New Delhi, 2009, pp.23-36 Falk, H. Be et, C. Ma edo ia I te ala vol.70, 2009, pp.197-216 Mo ths a d the E a of Azes Acta Orientalia R. N. Frye The History of Ancient Iran. Handbuch der Altertumswissenschaft ; 3. Abt., T. 7 (1984) Fuss a , G. Do u e ts pig aphi ues Kou ha s Bulleti de l E ole F a çaise d E t e e Orient vol.61, 1974, pp.1-79 Fussman, G. Chronique des études kouchanes (1975pp. 419-36. ) Journal Asiatique, vol.266, 1978, Fuss a , G. Do u e ts pig aphi ues kou ha s II Bulleti de l É ole f a çaise d E t Orient vol.67, 1980, pp.45–58 Fussman, G., Chronique des études kouchanes (1978pp.333-400 e- ) Jou al Asiati ue, vol.275, 1987, Fuss a , G. Co t i utio s des sa a ts f a çais au etudes kou ha es in G.N.Sear, ed, The Collection of Essays Read out in the fifth International Seminar of Kushan Studies, Kabul 8-15 November 1983, Kabul, 1987, pp.40-46 Fuss a , G. A de isi e stage i the stud of Kusha a oi age Numismatic Digest Vol.14, 1990, pp.73-101 Gö l, ‘. Die Mü zp agu g de Kušā o Vi a Kadphises is Bah a IV i F. Althei Stiehl, Finanzgeschichte der Spätantike, Frankfurt, 1957, pp. 173-256 a d ‘. Gö l, ‘ ) ei eue Te i i fü ei )e t ales Datu de Alte Ges hi hte Mittelasie s, das Jahr I des Kušā Kö igs Ka iška i Anzeiger der phillist. Klassed der Osterreicheschen Akademie der Wissenschaften, So.7, 1964, Vienna, pp.149-51 Göbl, ‘ Nu is ati E ide e ‘elati g to the Date of Ka iṣka i A.L.Basha , ed, Papers on the Date of Kanishka, Leiden, 1968, p.103-113. Gö l, ‘ The ‘o a -Kusha ia Medaillo i the B itish Museu Numismatic Society of India, 1976, vol.xxxviii, pp.21-26 i Journal of the Göbl, R. S ste u de Ch o ologie de Mü zpg ägu g des Kuša ei hes, Vienna, 1984 Göbl, R. Donu 1993. Bu s: Die Kušā ü ze i Mü zka i ett Be u d die Ch o ologie, Vienna, Gorin, A N (2011) 'Klad Mednyx Monet Pozdnekushanskogo Vremeni iz Iuzhnogo Uzbekistana (A late Kushan period hoard of copper coins from Southern Uzbekistan)' Scripta Antiqua, Vol.I, Russian Academy of Sciences, pp.369-405. Gos a i, K.G. Kosa I s iptio of the eig of Ka ishka: Yea 1937-38, pp.210-212. Epigraphica Indica Vol.24, Gupta, P.L. 'Eastern Expansion of the Kushan Empire' Indian Historical Quarterly vol.XXIX no.3, 1953, pp.205-221 [reprinted in Gupta & Kulashreshtha, 1994] Hill, J. E. Through the Jade Gate to Rome: A Study of the Silk Routes During the Later Han Dynasty 1st to 2nd Century CE, 2009 Hulsewe, A. F. P. China in Central Asia, the Early Stage: 125 B. C. -A. D. 23. An Annotated Translation of Chapters 61 and 96 of the History of the Former Han Dynasty, Sinica Leidensia, vol. XIV, Leiden, 1979 I e , “.“. T o B ah i I s iptio s f o Mathu a Epigraphia Indica 40, 1973-74, pp.19-22. Janert, K.L., ed, Hei i h Lüde s; Mathu ā I s iptio s: U pu lished Pape s, Göttingen, 1961. Khan, G. Studies in the Khalili Collection Volume V: Arabic Documents, London, 2007. Kha , G. ‘. Coppe oi s of Vasude a a d “u esso s f o vol.4, pp.39-161 Ta ila Gandharan Studies Kha , M.N. B āh ī i s iptio of the ea f o “pi a, “outh-Wazi ista , Pakista Indo Kōko Ke k ū No.22, Indian Archaeological Society, Tokyo, 2001, pp.89-93 Loe e, M. Chi ese ‘elatio s ith Ce t al Asia, - Bulletin of the School of Oriental and African Studies, University of London, Vol. 32, No. 1, 1969, pp. 91-103. Lohuizen-de-Leeuw, J.E. van, The Scythian Period: An Approach to the History, Art, Epigraphy and Palaeography of North India from the 1st Century BC to the 3rd Century AD, Leiden, 1949. Lohuizen-de-Leeu , J.E. a , Ne E ide e ith ‘ega d to the O igi of the Buddha I age South Asian Archaeology, ed. Herbert Härtel, Berlin (Reimer), 1979, pp.376-400 Lohuizen-de-Leeu , J.E. a , The “e o d Ce tu vol.2, 1986, pp.1-9. of the Ka ishka E a South Asian Studies Ma Do all, D.W. “ote Megas, the Ki g of Ki gs, the Kusha a Journal of the Numismatic Society of India, vol.XXX, 1968, pp.28-55 Ma Do all, D.W. The ‘efe e es to the Kuṣā ̣ as i the Pe iplus a d fu the numismatic e ide e fo its date Numismatic Chronicle 7th Ser. vol.10, 1970, pp.221-240. MacDowall, D.W. I pli atio s fo Kusha a Ch o olog of the Nu is ati Co te t of the Na eless Ki g Journal of the Numismatic Society of India vol.XXXVII, 1975, pp.44-71. Ma Do all, D W Be ell, A “ “hah azi A “h. A D H Bi a Bulleti of the Asia I stitute: Iranian Studies in Honour of A D H Bivar Vol.7: 1-8 Ma shall, J.H. The Date of Ka ishka , Journal of the Royal Asiatic Society, pt. 2, 1914, p.973997. Melze , G. “a de , L. A Coppe “ oll I s iptio f o the Ti e of the Al ho Hu s i J.Braarvig, ed, Manuscripts in the Schøyen Collection vol.3, Oslo, pp.251-279. Mi ashi, V.V. Epig aphi al Note Epigraphia Indica vol.37, 1967-8, pp.70-3. Mitchiner, M. Indo-Greek and Indo-Scythian Coinage Vol.8: The Indo-Parthians, London, 1976. Mitte all e , G. . P o le s of Ch o olog i the A t of the Kusa a Pe iod u pu lished paper given at I te atio al Se i a o the A t of the Kuṣā ̣ a age, De e e -22, 1984, State Museum Lucknow, pp.94-122, consulted at the IIRNS Nasik Feb 2015 Mitterwallner, G.v. Kuṣā ̣ a Coi s a d Sulptu es, Mathura, 1986. Mukhe jee, B. N. Ka 1973, pp.111–8. a I s iptio of Vajheshka vāsishka Indian Museum Bulletin vol.8, Mukherjee, B. N. The Ya a ajātaka of “phujidh aja, the ”akakāla a d the Ka ishka E a Journal of the American Oriental Society vol.102 no.2, 1982, pp.361-64. [JSTOR 602536] Nash, “.E. A haeologi al T ee-‘i g Dati g at the Mille iu Research vol.10 no.3, 2002, pp.243-75. Journal of Archaeological Nikitin, A. Shkoda, V. Alram, M. ‘In Memoriam Evgeny Vladislavovich Zeymal’ Bulletin of the Asia Institute Vol.11, 1997, pp.1-2. O B ie , A.G. The Ancient Chronology of Thar: The Bhāttika, Laukika, and Sindh eras, Delhi, 1996. Odd , W A. C i J. De ase e t a d “e ue e of late Kusha Gold Coi s i A al K. Jha a d Sanjay Garg, eds, Ex Moneta Essays on Numismatics in Honour of Dr David MacDowall, New Delhi, 1998 pp.275-292 Pi g ee, D. The Ya a ajātaka of “phujidh aja Journal of Oriental Research, Madras vol.31, 1961-2, pp.16-31. Pingree, D. The Yava ajātaka of Sphujidhvaja, 2 Vols, Cambridge, 1978. Plaeschke, H. Die Mathu ā-S hule, ei paläog aphis he u d ku sthisto is he Beit ag zu Lösu g des Ka iṣka – Problems, Halle (Saale), 1971 (unpublished dissertation) Puri, B.N. Kusana Bibliography, Calcutta, 1977. Pu i, B.N. The Kusha s i J.Ha 1994, pp.247-264. atta, ed, History of Civilizations of Central Asia Vol.II, Paris, Quintanilla, R.S. History of Early Stone Sculpture at Mathura, Leiden, 2007 ‘ajgo , D. A hoa d of late Kusha Coppe s f o Bhuj I dia Actes du XIe Congres International de Numismatique, vol.III, Brussels, 1991, pp.233-37. ‘a e , E. Gupta Gold Coi s ith a Ga uḍ a-Banner, Groningen, 1994 Ro he , L. ‘e ie : The Ya a ajataka of “phujidh aja 1980, pp. 173-174. Da id Pi g ee Isis vol. 71 no. 1, Rosenfield, J. Dynastic Art of the Kushans, Berkeley, 1967. Salamon, R. Indian Epigraphy: A Guide to the Study of Inscriptions in Sanskrit, Prakrit, and the Other Indo-Aryan Languages, Oxford, 1998. Salomon, R. 'Three Kharoshti Reliquary Inscriptions' Silk Road Art and Archaeology vol.9, 2003, pp.39-69. “alo o , ‘. The I do-G eek E a of / B.C. i a Buddhist ‘eli ua I s iptio . I Os u d Bopearachchi and Marie-Françoise Boussac, eds, Afgha ista : a ie a efou e t e l Est et l Ouest: a tes du ollo ue i te atio al o ga is pa Ch istia La des & Os u d Bopearachchi au Musée archéologique Henri-Prades-Lattes du 5 au 7 mai 2003. Indicopleustoi: archaeologies of the Indian Ocean 3, 2005 Turnhout, pp. 359–401. “alo o , ‘. “ hope , G. The I d a a a A a a Casket I s iptio ‘e o side ed: Fu the Evidence for Canoni al Passages i Buddhist I s iptio s The Journal of the International Association of Buddhist Studies, vol.7, no.1, 1984, pp.107-124. “ hi del, N A dashi II Kusha shah a d Hu ishka the Kusha : Nu is ati E ide e fo the date of the Kushan King, Kanishka I JONS 189, 2009, pp.12-14. “ hi del, N A dashi I Kusha shah a d Vasude a the Kusha : Nu is ati E id e e fo the Date of the Kusha Ki g, Ka ishka I JONS 220, 2014, pp.27-30 “ hope , G The I s iptio o the Kuṣā I age of A itā ha a d the Cha a te of the Ea l Mahā ā a i I dia Jou al of the I te atio al Asso iatio of Buddhist “tudies . , , pp.99[ ep i ted i “ hope , G Fig e ts a d F ag e ts of Mahā ā a Buddhism in India: More Collected Papers, Honolulu, 1987, pp.247-277 references are to the reprint] “e io , ‘.C. The I do-Greek and Indo-Scythian King Sequences in the Second and First Ce tu BC , “upple e t to Journal of the Oriental Numismatic Society no.179, 2004. Senior, R.C. A Catalogue of Indo Scythian Coins, 4 Volumes, 2000-2006 “ha a, ‘.C. Ne I s iptio s f o Heritage, 1989, pp.308-315 Mathu a i “ i i asa , D. Mathura: The Cultural Sharma, R.C. Buddhist Art: The Mathura School, New Delhi, 1995. Shrava, S. The Kushā a Nu is ati s, New Delhi, 1985. Shrava, S. Kushan Dated Inscriptions, New Delhi, 1993. Simonetta, A. ‘An Essay on the So Called "Indo-Greek" Coinage’ East and West, Vol. 8, No. 1 (APRIL 1957), pp. 44-66. Sims-Willia s, N. Ba t ia La guage E clopaedia Iranica http://www.iranicaonline.org/articles/bactrian-language, 1988, last updated 2011, consulted Mar 2015. Sims–Willia s, N. A e Ba t ia i s iptio of Ka ishka the G eat pt. : The ‘a atak I s iptio , Te t a d Co e ta . Silk Road Art and Archaeology vol.4, 1996, pp.77–97. Sims-Willia s, N. Fu the otes o the Ba t ia i s iptio of ‘a atak i Ni holas “i sWilliams, ed, Proceedings of the Third European Conference of Iranian Studies, Part 1: Old and Middle Iranian Studies, Wiesbaden, 1998, pp.79–92. Sims–Willia s, N. The Ba t ia i s iptio of ‘a atak: a e Institute. New Series vol.18, 2004, pp.53–68. eadi g. Bulletin of the Asia Sims-Williams, N. Studies in the Khalili Collection: Bactrian Documents, 3 volumes, London, 2000-2007. Sims-Willia s, N Ba t ia Histo i al I s iptio s of the Kusha Pe iod Silk Road Foundation Newsletter No.10 http://www.silkroadfoundation.org/newsletter/vol10/SilkRoad_10_2012_simswilliams.pdf, 2012 consulted Apr 2015 Singh, U. A History of Ancient and Eary Medieval India: From the Stone Age to the 12th Century, Pearson-Longman, Delhi, 2008. Singh, S P 'Numismatic Evidence of Kushana-Murunda Rule in Eastern India' JNSI XXXV, 1973, pp.127-41123. “i a , D. The Kusha as i ‘.C.Maju da , ed, The Age of Imperial Unity, Bombay, 1951, pp.136-153. “ope , A. ‘e e t “tudies i ol i g the Date of Ka iṣka Artibus Asiae vol.34 no.1, pp.102-5 & 107-13. Srivistava, A K. Coin Hoards from Uttar Pradesh, The State Museum Lucknow, 1980. Stavisky, D.Y. Vainberg, B.I. Gorbunova N.G. Novgorodova, E.A. Soviet Central Asian Archaeology and the Kushan Problem: An Annotated Bibliography, Moscow 1968. Sundermann, W. Hintze, A. de Blois, F. Exegisti monumenta: Festschrift in Honour of Nicholas Sims-Williams, Wiesbaden, 2009. 123 This should not be confused with a paper of similar title Gupta and Kulashrestha Kusana Coins and History which claims to be a reprint of a paper in JNSI XXXVIII but does not in fact appear in that volume. Thie , F. Yuezhi et Kou ha s Pièges et Da ge s des “ou es Chi oises in Osmund Bopearachchi and Marie-F a çoise Boussa Afgha ista A ie Ca efou e t e l Est et ; Ouest, Tu hout, , p.421-539 Whitehead, ‘.B. The Fi st Co fe e e o the Date of Ka iṣka i A.L:Basha , ed, Pape s o the Date of Ka iṣka, Leide , Wide a , F. U e su f appe de Go dopha es su He aios et u e aut e de Kozoulo Kadphises sur Gondophares qui apportent deux jalons numismatiques à la chronologie entre les Indo-G e s et le d ut de l e pi e Kou ha Bulletin de la Société Française de Numismatique vol.27 no.1, 1972, pp..147-151 Willia s, J.G. The Case of the O itted Hu d eds: “t listi De elop e t i Mathu ā “ ulptu e of the Kuṣā ̣ a Pe iod i Do is “ i i asa , ed, Mathura: The Cultural Heritage, New Delhi, 1989: 326-331 )e al, E V Na hal a a data Ka ishki – 278 g. n. Central Asia in the Kushan Period, Vol.1 pp.292-301. Zeymal, E. V. D ev ie o et Tadžikista , Dushanbe, 1983 )e al, E.V. Till a-Tepe ithi the o te t of the Kusha Ch o olog i M. Al a Klimburg-Salter Coins, Art, and Chronology Vol.I, Vienna, 1999, pp.239-244 Zurcher, E [full reference needed] a d D.