ãëàâíàÿ ñòðàíèöà / áèáëèîòåêà / îãëàâëåíèå êíèãè

Å.Â. Çåéìàëü

Êóøàíñêàÿ õðîíîëîãèÿ (ìàòåðèàëû ïî ïðîáëåìå)

// Ìåæäóíàðîäíàÿ êîíô. ïî èñòîðèè, àðõåîëîãèè è êóëüòóðå Öåíòðàëüíîé Àçèè â êóøàíñêóþ ýïîõó (Äóøàíáå, 1968). Ì., «Íàóêà», ÃÐÂË. 1968. 186 ñ.


E.V. Zeymal

Kushan Chronology: Review of Data Bearing on the Problem

// International conference on the history, archeaology and culture of Central Asia in the Kushan period (Dushanbe 1968). M.: 1968. 186 p.



(.pdf, Summary ïîëíîñòüþ, 5 Ìá)




1. It is perfectly clear that a solution of the problem of Kushan chronology will ultimately depend on the authenticity and reliability of the argumenta adduced rather than on their number. Yet authentic facts and conjectures are so closely interlocked in the literature on the Kushan kingdom that surmises and multi-stage assumptions as a rule obscure the existing evidence. The data cited in connection with the problem of Kushan chronology are far from always subjected to a proper analysis, based on the study of source materials. In substantiating some hypothesis, researchers often cite not only debatable evidence (whose use is permissible but calls for interpretation) but also doubtful evidence (from the point of view of its authenticity) and even material which has no logical bearing on the problem (and certainly cannot help to solve it). The evidence usually drawn upon in connection with the problem is discussed in this paper under the following angle: whether its employment in determining the absolute and relative dates of Kushan history is justified. Information on events in Kushan history (ch. I) is considered in the light of establishing their relative chronology. Events in the history of other countries which have, or may have, absolute dates and can be linked, with a greater or lesser degree of certainty, with some points of Kushan relative chronology are discussed according to regional time-scales: ch. 2, India; ch. III, China; ch. IV, Iran; ch. V, Rome.



2. Fluctuations la Kushan absolute chronology, admissible under the existing level of our knowledge of the problem, are far from limitless. In establishing them one naturally has to proceed not from the existing hypotheses but from the totality of facts known today. «Hou-Han Shu» («History of the Later Han»), which covers events from A.D. 25 to 221, tells us that Ch'lu-chiu ch'üeh (Kujula Kadphises) became the king of the Kuei-shuang (Kushan) state (see § 4). Comparison with an earlier chronicle, «Han-Shu» — «History of the (Former) Han» — which brings events up to A.D. 25 but fails to mention this fact, shows that the Kushan kingdom did not exist before A.D, 25. If it is to be assumed that the authors of «Han-Shu» (second half of the 1st century A.D.) did not yet know of the Kushan kingdom though it existed, then the time limit prior to which the Kushan state could not have existed, may be shifted to the threshold of the new ara or even further back. But there is no direct evidence (i.e., evidence which does not call for supposition) in favour of a date earlier than A.D. 25.


The limit after which the existence of the Kushan kingdom would be impossible is the last quarter of the 4th century A.D. In the eastern part of the Kushan kingdom, there is the Mathura inscription of Candragupta II, dated in the 61st year of the Gupta era (A.D. 380/381). In the west of the kingdom, the Sassanian governors issued coins bearing the titles «kwš'n MLK'» and «wozorko košono šoho»; a comparison with coins of Shapur II (309-379) places the beginning of their emission in the middle or in the third quarter of the 4th century A.D. Thus, the possible limits for the existence of the Kushan kingdom acceptable today are A.D. 25 to the 70's or 80's of the 4th century.


3. One can approach the necessary degree of objectivity essential in an analysis of material on Kushan chronology if the authentic elements (deducted directly from sources) and hypothetical elements (resulting from interpretation) are clearly delimited. Such a demarcation under the present level of knowledge of the existing data has enabled


me, in offering this compendium, to keep to my own conclusions, to take them into account along with other individual points of view. Specifically, the year 278 — which I regard as the most probable for the initial Date of Kanishka (DK) — is considered in this paper on a par with the other dates suggested. I hope that my opinion has not been given undue prominence, and, which is still more important, that it has not affected this analysis of objective data. Another unavoidably subjective feature of the present compendium is due to the fact that, while having a free command of numismatic and partly archeological material, I could not dispose as freely of the other necessary sources. This accounts for my paper being a piece of reference rather than research: a fall study of the source material available on the problem of Kushan chronology can be accomplished only by a large group of researchers specialising in different fields. Of great benefit to me in compiling the present work was the expert advice of K.V. Vasilyev, M.I. Yorobyova-Desyatovakaya, V.A. Livshits and B.A. Litvinsky — and especially of V.G. Lukonin, whose material, partly unpublished, forms the substance of Ch. IV.


Apart from the possible chance omissions or gaps due to lack of information, it was impossible to take into account all archeological evidence, since the available publications (particularly of 19th-century excavations) do not always give a clear idea of archeological complexes accompanying one or another find, of the stratigraphy, etc. As a compendium on the problem of Kushan chrolology, this work does not pursue historiographic aims, and this accounts for the fact that the bibliographical references do not include all the existing literature on the subject.


Ch. I. Relative Chronology.


Different kinds of sources giving an insight into the relative dates of Kushan history are of different value but they complement one another. Inscriptions, containing as they do dates and mention of Kushan kings, give a specific quantitative scale of relative chronology; information gleaned from other sources, above all the coinage of the Kushan


kings and references in the records, should be brought in correspondence with the inscriptions. With the help of numismatic material, these data provide the groundwork for appraising archaological finds over a vast territory from the Aral Sea to the Indian Ocean. The dated inscriptions also provide the groundwork for a relative chronology of art relies.


4. The only direct evidence of the rise of the Kushan kingdom is a fragment from «Hou-Han Shu», Ch. 118. When the Yueh-chih had moved to Ta-hsia — Bactria, «more than a hundred years later, the yabgu of Kuei-shuang (named) Ch'in-chiu-ch’üeh [Kujula Kadphises] attacked and destroyed the (other) four yabgu [of the YUeh-chih] and established himself as (their) king; the kingdom was named Kuei-shuang. (This) king invaded An-hsi, took the country of Kao-fu, and, moreover, destroyed P'u-ta and Chi-pin and completely possessed their territory. Ch'iu-chiu-ch’üeh died at the age of more than eighty years, and his son Yen-kao-chen [=Vima Kadphises] succeeded him as king. He in his turn destroyed T'ien-chu (India) and placed there one general to control it. Since then the Yüeh-chih have been extremely rich and strong. In the various (Western) countries (their ruler) is always referred to as 'the King of Kuei-shuang', but the Han, basing themselves upon the old appellation, speak about 'the Great Yüeh-chih'» (Zürcher 1960, p. 5). This fragment, giving a detailed sequence of events, conveys but a hazy idea of the periods of time between the events described. Particularly vague is the expression «more than a hundred years» (cf. Pulleyblank 1960, p. 2.). Ch'iu-chiu-ch’üeh who died at the age of «more than eighty years» may hare established himself as king when he was 20 — or 60, for that matter. This precludes us from establishing with certainty the duration of his rule as king of the Kushan.


5. Inscriptions with dates of the «unknown era» (over 40 Inscriptions with dates from 58 to 399 of that era; referred to below as «U») mention Kushan kings in three cases


only. The inscription U=122 from Panjtār and U=136 from Dharmarajlka give only of the king's title, without his name. Inscription U=187 (184?) from Khalatse was made daring the rule «of the great king Uvima Kavthisa» (Vima Kadphises). The titles in inscriptions U=191, U=299 (or 199) and U=303 may also be interpreted as indirect allusions to Kushan kings. Thus, the inscriptions of the «unknown era» warrant the following statements:

= in 122 and 136 of the «unknown era» a Kushan king (presumably one king) was in power; we do not know his name;

= in 187 (or 184) of the «unknown era» the Kushan kingdom was ruled by Uvima Kavthisa (Vima Kadphises), A comparison with the text of «Hou-Han Shu» shows that inscriptions U-122 and U-136 might date from the time of the rule of Ch'ia-chiti-ch’üeh (Kujula Kadphises).


6. For a list of inscriptions referred to the «unknown era» see the Russian text.


7. That the inscriptions are dated in the «unknown era» is often a conjecture, for the inscriptions themselres give no indication to this effect. As a rule, one has to proceed from a collation of these inscriptions with others, placed in the «unknown era», using simultaneously different (but not always universal) criteria. Of prime importance is paleographic research, though, as convincingly demonstrated by A.H. Dani (1963), the different writing may have nothing to do with chronology, being due to the different material used and the specific features of the local styles and schools of scripts. In this context, even such inscriptions as U=103 and U=78, whose content is of great importance, are debatable. There is no irrefutable proof in favour of placing them in the «unknown era»; neither are there any sufficiently convincing objective data (independent of the accepted version of chronology) to the contrary.

8. Inscriptions carrying dates of the 4th century of the «unknown era» are directly linked with art relics and to some extent they give a clue to the chronology of Gandhara art. The range of hypothetical absolute dates for these relics (see p. 28) demonstrates the unreliability of dating according to style features.



9. The name of another era, whose use in the territory of the Kushan .kingdom is attested by inscriptions — «the Kanishka era» — is likewise conventional, inasmuch as the inscriptions do not define it. Still, the names of Kushan kings and dates occurring in the inscriptions make it possible to place certain inscriptions in this era with greater accuracy than in the case of the «unknown era». Doubts may arise only in cases when an inscription bears a date but omits the name of the king or his title. Beginning with the 1st (or 2nd) and until the 23rd year of «the Kanishka era» the inscriptions speak of King Kanishka (I); from 24 to 28, Vasishka; from 28 to 60, Havishka) and from 64 to 98, Vasudeva. Moreover, the inscription from Ara (K=41) speaks of King «Kanishka, son of Vadjeshka» (Vasishka) — possibly Kanishka II — whose relationship with Huvishka is unclear. The assumption that part of the inscriptions dated in the «Kanishka era» belong to «the 2nd century» of this era (with the figures denoting hundreds omitted) has been based on shaky paleographic arguments (for counter-arguments, cf. Dani, 1963, p. 86) and is hardly acceptable.


10. The list of inscriptions dated according to the «Kanishka era» (see the Russian text), includes more than 130 inscriptions in Brāhmī (they are in the majority) and Kharoshthī, and one Bactrian inscription K=31c from Surkh Kotal. The list also includes inscriptions which may be placed in the «Kanishka era» according to some indirect indication (i.e., inscriptions which fail to mention Kanishka or his successors).


11. Numismatic data on Kushan relative chronology make it impossible, at the present level of our knowledge, to accept the hypothesis that another king (or even two kings) ruled between Kujula Kadphises and Vima Kadphises. The typological gap between the coins of the founder of the Kushan kingdom and his successor is most convincingly (but again hypothetically) explained by the specific features of Kushan coinage which was taking shape at the time. The interregnum between Vima Kadphises and Kanishka (I), which was emphasi-


sed in the outline of relative Kushan chronology by R. Ghirshman (1946), is not borne out by numismatic data either. R. Göbl (1957, p. 186) showed that the images on the latest gold coins of Vima Kadphises and the earliest gold coins of Kanishka were engraved by the game engravers. Moreover, these coins have a stable common feature: the manner of writing the Greek legend (, but not ). The junction between Vima Kadphises and Kanishka completely upsets, from the point of view of relative chronology, the long-abandoned hypothesis of DK = B.C. 57. The reign of Vasishka (Inscriptions dated in 24-28 of the «Kanishka era») is not attested by coins. Moreover, the early gold coins of Huvishka are very closely linked with the latest emissions of Kanishka I (of. Göbl 1957, Nos. 51, 52 and 60, 61). «Kanishka, son of Vadjeshka» (Kanishka II) is likewise unrepresented by authentic numismatic data, but several groups of copper coins, usually ascribed to Kanishka I, might be associated with this king (Zeymal 1967, pp. 69-74). There have been attempts to distinguish, on the basis of numismatic data, coins of «Vasudeva I», «Vasudeva II» and «Vasudeva III» (Bachhoffer 1936, and others). Such a division is based only on individual and most distant (in time and style) specimens bearing the name of this king. However, if we take the entire coinage of this king, we can see the gradual accumulation, from one series to another, of the distinctions which mostly stemmed from the minting techniques (including the engraver's artistic proficiency and literacy). But the coins of «Vasudeva I», «Vasudeva II» and «Vasudeva III» do not evince any distinctions that could be linked with political changes; neither can any interruption in their emission be ascertained. Even if there were more than one Kushan king by the name of Vasudeva, these namesake kings must have succeeded one another: the coins give no grounds whatever for believing that their rule outlasted the year 98 of the «Kanishka era». The gold and copper coinage of Kanishka III (with a seated Ardokhsh on reverse) cannot be juxtaposed with the inscriptions since the latter make no mention of this king. The coins inducate that the emissions of Vasudeva


and Kanishka in were roughly contemporaneous. It has been suggested that in that period tht Kushan kingdom divided into two parts, each of which was ruled by its own king.


12. R. Göbl offered a more intricate outline of Kushan relative chronology, based on the assumption that the Kushan kingdom had the institution of «senior» and «junior» kings, analogous to the Roman «Sĕnior Augustus» and «Jūnior Augustus». The rule of Kanishka, then, lasted from 1 to 41 of the «Kanishka era»; Vasishka (24-26) and Huvishka (28-40) ruled at that time as «junior kings». From the year 41 of the «Kanishka era» Huvishka became «senior king»; after that, Göbl's outline is practically the same. Göbl's assumption explains the absence of the coinage of Vasishka — the «junior king» (although Huvishka as a «junior king» coined gold and silver with «Kanishka's symbol») and the mention of Kanishka'a name in the Ara inscription of the year 41. But if Göbl's outline is consistently applied not only to these two facts, bat to all data available on the Kushan kings, it cannot be accepted without «trimming» many facts to fit it with the help of multistage assumptions.


13. Archeological data, in the current stage of publication, can be used for establishing Kushan chronology to a very limited degree. Coins remain the main item for dating all the other objects found in layers and complexes. Even the dating of such a remarkable monument as Surkh Kotal, uncovered comparatively recently and strictly according to modern methods, fully depends on inscriptions and coins. The same is true, to a still greater degree, of other sites excavated on the territory of the Kushan kingdom. Only excavation of whole large sites and a comparison of a mass of material from different sites (given impeccable stratigraphy) will probably help to tip the scales in favour of archeology. To evolve a summarising relative chronology of archeological material it is quite important to study the stratigraphic location of the coins discovered. Such work has been successfully launched in India (S. Ray 1959) but it is essential to have similar (or more detailed) researches covering other parts of the Kushan kingdom.



14. In connection with the above surrey of material on the relative chronology of the Kushan kingdom it is necessary to discuss the hypothesis according to which the «Kanishka era» ia the 3rd century of the «unknown era» (with the symbols of hundreds omitted). Advanced by D.R. Bhandarkar (1900) for DK = A.D. 278, this hypothesis was subsequently adopted by R. Ghirshman for DK = A.D. 144 (1946, 1957) and van Lohuizen de Leeuw: DK = A.D. 78 (1949). This hypothesis is supported by the following arguments: 1) the inscription of Vima Kadphises is dated by 187 (or 184) of the «unknown era»; 2) the coins show that Kanishka come into power immediately after Vima Kadphises, without any interregnum, though we do not know how many years after 187 of the «unknown era»; 3) having at our disposal a large number of inscriptions of the 1st, 2nd and 4th centuries of the «unknown era», we have hardly any inscriptions of the 3rd century of this era (with the exception of U=299 — after the last inscription bearing the name of Vasudeva and dated in 98 of the «Kanishka era»); 4) inscriptions of the «Kanishka era», dated from 1 to 98, make up a single group, which corresponds to the period 187 of the «unknown era» + X years, and which does not exceed the gap in the dates of the «unknown era», of which we are aware (200 to 299). Historically, the return to the old chronology after the end of the dynasty in 98 of the «Kanishka era», is fully justified. To accept the hypothesis that «the Kanishka era» corresponds to the 3rd century of the «unknown era», one has to assume that the rule of Vima Kadphises continued after 187 (or 184) of the «unknown era» till 200. At the current stage of our knowledge (even if this assumption proves incorrect), the risk is arithmetical rather than historical.


Ch. II. Indian Synchronisms.


Only a few dates in the history of India in the first centuries A. D. can be brought in correspondence with the existing chronology. As a rule, these dates occur in insc-


riptions or on coins and have as their starting-points such eras whose existence is attested by later records (the Vikrama era, the Śaka era). The puranical and the Buddhist tradition is far less reliable in this respect.


15. Reckonings used in inscriptions of the Kushan period may be tentatively identified with certain historical eras (i.e., systems of chronology that existed in reality and survived up to the Middle Ages). This way of solving the problem of Kushan chronology gives no direct sychronisms (neither does it give a final answer to the question of «the date of Kanishka»), but it makes it possible to ascertain the probability — or improbability — of one or another variant of identification along a whole chain of hypothetical synchronisms that arise in the process and are sometimes not connected directly with the history of the Kushan kingdom. Identification with historical eras evidently can be regarded as an independent argument, in contradistinction to identification with the «reconstructed» eras, i.e., speculatively selected points for counting off the «unknown era» and the «Kanishka era». «Reconstructed» eras are not considered in this compendium: this task involves historiographic research rather than the study of sources.


16. The Vikrama era (according to al-Bīrūnī, 1088 of this era corresponds to 400 of the «Yazdagird era») was counted from 58/57 B.C. Mention of this era, under the name «Vikrama-Samvat» or «Vikrama-kāla», does not occur before the 9th century. It is assumed that before this it was known as «Kŗta» and «Mālava». Medieval tradition ascribes the founding of that era to King Vikramāditya of Ujjayini, who inaugurated it on the occasion of victory over the Sakas. This legendary king may be associated only with Candragupta II (A. D. 376-414), who was the first among the historical personages we know to bear the title of Vikramāditya. He also terminated the rule of the Western Kshatrapa dynasty which had lasted for more than 300 years. Since there are no undoubted cases of the use of the Vikrama era before the 4th century, we cannot rule out the possibility


that before Candragupta II this era had not existed; what is more, an ancient date may hare been selected retrospectively as the starting-point for reckoning — a date that was not contemporaneous with the time of the introduction of the reckoning. This circumstance minimises the possibility of associating the Vikrama era with the Kushan eras. The suggested identification of the «Kanishka era» with the Vikrama era is at variance with the objective data not only on absolute but also on relative Kushan chronology. Identification of the Vikrama era with the «unknown era», and the Śaka era with the «Kanishka era» had been suggested by the most consistent supporters of the identification of Kushan eras with Indian historical eras. But in this case, 135-200 of the «unknown era» would have coincided with 1-65 of the «Kanishka era», and the inscription of 187 (or 184) of the «unknown era», which makes mention of Vima Kadphises as king, would have to be dated in 53 (or 50) of the «Kanishka era»; this is wholly improbable if S. Konow's reading of this inscription is to be accepted. Another hypothesis, based on the identification of the 3rd century of the «unknown era» with the «Kanishka era» assumes that the «unknown era» coincided with the Vikrama era, and DK coincided with A. D. 144. The acceptance of this identification, as it has been substantiated in special literature, is complicated because of a poorly grounded (at some points, totally untenable) outline of Kushan relative chronology. But the main thing that prevents us from definitely accepting one or another variant of associating the Kushan eras with the Vikrama era is that the existence of the latter in the first centuries A. D. is supported only by the legend mentioned above.


17. The starting-point of the Śaka era (according to al-Bīrūnī, 953 of this era = 400 of the «Yasdagird era»; 947 of the Śaka era = 416 of hejira) was A. D. 73. Though its first designation as the Śaka era occurs in 380 of this era, there is no doubt that it did exist also during the first centuries of the Śaka era (cf. the evidence of


inscriptions and Western Kshatrapa coins). The hypothesis of Fergüsson-Oldenberg that Kanishka was the founder of the Śaka era (DK = A. D. 78), exists today in a variant modified by Tan Lohuisen de Leeuw (1949): the «Kanishka era» is regarded as the 3rd century of the «unknown era»; the «unknown era» has been called «the old Śaka era» (the starting point is established by counting back 200 years from A. D. 78); it has been suggested that there existed «a second century of the Kanishka era» (which corresponds to the 2nd century of the Śaka era and the 4th century of the «old Śaka era»). The very fact of the identification of the «Kanishka era» with the Śaka era (if we disregard the concrete synchronisms) seems to raise no objection although it is still unclear why «the second century of the Kanishka era» and the 4th century «of the old Śaka era» existed in the same territory. The slight discrepancy with the tentative date of the rise of the Kushan kingdom — A. D. 25 (see § 2) obviously may be overcome if we assume that the compiler of «Han-Shu» did not yet know of the existence of the Kushan kingdom. Another suggestion (which is also based on the assumption that the 3rd century of the «unknown era» corresponds to the «Kanishka era») identifies the «unknown era» with the Śaka era, with DK = A. D. 278. As the latest variant among those offered for the solution of the problem of Kushan absolute chronology, this identification minimises to the utmost the gap between the Kushan kingdom and the florescence of the Gupta state (the earliest inscription of Candragupta II from Mathura is dated in 61 of the Gupta era, or A. D. 380/381).


18. Identification of the «Kanishka era» with the CedI (or Kalacuri) era, whose starting-point is A. D. 249 (B. S. Madjnmdar 1920) is questionable from the standpoint of both geography and history. The use of the Cedī era is attested by inscriptions (from the 5th to the 13th centuries) found in the territory to the south of the Kushan domains — in the Narbada River valley. The assumption that the Cedī era was founded by Iśvarasena of Ābhīra tallies with historical facts better than the hypothesis of R. S. Madjumdar.


19. Mention of Kanishka in the Buddhist tradition, attested by Chinese pilgrims (5th-7th centuries), by transla-


tions and Tibetan sources occurs fairly often, but the authenticity of this information is very doubtful, and the chronological allusions to the time of Kanishka's rule are inaccurate and contradictory (Zürcher 1960; Wilhelm 1960; Petech 1960; Eggermont 1960). In the northern tradition, Kanishka's patronage of Buddhism evidently became with time a standard ideal for a king's attitude and it was cloaked in legends and details which have no bearing whatever on the real historical personage.


20. The chronology of the Western Kshatrapas is based on a sequence of dates which researchers unanimously believe to start from A. D. 78 (the Śaka era). The fall of this dynasty is associated with the conquests of Candragypta II, who issued silver coins patterned on those of the Kshatrapas; these coins, however, were dated not according to the Śaka era (as the Western Kshatrapas' from 100 to 310 + X of the Śaka era), but according to the Gupta era (90 + X of the Gupta era = A. D. 410 + X), Under any of the existing variants of Kushan absolute chronology, the Western Kshatrapas ruled at a time when the Kushan kingdom definitely existed, but synchronising the history of these two dynasties is only possible if we adopt a certain variant of the «date of Kanishka». The question of whether the Western Kshatrapas were an independent, fully sovereign dynasty or whether at certain periods these rulers were vassals of the Kushan kings, is still unclear (and it depends directly on the Kushan chronology). The simultaneous reign of two Western Kshatrapa rulers (with the title kshatrapa — «junior», and mahākshatrapa — «senior») attest to the hereditary nature of the Western Kshatrapas' power. Brief intervals (from one to five or seven years), when the coins fail to show the use of one (or both) titles may be explained either by gaps in the emission of coins, or by lacunae in the extent emissions. The only periods that may be associated with some changes in the political situation are the interval 196-211 of the Śaka era (A. D. 274-348), and the interval 218-270 of the Śaka era (A. D. 296-348) when only the title kshatrapa was in use; the title mahākshatrapa evidently was not in use during that period.



21. A direct juxtaposition of Kushan chronology with absolute dates in the Gupta history is possible only if later variants of the «date of Kanishka» are accepted. The earliest direct evidence of the spread of Gupta power over the territory of the Kushan state is the Mathurā inscription of Candragupta II, dated in 61 of the Gupta era (A. D. 380/381). The image of seated Lakshmi on the coins of Samudragupta is clearly traceable to the images of Ardokhsh on the coins of Kanishka III, but it seems to shed little light on the problem, since Kanishka III may have lived at the same time as Vasudeva or may have been his successor.


22. Line 23 of the Allahabad Inscription gives the title of the Kushan kings as «Daivaputra-Shāhi-Shāhiānushāhi» (it occurs in the list of peoples and rulers on Samudragupta's frontier lands). This title can be regarded as evidence of the existence of the Kushan state at the time when the inscription was compiled. However, the date of the Allahabad Inscription has not been established so far (approximately A. D. 350-380). If early variants of «the date of Kanishka» are accepted, the title can be interpreted in different ways. This is obstacle to using the Allahabad inscription for determining Kushan absolute chronology.


Ch. III. Chinese Synchronisms.


23. Only a rough idea of the date of the formation of the Kushan kingdom is given by the above-mentioned fragment from Ch. 118 of «Hou-Han Shu» (see § 4). The period of «more than a hundred years» can be counted off from the approximate date of the first sojourn of Chang Ch'ien among the Yüeh-chih, whom he found «north of the Kui-shui River» (Amu Darya), when they had already subjugated Ta-hia (Bactria). In this case, the Kushan kingdom may have arisen after 30 B.C. (how much later, is a ratter of guesswork). Yet the evidence of this source does not contradict the idea of counting off «more than a hundred years» from the time of Bactria's complete subjugation (i.e., the settlement of the Yüeh-chin so-


uth of the Amu Darya). Since «Han-Shu», covering as it does events up to A.D. 25, says nothing about it, we cannot rule oat the possibility, on the strength of Chinese sources, that the settlement of the Great Yüeh-chih south of the Am Darya may have taken place after A.D. 25. The formation of the Kushan kingdom which occurred «more than a hundred years» later accordingly has to be shifted in time. Archeological and numismatic material from the sites of northern Bactria unfortunately does not help to minimise this fluctuation range in the interpretation of the «Hou-Han Shu» fragment.


24. Events in Eastern Turkistan in the last quarter of the 1st century A.D., connected with the activities of Fan Ch'ao, are often drawn upon in studying the history of the Kushan kingdom. Howerer, information on Pan Ch'ao and hie contact a with the Yüeh-chih should be used very carefully. First, one or another variant of the interpretation of these events fully depends on the hypothesis adopted with regard to Kushan absolute chronology, but it cannot be used, on any account, for determining «the date of Kanishka». Second, the authenticity of many statements of Pan Ch'ao himself and of his son Pan Yung cannot but be questioned by modern researchers — since they were put to doubt even daring Pan Ch'ao's lifetime.


25. Attempts to link the separation of the Western Region from China in 105-107 with ascending Kushan influence there (with all the ensuing conjectures on Kushan chronology) can hardly be regarded as well grounded, for «Hou-Han Shu» says in so many words that at that time the Western Region was seised by the «northern Haiung-nu». A relied allusion to a nameless Yüeh-chih king in connection with the events of 116 in Šu-le (Kashgar) was associated with the legend, cited by HaUan-tsang, about Kanishka and the hostages of «the frontier tribes in the region “West of the (Yellow) River”» (Ho-hsi = Shenai and Hansu). Howerer, there are no firm grounds either for such a juxtaposition, or for using it as a basic argument in favour of DK = A.D. 110-115.



26. «Hou-Han Shu» mentions only two Kushan kings — Ch'iu-chiu-ch'üeh (Kujula Kadphises) and his son Yen-kao-chen (Visa Kadphises), but it does not give the names of the later Kushan kings. This is sometimes referred to when trying to solve the problem of Kushan chronology, bat it is hardly justified. The «Hou-Han Shu» chronicle, compiled between 424 and 446, covers events from A.D. 25 to 221. This omission, without any farther substantiation, can be used only by those who place DK in the 3rd century A.D., bat the argument ad silentio is hardly fit to use in trying to solve such a major problem.


27. Identification of Po-t'iao with Vasudeva is suggested on the strength of the «Memoirs of the Three Kingdoms» (compiled at the end of the 3rd century and covering events from 220 to 265). Under the year 229/230, «The memoirs» say: «On the day kuei-mao [January 26, A.D. 230] the king of the Great Yüeh-chih, Po-t'iao sent an envoy with tribute. (Po)-t'iao was made (given the honorary title of) “King of the Great Yüeh-chih (who shows) Affection Towards the Wei”» (Zürcher 1960, p. 7). This fragments as well as the fragment from ch. 118 of «Hou-Han Shu» quoted in § 4, is a direct reference to a Kushan king in a Chinese source. The identification of Po-t'iao with Vasudeva is not, however, beyond doubt: «If puâ is the correct reading, Po-t'iao can hardly represent Vasudeva bat Pelliot's conjecture may well be right, and, in any case, it is difficult to suggest any alternative equivalence for Po-t'iao from among the known Kushan kings» (Pulleyblank 1960, p. 6). Last but not least, it is difficult to imagine that a Kushan king actually became «King of the Great Yüeh-chih (who shows) Affection Towards the Wei».


28. Chinese data on the Yüeh-chih in the 2nd and 3rd centuries have been considered in the context of Kushan absolute chronology by Pulleyblank in 1960. The hypothesis on the Kushan occupation of Khotan (if it could be substantiated by archeological or numismatic data) is tenable only for the period from 175 to 202, when envoys from Eastern Turkistan did not arrive to the Chinese court. But even if


we do establish tracts of Yüeh-chih — Kushan penetration into Central Asia (or merely traces of Kushan influence there) we shall still be a very long distance from a solution of the problem of Kushan absolute chronology.


Ch. IV. Iranian Synchronisms.


Events in the history of the Kushan kingdom can be linked with the Iranian time-scale chiefly in connection with Sassanian conquests in the east of ban. This can be done using the data of records (which as a rule are far removed in time from the events they describe), of Sassanian inscriptions and the coinages of the Kushan-Sassanian rulers.


29. Th. Nöldeke in his time mistrusted the statement of Tabari about Ardashir's conquest of Balkh and Khoresm and the subjugation of the «king of the Kūšān». According to V. G. Lukonin, in this part of hie narrative Tabari followed a source close in character and content to «Karnamak», and his description of the conquests of Ardashir I amounts to fitting the legendary (or leterary?) ideas on the conquests that behoved the first "king of kings" of Iran, into Iran's later geographical boundaries. The inscription of Shapur I from Kaaba-i-Zardusht (ŠKZ) on the composition of the court under Ardashir does not confirm Tabari's statement. The evidence of Western (Roman) sources makes it possible to speak of arduous wars in the west of Iran, conducted by Ardashir with brief intervals up to the time of accession of Shapur I (April 243); these sources also give indirect evidence which casts doubt on the authenticity of Tabari's statements about Ardashir's conquest of Balkh and the subjugation of the Kushan king.


30. The identification of Vehsadjan with Vasudeva has been repeatedly resorted to in attempts to solve the problem of Kushan chronology, Movses Xorenaci's «History of Armenia» mention Vehsadjan from Balkh who waged war against Ardashir. This identification, highly questionable from the point of view of phonetic correspondence, proves untenable also in its essence, since Vehsadjan (who, incidentally, is not referred to in this source as king of the Kushan or simply as a king) advanced with his own Karen


Pahlav branch» from Pahl against Ardashir (who should have been succeeded by Shapur by that time); Vehsadjan was «a kinsman of the Armenian king», and, finally, an uncle of Gregory the Educator. The whole episode is given in Movses' «History» in connection with Gregory's descent. We can hardly take the word of Movses Xorenaci to apply all this information on Vehsadjan to Vasudeva.


31. Kushanshahr is mentioned in the ŠKZ list of lands subject to Shapur I. This is chronologically the first reliable testimony of Sassanian penetration to the territory of the Kushan kingdom. However, it cannot be employed in determining the «date of Kanishka»: the data of the inscription may be interpreted as evidence of the final collapse of the Kushan kingdom by 262 only tentatively. The inscription itself gives no answer to the question whether the entire territory of the Kushan kingdom was conquered and whether this conquest was stable and final.


32. The titles of Sassanian governors in the east of Iran, as borne out by inscriptions, show that the title «sk'n ML'K» existed even under Ardashir I. In 262 Shapur's son Narse, who had the title of king, ruled over the territory of «Hind, Sakastan and Turestan down to the seaboard». The existence of this title (and domain) can be traced till 327 (the Persepolis inscription dated in the 18th year of the rule of Shapur II). The only mention of the title «king of the Kushan» occurs in Narse's inscription from PaikulI, but in a context which has not been deciphered — in a list of «allied kings» and semi-dependent rulers: moreover, the name of the king is not given. Thus, the titles of the Eastern governors of the Sassanian kings give no direct evidence of the subjugation of the Kushan kingdom to the Sassanidae (or some other form of its dependence ).


33. Kushan-Sassanian coins alone (whose legends contain the titles «king of the Kushan», «great king of the Kushan», «king of kings of the Kushan») can be definitely regarded as evidence of the incorporation of the western parts of the Kushan kingdom into Sassanian Iran. Employment


of these coins for solving the problem of absolute Kushan chronology puts forward two independent questions: (1) their dating and (2) whether this dating makes it possible to obtain the absolute chronology of the fall of the Kushan kingdom.


A new dating of these coins (R. Göbl 1967; V. Lukonin 1967) seems to be better grounded than the one offered by E. Herzfeld (1930). The new theory is based on a comparison of early Kushan-Sassanian emissions with the coins of Shapur II (309-379), as well as on associating one of the Kushan-Sassanian rulers (whose legend is «Lord Varahran, great king of the Kushan») with the future shahanshah of Iran, Varahran IV (388-399), who issued Kushan-Sassanian coins prior to his accession in 388.


Early Kushan-Sassanian gold coins (of. Göbl 1957, Nos. 319 and 320) typologically follow the latest (under the existing relative chronology of Kushan gold coinage) coins of Vasudeva. This similarity, however, evidently cannot be accepted as a direct argument in solving the problem of Kushan chronology, since we do not know the duration of the gap between the end of Vasudeva's emissions and the beginning of Kushan-Sassanian coinages. The conjecture assuming the existence of several kings named Vasudeva is easier to disprove, as it is not borne out by, numismatic data — provided we use the totality of available coins and not certain pieces taken out of their numismatic context. But even if the Kushan-Sassanian coinage is assumed to have started in the 370's or 380's (or, on a wider scale, during the rule of Shapur II) and the artificial «lengthening» of the history of the Kushan kingdom (by presuming the existence of two or three «Vasudevas») is ignored, we still can get only a rough idea on the probable duration of the gap between the last coin of Vasudeva and the first gold Kushan-Sassanian coin.


34. Mention of the Kushan in Faust Buzand's «History» is connected with the events of the 370's when «the Persian King Shapuh of the Sassanidae» (Shapur II) waged war against


«Arshakuni, king of the Kushan, whose seat was in Balkh»; according to Faust Busand, «the great king of the Kushan» scored several victories. We do not know whether this information is reliable. Though Faust Buzand wrote about events which had taken place only a hundred years before, the war between the Kushan king and Shapur II Interested him merely because people of Armenian origin had taken part in it. The overt anti-Iranian bias of this author should also be taken into account. Faust Busand's statements can hardly be directly linked with those of Ammianus Marcellinus (XVI, 9) about the arduous campaign of Shapur II in 338 (as well as J.Marquart's tentative correction cuseni, i.e., Kushan, instead of euseni). The war itself, mentioned by Faust Busand, is quite «neutral» in the context of the problem of Kushan chronology, and it is easily accounted for under any of the existing hypotheses.


35. Equally obscure and negative for determining the «date of Kanishka» is the mention of Kushan country and «the frontier post of Kushan» in a Manichaean text describing the journey of Mār Ammō (circa 263-270).


Ch. V.  Roman Synchronisms.


Written sources say nothing on the diplomatic (inter-government) contacts of the Kushan kingdom and the Roman Empire. Mention of Bactrians and Indians by Roman authors in connection with different events of the 1st-4th centuries cannot be used for synchronising Kushan history with Roman chronology. Kushan-Roman indirect synchronisms can be established only on the basis of material evidence of contacts between the Kushan kingdom and Rome (in a broad sense). Roman articles, including coins, imported into the Kushan kingdom or the contiguous lands, as well as evidences of the impact of Roman-world culture, traceable in Kushan art relics (sculptures, reliefs, toreutic objects) and coins — all this barely helps to establish the dates after which certain contacts of the Kushan kingdom with the West could hare taken place. Indirect synchronisms, naturally, give no opportunity of saying how long it took


for a certain Roman object (as each or as a prototype) to reach the Kushan kingdom and how long it could have persisted in Kushan territory. It is impermissible to use Roman dates thus deducted as terminus-ad-quem. Archeological material may have served as the most reliable means of establishing terminus-ante-quem for objects discovered in favourable (and indubitable) stratigraphic conditions, bat so far I do not know of any such finds (cf. Stavisky 1964; Allchin 1960).


36. An important place in constructions of absolute Kushan chronology is often held by the identification of the «Indo-Parthian» king Gondophares, known from coins and the inscription U=103, with king Gudnafar, mentiones in a Christian legend describing St. Thomas' journey to the East. But the vague historical and geographical information in the legend is patently inadequate for proceeding from a similarity of names to the date of Gondophares (placing him approximately in the 40's of the 1st century A.D.). The earliest date offered for the Syrian (primary) version of St. Thomas' Acts is the second quarter of the 3rd century. The only chronological conclusion that follows from this coincidence of names has been aptly formulated by van Lohuizen: «The only thing that can be said again, is that the name of Gondophernes was already known in the West in the 3rd century A.D., and that the Indian association with his name was the cause that he was connected with the legend of St. Thomas» (1949, p. 355). This conclusion is fully compatible with any of the existing variants of Kushan absolute chronology: from DK = A.D. 78 to DK = A.D. 278.


37. Finds of Kushan (or pre-Kushan) coins together with Roman coins are few. The optimal case, from the point of view of the circumstances under which the find was made and the synchronous withdrawal of the coins from issue, is undoubtedly the Ahin Posh Tope find of 1879. Ten dinars of Vima Kadphises, six dinars of Kanishka, one dinar of Huvishka were discovered during the excavations of a stupa, together with the aurei of Domitian (81-96), Trajan (98-117) and Sabina, the wife of Hadrian (117-138). The coinage for Sabina began in 128; she died in 137. Thus, the coin of Huvishka is the «youngest» of Kushan coins, and the coin of Sabina


is «the youngest» of the Roman. Consequently, all the coins in the hoard were withdrawn from circulation after 128. This date does not contradict any of the existing variants of Kushan absolute chronology, for it can be assumed with an equal degree of certainty that the Kushan coins found together with the Roman were issued either before 128, or after 128. The opinion voiced with regard to Sabina's coin, that it did not come from the imperial mint, being an ancient cast forgery, adds to the uncertainty of any chronological conclusion about the Ahin Posh Tope find. Obviously, this hoard, as well as all the other finds of Roman and Kushan coins together, cannot be of any help in solving the problem of Kushan chronology. Suffice it so say that at one time the Ahin Posh Tope hoard was brought in correspondence with DK = 57 B.C. (Kennedy 1913b. p. 371; «Discussion», 1913, pp. 936-937), and then cited as an «argument» in favour of DK = A.D. 78 (Thomas 1913, pp. 644-645); recently it was discussed in connection with the hypothesis DK = A.D. 225/230 (Göbl 1967, II, p. 272). The decisive factor here should be not the date of the Roman coin but an independent archeological dating of the stupa where the hoard was found. Naturally, it is difficult to judge of the matter today, but in 1879 R. Hoernle wrote that the construction of the stupa could not be placed earlier than the 3rd century A.D.; later dates were ruled out by him for the only reason that otherwise the Kushan dinars would have to be credited with too long a circulation (accepting the early variants of the chronology of Kushan kings) — over 200 years (Hoernle 1879, pp. 16-17).


38. The «Roman head» coins of Kujula Kadphises are the earliest case of Roman-Kushan iconographic parallelism in numismatics. If we assume that the coins of Augustus were the prototype for the head, then this group of Kujula Kadphises coins was manufactured after 29-27 B.C. Even if we subscribe to the opinion of J. Allan (cf., however, Göbl 1960a, pp. 80-81) that the coins of Claudius (41-54) were the prototype for these Kushan coins, the deducted date — after A.D. 41 — does not exclude the earliest of the existing variants of DK = A.D. 78.



39. A comparison of images on Kushan coins with those on Roman coins (Roman is used in a broad sence) has currently been made in connection with the problem of Kushan chronology (in favour of DK = A.D, 144, Göbl 1960a, Göbl 1960á; in favour of DK = 225/230, Göbl 1967, II, pp. 273-275; in favour of DK = 278, Zeymal 1964). Prototypes in Roman coinage make it possible to establish the time after which Kushan coins patterned on these prototypes were manufactured. Therefore, we must have proof that there were no earlier specimens that could have served as prototypes (with a corresponding alteration of terminus-postquem). It is equally important to show that the images traceable to Roman prototypes could not have been fashioned in the Kushan milieu independently of Roman numismatic iconography. These two methodological premises are far from always strictly adhered to in the Roman-Kushan iconographic parallels by R. Göbl (1960a). Having artificially limited the range of prototypes to the coinages of Rome and of Alexandria, R. Göbl obviously has not made full use of the opportunities offered by this method.


40. The allusion to the Roman-Kushan medallion (the British Museum), with the image of Constantine the Great (obverse) and the Kushan goddess Ardokhsh (reverse), in solving the problem of Kushan absolute chronology is hardly justified. A strict approach to its datirg indicates that the medallion could have been fashioned after A.D. 326 (+ X years). The image of Ardokhsh which is undoubtedly traceable to the reverse sides of Huvishka's coins, could have been executed either earlier or later than 326. The corrupt legend on the obverse suggest that the medallion was not likely to have been fashioned during Constantino's lifetime. There are no other signs that would lead us to establish how much later than 326 the medallion was made. Apparently it cannot be ruled out that the date of the medallion is the one suggested by R. Göbl for its setting — 5th century A.D. (Göbl. 1967, II, pp. 301-302).



41. Still more questionable is the tenability of drawing upon the so-called «unique Kushan-Roman gold coin» in solving the problem of kushan chronology (Altekar 1930). Highly hypothetical is also the date suggested for this «coin» (which is sooner a pendant) — 3rd century A.D. According to Altekar, it was made later than A.D. 200, for it is patterned on Vasudeva's coin, but it «must be earlier than the time of Candragupta I, circa 310 A.D., when the new fold currency was introduced in Magadhā by the Guptas».


42. The Gandhara sculptures which show clear traces of Roman influence, are so far potential rather than real evidence as regards the problem of Kushan chronology. Questionable, too, are the problems of the correlation of the Gandhara relics with Kushan relative chronology. The currently existing viewpoints (those formulated and not finally disproved) support all the possible variants of this correlation: Gandhara art under Kanishka was (1) either taking shape (Smith 1930, Rowland 1936, a.o.), (2) or was flourishing (Foucher, 1923, s.o.), (3) or was on the decline (Marshall, 1947. a.o.). The majority of the Gandhara relics are dated in keeping with the concept selected, since only few specimens of Gandhara art are directly linked with relative Kushan chronology. But these few relics, too, admit of different interpretations. Thus, the «relic-casket of Kanishka» (with the K=[1] inscription) was regarded as a specimen of archaic — vis-à-vis Gandhara — art and also as evidence of the retrogression of the Gandhara school under Kanishka. There is some reason for bringing the Gandhara sculptures with inscriptions and dates of the 4th century of the «unknown era» in correspondence with Kushan relative chronology, but the determination of the era mentioned in these inscriptions has given rise to much controversy among the students of Gandhara art (see Deydier's compendium 1950, p. 223-224; cf., for example, Dobbins 1967). The gold coin of Kanishka with a standing Buddha image on reverse, likewise does not rule out any of the suggested variants of the correlation between Gandhara art and Kushan relative chronology. Therefore, questions of the reliability and certainty of Roman prototypes, sug-


gested for the Gandhara relics (and the corresponding absolute chronology) cannot be satisfactorily — and convincingly — settled as long as the problem of the relative chronology of Gandhara art remains unsolved.




43. Material on Kushan chronology which we command today, fails to give a direct and reliable solution of the problem. This is the reason for writing the present compendium rather than the conclusion following from it. This has to be stressed inasmuch as the authors of both special works and summarising works (on history, archeology, history of art and architecture, etc.) often adopt one or another of the existing hypotheses and develop a tendency to regard Kushan and pre-Kushan absolute chronology as a settled problem. This approach, while conveying a simplified historical picture and giving rise to numerous «filial» (and «grand-filial») conjectures, actually precludes us from arriving at a valid solution of the problem of Kushan chronology.


Apparently, the range of data described above, which are today drawn upon to solve (or rather, to discuss) the problem of Kushan chronology is going to be modified after a more profound analysis based on sources. If data of doubtful authenticity (see above, §§ 19, 29, 30, 34, 36, etc.), as well as data used in discussing Kushan chronology without sufficient logical grounds (§§ 24, 25, 35, 37, 40, 41), are ignored it will hardly «impoverish» material on the problem or hinder its solution.


The debatable data, whose employment for a solution of the problem is impossible without a definite interpretation, also include materials which are not of equal value. Most active use can be made today of the data of the Chinese dynastic chronicles (§§ 23, 27), Sassanian (§§ 31, 32) and Gupta (§§ 21, 22) inscriptions. The question of whether they should be employed in discussing the problem directly depends not so much on a more profound study of these sources but on an accurate and objective interpretation of these


data as they apply to the history of the Kushan kingdom. Another group of debatable data (§§ 13, 28, 33, 38, 39, 42, etc.) can be of use in settling the problem of Kushan chronology only after thorough researches based on sources — researches that should not hinge on the existing conjectures on Kushan chronology. Meanwhile, this material can be regarded as a potential source. To a certain extent, the solution of the problem could be also promoted by the provision of reliable, well-grounded paleographic criteria of dating (both relative and absolute).


If we do not gloss over the fact that the problem of Kushan chronology has not been solved to this day, pertinent archeological material is likely to be accumulated more intensively. Individual finds of coins or imported articles that can be dated will hardly be of much use, but a systematic accumulation of information of such finds over the entire territory of the Kushan kingdom may in time lead to the archeological material being resorted to at least on a par with other data in working out a solution.






ãëàâíàÿ ñòðàíèöà / áèáëèîòåêà / îãëàâëåíèå êíèãè