The Date of Kanishka

Has the Date of Kanishka been solved?

In 2001 Harry Falk published a translation of a passage in the Yavanajataka (written by Sphujiddhuaja), which, he claimed, settled once and for all the date of Kanishka's ascension to the throne. The passage stated that the present year of the Saka Era could be calculated by adding 149 to the Kushan Era. If this was the case then the Kushan Era must have begun 149 years after the Saka Era (whose starting date, 78AD, is beyond dispute). This gives a date of 227AD for the Kushan Era. Falk argued that this date must refer to the second century of the Kushan era which is often written in inscription with the character for 100 missing. This would give a date of 127AD for year one of the Kanishka era. 

This argument has been widely accepted in Kushan studies (as of 2004 I am the only person in Kushan studies to have publicly challenged it). It is therefore of central importance to establish if it is a sound argument.  The interpretation of the passage is fairly certain, with no challenge so far being made on linguistic grounds. However it is the second stage in the argument that is problematic. The source certainly says "Saka Year equals Kushan year plus 149" but what this means, in chronological terms is not so clear. 

Falk's argument is based upon the idea that Kushan inscriptions are spread over two centuries and that in the second century the dates were abbreviated with the centuries being dropped. There is considerable agreement that the dates for Kushan inscriptions do form two sequences, one running from 1 to 98 and the other from 1 to 41. 

"This dropping was nothing but a theoretical possibility until it was given a solid basis in 1949 by J. van Lohuizen-de-Leeuw, strengthened with different arguments by Hartel 1996: 102f. It is now almost generally accepted by art-historians as well as numismatists." (Falk: 121)

Falk's statement is a serious misrepresentation. Prior to his article it was an open question whether this represented two centuries or two separate eras (Rosenfield 1968: 266; Puri, 1993: 253; Cribb, 1997: 222) . The problem was essentially that there was no evidence one way or the other. So the level of certainty implied by Falk is inappropriate. If we are talking about two eras rather than two centuries, then Falk's argument falls down almost immediately.

However, let us presume, for the sake of argument, that the 'dropped hundreds' argument might hold true. Now, imagine a contemporary of Sphujiddhuaja who has just been informed that an event took place in the reign of King Huvishka in Kushan year 28. He tries to calculate when this was by adding 149, and gets Saka year 177 (255 AD). He is then told that later in the reign of Vasishka in year 24 another took place. Again he adds 149, getting Saka year 173 (250 AD). The result is clearly nonsense. Of course, no-one would have said that Vasishka ruled in year 24, they would have said year 124. After all it is only an abbreviation used on inscriptions due to lack of space (like 'Sam' for 'Samvatsare').

Yet, Falk's argument depends upon Sphujiddhuaja formula applying to the second Kushan century, and not to the Kushan centuries inception. But this would lead to nonsense if a calculation were made that involved a Kushan date more than 40 years old, or if the calculation were made more than 60 years after the book were written. And more to the point it would require that Sphujiddhuaja was very badly informed about the Kushan era, after all he writes 'samvatsare' instead of 'sam' so why would he abbreviate the date, he is hardly under the pressure of space used to justify it in inscriptions. The attempt to rescue Falk's thesis from this problem has involved turning Lohuizen's original argument about abbreviation into one in which the Kushan era becomes cyclical with the dates being 'reset' after 100. This explains Sphujiddhuaja's formula but does nothing to stop the calculations becoming nonsense for anything other than immediate contemporary events. More to the point, prior to Falk no-one had ever suggested that the Kushan era might be cyclical, and there is absolutely no evidence for it.

It is clear that while there was no evidence before Falk's article to decide between two eras or two centuries, that the idea of Sphujiddhuaja talking about a second century is clearly untenable.

Conclusion: What does Sphujiddhuaja really tell us?

Now that we have highlighted the difficulty in taking a simple and naive reading of this source and re-examined the evidence for the dating of Kushan inscriptions we are in a position to say what it really tells us. Though we cannot exclude the possibility that Sphujiddhuaja is just very badly informed, and has either got the name of the era, or its date, wrong we have no particular evidence to support that conclusion.

Presuming he can be relied upon, he is clearly talking about an era that begins in 227AD. This could again be another era to which the name Kushan became attached (such as one founded by the Kushano-Sassanians, such as that proposed by Simms-Williams (1997)), or it could be that Kanishka's era began in the 3rd century as proposed by Gobl and others. However it is absolutely clear the passage is talking about an era which begins in 227AD, there is simply no way to support Falk's assumption that it could be talking about an era that is a hundred years old.

If the apparent certainty which Lohuizen's flawed analysis had previously given us is abandoned we can now see that the most plausible solution is to take this as evidence that a second series of Kushan dates represent a separate era. Of course the commencement of a second era by Kanishka II is an interesting 'fixed point' but because we cannot be certain of the gap between Vasudeva's last inscription and Kanishka II's first (in fact we presently have no way of separating the inscriptions of Kanishka I and Kanishka II), we are unable to use this as evidence for Kanishka's era. Other than to say it is compatible with the previously suggested dates of 115AD and 128AD.

Though we must be clear that given the uncertainty involved Falk's argument on its own does not exclude a date near 78, 144AD or 227AD itself for the commencement of Kanshka I. A solution to this problem will only present itself by reading the various sources in context, as a proper understanding of this source only really becomes possible by reading it in context.


Contents Page Evidence for Kanishka's Era The Second Kushan Century Lohuizen's Dates Bibliography

Robert Bracey.