The Minor Indo-Parthian Eras

There is an old debate in taxonomy between 'lumpers' and 'splitters'. The lumpers combine similar animals together to keep a minimum number of species, recognizing separate species only when their are clear and marked differences. The splitters are much more sensitive, seeing new species wherever they find variation. The problem isn't quite as intractable in Indian Chronology but their is a similar debate over eras. There are those, such as Fleet or Lohuizen, who have an almost pathological desire to ascribe inscriptions to the minimum number of eras. In their cases trying to date all the inscriptions of the north-west to a single reckoning, often employing dropped hundreds as a method of explaining apparent discrepancies. Then there are the 'splitters', who tend to see new eras wherever they find dates which do not immediately fit. A new date is uncovered and a new era is created to match.

There have been two reasons to prefer 'lumping' to splitting in Kushan studies, one philosophical and one practical. 

Occam's Razor

The idea that the simplest solution is usually the best is often called Occam's Razor after the Medieval theologian William Occam. It is a central part of western philosophy, and is central to the hugely important area of genetics known as cladistic analysis, where it is also called parsimony.  

It is not that the simplest solution is more 'true', after all the past is almost inevitably more complicated than our sources represent. However, Occam's Razor minimizes the number of unsupported assumptions, and so the solution is often 'least wrong'. 

The philosophical reason is that the fewer the number of eras employed to explain the inscriptions, the simpler the solution, and it is a general principle that the simplest solution is usually the best. It is also fairly easy to identify the need for new eras, which become obvious as anomalies arise, and these anomalies give a clue as to which inscriptions need to be explained by increasing the number of eras. On the other hand 'splitters' find it very difficult to work in the opposite direction and identify eras that should be combined.

The practical reason is that it has been believed that eras were pretty rare things. Only two eras had survived from ancient times to modern times; the Saka era of 78 AD, and the Vikrama era of 57 BC. So it seemed at the start of this century that eras were founded by great kings, and endured for a long time, that it was therefore incongruous to start inventing new ones unless there was very good reason to do so. For a long time it was believed that only very powerful and important kings founded eras.

However, a great many eras have been uncovered by historians. The earliest, the Greek era which belongs to c.185 BC; the Azes era in the first century BC (and may be the same as the Krita/Vikrama era); the Saka era, mentioned above, began in 78 AD; and the Kanishka era has been known, since the 1960s, to be distinct from the Saka era and have a starting point of 105 - 130 AD; this may be the same as, or distinct, from the Kushan era of 227 AD; it has also become clear that from documents of Bactria and Central Asia that a Kushano-Sasanian era began in 233 AD; and in India the Kalachuri-Chedi era of 248 AD; the Gupta era of 319 AD; and the Ganga era of 420 AD, which are clear demonstrations of the proliferation of eras (see the map below).

Beyond this list of eight well-known eras, there are also three eras which have been postulated for the period of the early Kushans.

Maues Era

This era is known from only a single inscription, the Taxila Copper Scroll (insc. 346). Despite being called 'Taxila' the actual find spot is unknown. The inscription begins "In the seventy eighth, 78, year the Great King, the Great Moga, on the fifth, 5, day of the month Panemos, on this first, of the Kshaharata and Kshatrapa of Chukhsa - Liaka Kusuluka by name - his son Patika - in the town of Takshasila..."

The Satrap Patika is known from the Mathura Lion Capital (insc. 348) which belongs to the early part of the first century, preceding or contemporary with Gondophares (c.20 to 40 AD). Patika  is also believed, on the basis of coins to proceed Jihonika (Zeioneses) whose coins proceed Gondophares, and who is known from a slightly dubious inscription of year 191 (insc. 364). This date is assumed to be in the Greek era and gives a date of c.10 AD. If the date of 346 is ascribed to the Azes era this would give 21 AD for Patika. This is a perfectly reasonable date for Patika, though it might have been expected to be two to three decades earlier. 

However the statement year 78 of Moga is hard to reconcile. It cannot be the reign of a king, so it must be the name of an era. Since it is clear that the era in other inscriptions was called the Azes era it cannot be that, and it should be attributed to Maues, an Indo-Parthian who proceeds Azes. So it would seem that this inscription testifies to the existence of an era begun in the first half of the first century BC, which is in use in the Taxila region at the start of the first century AD.

The Gondophares Era

The suggestion that Gondophares founded an era (probably as a continuation of his regnal years) has been made by Joe Cribb (1999). It revolves mostly around an inscription of Sodasa (the son of Rajavula) satrap of Mathura (insc. 857), and an inscription of the Western Satrap Nahapana dated in the year 45 (insc. 859).  The Nahapana inscription will not be discussed here as it does not really impact on the argument.

The Sodasa inscription is dated in the year 72, though it should be noted it is very hard to distinguish 4 and 7 in some inscriptions. If this inscription is referred to the Azes era then Sodasa would have to rule in c.10 BC or c.25 AD (give or take due to the difficulties in fixing the Azes era). This is far, far too early as he would have to proceed Gondophares by some time. Likewise if the date is ascribed to the Kushan era it would be incompatible with Kushan rule at Mathura. Another possibility would be to ascribe the date to the Saka era of 78 AD, this would give a date of 120 AD or 150 AD. This not plausible, as it would posit a combined rule of nearly a century for father and son, Rajavula and Sodasa, and again would put Sodasa in Mathura when it is believed he had been deposed by the Kushans.

Gondophares has left an inscription (insc. 353) dated in year 103 of the Azes era and in year 26 of his own reign. So if Sodasa is using an era which is a continuation of Gondophares reign this would mean that Gondophares Era = Azes + 77 or c.20 AD. So Sodasa's inscription would be 62 AD or 92 AD. These are roughly compatible with Sodasa's dates but 92 AD is fairly late (though possible). What is clear is that Sodasa cannot be dated using the Azes or the Saka era, and that 42 and 72 are too high to be the reign of the king. So even if it is not a Gondophares era the Sodasa inscription does imply the existence of another era in the North of India.

The Odi Era

This tentative suggestion has been made by Richard Salamon (2003: 48-9). Unlike the Maues and Gondophares era this suggestion could be disproved any day by a new publication. Salamon points out that three dated inscriptions are known for the kings of Odi. Year 4 of Ajitasena, year 5 of Varmasena (insc. 572), and year 14 of Senavarma (insc. 568). These could just be regnal years, but it is interesting to speculate that they could also be an era founded by Ajitasena. For the moment it is best to view these as regnal years, and await further evidence.


Once these tentative eras are added to the Eras already known it begins to break down the appearance that eras are rare or unusual things. They give the impression that anyone and everyone invented eras, and what is important is not who might found one (because almost anybody could, and did) but why particular eras continued in use, and others vanished very quickly. Though it does not undermine the basic method, that chronological enquiries should always minimize the number of eras it suggests that more attention should be given to the idea of 'splitting'. So from the 'splitters' point of view here is the full list of eras from the first to the fourth century AD.

Data extracted on 20 April 2005, from a sample of 873 inscriptions. A higher resolution image is available.

Inscriptions in an unknown era are distinguished from the reigns of kings by a question mark. And with the exception of Mathura and Bandogarh, where multiple inscriptions in the same year are not shown, all inscriptions are marked.

Indian Eras

India had a large number of eras which operated in the early historical period. Though there are sporadic dated inscriptions before the first century, it seems to be a fairly new fashion (and a fashion it certainly is, which comes and goes for reasons that are not understood). The first dates are in the Azes era (red) and Greek (green) in the North, the Saka era in the West (olive green) and the use of regal years (black), in the south. In the second century the Kushan era begins (which dominates the dated record at this period, vastly outnumbering dates in any other era). In the third century the Vikrama era starts to be used in Rajasthan (shown in a dark yellow), and the Kalachur-Chedi era in Western India. These are followed by the Gupta era in the fourth century (which achieves a very wide distribution), and the Ganga era in the fifth. 

The map on the left shows how complex, and incomplete, our picture of Indian eras is. There are extensive areas where it is unclear what era was in use, and it is impossible to know if this is because inscriptions were not made or simply because none have been discovered and published. For example, inscriptions are attributed to the Saka era both in West India and in the Gangetic region. It is known (and can be seen clearly on the map) that the Saka era spread very quickly along the west coast, but there are no inscriptions to confirm that it did spread along the trade routes from Nasik to Kosam. In fact ignoring the Vikrama, Kalachuri, and Gupta inscriptions (which are all third century or later) there are no inscriptions at all between the western Saka inscriptions and the eastern inscriptions attributed to the Saka era. So no way of confirming or denying that these inscriptions belong to the same era.

Another salutary lesson is that not a single date of the Kushano-Sassanian (233 AD) era appears on the map. These are known exclusively from letters and other documents on perishable materials, and if it were not for the peculiar circumstances of preservation of Central Asia and recent luck in discovery, no record of that era would exist for today's historians. A reminder of the old archaeologist's dictum that 'absence of evidence is not evidence of absence'.

Contents Page The Kanishka Era   The Azes and Greek Era Inscription List Bibliography

Robert Bracey.