review article Indian script

[falk at zedat.fu-berlin.de]

There were some contributions to the origin of Brahmi lately, with
statements which should not remain without comment. 
In Richard Salomon's review article the recent finds from
Anuradhapura are not mentioned, simply because the books under
review did not deal with them. 

> I scanned through the review article about Brahmi. I did not notice
> any reference to the `recent' finds of potsherds from Sri Lanka which
> are supposed to marked with Brahmi letters. Due to limited library
> facilities, I do not have access to the primary literature. My only
> reference is ``Archaeology of Early Historic India'' by Allchin et al.
> On pp 176--177, there is a reference to these potsherds. 
> Radiocarbon dating is said to put these in the period of 450BCE--350BCE,
> with the original excavator dating them to 600BCE--500BCE, presumably
> on archaeological grounds.
The author S. Deraniyagala, being the excavator mentioned above, 
had several very early dates on stock, at least when he was reporting
on his findings at Berlin, about 2 years back. These included: horses
in Ceylon around 1000/900 BC, rouletted ware around 200 BC (although
invented in Italy around 50 BC!), and Brahmi around 500 BC. These
findings, based on C14, are either a series of sensational discoveries,
or something is wrong with C14-dates from Ceylon. I firmly believe
in the latter possibility, mainly on the basis of my studies of the
development of Ceylonese Brahmi (please read the long chapter "Die Brahmi
Ceylons" in H.F. "Schrift im alten Indien", Tuebingen 1933, pp. 205-218.
There cannot be the least doubt: the developments of Brahmi regarding
the shape of characters, the orthography, layout of text etc. all point
to North India, followed by the adaptation in Tamil/Telugu speaking areas 
to the
needs of Dravidian languages, and then the adapted forms come to Ceylon.
Thinking of any alternative direction would imply that Dravidians in
Ceylon developed the script for their language, but never used it there
for that purpose. Rather unlikely.
Now for the publications of Profs. Allchin and Coningham. I talked to
both of them and they firmly believe in their C14 dates. But if you
check the latest article in Cambridge Archaeological Journal 6,1 (1996),
73-97, you have to realise that their material is rather scanty: 8 sherds
with 2 to 4 characters each. Their ideas about  which character looks
old and which does not displays a not too deep acquaintance with that
script at all. They make a whole set of conclusions depend on
their dating and their interpretation of the sequence of their sherds.
There is something I cannot evaluate. On page 79 they state that
"no Southern Hemisphere correction was used as the validity of such
a correction in the latitude of Sri Lanka has yet to be
established". What does that mean? Would such a correction change
the dates to something earlier or later? What does that
correction mean in those areas where it has been established?
To make it short: I have severe doubts about the conclusions reached
by Allchin and Coningham, not to speak of the dates forwarded by
Deraniyagala. And I am not the only one to doubt. In addition I can
say that I am working on the sherds coming from the Cultural triangle
in Sri Lanka. There are hundreds coming from all levels, not just
8. There will be a series of C14 dates dating them. After finishing
the new batch, which just arrived from Sri Lanka 2 weeks back, we
will know more - sooner or later. Inthe meantime it should be made
clear that all epigraphists, European or American, who can
distinguish between the characters, are not willing to change their
conclusions regarding a most likely young date for the development
of Brahmi. As I have made it clear more than once: there is nothing
either literary or circumstantial that would favour a date before
Ashoka. It is tiresome to realise that old arguments are repeated
time and again - without putting new flesh on old bones.


> The first publication of Deraniyagala has been mentioned by Falk in his 
> bibliography (Schrift in alten Indien, Tuebingen 1993), but, as far as I could 
> see, he did not take it into consideration in his conclusions regarding the 
> date of origin of script in India. 
no, it came too late. And a reply needs material I can check myself, like
the sherds from the Cultural triangle.

> and the recent works discussed (Falk's, von Hinueber's, Fussmann's, Norman's) 
> are already now outdated with regard to their main conclusion (to cite Salomon: 
> the "later date of origin for the Indian scripts), although these works will 
> remain valuable as collections and overviews of decades of research on this 
> topic. 
How fast can one decide without checking the evidence?

Harry Falk