The software, being designed at the Institut National de Recherche en Informatique et Automatique (INRIA), Paris, in collaboration with engineers from La Sapienza, Rome, is to be used for manuscripts collated by the Bhandarkar Oriental Research Institute (BORI), and the Mahendra Sanskrit University, Nepal. The last, though, seems to have dropped out of the project.

“This is part of a project titled IT Archaeology of Asian Texts, initiated by INRIA and funded by the European Commission. BORI was the obvious choice for the project because of the large number of Sanskrit scholars and manuscripts it possesses,’’ says Francios Patte, a representative of INRIA.

A Sanskrit scholar himself, Patte came to know that Dr Saroja Bhate of BORI was working with Paris-based scholar Johannes Bronkhort to create a critical edition of Kasika Vritti a couple of years ago. “Bronkhort was desperate because there were just so many manuscripts to go through. Then I mentioned the project to people at INRIA and they got interested,” says Patte.

TRADITIONALLY, an ancient text was copied time and again by various scribes, resulting in idiosyncratic additions, unintentional omissions, and mistakes. Consequently, there exist hundreds of versions of these texts across the length and breadth of the country in a number of scripts, adding to the general confusion of linguists, lexicologists, sociologists and anthropologists around the globe.

This is the precise problem the INRIA project seeks to tackle. According to Patte, one software will compare all the versions of a text and analyse the discrepancies, while a second software will propose a genealogy, which will enable the scholars to trace the manuscripts from which the copies originate. The algorithm used for the comparison will be very similar to the one used to compare DNA in biology.

“Once all the differences are known, the editors of the text will choose the version they believe to be correct and publish it as the critical edition. The omissions will also be mentioned. With many copies being made through the ages, we can find families and reconstruct the history of the text using the second software,” says Patte.

But why the Kasika Vritti? Bhate, the academic supervisor of the project, explains: “Besides being written originally in a meta-language that makes for easier computer-adaptability, Kasika Vritti is also a voluminous project, involving 200 manuscripts. We want to see if the computer can take on the challenge of tackling the large number of manuscripts and scripts.”

Underway since last June, the project will be completing its first phase by the month-end. “Till now we have collated only 50 manuscripts. These are just part of the alphabet sutras, we still have to venture into the grammar sutras,” says Bhate.

Though the language will be Sanskrit, the script will be Roman, which is more computer-friendly. Once the work on Kasika Vritti is completed the project will be extended to other Asian texts as well.

As part of the agreement, the funds given by the European Commission will be collected by INRIA and distributed to La Sapenzia and BORI, after verification of the accomplishment of each partner’s task as listed in the contract. While INRIA will be getting 70,629,07 euros for management workshops and 76,022,84 euros for other research work, BORI will get 44,764,23 euros and La Sapenzia will get 22,520,00 euros.