Vishasita started off as a desire to apply Machine Learning and NLP techniques to the Indus Script. However, there was no electronic corpus of seal text that was available to the public. So we decided to make one from scratch and share it with the public.
One is unfortunately led down the path of trying to figure out what it takes to decipher the seals. Even the act of mapping signs on seals to characters is an interpretation where one needs to do the equivalent of deciding between (a vs A) or (a vs b). When one cannot read the text, these choices are taken blind.
Another problem is that it is not possible to know what is written. One does not know whether the text on the seals represent addresses on mailing receipts, words in a quotation, etc. Any of these could exhibit similar information signatures and all of them are valid interpretations. There is no question of finding the right decipherment because that is not possible. With a Rosetta Stone one can learn from someone who knew. In its absence, the best that one can do here is guess.
Given this situation, a scientific approach is better than a logical one. Open it up so that anyone can come up with a theory. And chances are that the most elegant one, the theory which requires the least leap of faith but explains the maximum number of the facts, is most likely to be correct. At least until something better comes along.
Vishasita aims to make this process more efficient by cracking open the relevant pieces of the puzzle and making it available to the public. It will also attempt at creating a new conceptual framework for decipherment and render its own interpretation so that people who are interested can take a look at it and see what works for them. The hope is to create a sufficiently relevant context so that others can reasonably leverage it in their attempts at decipherment.
to vishasita at gmail dot com.