Translating ancient scripts is difficult, especially when the civilization they belonged to is long gone.
We lucked out with the ancient Egyptians when we found the Rosetta Stone, which had the same passage translated into three different scripts: Ancient Egyptian hieroglyphics, demotic Egyptian and classical Greek. Since linguists could read classical Greek, they were able to use this knowledge to understand the hieroglyphic script on the stone.
However, there is no similar artefact available for ancient scripts such as the hieroglyphics used by the Indus Valley civilization. These people lived approximately 4,000 years ago, along what it now the Indian-Pakistani border. They were very technologically advanced for that time, living in cities equipped with the first known urban sanitation systems in the world.
They were also excellent traders who developed an extremely accurate, standardized system of weights and measures. But could they write? Many of their artefacts are decorated with symbols, but nobody knows what these symbols mean. In fact, some researchers doubt that they even represent a written language at all.
So, researchers at the University of Washington have teamed up with researchers from India to try to translate the script using computers. The computer program looks at existing examples of the script and tries to perceive patterns in the order of the symbols.
Using a statistical method called the Markov model; the program has been able to demonstrate that the placement of symbols follows a logical pattern, supporting the theory that they represent a language. As one of the researchers noted in the article referenced above, “The finding that the Indus script may have been versatile enough to represent different subject matter in West Asia is provocative. This finding is hard to reconcile with the claim that the script merely represents religious or political symbols.”