Computer explores the secrets of the ancient Indus script

The team, led by a University of Washington scientist, used computers to decode symbols in ancient figurative Indus. The results of the study are published this week in the Proceedings of the National Academy of Sciences, which show the types of sequences of sequences and publish statistical models for the secret language that families science has created.

'The statistical model helps to better understand the grammatical structure of the Indus script,' said study leader Rajesh Rao, a professor of computer science at the University of Washington. 'Such a model can be very useful for decoding, since any meaning assigned to a symbol must match the context of the context with the letters standing in front of and behind it.'

The co-authors of the study are Nisha Yadav and Mayank Vahia scientists from Tata Basic Research Institute and Mumbai Basic Science Research Center; R. Adhikari from Chennai Institute of Mathematics; and Iravatham Mahadevan from Indus Research Center Chennai.

Despite previous efforts, mankind has never successfully deciphered the Indus script. Characters often appear on tiny seals, carved wooden plates or amulets created by residents living in the Indus Valley 2,600 to 1900 BC. Each of these artifacts usually contains a sequence of 5 or 6 characters.

Some even wondered if these characters were actually a language, or were they merely a symbol or religious symbol.

The researchers used computers to decode the type of pictograms in the ancient Indus script. (Photo: JM Kenoyer / harappa.com)

The new research looks for mathematical types of arrangements in character sequences.Calculations show that the order of characters here is significant; If taking a character from a string on artifact and changing its position, the new string obtains very little sense of meaning in the language system under consideration. The authors argue that the existence of these character sequencing laws again supports the team's findings, published in Science earlier this year, according to which the Indus notation system may actually be. a language.

'These results allow us to claim a clear logic in the Indus script,' Vihia said.

The seals with the Indus character sequences were also found forever in Western Asia, in the Mesopotamia region of present-day Iraq. The statistical results show that the West Asian sequences are ordered differently from the sequences on the artifacts found in the Indus valley. This supports the hypothesis that this system could be used by Indus merchants in Western Asia to have a different expression meaning when used in the Indus region.

'Discovering the Indus script is flexible enough to express another meaning in West Indies is very interesting. This finding is completely against the idea that the Indus notation system is actually just political or religious symbols, ' Rao said.

Researchers have used a Markov model, a statistical method that estimates the likelihood of a future event (ie, writing a specific character) based on a known type of arrangement. This method was developed by Russian mathematician Andrey Markov a century ago and is increasingly used in economics, genetics, speech recognition and many other research areas.

'One of the main purposes of this article is to introduce Markov models, and statistical models in general, as computing tools for surveying ancient writing systems,' Adhikari said. .

An application described in the article uses a statistical model to fill in the blank of missing characters in archaeological objects . These characters can enhance the data system for decoding the ancient Indus writing system, Rao said.

The research was funded by Packard Foundation, Sir Jamsetji Tata Foundation, Washington University and Indus Research Center.