Blog: Computer Automatically Deciphers Ancient Language

Computer Automatically Deciphers Ancient Language
MIT News (06/30/10) Hardesty, Larry

Researchers at the Massachusetts Institute of Technology (MIT) and the University of Southern California have developed a computer system that deciphered much of the ancient Semitic language of Ugaritic in a matter of hours. The researchers say archeologists can use the system to decipher other ancient languages and it also could help expand the number of languages that automated translation systems can handle. "We iterate through the data hundreds of times, thousands of times, and each time, our guesses have higher probability, because we're actually coming closer to a solution where we get more consistency," says MIT's Ben Snyder. To decipher a language the system makes several assumptions, including that it is related to another language, that there is a way to systematically map the alphabet of language onto another, and that the language's words share at least some roots with each other. "Each language has its own challenges," says MIT professor Regina Barzilay. "Most likely, a successful decipherment would require one to adjust the method for the peculiarities of a language."

View Full Article