Simple computer program decodes lost Biblical language

The new program was able to translate the 3,000-year-old language using the computing power of a laptop.

By Bryan NelsonMon, Jul 19 2010 at 6:34 PM EST
 67

UGARITIC: Related to Hebrew, deciphering the Ugaritic language has been crucial to clarifying Old Testament text. (Photo: Wiki Commons/CC License)
A project led by professor Regina Barzilay of the Massachusetts Institute of Technology may be the first to show how ancient, lost or unknown languages can be decoded using a computer program, according to National Geographic.
 
  
Earn Points
What's this?
 
The MIT team was able to decode the "lost language" of Ugaritic, an ancient Semitic language used in Old Testament times, using no more computing power than that of a laptop. The program took no longer than a few hours to link most Ugaritic symbols to their Hebrew equivalents.
 
Ugaritic text was nothing more than a series of dots and wedge-shaped marks to linguists and scholars when it was first discovered on clay tablets in 1928, excavated from the rubble of the ancient city of Ugarit by French archaeologists. Even though the language is closely related to Hebrew, experts did not decipher it until 1932.
 
It took only hours to accomplish what took linguists years to complete, leading scholars to hope that the new computer program can be a prototype for a more powerful system to decode ancient languages that remain a mystery to scholars. In other words, it may not be long before computers become modern day versions of the Rosetta Stone.
 
"Traditionally, decipherment has been viewed as a sort of scholarly detective game, and computers weren't thought to be of much use," Barzilay said. "Our aim is to bring to bear the full power of modern machine learning and statistics to this problem."
 
But some experts remain skeptical. Richard Sproat, an Oregon Health and Science University computational linguist, notes that "in the case [of Ugaritic], you're dealing with a small and simple writing system, and there are closely related languages. It's not always going to be the case that there are closely related languages that one can use."
 
For example, a language like Etruscan, which was used by ancient Italians around 700 B.C., is known today from scant written examples and shares no relation to any other known language, except for a few words adopted by the Latin language (e.g., the name of the city of Rome comes from Etruscan). Deciphering Etruscan symbols could potentially give historians invaluable contextual clues about the region before Latin superseded the earlier language.
 
Barzilay thinks the MIT program can be upgraded to decode languages like Etruscan by scanning multiple languages at once and taking contextual information into account. At the very least, such a program could reveal new, obscure clues that scholars can use to learn more about ancient unknown languages.
 
CLOSE link:

Comments

Follow this conversation View:

anonymous
Near Eastern Grad 09/02/2010 19:12 PM

This article confuses me. Over 30 years ago, friends of mine were studying it at UC Berkeley. Ugaritic has been translated. You can even study it online! http://www.theology.edu/ugraintr.htm, see:
Stanislav Segert. A Basic Grammar of the Ugaritic Language: With Selected Texts and Glossary University of California Press, 1984
M. Dietrich. The Cuneiform Alphabetic.... More

anonymous
Nathan 07/22/2010 07:51 AM

It would be helpful to work out what on earth people say to me on facebook.

anonymous
Ali Clarka 07/22/2010 01:24 AM

Hopefully everyone will be able to speak it soon someday!

anonymous
Ursula Koons 07/22/2010 00:39 AM

I want to know if the translation differed... While this is cool in a 'it can be done' way it would be real news if it confirmed the original human translation.

anonymous
Ursula Koons 07/22/2010 00:38 AM

I want to know if the translation differed in anyway... While this is cool in a it can be done way it would be real news if it confirmed the original human translation.

anonymous
tjbv 07/21/2010 18:16 PM

Linear A? The Voynich manuscript? Rohonc Codex? Rongorongo?

I'd love to see them translated.

anonymous
Seth 07/21/2010 17:38 PM

:) I think it's great. Maybe the idea of an universal language decoder is not so far fetched.

anonymous
Robert Maxwell 07/21/2010 17:02 PM

Just the fact that it will read is astonishing. They should feed in the Arizona Immigration Law and translate that for Eric Holder.

anonymous
Dean Kaflowitz 07/21/2010 16:46 PM

The comparison of the computer to the Rosetta Stone makes no sense. The Rosetta Stone was a an artifact that had a large number of hieroglyphs and was deciphered largely because it was bilingual. It was, in this context, similar to a clay fragment containing Ugaritic symbols, not to a laptop computer used to decode the symbols.

anonymous
Jerod 07/22/2010 13:05 PM

They didn't say THE Rosetta Stone, they said Rosetta Stone... look it up. How have you no idea what it is?

anonymous
they did say THE Rosetta Stone 07/28/2010 02:38 AM

no, they did say "the Rosetta Stone" not rosetta stone; and it's a good analogy. Think about it again

anonymous
Mark 07/21/2010 16:19 PM

This is not news. Read the third paragraph.
"Ugaritic text was nothing more than a series of dots and wedge-shaped marks to linguists and scholars when it was first discovered on clay tablets in 1928, excavated from the rubble of the ancient city of Ugarit by French archaeologists. Even though the language is closely related to Hebrew, experts did not decipher it until 1932"

MIT has written a computer program to to what has already been done.

anonymous
Brent 07/21/2010 16:34 PM

Its not the fact that it was already deciphered that is important, it is the fact that a computer program was able to do in hours what took years, and has never been done before by a machine. Also the professor from Oregon is just mad that his group was beat to the punch on this.

anonymous
George 07/21/2010 20:34 PM

I'm forced to agree that there is probably no story here. It seems very likely that MIT wrote the program specifically to tackle this particular language. However, it could be a first step toward a program that could do this for real. That would be a story.

anonymous
carolyn 07/21/2010 16:11 PM

Bloody cool. OK, again what does it say?

anonymous
robert Cleveland 07/21/2010 16:38 PM

What change /differences does this particular document offer? If any.

anonymous
Sereneathena12 07/21/2010 16:03 PM

MIT deserves a lot of credit. It might not seem like a big deal, but computers are not good with language. It's very difficult to create a program to translate written language and even more difficult for spoken. Just type a complex sentence into google translate and see what it gets translated into. If this is as successful as this story leads readers to believe, this is a big deal (at least to linguists like me).

anonymous
Greatest Programmer in the World 07/22/2010 16:45 PM

computers, computer languages, computer programs are an extension of the human mind. A human computer programmer can write a decyphering program that works well or not at all.

As for being the GPINW and having written programs that interpreted the logic of several computer programs into other computer program languages successfully, i would find this an interesting challenge and valuable tool for linguists like you :)

anonymous
whocares 07/21/2010 15:47 PM

Speaking of language--did the author really use the phrases, "no more," "no longer," and "nothing more," within a few sentences of one another, in a NEWS piece? Trying to stretch the word count, eh? Wow.

anonymous
michele in mississippi 07/21/2010 17:21 PM

i wondered if anyone else noticed that.....

anonymous
Roman 07/21/2010 15:37 PM

on decoding the new health care law

anonymous
Richard 07/21/2010 16:22 PM

the thought patterns of the modern Republican?

anonymous
Paul 07/21/2010 15:51 PM

Very, very funny. I have not stopped laughing since I took it in

anonymous
Judy 07/21/2010 15:33 PM

On the Etruscan Language, read the grammar by Josef Pfiffig (in German).
The Voynich Manuscript has at least one level of encryption, this is a different technique from translation
Sumerian is an isolate, not a semitic language, but Akkadian is a semitic language.

anonymous
Moirwen 07/21/2010 15:06 PM

I wonder if this process could be used in decoding the Voynich Manuscript.

anonymous
Don F. 07/21/2010 14:58 PM

I strongly suspect that this article overstates the accomplishment. In all likelyhood, the "simple program" through the process of pattern matching was able to build on the similarities between Ugaritic and ancient Hebrew or perhaps other ancient semetic languages such as Sumarian. This would enable them to take an string of characters and relate them to the semantic and syntatic patterns of a previously known language.

This is not taking an arbitrary language and translating it into.... More

anonymous
Old McDonald 07/21/2010 15:22 PM

Please proof your comment before putting it online; the misspellings are ironic, indeed, considering the subject matter!

anonymous
Rosetta Stone 07/23/2010 02:10 AM

Please point out his misspellings because I cant see any...... But then again I am not the grammer police either so I better leave that kinda work to the "pro's". If you have to be a doushe at least try being somewhat relevant to the topic

anonymous
Hillbillet 07/26/2010 17:19 PM

I'm not the GRAMMAR police either, but you might want to actually check your spelling before saying something like that. And it's DOUCHE. Sorry, couldn't help it.

anonymous
supertramp 07/30/2010 09:13 AM

Oh, so funny! Spelling & grammar & word count are very important points when reporting and discussing "scholars" and their research concerning linguistics.

anonymous
Others 07/21/2010 14:54 PM

They should be able to test it on other obscure languages. Many north american indian, Inuit, etc languages are still used today, so we could see it it works on them.

anonymous
David 07/21/2010 15:23 PM

I don't think this program would work with most Native American languages ... the reason being that most of the tribes never developed a written language.

anonymous
ME 07/21/2010 14:39 PM

Why do these articles never say the whole scoop?? What did the decoded script say??

anonymous
Clue 07/21/2010 15:49 PM

it said it will be decoded one day by MIT folks.

anonymous
Curtis 07/21/2010 14:31 PM

When translated, the ancient hieroglyphic writing revealed clues to the daily lives of the Ugaritic people at a time before Christ. The phrase reads "U R teh gay. lol @ u"

anonymous
grant ellsworth 07/21/2010 14:12 PM

I read a book several years ago where it made a very strong case for ancient Etruscan being related to modern Albanian. I think this was supported by further research. Maybe the article's writers should look further into this.,

anonymous
grant ellsworth 07/21/2010 14:19 PM

Sorry for the omision - Book title was "The Etruscans Learn to Speak" - publilshed in late 1960s (?)

anonymous
Guest 07/21/2010 15:43 PM

It looks like the book is: Zĕchariă Mayani. The Etruscans Begin to Speak. [London] : Souvenir Press, [1961]. Search for the title at WorldCat (dot) org.

anonymous
Steve 07/21/2010 13:38 PM

Language has the mechanical element of patterns and syntax. Simple object identification and practical communication are easy to deduce. But when you get into semantics and meanings, the computer is virtually worthless.
Human consciousness is infinitely complex. Innuendo, sarcasm, subterfuge and irony are expressions of underlying emotions. Intonation, emphasis and expression are infinitely complex factors that influence language.
The computer will help us with the.... More

anonymous
David 07/21/2010 13:17 PM

Whether out of interest or Religious compulsion, understanding Ugaritic aids in understanding the Jewish Bible. Fundamental human insights are fundamental human insights - enjoyed by ancients and moderns alike (well, some...) Technology, not so much. Moderns have ancients beat hands down. We have stood on too many of those giants' shoulders.

anonymous
Mathew 07/21/2010 13:12 PM

The algorithm is a simple inversion my friend. What the wife says is the opposite of what she means, unless she is testing you with a paradoxical question like "Does this dress make me look fat?" to which there is no correct answer.

anonymous
Kevin Handy 07/21/2010 13:04 PM

I've done hand translations from English to German and from German to English as well as Spanish to English and so on. What you tend to find is that patterns are evident. It wouldn't matter if the target language was European or non-European the point is that in every language there is syntax, there are patterns - and I know someone will point to an example where there appear to be none. But if you think about language and how much you can do with 500 - 700 words of knowledge - that tells you.... More

anonymous
is this legal? 07/21/2010 12:55 PM

But can it translate what my wife says into what she really means?

anonymous
ZeroCoast 07/21/2010 12:54 PM

Ancient men may have been more intelligent than many people give them credit for. However a major difference that can be measured is the access of information that modern people enjoy, that hardly anybody throughout history had, up until the past century.

anonymous
richard 07/21/2010 12:25 PM

C-3PO?

anonymous
Ronin 07/21/2010 12:23 PM

Elvis and god reflect the mindset of a lot of researchers. Beginning from the perspective that everyone besides the researcher is less intelligent or somehow less sophisticated.

Ancient man designed and built structures that we are still incapable of building today. Ancient man knew enough to chart the stars and understand the workings of nature...but yet he had "far less knowledge and intelligence." What short sighted drivel.

anonymous
Don F. 07/21/2010 14:45 PM

I am not sure your statment "we are still incapable of building today". I think we are most capable of building today virtually any ancient structure. What do have, in many cases, is the knowledge as to how the ancients built them in the context of their culture its technological limitations. We also don't, in many cases, know by what processes they developed their technological prowess.

anonymous
Brennen 07/21/2010 12:34 PM

All of those structure were accomplished only because everyone spoke the same language. Imagine it that were the case today. What humans could build.

anonymous
Don F. 07/21/2010 14:48 PM

I think you are over stating things. I am assuming you are refering to the tower of Babel story which only refers to the tower of Babel. Its relevence either chronologically or situationally to the construction of any other ancient marvel is questionable at best.

anonymous
Biblical Scholar 07/21/2010 12:58 PM

What are you talking about? Are you buying into the Tower of Babel myth?

Add your comment

Sign in with one of these accounts or just add your comment below.
Used only for emailed comments and will not be displayed with your post
Notify me with an email when other people comment on this article.
The posting of advertisement, profanity or personal attacks is prohibited.
Click here to review our Terms of Use

ADVERTISEMENT

TOP MEMBERSJoin Now

ADVERTISEMENT