Ancient languages reconstructed by computer program

Good to see technology utilized for a good cause...for a change I'm sure there are several language softwares and companies out there but this would definitely help the much needed help in decoding and/or understanding the languages of the ancients. It still has some quirks (as expected with any new system).

A new tool has been developed that can reconstruct long-dead languages.

Researchers have created software that can rebuild protolanguages - the ancient tongues from which our modern languages evolved.

To test the system, the team took 637 languages currently spoken in Asia and the Pacific and recreated the early language from which they descended.

LINK1
LINk2

Now this is cool. Thanks for posting this. This is one example of how computers are being incredibly useful to us. They are resurrecting our past for us in a way that traditional archaeology can't.

Starred and flagged as something for history and etymology buffs to follow.

Code breaking...sort of.
I wonder, with the rapid extinction of existing languages could this be a tool for preservation?
www.rferl.org...

It doesn't do anything that the experts haven't been doing since language reconstruction started in the 19th century.

Just does it a lot quicker.

cool thing, would be cool if they represent some of these old languages, i wonder how to curse someone in x000year old language

offtopic: i do like chemistry, they do cool stuff,
crawley: chemistry?! our bodies does chemistry the whole time, everyone can "do" that, not impressed at all
tony: :/

erm, is it magic or basically a specialized computer that just does one task? Copy /replicate or something?


Or is it interpretation what the languages "might" be I wonder if they put pig Latin into it what or how its "predecessors" look since its a fake/ fun pretend language,.


Anyways I guess what IM asking is it working of predefined set of circumstances "guidelines" or are we just putting languages into it and hoping they are correct.


I mean how do we even know that the human versions are correct?

hmmm.... i'd like to see one that can accurately read sumerian glyphs



until then, suppose ill just watch ancient aliens hehe

reply to post by ~widowmaker~

erm, is it magic or basically a specialized computer that just does one task? Copy /replicate or something?

It is a program that applies rules of "phonetic drift" worked out by German philologists in the 19th Century. The idea is that modern languages share "cognates," similar words, that are derived from an earlier parent language. For example, "water" in English, "Wasser" in German and "voda" in Russian. All (presumably) came from a word like "vada" in some earlier language.

Or is it interpretation what the languages "might" be I wonder if they put pig Latin into it what or how its "predecessors" look since its a fake/ fun pretend language,.

It works by comparing several related languages. It might be fun to run Pig Latin versions of English, French and Spanish just to see what comes out.

Anyways I guess what IM asking is it working of predefined set of circumstances "guidelines" or are we just putting languages into it and hoping they are correct.

It applies rules that were worked out by scholars, hoping they are correct.

I mean how do we even know that the human versions are correct?

A fair question. There are no surviving Indo-European or Proto-Semitic speakers around to test the theory. Nevertheless, it is a good one that is generally accepted by linguists.

Seems pretty cool. Any chance this could be/has been applied to the Voynich Manuscript? Its text follows the structure of a language, but none currently known.

Voynich Manuscript on ATS

reply to post by Phage
Well didn't UK stored all the library material underground in a fortified bunker after scanning the documents? Much of history has been lost so I wouldn't be surprised if the software can do just that. Preserve that is.

Originally posted by CJCrawley
It doesn't do anything that the experts haven't been doing since language reconstruction started in the 19th century. Just does it a lot quicker.

Experts cannot be trusted upto certain extent (atleast in the intelligence community ). Softwares are much more reliable (when programmed correctly) for translation services.

reply to post by ~widowmaker~
Precisely as per what is a correct way of pronouncing After all, every language has a set of rules.

reply to post by johnsequitur1221
Good question. Never know but who knows if it is being tested in the background on it.

reply to post by johnsequitur1221

Seems pretty cool. Any chance this could be/has been applied to the Voynich Manuscript? Its text follows the structure of a language, but none currently known.

No. There is no other example to compare it with. Also, it is not clear whether or not the random characters in the manuscript reflect and underlying language. Most likely it is a 16th Century forgery intended to defraud Emperor Rudolf.

I'm... actually going to want to see some names and papers here.

Now, I do understand what's going on -- but the "Monte Carlo Markov Chain" (which, yes, I do understand) process applied to linguistic phonemes really seems the wrong way to go about it. It *may* be able to approximate a set of base sounds, but I am not confident of its ability to reconstruct (for example) Latin out of the languages that evolved from it.

Playing with linguistics without incorporating grammatical (structural) rules is a very well known method of producing bogus results.

If I sound outrageously suspicious, remember that linguistics is a subdiscipline of anthropology and "sensemaking" is a subdiscipline of Information Science. I'd like to be reassured that this is something more than "let's play with phonemes."

But, I'm curious. So a bit later, I'm going to go hunt up names and papers.

Actually, after looking at the paper, I'm even more dubious. The authors are statisticians, programmers, and psychologists -- nobody appears to actually have any linguistic background, other than having a linguist who specializes in Pacific languages confirm that something they did worked 85% of the time.

Yeah, I'm an old stick in the mud. I think I'll cruise LingList and see what the linguists say.

My Kindle comes equiped with a dictionary and if I move the curser next to a word in the text I am reading it gives the definition of the word and the origin. Most of the english words go back to either greek, french or latin. These languages evolved from the earliest languages. There are only so many sounds that the human mouth can make. Very interesting to know what our ancestors saying and how they were saying it.

Originally posted by hp1229

Originally posted by CJCrawley
It doesn't do anything that the experts haven't been doing since language reconstruction started in the 19th century. Just does it a lot quicker.

Experts cannot be trusted upto certain extent (atleast in the intelligence community

). Softwares are much more reliable (when programmed correctly) for translation services.

Wait..... who is doing the programing again? oh would that be the "experts". Me thinks so. ( Taken from old english meaning I think this is true) Big LOL

I'm gonna post a reply just because i'd be shocked if it could translate this one sentence into old english.

Originally posted by hisshadow
I'm gonna post a reply just because i'd be shocked if it could translate this one sentence into old english.

It actually didn't say that... and which flavor of Old English would you like your sentence translated into? Try this site (but you may not be able to interpret the results)