Wow! some scientists created a computer program to reconstruct vocabularies of ancient languages using only their modern language descendants. A linguistic Time Machine!
The program, a linguistic time machine of sorts, can quickly crunch data on some of the earliest-known “proto-languages” that gave rise to modern languages.
When I met with Romani people in Europe, I wanted to hear their language, I right away found many Indian words in Romani language. They migrated from India 1500 years ago. Still you can trace the language they brought from their homeland. Without the help of a linguistic time machine we know about the origin of Romani language.
This time machine is just awesome. Will they be able to find out the origin of Basque language using modern Basque descendants? I have always been curious to know about the origin of Basque. Where did the Basque language come from? Nobody could give me a satisfactory answer. Hope the linguistic time machine will do. When I was in Bilbao, I learned some Basque words, which were not similar with any other words exist on Earth. Ear-belarria, eye-begi, nose-sudurra, mouth-ahoa, tooth-hortza, arm-besoa,hand-eskua,heart-bihotza,stomach-sabela,leg-hanka, foot-oina,head-burua,face-aurpegia,hair-ilea. I am eagerly waiting for the solution the linguistic time machine gives.
We look to the future. Sometimes it is necessary to look back to the past.
The trouble with Basque is that it has no relatives, so the only way for the algorithm to work would be to compare modern Basque to words from Basque that jumped to other languages at earlier times in history. And my suspicion is that there are few enough words (probably mostly if not entirely in French and Spanish) that that analysis has already been done pretty thoroughly by linguists.
I think Taslima was being a bit sarcastic, if I get the tone of the post right. The linguistic “time machine” has been (as scientific news always is) been greatly exaggerated and misrepresented by the main stream media. Basically it is an automated version of comparative linguistics, taking a group of related languages, and by pattern matching try to find out by which rules it has changed, thus reconstructing a prior version of the language. I’ve read that they reconstructed a single word from some Polenesian language, and claim it is close to the word that linguists have reconstructed. I’d be more impressed if by inputting French, Spanish, Portuguese, Italian and Romanian they were able to near-perfectly reconstruct 1st century Vulgar Latin (of which we have many documents, so we could actually check whether it were near-perfect).
I don’t know why the researchers selected the Austronesian languages. Could it be because there are a lot of them and because they have gotten a lot of study? That would make possible good statistics and good checking of their results.
I would have expected them to do Indo-European, the best-studied of the larger language families. It would be interesting to see what the software comes up with for (say) the ancestral Indo-European stop consonants. Would it recover the grid of 3 voicings and 4 or 5 articulation points? (t, d, dh; t, p, k, kw, another k) What would it conclude about their pronuniciation?