While this would seem to be an almost impossible task, linguistics researcher Dr. Aleks Steinbergs and a team of 21 researchers around the world are doing just that.
"What we are doing is a collaborative project; with scholars over North America and Europe we are working backwards in time to try and reconstruct what the words of Proto-Baltic -- a language spoken between 500 BC and AD 500 -- must have been," said Dr. Steinbergs, a member of Memorial's Department of Linguistics.
Proto-Baltic was a contemporary of Latin. Its linguistic grandparent was Proto-Indo European and its grandchildren are, among others, modern Latvian and Lithuanian.
"It's as if Latin had never been written down, if it had been only a spoken language the way Proto-Baltic was, and if you tried to reconstruct what specific Latin words would have been, having as the only evidence the languages descended from it," she said.
Dr. Steinbergs and her team take older forms of Latvian and Lithuanian words and trace them all the way back to their Proto-Baltic forms. This process of reverse evolution is guided by what linguistic research has revealed about language development.
"For example, if you knew what the word for 'milk' was in French, Spanish, Portuguese, Italian and Romanian -- all the languages that have descended from Latin -- then you could work backwards and say the common ancestor must have been 'X'. This would be based on what linguists know generally happens to pronunciation and meaning, and what is likely to happen, and what never happens in language formation."
Researchers have each been assigned "semantic fields" such as agricultural terminology (field, planting, dig, etc.), natural phenomena (earth, moon, etc.), or the human body, which is Dr. Steinbergs' own field. Using such fields ensures most words get covered and provides a source of words that were most likely in use 1,400 years ago. "Words for parts of the body, words like eye or mouth, tend to last; there is very little reason to change them."
In addition to formal training in linguistics, Dr. Steinbergs' background provides her with an advantage in this project. "I am a native speaker of Latvian. I've spoken Latvian all my life and I have been involved in doing research on Latvian basically all my career, including a PhD dissertation on the sound system of Latvian."
Her research led her to this project when she realized that quite a gap existed in her area of study.
"About eight years ago I was talking to somebody about the fact that people had done work on a Proto-Algonquian lexicon, the ancestor languages of a number of American Indian languages, and similar lexicons had been done for many non-European languages, yet in the Baltic family this had not yet been done, and it was a significant gap as far as scholarship goes."
While the idea for the dictionary was born nine years ago, work began four years later with a grant from the Social Science and Humanities Research Council. Dr. Steinbergs is roughly halfway through the project, expecting to publish about three years from now. Her research to date has yielded just over 700 entries, or about one-third of the projected 2,000 words needed to complete the dictionary.
The dictionary can't arrive soon enough. Dr. Steinbergs notes that the Proto-Baltic dictionary will be an important, if overdue, contribution to linguistics. "Baltic is a branch of Indo-European, it's a very important language family, yet this is one branch that for some reason the scholarship just hasn't been done. Whenever I tell a linguist what this project is about I get one of two responses. One is 'Oh! What a good idea!' and the other is 'You mean it hasn't been done yet?'"