Canadian computer scientists say a computer can reconstruct lost languages by analyzing the sounds uttered by those who speak their modern successors.

Alexandre Bouchard-Cote at the University of British Columbia in Vancouver and colleagues report a machine-learning algorithm can suggest the most likely phonetic changes behind a language's evolution as words mutate and sounds shift.

As an example of sound shifting, the researchers cite a recent change known as the Canadian Shift, where many Canadians now say "aboot" instead of "about."

"It happens in all words with a similar sound," Bouchard-Cote told NewScientist.com.

The researchers used the computer algorithm to analyze thousands of word pairings in 637 Austronesian languages, a language family that includes Fijian, Hawaiian and Tongan.

The system was able to suggest how ancestor languages might have sounded and also identify which sounds were most likely to change, they said.

The technique could help preserve endangered languages if they are phonetically related to more widely spoken tongues, Bouchard-Cote said.