Page last updated at 09:23 GMT, Thursday, 26 February 2009
'Oldest English words' identified Medieval manuscripts give linguists clues about more recent changes Some of the oldest words in English have been identified, scientists say.
Reading University researchers claim "I", "we", "two" and "three" are among the most ancient, dating back tens of thousands of years.
Their computer model analyses the rate of change of words in English and the languages that share a common heritage.
The team says it can predict which words are likely to become extinct - citing "squeeze", "guts", "stick" and "bad" as probable first casualties.
"We use a computer to fit a range of models that tell us how rapidly these words evolve," said Mark Pagel, an evolutionary biologist at the University of Reading.
"We fit a wide range, so there's a lot of computation involved; and that range then brackets what the true answer is and we can estimate the rates at which these things are replaced through time."
Sound and concept
Across the Indo-European languages - which include most of the languages spoken from Europe to the Asian subcontinent - the vocal sound made to express a given concept can be similar.
New words for a concept can arise in a given language, utilising different sounds, in turn giving a clue to a word's relative age in the language.
At the root of the Reading University effort is a lexicon of 200 words that is not specific to culture or technology, and is therefore likely to represent concepts that have not changed across nations or millennia.
"We have lists of words that linguists have produced for us that tell us if two words in related languages actually derive from a common ancestral word," said Professor Pagel. When we speak to each other we're playing this massive game of Chinese whispers
Mark Pagel, University of Reading
"We have descriptions of the ways we think words change and their ability to change into other words, and those descriptions can be turned into a mathematical language," he added.
The researchers used the university's IBM supercomputer to track the known relations between words, in order to develop estimates of how long ago a given ancestral word diverged in two different languages.
They have integrated that into an algorithm that will produce a list of words relevant to a given date.
"You type in a date in the past or in the future and it will give you a list of words that would have changed going back in time or will change going into the future," Professor Pagel told BBC News.
"From that list you can derive a phrasebook of words you could use if you tried to show up and talk to, for example, William the Conqueror."
That is, the model provides a list of words that are unlikely to have changed from their common ancestral root by the time of William the Conqueror.
Words that have not diverged since then would comprise similar sounds to their modern descendants, whose meanings would therefore probably be recognisable on sound alone.
However, the model cannot offer a guess as to what the ancestral words were. It can only estimate the likelihood that the sound from a modern English word might make some sense if called out during the Battle of Hastings.