Geoffrey Chaucer's "Canterbury Tales" may have influenced European literature for centuries, but scholars have been arguing for nearly as long about what Chaucer's exact words were.

But now biochemists at Cambridge University are using computer programs to analyze different versions of Chaucer's bawdy tales to reconstruct the text from the original 14th century manuscript, which has not survived.A report published in the scientific journal Nature on Wednesday said programs normally used for genetic research had been adapted to analyze various manuscripts of one of the seminal works of English literature.

Chaucer's original text has been lost and later manuscripts often vary considerably, leading scholars to argue over which versions were copied from the original and which are most accurate.

"The sheer quantity of information in . . . the `Canterbury Tales' defeats any system of manual analysis," said the report by Christopher Howe.

Using the principles of genetics, the computer compares different versions of the text, analyzing small changes made when books were copied by hand before the advent of printing.

Previously, such research had to be done manually, involving hours of painstaking work even for short texts.