Descendants of the Wife of Bath

The study of medieval literature may seem far removed from the rigor of 20th-century science. But techniques from evolutionary biology have helped scholars to sort out the relationships among medieval manuscripts of The Canterbury Tales by Geoffrey Chaucer. The results, reported in Nature tomorrow, should help scholars reconstruct Chaucer's long-lost original text.

One of the puzzles for scholars of ancient literature is determining what the authors actually wrote. For many years, researchers have tried to identify manuscripts closest to the missing original by comparing the changes and errors that were copied from one manuscript to another--just as evolutionary biologists construct biological family trees by comparing inherited changes in DNA.

After hearing a friend's lecture on ancient manuscripts, biochemist and evolutionary biologist Christopher Howe of the University of Cambridge says he suddenly realized, "hang on, that's what I've been doing all morning." He wondered if the computer techniques he uses to determine the relationships among chloroplasts in plants might help the classicists. He teamed up with Chaucer scholar Peter Robinson of De Montfort University in Leicester, U.K., who heads the Canterbury Tales project, an effort to determine what Chaucer originally wrote.

Robinson and his colleagues had already used evolutionary biology techniques to analyze their computerized database of 58 copies of "The Wife of Bath's Prologue," from The Canterbury Tales, all transcribed before 1500. When Howe introduced them to a much faster and more sophisticated program, the computer identified 11 copies that seem to have no close relatives. The unique changes and errors suggest that these may have been copied directly from Chaucer's lost original--which was likely unfinished--or early copies of the original. A careful analysis might help scholars determine whether the particularly lascivious passages in some versions were Chaucer's or a later scribe's.

Although Robinson says his field "has a large number of people who don't like computers," and who are suspicious of the analysis, he says others are picking up the technique, including a group in Germany hoping to analyze 5000 copies of the Greek New Testament, and another group working on Dante's Divine Comedy. Says Howe, "it has been fascinating to see how the problems [in the two fields] are the same, just with different terminology applied to them."