The 17th century playwright Molière is as important to French literature as William Shakespeare is to the English canon. But for the past 100 years, a question has swirled around him: Did Molière really write his plays? Or was Pierre Corneille, another famous French playwright of the time, the true author? A new study uses computational methods to analyze subtle, unconscious elements of both authors’ writing and concludes that Molière did indeed write the plays attributed to him.
Moliere is best known for his comedies, such as Tartuffe and Dom Juan. Yet in the early 20th century, some academics began to question his authorship. In 1919, for example, French writer Pierre Louÿs noted that Molière spent most of his life as a traveling actor and suddenly started to write masterpieces at about age 40. What’s more, scholars have never found an original manuscript signed by Molière.
Louÿs proposed the more educated Corneille as a possible ghostwriter, suggesting he composed plays that Molière would affix his name to and promote using his fame as an actor, in a mutually beneficial relationship. “It’s easy to be suspicious,” says Florian Cafiero, a computational linguist at CNRS, the French national research agency in Paris. (Similar doubts have occasionally been cast on Shakespeare’s authorship because of his lack of formal education.)
Suspicions deepened in the early 2000s after researchers noticed significant overlap in the word choices of the two playwrights. These linguists concluded that Corneille must have been the true author of Molière’s plays.
Cafiero and Jean-Baptiste Camps, a computational philologist at the École Nationale des Chartes, part of PSL Research University in Paris, have now brought a new technique to the debate. They compiled the text of comedies attributed to Molière, Corneille, and 10 of their contemporaries and used a sophisticated computer program to analyze and compare the linguistic features.
Rather than just vocabulary, the duo homed in on subtle characteristics like the frequency of “function words”—such as “the,” “that,” and “of”—which serve to create relationships between the other words in a sentence. They also looked at each author’s preferred grammatical structures and other linguistic patterns that unconsciously sneak into people’s writing. “They’re looking at lots of different indicators and seeing if they all say the same thing,” says Patrick Juola, a computational linguist at Duquesne University in Pittsburgh, Pennsylvania. “That’s absolutely the way to go.”
The computer program grouped together plays that had similar linguistic traits. If Corneille wrote the plays attributed to Molière, the two authors’ works would have ended up in the same group. But “whatever feature we study, Molière is clustered with Molière and far from Corneille,” Cafiero says. That’s good evidence that Molière did, in fact, write his masterpieces, the team reports today in Science Advances. “We are really, really convinced,” Cafiero says.
Literature experts aren’t surprised. Despite the fact that 17th century French theater had strict genre conventions that resulted in superficial similarities between authors, Molière does have a unique, identifiable voice, notes Joan DeJean, a professor of French literature at the University of Pennsylvania.
Adam Hammond, a literature professor and digital humanities expert at the University of Toronto in Canada, wonders why 20th and 21st century readers care so deeply about authorship in the first place. Ongoing, uncredited revisions, especially in theater, and retelling common stories were the norm in the past, he notes. “We care a lot more about who wrote these plays than Molière or Shakespeare did.”
*Correction, 2 December, 12:05 p.m.: This article has been updated to correct the spelling of the play Dom Juan.