|Petr Plecháč used Shakespeare's works such as Henry VIII to train the AI model / Photo Credit: Anton_Ivanov (via Shutterstock)|
According to Emerging Technology from the arXiv, via magazine MIT Technology Review, a scientist claimed he knew where Shakespeare ends and his long-suspected collaborator begins, as cited by Caroline Delbert of popular science and technology news website Popular Mechanics. Petr Plecháč employed machine learning to train an algorithm on William Shakespeare and John Fletcher’s works. Literary scholars have speculated on the possibility of shared authorship in 1860 and Plecháč believes that their theories have been confirmed.
“Our results highly support the canonical division of the play between William Shakespeare and John Fletcher proposed by James Spedding,” he said. James Spedding was the scholar who put forth the theory of Shakespeare collaborating with Fletcher on Henry VIII. Plecháč’s approach is straightforward. Machine-learning algorithms were utilized to identify distinctive patterns in the authors’ writing styles. Plecháč used Shakespeare’s work to train the algorithm, namely “Henry VIII,” “The Tragedy of Coriolanus,” “The Tragedy of Cymbeline,” “The Winter’s Tale,” and “The Tempest.”
Every work should have the same literary style, as a writer’s style can evolve throughout their lifetime. Then. the algorithm was trained to recognize Fletcher’s work such as “Valentinian,” “Monsieur Thomas,” “The Woman’s Prize,” and “Bonduca.” The result? The algorithm confirmed Spedding’s theory. Plecháč found that “Fletcher wrote scenes amounting to almost half the play.” Interestingly, the algorithm also found changes in authorship in some scenes, and even to towards the end of previous ones. For example, the model suggested a mixed authorship after line 2081 in Act 3, Scene 2. It also found that Shakespeare took over at line 2200, before the beginning of Act 4, Scene 1.
Plecháč also trained the model to recognize Phillip Massinger’s works, another of Shakespeare’s long-time, suspected collaborator. He concluded, “The participation of Philip Massinger is rather unlikely.” But AI won’t be writing its own Shakespeare, Fletcher, or Massinger anytime soon. AI can point a correlation between one evidence to another, but it won’t create original content or formulate its own theories.