No, computers can’t prove what Shakespeare really wrote

Did Shakespeare write Shakespeare? It’s unsurprising that Darren Freebury-Jones, a lecturer at the Shakespeare Birthplace Trust in Stratford-upon-Avon, asserts that he did. What’s more surprising is that, as this new book expertly shows, he did not write alone. Shakespeare’s Borrowed Feathers is less about the “upstart crow” who upset the gossipy, close-knit world of the Elizabethan theatre, and more about those other writers whose influence, rivalry and collaboration shaped the canon we attribute to Shakespeare solo.

Here we meet some familiar figures, including the blazing star Christopher Marlowe, who transformed the stage with his poetic ambition, and the fashionably Italianate John Fletcher, with whom Shakespeare wrote his last plays. The verbal wit and gender fluidity of his comedies owe much to John Lyly’s courtly style, and traces of some of Ben Jonson’s plays, in which Shakespeare acted, show up in his writing. Noting some phrases repeated across Every Man in His Humour and Othello, Twelfth Night and Hamlet, Freebury-Jones makes a cogent case that Shakespeare was cast as Jonson’s character Matheo, a chump who writes terrible poetry.

First, however, we meet Freebury-Jones’s hero: Thomas Kyd. For the first time in over four centuries, Kyd is currently having a moment. Alongside his revenge-tragedy blockbuster The Spanish Tragedy, there is a large-scale reattribution to him of numerous currently anonymous plays. A critical tug-of-war over his relationship to Shakespeare is focused in particular on Arden of Faversham, a true-crime story in which the wayward Alice and her nogoodnik lover Mosby murder Alice’s husband. Authorship of this play is highly contested ground among scholars. An unkind assessment might recall Jorge Luis Borges’ two bald men fighting over a comb: Freebury-Jones explains both depilated arguments politely while making his view very clear that the prize belongs not to Shakespeare but to Kyd.

The Elizabethan theatre world was small: Shakespeare must have known Kyd, and Freebury-Jones shows shrewd literary judgement when discussing what he learnt from him. He revives the old idea about the so-called Ur-Hamlet, an earlier, now-lost version of Shakespeare’s most famous tragedy, and contends that it was by Kyd. He makes a persuasive case that the incomplete framing device around The Taming of the Shrew draws on Kyd’s stagecraft in The Spanish Tragedy, and sees echoes in structure and characterisation of Kyd’s other tragedies in Othello and Julius Caesar. The question, however, of when such correspondences become not evidence of influence, but of authorship, is almost impossible to definitively answer.

Freebury-Jones is alert to these difficulties, but keen to nail authorship nevertheless. You might suspect him of having a special agenda to build Kyd’s canon. Freebury-Jones’s extensively technocratic process relies on an online play text database called, unpromisingly, Collocations and N-grams, which he describes as “a revelation” that has “raised the discipline [of attribution studies] to a new level”. It’s a much less compelling method than his sensitive literary analysis. N-grams are continuous strings of words that computerised searches can find and compare across texts: the database makes a spreadsheet for each play ranking its n-gram parallels with other works. If reading by spreadsheet sounds soulless – well, it is. It highlights not exceptional phrase-making but repeated, often banal snippets: it’s hard to register duplicates such as “Heaven bless thee” or “cause I cannot” as meaningful authorial signatures.

N-gram results are also murky. Freebury-Jones presents, triumphantly, the list of the top 10 parallels for Macbeth, which indicates that four Shakespeare tragedies show significant verbal overlap. But that leaves six other plays which are not by Shakespeare. And, in a later chapter on Shakespeare’s work with his collaborator Thomas Middleton, which many scholars believe includes Macbeth, Freebury-Jones has to admit that the database is unhelpful in identifying common forms of co-authorship such as revision or collaboration.

On the other hand, a low n-gram score is used to deny significant relationships between plays. John Marston’s Hamlety (and extant) Antonio’s Revenge gets short shrift compared to Kyd’s lost Hamlet. Too often, “computer says no”. The problem is that N-gram parallels assume that the most important connection between plays is verbal, but playwrights for the commercial theatre were surely more on the look-out to snaffle bits of stage business, clever uses of props and other features of performance: all of which, happily, elude easy quantification.

Freebury-Jones is the first writer to bring the mysteries of computerised stylistic analysis to general readers, although his prose retains more than a touch of the lecture room. His account is mostly clear and generous, if sometimes credulous. But despite his modern techniques, the overall tendency of the book is conservative. Shakespeare’s redeployment of Marlowe’s innovative iambic pentameter, for instance, demonstrates “more flexible verse, more nuanced rhetoric and an overall suppler dramatic style”. To put it another way, Shakespeare was better than Marlowe. Although Freebury-Jones ends with the injunction to celebrate “the fact that in his lifetime Shakespeare was one member of a broad and brilliant community of playwrights”, his argument makes this pursuit entirely Shakespeare-centric. According to TS Eliot, “mature poets steal” and “make it into something better, or at least something different”. As this book reveals, that was how the once upstart crow became the swan of Avon.

Emma Smith’s books include This is Shakespeare. Shakespeare’s Borrowed Feathers is published by Manchester UP at £25. To order your copy for £19.99, call 0808 196 6794 or visit Telegraph Books