When the Raw Data Is Hidden, the Conclusions Mislead
MacDonald P. Jackson. Defining Shakespeare: Pericles as Test Case. 272pp, hardback, 1 line illustration. $180. 6X9”. ISBN: 978-0-199260508. Oxford: Oxford Scholarship Online: Oxford University Press, 2003.
*****
My own linguistic analysis determined that Pericles’ “Shakespeare”-attributed Acts 1 and 2, and its frequently contested Act 3 were both written by Munday. Acts 1 and 2 are slightly more collaborative including signatures matching Fletcher and Munday as well as other highly collaborative works. As I mentioned earlier in this set of reviews, I determined that Wilkins’ Enforced Marriage play was ghostwritten by Drayton, so it is likely this study used illogical, flawed, biased or otherwise misleading methods to reach its proposed conclusions. When Wilkins’ Enforced is compared with Pericles’ Acts 1 and 2, they only match on 2 out of 28 tests, indicating extreme dissimilarity, or a certainty they are by divergent authors. If scholars believe the attribution conclusions of this particular study, they would be believing the opposite of what the math indicates regarding these texts.
Here is what it attempts to prove: “‘That very great play, Pericles’, as T. S. Eliot called it, poses formidable problems of text and authorship. The first of the Late Romances, it was ascribed to Shakespeare when printed in a quarto of 1609, but was not included in the First Folio (1623) collection of his plays. This book examines rival theories about the quarto’s origins and offers compelling evidence that Pericles is the product of collaboration between Shakespeare and the minor dramatist George Wilkins, who was responsible for the first two acts and for portions of the ‘brothel scenes’ in Act 4. Pericles serves as a test case for methodologies that seek to define the limits of the Shakespeare canon and to identify co-authors. A wide range of metrical, lexical, and other data is analysed. Computerized ‘stylometric’ texts are explained and their findings assessed. A concluding chapter introduces a new technique that has the potential to answer many of the remaining questions of attribution associated with Shakespeare and his contemporaries.”
It is troubling how trusting “Shakespeare” literature scholars have been in the computer scientists running these types of “stylometric” studies that are leading them further from the truth of the matter. I look forward to examining what went wrong in this study as I continue my research. Other scholars who are also interested in correcting past missteps in this field might benefit from reading this project, but curious or casual readers of “Shakespeare” should avoid it as it will misdirect them towards erroneous conclusions without realizing they are being misguided.