SAT Conference 2017 - 3 - Ros Barber - Stylometry: Hard Evidence or Pseudo-Science?

www.shakespeareanauthorshiptru...
Dr Ros Barber at the Shakespearean Authorship Trust conference 2017 delivers a talk on the methods behind the recent attribution, by the New Oxford Shakespeare, of a co-authorship credit to Christopher Marlowe for all three of Shakespeare's Henry VI plays.
Apologies for problems with the sound.
Filming by Tim Pieraccini.

[@anthonymccarthy4164]

This is a difficult but important lecture. I'd say stylometrics are probably trying to test things too complex to really come up with an accurate match of hypothesis and results and, as Dr. Barber points out, the practice in this case clearly discredits the practice. It's quite brave, especially as it has been used to get a result which particular groups like. I'm very impressed with Barber's scholarship, on a par with that of Diana Price in importance in this issue.

[@Jeffhowardmeade]

And yet, under controlled circumstances, styolometry gets it right time after time.

[@daver8521]

I am curious why no one has ever done a stylometric comparison of "Hero and Leander" with "Venus and Adonis" and "The Rape of Lucrece." It would seem that these works would be much better candidates for meaningful stylometric analysis than the plays. They are all written in the same medium, and are more or less contemporary works, whose texts have not been altered.

[@MandyJMaddison]

I have been wondering exactly the same thing.

[@tracik1277]

Thank you very much. I cannot think of a single reason to use the word death in this sentence x

[@ethelburga]

"I have absolutely no idea" Never a truer word spoken.

[@Jeffhowardmeade]

The irony is that Anti-Stratfordians were big on stylometry back when they thought it would prove their case. Marlovians especially touted an early study as proof that early Shakespeare was indistinguishable from Marlowe. As they started adding factors in, though, the patterns diverged. Now stylometry is a "pseudo-science", because it no longer produces the desired result. Surprise, surprise.

[@anthonymccarthy4164]

I wonder what the fact that the number of words in the "Shakespeare" corpus is so large does to the identification of particular words being "typical" or "atypical" of a "Shakespeare" style as opposed to someone who used a far smaller vocabulary. It would seem to me that if you use a 30% or 40% more words than another author that this attempted method of comparison runs into a real problem of comparing two different "things" with quite different characteristics. I've always been skeptical of stylometric and similar attempts at the scientific determination of things like authorship. This lecture is even better the second time listening to it.