Some Problems in the Non-Traditional Authorship Attribution Studies of the Dramatic Canon of William Shakespeare: Are they Insurmountable

William Shakespeare is arguably the greatest dramatist of all time. Yet, the man and his works are shrouded in mystery and uncertainty. This paper posits that a man named Shakespeare wrote the First Folio. It is this First Folio that provides the “most certain” body of Shakespeare’s plays.
The makeup of the Shakespeare dramatic canon has prompted more attribution studies (traditional and non-traditional) and caused more controversies than any other canon – by far. This paper looks at the several hundred non-traditional studies (and concomitant “flame wars”) and points out some of the more serious problems. There is no doubt that it is a canon in disarray. Most of the scholars involved in the controversies are the “heavyweights” of Shakespearean studies – e.g.:

Rasmussen (1977) vs. Hope (1994)
Vickers (2011) vs. Craig and Kinney (2009)
Taylor (2015) vs. Stern (2004)
Craig vs. Vickers and Jackson (Hirsch and Craig, 2014)

But the most famous controversies involve (1) Donald Foster vs. Ward Elliott and (2) Robert Valenza and Donald Foster vs. the world.
In a recent article (Rudman, 2016) I pointed out many caveats to scholars working on authorship attribution on the canon of William Shakespeare. Among these are:

Reproducibility
Input Texts
Genre
Editing
Controls
Isolation of Variables
Choice of Style Markers
Statistical Tests
Sample Selection and Size
Treatment of Errors
Collaboration

In this paper, I expand on one of these caveats (Genre), look at and cite examples from many of the non-traditional studies of the Shakespeare canon in order to highlight these problems, and suggest solutions.

Each genre is governed by different linguistic rules and rhetorical purpose – a practitioner should not mix genre.
Drama – Comedy, History, Tragedy, Romance – how far down should these be catagorized
Verse within the drama – rhymed verse within the verse – how far down should this be broken
Music
Dialogue vs Dramatic Monologue

Also in this paper, I address the conundrum of using what I deem as seriously flawed studies to show problems with other studies – e.g. if a practitioner mixes genres (history and tragedy) but shows that Shakespeare's style changes over time, I cite this change as evidence that chronological constraints must be employed.
No matter how flawed I consider a study, there are parts of that study that are correct – there is no non-traditional study (of the hundreds conducted) that is completely without merit. By invoking the
etiam si non est verum paradigm, I show how almost all of the necessary steps in a valid study exist in the literature. We can look at all of the links in he chain (even the broken ones) and try to piece together what we should do to move the field forward. The evidence in this paper reinforces my conclusion from the
JEMS article that any attribution results are problematic at best.

The Conclusion reached (and I believe supported by very strong evidence) is that all of the non-traditional studies are seriously flawed and that as of today we do not have a valid Shakespeare text to conduct non-traditional attribution studies.
The following bibliography is only representative. There are more than 200 non-traditional studies on the Shakespeare canon. My working bibliography for this study is well over 1,000 entries.

Bibliography

Bullough, G. (1957).
Narrative and Dramatic Sources of Shakespeare, Vol. 2(8), London: Routledge and Kegan Paul.

Burrows, J. (2012). A Second Opinion on Shakespeare and Authorship Studies in the Twenty-First Century.
Shakespeare Quarterly,
63(3): 355–92.

Busse, U. (2002).
Linguistic Variation in the Shakespeare Corpus: Morpho-syntactic Variability of Second Person Pronouns. Amsterdam–Philadelphia: John Benjamins.

Hirsch, B. D. and Craig, H. (2014). Mingled Yarn: The State of Computing in Shakespeare 2.0. In Bishop, T. and Huang, A., Hirsch, B. D. and Craig, H. (eds.),
The Shakespearean International Yearbook. Aldershot: Ashgate Publishing, preprint edition courtesy of publisher, 14: 3-35.

Hope, J., and Whitmore,
M. (2014). Quantifiaction and the Language of Later Shakespeare.
Actes des Congrès de la Société Française Shakespeare,
31: 123–49.

Horton, T. B. (1987).
The Effectiveness of the Stylometry of Function Words in Discriminating Between Shakespeare and Fletcher. Ph.D. Thesis, University of Edinburgh, 1987.

Lancashire, I. (2002). The State of Computing in Shakespeare. In Elton, W. R. and Mucciolo, J. M. (eds.),
The Shakespearean
I
nternational Yearbook.
Where are We Now in Shakespearean Studies?. Aldershot: Ashgate Publishing, 2: 89–110.

Mowat, B. A. (2003). Whats in a Name? Tragicomedy, Romance, or Late Comedy. In Dutton, R. and Howard, J. E. (eds.),
A Companion to Shakespeare’s Works.
The
P
oems, Problem Com
edies, Late Plays. Oxford: Blackwell Publishing, 4: 129–49.

Rudman, J. (2016). Non-Traditional Authorship Attribution Studies of William Shakespeare’s Canon: Some Caveats.
Journal of Early Modern Studies, forthcoming.

V
ickers, B. (2009).
Shakespe
are and Authorship Studies in the Twenty-First Century.
Review Essay.
Shakespeare Quarterly,
62(1): 106–42.

Witmore, M. (2009).
A Genre Map of Shakespeare’s Plays from the
F
irst Folio.
Wine Dark Sea, http.//www.winedarksea.org/?p=40, (Accessed May 6, 2015).