The Shakespeare Conference: SHK 17.0653 Thursday, 13 July 2006
From: Martin Mueller <
Date: Thursday, 13 Jul 2006 08:02:36 -0500
Subject: Gender, Genre, Pronouns, and WordHoard
Do Shakespeare's men and women talk differently? The latest version of
WordHoard (http://wordhoard.northwestern.edu) lets you get at some of
the answers, and it is probably the only digital tool that lets you do
so with reasonable ease. WordHoard now has 'word sets' of words spoken
by men and women in the comedies, romances, tragedies, and histories,
conventionally defined. Some results are obvious, others puzzling.
Female speakers use words like 'alas', 'sorrow', 'woeful', 'prithee' and
'love' a lot more and do so consistently across genre. 'The' is a
consistent marker of male speech across genre. Pronouns, always an
interesting topic of inquiry, tell different stories by genre. Forms of
the first person singular (I, me, my, mine) are more common in female
speech of all genres. The statistical differences are astronomical. By
equally wide margins, forms of the first person plural (we, us, our,
ours) are more common in male speech in the tragedies, histories, and
romances. But 'we' is not a discriminator between male and female speech
in the comedies.
Forms of 'he' are somewhat more common in male speech of the tragedies,
while forms of 'she' are less common in male speech of the histories and
tragedies. Amusingly enough, the comedies behave quite differently. By
equally astronomic margins, 'she' is more common in male, and 'he' more
common in female, speech. 'You' is a female speech marker in the
comedies and romances, but not in the tragedies or history. In the
comedies, 'thou' is strongly associated with male speech. In the
histories 'thou' is a very strong, and in the tragedies a fairly strong,
marker of female speech.
It remains to be seen whether any of this really adds up to something.
Variations of individual character or different ratios of poetry and
prose may be as important as differences in gender and genre. A group of
computational linguists have recently claimed that there are robust and
simple speech markers that let you assign written documents to male or
female authors and get it right in 70-80% of the cases. One of them,
Shlomo Argamon at IIT, has asked whether these differences are preserved
in Shakespeare's imitation of male and female speech (see the abstract
of a talk at the Digital Humanities 2006 conference at
http://www.allc-ach2006.colloques.paris- sorbonne.fr/DHs.pdf). His
tentative findings are suggestive.
With relatively little investment of time you can now do a fair amount
of interesting work along these lines. In the old days before computers,
it would have taken weeks or months to assemble the relevant evidence.
In the early days of computers, you could cut that time to days provided
you had the programming skills. With new tools the time cost of
assembling the evidence is measured in minutes or hours. You can now set
a group of undergraduates to the task of figuring out what you can learn
about a particular Shakespeare play by attending carefully to the
distribution of 'little' words such as pronouns. More often than not
they may learn quite a bit, not to speak of the meta-lesson of what you
learn about 'big' questions by attending to small and humble detail.
S H A K S P E R: The Global Shakespeare Discussion List
Hardy M. Cook,
The S H A K S P E R Web Site <http://www.shaksper.net>
DISCLAIMER: Although SHAKSPER is a moderated discussion list, the
opinions expressed on it are the sole property of the poster, and the
editor assumes no responsibility for them.