Make a Donation

Consider making a donation to support SHAKSPER.

Subscribe to Our Feeds

Current Postings RSS

Announcements RSS

Home :: Archive :: 2004 :: January ::
Shakespeare et al. and the New "Science"
The Shakespeare Conference: SHK 15.0197  Monday, 26 January 2004

[1]     From:   R. A. Cantrell <
 This e-mail address is being protected from spambots. You need JavaScript enabled to view it
 >
        Date:   Friday, 23 Jan 2004 10:42:37 -0600
        Subj:   Re: SHK 15.0170 Shakespeare et al. and the New "Science"?

[2]     From:   Ward Elliott <
 This e-mail address is being protected from spambots. You need JavaScript enabled to view it
 >
        Date:   Friday, 23 Jan 2004 18:54:12 -0800
        Subj:   RE: SHK 15.0170 Shakespeare et al. and the New "Science"


[1]-----------------------------------------------------------------
From:           R. A. Cantrell <
 This e-mail address is being protected from spambots. You need JavaScript enabled to view it
 >
Date:           Friday, 23 Jan 2004 10:42:37 -0600
Subject: 15.0170 Shakespeare et al. and the New "Science"?
Comment:        Re: SHK 15.0170 Shakespeare et al. and the New "Science"?

 >I'm working at the connection between Shakespeare and the new internet
 >world. I 'm looking for a list of new technologically most used internet
 >resources in the studying and analyzing of humanities and literature.
 >Also articles, opinions, reviews, methods, way of looking forward this
 >new "science".
 >
 >Can anyone give me a hand?

You've hit upon a happy phrase "New Science." It may be a few years
coming, but Stephen Wolfram's "New Kind of Science," will, I think, have
a staggering impact on stylometrics and other quantitative analytical
approaches to Shakespeare (inter alia).

[2]-------------------------------------------------------------
From:           Ward Elliott <
 This e-mail address is being protected from spambots. You need JavaScript enabled to view it
 >
Date:           Friday, 23 Jan 2004 18:54:12 -0800
Subject: 15.0170 Shakespeare et al. and the New "Science"?
Comment:        RE: SHK 15.0170 Shakespeare et al. and the New "Science"?

I can't tell whom I'm supposed to respond to on this, so shall just hit
the reply button.

Rob Valenza and I have experimented several times over the years testing
various student groups' powers to guess authorship by intuition alone.
In 2002, as if in anticipation of the many warnings we have since gotten
from Shaksperians to ditch our unseemly crunching and counting and get
back to reading, we tested a class of 12 Claremont McKenna College
undergraduates with a selection of 28 sonnet-length passages, half by
Shakespeare, half not.  As individuals, the students were all over the
place, doing, on average, modestly better than chance, which would be
around 50%.  Half of the students, however, did much better than chance,
getting 70-79% of the ascriptions correct.   Moreover, as a group, the
whole class guessed better than the best of its individual Golden Ears,
better, in fact, than the aggregated vote of all six of its Golden Ears.
  Some accuracy rates:

Worst individual:               54% correct
Best individual                 79% correct
6 best combined                 84% correct
All 12 combined                 89% correct

18 of the group's 25 successful identifications (72%) were by lopsided
votes, 8-4 or higher.

How does this accuracy compare with that of our best quantitative tests?
  "Far higher" would be a persuasive answer for such short,
Sonnet-length samples.  All of our quantitative tests are sensitive to
sample length because longer samples average out more variance than
shorter ones, giving us tighter ranges and higher discrimination for
long samples than for short.  Most of the samples we used in our Golden
Ear test have no more than 150 words, far shorter than any for which we
have dared to validate any of our quantitative tests.  For comparison,
our current estimated composite accuracy rates for longer,
single-authored passages look something like this:

Text                    Shakespeare     Non-Shakespeare

Whole plays                     100%            100%
Poems, 3000 words               100%            100%
Play Verse, 3000 wd             95%             100%
Poems 750 words                 93%              71%
Play Verse 750 words            97%              75%
Poems, 470 words                92%              73%

These figures say that 470-word passages are at or below the rock bottom
of our comfort zone, with a combined accuracy of around 83% -- yet our
students, in aggregate, have achieved accuracy of 89% on pure intuition.
  It's true that, by itself, these results would not be quite enough to
validate CMC students' intuition by the same rules that we use to
validate computer tests.  Because we rely on negative evidence ("silver
bullets"), our accuracy requirements have been asymmetrical.  We have
looked for much higher accuracy - normally 95% or better -- in saying
"could-be" to Shakespeare than in saying "couldn't be" to
non-Shakespeare, on the reasoning that, if you test by primarily by
negatives, false negatives are much more damaging to your case than
false positives, and you should keep them to a minimum.  Unfortunately
for us, the tested class got only 82% of their Shakespeare passages
right, far too many false negatives to make up for their phenomenal
accuracy of 96% with non-Shakespeare passages.  We wish it had been the
other way around.

Moreover, while it is conceivable that using longer passages would raise
our Golden-Ear accuracy (though many of our lit colleagues doubt it),
it's hard to imagine that we could ever get test-takers to sit still for
a test with, say, 28 3,000-word passages instead of 28 100+word
passages.  Such a test would be equal in length to Hamlet, Macbeth,
Romeo and Juliet, and The Comedy of Errors combined and would take more
than a day just to read, let alone analyze, in entirety.  In general,
the longer the passages, the fewer can be tested without performing
miracles of motivation.  From this perspective, Golden-Ear testing may
be almost as impractical for wholesale testing of long passages as
computers are for testing short passages.

These points aside, our pilot study still seems to us an impressive
performance by undergraduates and an impressive clue as to what might be
done with more and better panelists.  What if, instead of a dozen
undergraduates, it were every interested SHAKSPERian?  Or every
interested SHAKSPERian who scores above a certain level on a
standardized Golden-Ear test, whichever produces the best combination of
aggregation and selection?

Such analysis is screaming to be done on the net, and it might well cast
new light on authorship questions left unanswered by computers and
individual "sniff-testing."   To make it work, we, or someone else,
would have to put a Golden-Ear test on line for a group like Shaksper
and do some analysis of the results.  Last year, with quite a bit of
trial and error, we managed to get our test up and running on the net,
and Weston Thompson, CMC's web editor, is now working on ways to record
the results for analysis.  When it's ready to go, SHAKSPERians will be
among the first to know, and will be invited to give our test, and their
Golden Ears, a trial run.  We hope it will be sooner, rather than later.
   In the meantime, if someone is looking for innovative uses of the web
for teaching and learning more about Shakespeare, we've got one on the way.

Yours,
Ward Elliott

_______________________________________________________________
S H A K S P E R: The Global Shakespeare Discussion List
Hardy M. Cook, 
 This e-mail address is being protected from spambots. You need JavaScript enabled to view it
 
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.
 

©2011 Hardy Cook. All rights reserved.