The Shakespeare Conference: SHK 15.0663 Friday, 12 March 2004
Date: Thursday, 11 Mar 2004 07:26:16 -1000
Subject: Re: Stylometrics
Among of the difficulties of responding to stylometrical analysis is
that the extensive data are so detailed and checking them for accuracy
is both tedious and horribly time-consuming. Analysts' results however
are not always reliable. Take for example Alfred Hart's well-known data
for 'Words in Shakespeare beginning with Un-,' which Vickers
(Shakespeare Co-Author, Table 6.34, p. 427) cites as hard evidence. As
soon as we attempt to verify Hart's numbers we run up against questions
of definition. Should 'until, ' 'understand, ' 'unto, ' 'uncle,'
'under,' etc. be included? The answer seems clearly no, but what did
Hart take into account? The data implicitly represent negatived
positives, that is, existing words modified by 'un,' e.g,, 'unnatural,'
but what about marginal calls like 'unless' and 'unusual'?
Attempting to repeat Hart's exercise, that is, check his results, I
included 'unusual' but omitted 'unless.' I also left out stage
directions. like 'unveiling' and 'unmuffling,' since they are
theatrically silent. But did Hart?
Here are my results. Hart's figures are given first, followed by my count:
2 Henry VI: 34, 44; Romeo and Juliet: 44, 57; Richard II: 52, 61; 1
Henry IV: 39, 45; Twelfth Night: 33, 39; Hamlet: 71, 80; Lear: 55, 62;
Coriolanus: 48, 55; The Tempest: 20, 20 (Source: Alfred Hart:
Shakespeare and the Homilies and Other Pieces of Research into the
Elizabethan Drama (Melbourne and Oxford U. PP., 1934), p. 253.)
One could say that the statistical trends are roughly similar, and that
Hart and I agree in one case (though are the counted examples the
same?). But that's surely not good enough for a methodology claiming the
authority and precision of a science.
S H A K S P E R: The Global Shakespeare Discussion List
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.