The Shakespeare Conference: SHK 12.0772 Wednesday, 4 April 2001
From: Paul Maddox <
Date: Friday, 30 Mar 2001 22:19:56 +0100 (BST)
Subject: 12.0728 Re: Shakespearean Authorship Research
Comment: Re: SHK 12.0728 Re: Shakespearean Authorship Research
> > However my program does not compare words, it compares word catagories.
> > Hence such things as spelling errors and rare words are minimised.
> Why is this desirable? If Shakespeare, to cite a common example, tends
Using word catagories rather that words makes the comparison more
generic. This is particularly useful for comparing documents with
slightly different content.
> to spell the word *Oh* with the first letter only (O), your program
> tends not to take this fact into account. Is that correct? If so, what
> is the advantage of this when making your authorship comparisons?
Interestingly, the tagger I use defines 'o' and 'oh' differently, hence
my program would take that into consideration. I think that's more
through luck than anything else.
> down in the specific, let's have a peek at the general. If in those 14
> lines you find several other poetic conventions common at the time, you
> will still get a low statistical number (meaning a high degree of
> linguistic commonality). True? If so, your program may still suggest
> common authorship, but might it not instead suggest two authors who used
Yes, you're right in saying this. I've yet to compare various sonnets by
hand to see if the tagger negates this problem. For instance it may be
that a common convention may be:
(determiner) (adjective) (adjective) (noun) (verb)
However, my program may classify this as:
(determiner) (adjective superlative) (noun plural) (verb, past
which is very different to:
(determiner) (adjective comparative) (noun singular) (verb past tense)
In literary terms however, both of my complex examples may be considered
similar. This is perhaps the advantage of my program: that it compares
on a purely scientific basis.
> About your comparison of Shakespeare with Shakespeare, Bacon with Bacon,
> etc., Did you compare every Shakespearean sonnet with every other?
> Every de Vere? Every Bacon? Then every poem by one author, with every
> poem by the others, or did you just sample?
I compared every sonnet with every other, for the sets you describe. I
thought I sent a mail out with the results of those tests, it may have
been moderated though.
On a general note, my apologies to anyone expecting a reply, I've not
been rude, I have replied to all emails I've received. I think replying
to multiple posters in one email may have come back to haunt me in the
All the best,
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 Webpage <http://ws.bowiestate.edu>