Date: Mon, 12 Apr 1999 19:13:35 +0100
From: Willard McCarty <Willard.McCarty@kcl.ac.uk>
Subject: authorship attribution problem
The following statement of an authorship attribution problem was sent to me
by a writer in N California who is investigating the subject described
below. Please send your replies both to Humanist AND to Michael Koepf
<email@example.com>. It sounds like an interesting if very difficult problem.
THE LORD'S AVENGER
In 1990 a letter appeared with specific information about a car bombing.
It is approximately an eleven hundred-word typewriter written document
with an additional eight biblical quotations set off from the main body
of text. The letter is written in the style of a Christian zealot, but
opinion seems unanimous that the letter was a hoax, albeit a hoax
containing details of two criminals acts verified by law enforcement.
Problem: To compare The Lord's Avenger letter of unknown authorship with
works of authorship from four known authors. The samples from the known
authors will be of at least equivocal length as the unknown letter and
all known samples will have been taken from unedited works that appeared
in a newspaper. One submitted sample of writing will be a known hoax
authored by one of two of the included authors.
The known authors will be identified by number rather than name to
protect their identities. The results of the problem are not intended to
prove certain culpability in a crime. Indeed, exoneration of authorship
is the most probable outcome for all four included known authors, but if
some statistical attribution for one particular author emerges it will
not be viewed as proof of guilt. It will merely augment a line of
investigation and research.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
Dr. Willard McCarty, Senior Lecturer, King's College London
voice: +44 (0)171 848 2784 fax: +44 (0)171 848 5081
Humanist Discussion Group
Information at <http://www.kcl.ac.uk/humanities/cch/humanist/>