[Humanist] 23.557 stylometric study wins
Humanist Discussion Group
willard.mccarty at mccarty.org.uk
Mon Jan 11 07:22:53 CET 2010
Humanist Discussion Group, Vol. 23, No. 557.
Centre for Computing in the Humanities, King's College London
Submit to: humanist at lists.digitalhumanities.org
Date: Mon, 11 Jan 2010 06:10:16 +0000
From: Willard McCarty <willard.mccarty at mccarty.org.uk>
Subject: prize-winning computational analysis of Agatha Christie's novels
U of T research tops N.Y. Times' 2009 Ideas list:
Analysis shows that Agatha Christie likely suffered from Alzheimer's
By Elaine Smith, posted Wednesday, December 16, 2009
University of Toronto News
Research by U of T professors Ian Lancashire and Graeme Hirst has
garnered top spot in the N.Y. Times' 9th Annual Year in Ideas.
Lancashire, a professor of English, and Hirst, a computer scientist,
provide evidence that famed mystery novelist Agatha Christie suffered
from Alzheimer's-related dementia during the final years of her life.
It's a conclusion some of her biographers have reached, but the U of T
duo offers proof.
The pair digitized 14 of her novels and used textual analysis software
to determine the richness and size of the vocabulary used, as well as
phrases often repeated and an increase in the use of indefinite words,
an indicator of the disease.
Their results, published in a paper titled Vocabulary Changes in Agatha
Christie's Mysteries as an Indication of Dementia, were statistically
significant. They showed that her final two books use a much smaller
vocabulary than her earlier works, with differences as large as 31 per
cent. Other later works compared with her last two volumes also show a
much richer vocabulary.
"This publicity -- and the honour it bestows -- reflects a hope that an
aging society has for ways to detect Alzheimer's disease, a human
scourge, earlier than possible now," said Lancashire. "People in all
walks of life can understand, and even become conscious of, a change in
their personal language. People have a horde of e-mail or blog entries
now that go back some years. The simple vocabulary measures used in the
poster, the graph and the brief paper can be grasped and applied by
anyone, privately, non-invasively. The findings astonished me when I
found them two years ago. If the N.Y. Times recognition brings more
medical researchers to study language, I'll be delighted.
Lancashire said the New York Times publicity is also a recognition of
the value of interdisciplinary research and the role the humanities have
to play in such projects.
"At Toronto, the N.Y.Times notice highlights the deep strength of this
university in interdisciplinary research. Even English professors may
have a role to play in practical research of broad public interest. I
could not have presented and interpreted my findings properly without
the collaboration of Graeme Hirst in computational linguistics and
Regina Jokel at Baycrest. I am so fortunate to work in a team now with
these colleagues and Graeme's student Xuan Le."
Lancashire and Hirst plan to continue their textual analysis work,
examining the writings of mystery novelist P.D. James, who continues to
be prolific as she ages, and mystery writers such as Ross MacDonald, who
is known to have suffered from Alzheimer's disease.
Willard McCarty, Professor of Humanities Computing,
King's College London, staff.cch.kcl.ac.uk/~wmccarty/;
Editor, Humanist, www.digitalhumanities.org/humanist;
Interdisciplinary Science Reviews, www.isr-journal.org.
More information about the Humanist