19.687 text-analytic fantasies and realities

               Humanist Discussion Group, Vol. 19, No. 687.
       Centre for Computing in the Humanities, King's College London
                   www.kcl.ac.uk/humanities/cch/humanist/
                        www.princeton.edu/humanist/
                     Submit to: humanist_at_princeton.edu

   [1] From: Ryan Deschamps <Ryan.Deschamps_at_Dal.Ca> (41)
         Subject: Text Analysis as a literary clock

   [2] From: Norman Hinton <hinton_at_springnet1.com> (7)
         Subject: Re: 19.684 an idle fantasy

--[1]------------------------------------------------------------------
         Date: Sat, 01 Apr 2006 08:47:27 +0100
         From: Ryan Deschamps <Ryan.Deschamps_at_Dal.Ca>
         Subject: Text Analysis as a literary clock

Hello fellow Humanists,

As someone whose education is too broad to be important, but just sophisticated
enuogh to make me dangerous, I thought I would share my own dream of text
analysis in hopes that someone would think of stealing it for their own benefit
(I am a public librarian with very little time to do "tight" research).

It seems to me that text analysis has been used mostly as a map for particular
works or groups of works -- a way of outlining the major features of a text and
project significant themes or habits of an author or group.

This should not be surprising, since (I believe) most of us have the continued
suspended belief that a book/text is, as Northrop Frye would say, "Frozen in
Time" -- meaning that the entire situation is static, unchangable.

Obviously, this is not the case, since it takes time to read and write a text,
particularly one the size of a novel. And, paragraphs and chapters are not
all written or even conceived in the order they are read. I think Derrida's
_Disseminations_ uses the example of a "preface" which is read before a work,
even though the author writes it _last_ (you cannot comment on your work until
it is finished). Other cases include authors who have a beginning and an end
at the start of their writing and use the rest of their time trying to fill in
the gaps. With Web-based technologies such as wikis, blogs etc. we may be more
aware of process over product than in previous times -- but humanists appear
doomed to treat the study of _Ulysses_ as if it appeared out of Joyce's hat on
one day in 1922.

It seems that a combination of text analysis and time-series
statistics could be
a useful tool to "clock" out the writing process of individual or groups of
authors, philosophers or thinkers. It would take an extremely detailed
examination of the authors' journals to determine which works were written when
(or perhaps a less perfect methodology would be to use someone who wrote in
serials such as Dickens or Trollope). Then you map out major tropes per time
period (weekly, monthly, annually?) and perhaps run tests to correlate these to
external stimulae (stock market? coal/wheat/coffee prices? birth/death/crime
rates? funding for arts?), tag it for personal events and perhaps even look
for seasonal trends (more or less depressed in december?). This could be an
interesting way to examine the impact of environment on artistic process (if
there is one) and perhaps could lead to some new areas of
understanding about the philosophy of aesthetics and design (does art
come from God/imagination or
are good art "habits" formed by social circumstances).

Anyway, just an idea for anyone interested in a post-doc or PhD from you
friendly neighbourhood public librarian!

Best of luck!

Ryan Deschamps MPA/MLIS

--[2]------------------------------------------------------------------
         Date: Sat, 01 Apr 2006 08:47:53 +0100
         From: Norman Hinton <hinton_at_springnet1.com>
         Subject: Re: 19.684 an idle fantasy

Willard, I guess I've traveled all of those roads and trails at one
time or another, and they are all fun to traverse -- you left out
CAI Turnpike (Computer Assisted Instruction), which I guess some
folks will say belongs on another map altogether.

Let me put out a no doubt ignorable plea for more hardy souls to
explore Statistical Analysis Trail. All you need to get started is a
very good friend who is a statistician, if you haven't had your own
course of instruction...
Received on Sat Apr 01 2006 - 03:20:14 EST