18.260 historical development of tools

               Humanist Discussion Group, Vol. 18, No. 260.
       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

         Date: Sun, 03 Oct 2004 08:21:21 +0100
         From: Øyvind Eide <oyvind.eide_at_muspro.uio.no>
         Subject: historical development of tools

Dear humanists,

Thank you for all the comments and suggestions for reading! To start
with the practical matters, I am very happy for Geoffrey Rockwells
offer to use the TAPoR Wiki for an information/discussion web for the
history of tools (Humanist 18.174). I will contact him directly and we
will hopefully post an address to the Humanist when something becomes
available. If this will eventually develop into a history of text
analysis, by someone with more knowledge and time invest in it than
me, so much the better! And if it eventually becomes a ruin, as
Francois Lachance suggests in Humanist 18.252, we would probably need
some virtual archaeologists?

Reading through articles, among them many of the ones suggested here,
it is interesting to note the two different traditions running almost
independently - the philologist/SGML/TEI/concordance based tradition
in the humanities on one hand, and the social science qualitative text
analysis/annotation/theory building systems on the other. And never
shall the twain meet, apart from some references in articles?

As historians (as a group) are in both of the worlds (in Norway, at
least, the history departments seems to be placed in the humanities or
the social sciences depending on the founding date of the various
universities), they could use tools from both the traditions. They
seldom use any.

Some suggestions for reasons have been put forward in the last weeks,
I will comment on one of Donald Spaeths suggestions from Humanist
18.136:

> In any case, historians rarely rely on intensive analysis of single
> texts, but are more likely to build up a corpus of information from
> many different sources. In most cases, they are extracting isolated
> information from much larger sources. Computerate historians are most
> often interested in entering sources into structured databases, a task
> for which textual analysis tools are not designed.

The use of a single base source is a common methods in many fields of
research. In literary studies, this is obvious - the text studied is
the object of study. As I understand, research in the social sciences
often follows a similar pattern, where the source is
e.g. (transcriptions of) interviews or fieldwork (notes).

In history, this is usually not regarded good research. As historical
research is more event oriented (trying to understand and describe an
event from every angle) one have to use several (in principle, all
relevant) sources of information. Thus, there have traditional been a
divided group of qualitative sources, some of them in digital form,
many not digital. In this source situation, using digital text analysis
might be seen as giving undue weight to some sources (the digital
ones), selected out of coincidence only. This might be different in
the future, and it should perhaps be different for contemporary
historians?

It could be that some historians would be interested in theory
building methods offered by social science software, but as this is
usually based on annotation of the source texts, they are hard to use
if e.g. 80 % of the sources are in print only. So they extract
information into databases or text documents instead, with weaker
links between source and analysis as a possible result.

Kind regards,

Øyvind Eide
Received on Sun Oct 03 2004 - 03:51:45 EDT