[1] From: John Merritt Unsworth (43)
<jmu2m@jefferson.village.virginia.edu>
Subject: Re: 9.251 more on layered hypertext
[2] From: rbh <R.B.Hardy@ukc.ac.uk> (36)
Subject: Re: 9.251 more on layered hypertext
[3] From: Language Technology <langtech@DGS.dgsys.com> (35)
Subject: Re: 9.251 more on layered hypertext
--[1]------------------------------------------------------------------
Date: Mon, 23 Oct 1995 20:59:51 -0400 (EDT)
From: John Merritt Unsworth <jmu2m@jefferson.village.virginia.edu>
Subject: Re: 9.251 more on layered hypertext
Willard,
> Fellow Humanists may remember my undoubtedly vague description of a "layered
> hypertext" system for representing annotations to a text. Roy Flannagan then
> asked me to clarify what I wanted out of my text that I needed such a system
> to represent it. I return by asking myself what use it is to dream of
> something no one has the time or money to build.
I think we might have built it, or at least a first pass at it.
>
> I guess one way of approaching the question Roy raised is to think about the
> bind New Critics (or at least their students) used to get into when
> attempting to make note of all the minute phenomena one could notice in a
> text. Another way is to think about any text you have studied intensively
> for a long time, and the edition of it in whose pages you have made detained
> marginal notes. I imagine that if one could virtually place a series of
> transparent layers over the image of the text, on each of them inscribing a
> particular set or kind of notes, one might achieve a more adequate
> representation of the text's complexity than otherwise possible. In other
> words, 3-dimensionalize the scholarly commentary.
>
Willard! At IATH, we have been developing a software package called Inote,
which takes any image, including an image of a page of text, and puts any
number of clear overlays on top of it, allowing you to draw details
(points, circles, rectangles) on those overlays and attach each detail to
either a textual annotation or another image. You can display multiple
overlays simultaneously, or you can turn them all off and just see the
image.
We've produced the first of several special classes of overlays--an
overlay that can take a page of text, identify each line automatically,
create a detail over each line, and number each detailed line (or, with
user input, identify top, middle, bottom, right, middle, left margin
areas plus each line of text).
One goal is to integrate Inote with SGML markup, so that one could go
from an SGML file to the relevant section of an image; another goal is to
port Inote to Java, so that it will be distributable to all the major
computer platforms and usable over the Web. We're actively working on
both of those things now. If anyone's interested in the current state of
the software (version 3.0, written in C++, X, Motif++, HDF), the source code
is available by anonymous ftp at jefferson.village.virginia.edu, where it
can be found in pub/software/iath, and if you have Xwindows or
X-emulation software on a PC or Mac, you can experiment with the software
using a forms-based demonstration on the IATH web site, at
http://jefferson.village.virginia.edu/software.html
Comments, design suggestions, and/or programming assistance welcome...
John Unsworth
--[2]------------------------------------------------------------------
Date: Tue, 24 Oct 1995 09:25:40 GMT
From: rbh <R.B.Hardy@ukc.ac.uk>
Subject: Re: 9.251 more on layered hypertext
This nascent discussion has got me thinking about putting on offer
a creative writing class to follow on from the one which I now
teach at the University of Kent. Call it `Professional Writing:
Hyperfiction' to match the title of the present course:
`Professional Writing: Prose Fiction'. (The professional
bit is an attempt to turn the dopeheads/alkies away.)
I find the layering idea fascinating, for it seems to me to represent
something of the way my mind works when I muse and meander over
possibilities, both real life and in the fiction I write. The
toggle-switches of the present system are interesting, but
ultimately limiting, from my point of view. I can
see no computational problems in layering, provided the
writer of the code/user of the code could agree on control
over how one would fade in/out two or more layers. We're
obviously talking about a system incorporating HTML in its
?stage three form, but going beyond it.
People might be interested to know that in the
computing-for-humanists course I run at this university
for FIRST YEAR undergraduates (100 of them) the students
will be writing their first hypertext today. (Non-Brits
should take note of the fact that this is the third week of the
British autumn term, and the students' 10th hour of instruction.)
Their first steps are to copy source for one of our humanities
departments' home pages, and, step by step, edit out the
department's stuff, and edit in theirs, checking out
the results on-screen. A more disciplined approach will be
introduced later under the generic heading: mark-up. It
will attempt to associate xroff, HTML, TEI and SGML stuff,
not to produce master-markers, but to try to show the
undergraduates a non-GUI way of dealing with text, (and
how the results can be subsequently processed to provide
an object for the technologically disadvantaged.)
(Should you object to the latter phrase, think of musicians
who can look at a score and `hear'; and think
of the musically disadvantaged who are limited to listening.)
I'll be following what I hope is a long thread with much interest.
Roger Hardy (rbh@ukc.ac.uk)
--[3]------------------------------------------------------------------
Date: Tue, 24 Oct 1995 11:23:32 -0400 (EDT)
From: Language Technology <langtech@DGS.dgsys.com>
Subject: Re: 9.251 more on layered hypertext
I think you have touched on an important point about software that I was
discussing with a colleague just this morning. While there is lots of
software development going on right now, there isn't much thought being
given to the notion of what kind of software we need. [NB: for 'we',
read whatever application one is concerned with; for Humanist it is
humanities computing, for Wall Street it is financial computing, and so
on.] I believe that is the job of the 'scholarly elite' [dirty words in
this ole US of A, these days], to think about what we WOULD be doing with
software, if we could make it do anything we wanted.
It's a sort of meta-software problem mostly dealt with by the Microsofts
and Lotuses of the world! That of course raises the question: do we
want the world of computer software to be designed by those with the
most to gain financially? or should we as humanists insist on being
heard wrt to these issues?
Software design essentially defines how we think about the concepts of
computing! [that's where the 'meta-' comes in!] So we need to discuss
the concepts of humanities research in order to evaluate and propose
humanities software. The software interface determines the way we use
(and think about) the processing, which is really backwards from the way
we as scholars would want it to be.
Many of us are old enough to remember when everything had to printed out
of any computer in all caps -- because that's the way it was done. It
took a lot of years of objecting to get the situation changed
enough that we could even ask the question, "how can I display phonetic
characters?" You don't worry about phonetics when you can't distinguish
upper and lower case!
So, Willard, as usual, you've hit on a very important topic: how do we
think about humanities thinking? Thanks again to the elitest of the
scholars!
Mary Dee
Mary Dee Harris, Ph.D. 202-387-0626
Language Technology, Inc. langtech@dgs.dgsys.com
2153 California St. NW mdharris@aol.com
Washington, DC 20008