5.0881 ALLC-ACH '92: Report 3 (1/111)
Elaine Brennan & Allen Renear (EDITORS@BROWNVM.BITNET)
Mon, 4 May 1992 19:07:40 EDT
Humanist Discussion Group, Vol. 5, No. 0881. Monday, 4 May 1992.
Date: Sun, 3 May 92 20:50:45 -0400
From: jdg@oz.plymouth.edu (Dr. Joel Goldfield)
Subject: Post-conference synopsis of ALLC-ACH talk
Accessing and Using Electronic Texts in Academia
by
Joel D. Goldfield
Associate Prof. of French
Plymouth State College/Univ. System of New Hampshire;
Associate Fellow in Computing and the Humanities
Institute for Academic Technology
Univ. of N. Carolina at Chapel Hill
10 Key Points Developed for the 1992 Annual Convention of the
Association for Literary and Linguistic Computing and the
Association for Computers and the Humanities
Oxford University.
11:15 Session on Thursday, April 9, 1992:
Electronic Texts: Policy, Acceptability, and Control
(Note: Many of the points given below were developed from comments made by
attendees at the IAT Colloquium referenced in #1. However, the elaboration
of each point is solely the responsibility of the author and does not ne-
cessarily reflect the position of any organization with which he is affili-
ated.)
1. Infrastructure & fees. The number #1 concern of attendees at the 1991
IAT Colloquium on Computational Approaches to Textual Studies was the pos-
sible inaccessibility of extant electronic texts owing to exclusionary res-
trictions imposed by those controlling a distribution infrastructure or by
those determining subscription fees for a product. These aspects need our
close attention.
2. Display standards. There was unanimous consensus, as in #1 above,
regarding the importance of this category, which consists of three com-
ponents: a) screen resolution; b) physiological ramifications; and c)
predictable (standardized) encoding for the appearance of text and graph-
ics. Starting with the first two items, some colleagues note that
reflected light, rather than direct light, would be preferable in their
work, especially where long hours of intense reading and editing are neces-
sary. Straddling both items, viewers would like a standard enlargement
feature for zooming in on any section of the screen. Researchers would like
more data on possible advantages of high-resolution LCD displays, both
monochrome and color, versus incandescent (i.e., CRT-produced) screen
displays under various lighting conditions for varying periods of high-
intensity reading. Regarding item c, colleagues are looking forward to
evaluating the impact of SGML and the Text Encoding Initiative on the use
of electronic texts.
3. CD-ROM and similar media. Software should be separated from the tex-
tual and graphical data. Scholars need to implement stylometric and other
programs to process the textual information in ways often unanticipated by
the producers. On-board software usually renders this pursuit impossible.
4. Professional organizations. We need professional organizations to for-
mulate a collective, cogent policy on the:
a) creation of e-texts;
b) reliability of e-texts;
c) distribution of e-texts;
d) legal use of e-texts.
5. Copyright Policy. Elaborating on #4d, a re-evaluation of copyright
restrictions on the use of e-texts for purely research purposes is strongly
recommended. This does not imply a naive assumption that scholars should
be able to acquire all e-texts gratis, particularly those texts not yet in
the public domain. Where stylo-statistical or linguistic research is
involved, for example, if the text is still in-print, an "archival" copy
might be purchased, then scanned or typed into machine-readable form by the
researcher's organization (unless such a medium could be purchased) with
the understanding that it would not be redistributed for reading purposes
in lieu of a student textbook. Legal conditions might vary if the book is
no longer in print.
Patrimony restrictions should be re-evaluated. For example, restrictions
on the use of out-of-copyright texts seem particularly noxious for literary
researchers desiring to download e-texts for local processing. As distri-
bution centers for nationally controlled e- texts do not always have the
particular software tool a researcher may require to analyze some feature
of a sub-corpus or corpus, more flexibility is needed in the relationship
between such sites and users.
6. Infrastructure of scholars. An infrastructure or consulting group of
scholars working with administrators should be created to evaluate and pro-
mote the production of e-texts in a timely manner for designated projects.
These projects would be evaluated and funded by grant organizations, par-
ticular institutions or private enterprises. Projects could be requested
by individual academic researchers or by institutions.
7. Peer-review. A procedure should be established by professional organi-
zations and the e-text center for the peer-reviewing of annotated e-texts
if they are tagged beyond screen mark-up (e.g., morphological and literary
tagging). This reviewing could take place prior to or following in-house
editing, depending on the expertise of the reviewers.
8. Technical assistance. High-quality technical assistance would be
required for determining the best data-entry process, encoding, distribu-
tion possibilities, in-house applications programs and interfacing with
client needs (e.g., statistical, audio, video).
9. Referral service and professional development. Academic institutions
would benefit from a regional or national referral service which would put
potential or current users of the e-text distribution center in touch with
the appropriate personnel for advice on research problems. Tutorial
software would be available on-line to assist researchers in learning to
use e-texts for particular projects such as determining authorship,
developing concordances, tracking changes in linguistic features over time
within a genre in a particular national literature, identifying semantic
webs, etc.
10. Mainstreaming literary computing. We need to move our
computationally-informed research into the mainstream journals and meet-
ings. Scholars submitting manuscripts to specialty journals might be
encouraged to contact particular mainstream journals for re- publication of
a revised version. Model studies should be publicized (see also #9 above).