TYPE OF PROPOSAL: poster or project demonstration.

TITLE: We Built it. Where are They?: Text Encoding and the Humanities Scholar

AUTHOR: Hope Greenberg

AFFILIATION: University of Vermont

E-MAIL: hope.greenberg@uvm.edu

KEYWORDS: TEI humanities etext

CONTACT ADDRESS:
Academic Computing
238 Waterman Building
University of Vermont
Burlington, VT 05405

FAX NUMBER: 802-656-0872

PHONE NUMBER: 802-656-1176

EQUIPMENT NEEDS: Mac or Win PC with network connection and web browser. Also an overhead display if done as a presentation.

--------------------

In The Diffusion of Innovations, Everett Rogers characterizes the adopters of new technologies as innovators, early adopters, early and late majority, and laggards. Building on this idea, and on Moore's conclusion that a chasm exists between the first two groups and subsequent groups, Geoghegan suggests that early adopters have a high comfort level with technology, value innovation, and are willing to experiment. The majority, however, values minimal risk, requires clear and compelling pedagogical value before attempting a new technology, does not wish current practices to be constrained, and eschews technology for technologies' sake.

Since its first 'official' appearance in 1994, the Text Encoding Initiatives Guidelines for Electronic Text Encoding and Interchange has had a profound impact on the innovators and early adopters of the electronic text world. Large text collections and scholarly projects already versed in SGML, like the University of Virginia's Electronic Text Center and The Rosetti Archive, and other projects at the Institute for Advanced Technology in the Humanities, have adopted it. Others, like the Model Editions Partnership and the Making of America project, are exploring ways to apply it to historical documents. The MLA recommends the TEI as the encoding method of choice in its Guidelines for Electronic Scholarly Editions. Text collections that focus on a particular topic or era, such as the Brown Women Writers, Victorian Women Writers, and the ORLANDO projects, as well as several thesaurus projects, have found that it supports building a collection that might not otherwise be possible in the print world.

A quick perusal of the list of projects using the TEI, as collected at their web site, shows that many are large scale projects hosted by libraries or consortia, and specifically funded. Given the technical difficulties associated with creating these documents, it is not surprising that early adopters tend to be those with resources to devote exclusively to these projects. Nor is it surprising that several groups are attempting to provide materials to assist scholars and students who are interested in undertaking projects of this type (cf. University of Virginia Electronic Text Center, Brown University Scholarly Technologies Group, etc.), providing information on their encoding practices (cf. British Women Romantic Poets Project, Victorian Women Writers Project, etc.), or formulating best practices for creating digital resources in general (cf. Arts and Humanities Data Service Guides).

Individual scholars or initially small-scale projects are not entirely absent. The fine work begun by Dr. Stuart Lee as a web-based tutorial on the poetry of Isaac Rosenberg and growing into the Virtual Seminars for Teaching Literature, as well as projects like The Walt Whitman Archive and others supported by IATH, all point to the pedagogical benefits to be obtained from individual scholars creating these texts.

But can the chasm between the TEI early adopters and the majority of scholars, be bridged? Is there a compelling pedagogical benefit to the individual humanities scholar in creating TEI-encoded electronic texts, or is this a job best left to large projects, consortia, publishers, and libraries? Should individual scholars be creators or simply consumers?

The Godey's Lady's Book site at the University of Vermont attempts to explore these questions. Before compelling benefits can be derived from the creation of individual projects, we must determine if the creation of such projects is feasible given limited time and resources. That is, can the creation of electronic texts fit within the requirements for adoption by the majority as described by Geoghagen?

Drawing on the models available, this project is creating lightly encoded texts from a popular mid-19th century American magazine. Like the Making of America project, the contents are page images backed by hidden OCR'd text. The accuracy of the OCR process is insufficient to provide reliable transcriptions but sufficient to provide reasonably good indexing for purposes of searching. The indexing and display software currently used are DynaText and DynaWeb, which were acquired through Inso Corporation's educational grant program. Along with the texts themselves will be ancillary documents related to the Book's contents, supporting documents describing the encoding, and a tutorial for creating similar documents.

In choosing to create the digital Godey's Lady's Book, several issues were considered:

• Issues of preservation and archiving: It is not enough to have a digital version, even a TEI-encoded platform-independent digital version if the resulting files are not archived. Even with the recent resurgence of interest in 19th century women writers, the majority of writers featured in Godey's Lady's Book are no longer recognizable names. Reviving these documents in a digital version with out making plans for their preservation would hasten their return to oblivion. In the first step of a plan for long-term preservation, the documents will be moving from a personal server to the university cluster in spring of 1999.
  
  
• Issues of fidelity and trust: In their session titled The Epistemology of the Electronic Text, presented at the 1997 ACH/ALLC joint conference, Flanders, Lavagnino and Barash discuss aspects of using images of the text instead of or in addition to a transcription of the text, in an electronic text edition. Several of the ideas they outline intersect in the Godey's Lady's Book documents. Godey's plates and images obviously provide important cultural artifacts integral to the magazine. Fortunately, at least in terms of making the decision to make this an image collection backed by indexed text, the very nature of that text makes this model the most practical--the magazine is set in 9 and 7 point serif fonts that tax the OCR process. Two other considerations led to the decision to present these documents in this fashion: the magazine was not meant to be read cover to cover, but selectively (although Godey asserts that it was certainly read completely). While maintaining an edition that can be searched or browsed through standard Table of Contents links, this model also provides a "page-turner" approach where the reader can "flip through" the magazine in a way similar to the original artifact. Also, comments collected from visitors who have seen both a small HTML-based edition that used transcriptions of the text and this text image edition suggest that the latter is more useful.
  
  
• Issues of economics: The digital Godey's Lady's Book has been created using free or low cost software and low-end hardware. The hidden costs have been in the time devoted to its creation and in the upcoming use of disk space on the university cluster. Faculty and students at the University of Vermont have access to the same computing resources, so project storage space for additional projects is not a constraint at this time.
  
  
• Issues of usability: The web provides a wonderful vehicle for disseminating and sharing projects like the digital Godey's Lady's Book. However, as David Chesnutt suggests in his work with the Model Editions Partnership, an infrastructure that facilitates cross-edition searching and indexing would be a boon to scholars. Nor does creation of TEI-encoded documents automatically ensure complete interoperability between texts. As Perry Willett points out, searching for "compatible encoding practices, in order to allow for future interoperability of collections" is another laudable goal. The digital Godey's Lady's Book is following these and other projects with the object of developing in concert with them. In the short term the texts are being lightly encoded to make future changes easier.
  
  
• Issues of scalability: While the contents of the digital Godey's Lady's Book is and will continue to be a growing resource for interested scholars, its parallel purpose, the creation process itself acting as a model for other individually-based electronic text projects at or beyond the University of Vermont, cannot be forgotten. Concurrent with the development of the GLB is an effort to introduce faculty and students to the TEI and text encoding through presentations (cf. Text + Context = Etext: Scholarship in the Information Age ). Further developments of this process will be discussed during the presentation.

At each phase in creating the digital Godey's Lady's Book the question kept at the forefront is: can this model be duplicated by individuals or small groups with limited resources while remaining in concert with, and informed by, the broader text encoding world? Unless the answer is yes, the chasm between early adopters and the majority of humanities scholars may well remain unbridgeable.

References:

Chesnutt, David. The Model Editions Partnership--Towards a National Database. Paper presented at the ACH/ALLC Joint Conference, 1997. Available online: http://www.qucis.queensu.ca/achallc97/papers/p036.html

Flanders, Julia, John Lavagnino and Carol Barash. "The Epistemology of the Electronic Text: Scholarly and Pedagogical Considerations" presented at ACH/ALLC Joint Conference, 1997. Available online at: http://www.qucis.queensu.ca/achallc97/papers/s002.html

Geoghegan, William H.Whatever Happened to Instructional Technology? Paper presented at the 22nd Annual Conference of the International Business Schools Computing Association, Baltimore, MD. 1994.

Moore, Geoffrey A.Crossing the Chasm. New York: HarperBusiness, 1991.

Rogers, Everett M. The Diffusion of Innovations. NY: The Free Press, 1995.

Willett, Perry. Issues in Project Cooperation II: Markup Issues. Presented at the ACH/ALLC Joint Conference, 1998. Available online: http://lingua.arts.klte.hu/allcach98/abst/jegyzek.htm

URLs for sites listed:

Arts and Humanities Data Service Guides: http://ahds.ac.uk/public/guides.html
British Women Romantic Poets Project: http://sunsite.berkeley.edu/MLA/guidelines.html
Brown University Scholarly Technologies Group: http://www.stg.brown.edu
Brown Women Writers Project: http://www.stg.brown.edu/projects/wwp/wwp_home.html
University of Virginia's Electronic Text Center: http://etext.lib.virginia.edu/
The Rosetti Archive: http://jefferson.village.virginia.edu/rossetti/rossetti.html
Institute for Advanced Technology in the Humanities: http://www.iath.virginia.edu/
Making of America: http://moa.umdl.umich.edu/moa/
MLA Guidelines for Electronic Scholarly Editions: http://sunsite.berkeley.edu/MLA/guidelines.html
Model Editions Partnership: http://mep.cla.sc.edu/
ORLANDO: http://www.ualberta.ca/ORLANDO
Text + Context = Etext: Scholarship in the Information Age: http://www.uvm.edu/~hag/presentations/etext
Text Encoding Initiative: http://www-tei.uic.edu/orgs/tei/
Victorian Women Writers: http://www.indiana.edu/~letrs/vwwp
Virtual Seminars for Teaching Literature: http://info.ox.ac.uk/jtap
Walt Whitman Archive: http://www.iath.virginia.edu/whitman