In studying the history of an institution, historians often look back at its administrative records. Today, more and more, those records are being created in electronic form and never even exist on paper.
Funded by a $123,928 grant from the National Historic Publications and Records Commission (NHPRC), archivists and computer systems specialists at Cornell University have embarked on an 18-month project to study new record-keeping technologies and recommend ways to ensure that electronic records are preserved for the future.
"There are things we have been collecting for over 50 years that suddenly aren't being generated on paper," said Elaine Engst, director of Cornell's Division of Rare and Manuscript Collections, university archivist and project director. "How do we preserve the history of the university in the 21st century?"
It's partly a hardware problem, she said. Because of the rapid changes in computer technology, reading today's computer record tomorrow may be akin to finding a machine to play a 78-rpm record. But that's only one of the worries: Even if the physical data can be retained, archivists will have to work with it in new ways.
For example, Engst pointed out, the university has kept student records in electronic form since 1982. This means, among other things, that there are no such things as transcripts of grades. "You don't have a single record listing all of a student's courses and grades; you have student records, courses, course information, grade information. You assemble data from those into a transcript when you need it," she said.
Another example is university policy manuals, which are no longer printed and distributed but maintained on a web site. "Web sites are four-dimensional since they vary in time," she said. When policies are changed, the computer files that make up the web site are changed. "How will you be able to know what the policy was years ago?" Engst asked.
The research project coincides with a major revamping of Cornell's administrative computer system, known as Project 2000, which will use new software supplied by the PeopleSoft Corp. The research team will work with administrators setting up the new system and recommend procedures that can be built in to preserve critical records. One of the goals is to create "metadata," or data about data. This data describes records, indicating what sort of hardware and software were used to create and store them as well as who created them, when and why.
The project focuses on two of Cornell's 12 colleges, the College of Arts and Sciences and the College of Human Ecology, chosen because they have very different administrative structures, Engst said. Project staff already have conducted extensive interviews with managers in both colleges and at the university administration level. Among other things, they will be determing the level off duplication of records.
"It is the intent of the project in focusing on Human Ecology and Arts and Sciences to produce results that can be implemented in a wide variety of academic institutions," Engst said. The work builds on the experience of other recent NHPRC-funded projects, especially those at the University of Pittsburgh and Indiana University, she said.
The project team brings together expertise in both library archiving and computer technology. Cheryl Stadel-Bevans, who holds graduate degrees in library and information science and in mathematics, is the full-time project archivist. Others on the team are Oliver Habicht, computer systems specialist with the Cornell Institute for Digital Collections, Philip McCray, technical services archivist with the Division of Rare and Manuscript Collections, and Eileen Keating, university records manager. Peter Hirtle, assistant director of the Cornell Institute for Digital Collections, formerly with the National Archives, is project adviser, and Lee Stout of Pennsylvania State University serves as consultant.