ETD: Statement About Publications
Universities are working together to build a National Digital Library of Theses and Dissertations. This initiative involves training graduate students both to publish electronically and use digital libraries, to better prepare them for their research and future careers. The electronic theses and dissertations (ETDs) they produce will be widely accessible, supported by powerful search and browsing software.
The ETD participants are research universities with graduate programs, whose libraries now keep copies of dissertations, theses, and special reports from undergraduate and graduate students. Copyright generally is retained by the authors, their institution, or both.
Historical Means of Distribution
Some theses and most dissertations have been archived with UMI, that has a repository of over a million works. A microform copy is archived there, and copies are available on demand as microfilm or paper. UMI aggressively advertises its holdings through Dissertation Abstracts. In some cases authors obtain royalties when there are a sufficient number of copies sold.
Many libraries use Interlibrary Loan to respond to requests for access to theses or dissertations. There often are multiple copies of a thesis or dissertation available at an institution, such as in the department where the work was done, so Interlibrary Loan can help broaden access to other locations.
New Types of Control
With the advent of electronic theses and dissertations, standard practices of restricting access to particular works can be expanded. At present, with paper copies, the approved work may be locked up --- the only method of restricting access --- which is important when there are issues of security or proprietary information or pending patents.
With electronic documents, access may be delayed, as before, by securing the electronic copy. In addition, access may be restricted to a certain group of readers, using passwords, encryption, or other schemes. In the most severe case, only the graduate committee may see the work. Access restrictions also can be based on access location (e.g., to the campus of origin, to campuses in the state, or to other special ranges of internet addresses).
By combining these schemes, access immediately after approval may be restricted to an institution, access commencing after 3 months may be restricted to other educational institutions, and widespread access may be limited for a longer period, such as a year. In addition, access may be limited by count, so that only 100 different people may gain access before special restrictions are put into play. Further, access can be limited by component, so that one chapter is secured while the rest of the work is made available. Some institutions may add other restrictions to defray costs, such as charging for copies, much like UMI does.
While many types of access restriction are possible, the special needs of authors, requirements of the research projects they have participated in, and the policies of the institution awarding their degree, must be considered in concert when deciding what type of access is suitable for a given work.
Research that led to a thesis or dissertation has often been discussed in other types of documents, such as conference papers, journal articles, or manuscripts. In some cases, those publications have been released before the graduate work is approved, but in many cases the related publications appear after the thesis or dissertation has appeared.
Timing is clearly important. For hot results it may be important for a journal article to appear before widespread access is allowed to a related dissertation, but making the dissertation available quickly thereafter may be of great value in stimulating further confirming or related research.
Related publications almost always are radically different from the theses or dissertations. Typically, conference or journal articles are much shorter, are organized differently, benefit from considerable reflection on what was most important during the period of graduate research, and are much better written. They benefit greatly from the comments of reviewers and the re-work and re-writing demanded by scholars in the field.
With ETDs, there is no fundamental difference from the paper based scheme currently in effect, regarding related publications. ETDs are not more published than current dissertations that are available through Interlibrary Loan or on sale from UMI. Rather, there is much greater possibility of controlling access to ETDs than is possible now with paper copies.
As the practice of preparing ETDs spreads, it is likely that special arrangements will arise with associations and other publishers. For example, ACM, the First Society in Computing, has endorsed plans for a National Digital Library of Theses and Dissertations. It has agreed to have its online WWW pages point to that digital library. ACM also has agreed to point to a subset of the digital library of ETDs that deals with computer science. Finally, in connection with its dissertation awards, it is willing to point to the highest rated dissertations in their online form, to promote access to those works.
Other associations and publishers are encouraged to endorse the ETD effort, and to work out ways to help their readers gain access to the emerging digital library of theses and dissertations. Possible examples include launching dissertation awards programs, developing practices of using citations (e.g., Universal Resource Names) that refer to large objects which can fit into a digital library but which would not fit into journals (e.g., data files, collections of color images, audio or video files), and publishing abstracts or reviews or annotated bibliographies of important theses or dissertations.
Working with Universities
Since the ETD effort is considered by universities to be of great importance regarding cutting costs and improving graduate education, it will be in the best interest of publishers to work with the academic sector as it builds a distributed digital library of theses and dissertations.
One important benefit to publishers may be to reduce their production costs. Through the ETD effort, all graduate students will learn about electronic publishing and digital libraries. They will work with both PDF and SGML. This last point is highly significant, since many publishers are having trouble shifting to use SGML, and if the ETD effort can support capture of any type of thesis or dissertation using SGML, publisher needs can surely be met! Accordingly, students should be able to prepare other documents, during the rest of their career, that can go directly into the digital libraries being developed by publishers or publisher groups. This should help reduce production costs throughout the industry, and increase quality relative to practices where works are re-keyed by typesetters.
Publishers interested in discussing these matters further should contact the ETD initiative group, whose email address is given below.