Electronic theses and dissertations in music.
A CONCISE HISTORY OF ETDS (2)
The history of electronic theses and dissertations begins in 1987 with a meeting convened by Nick Altair of UMI in Ann Arbor, Michigan, involving participants from Virginia Tech, the University of Michigan, and two fledgling software companies: ArborText and SoftQuad. The discussion focused on the latest approaches to electronic publishing and the idea of applying the Standard Generalized Markup Language (SGML, an ISO standard approved in 1986) to the preparation of dissertations, possibly as an extension of the Electronic Manuscript Project of the Association of American Publishers. That same year (1987) SoftQuad released AuthorEditor, the first graphical SGML editor.
In 1988 the first SGML Document Type Definition (DTD) (3) for theses and dissertations was developed by SoftQuad's Yuri Rubinsky with funding from Virginia Tech. With the appearance of Adobe's Acrobat software and Portable Document Format (PDF) in the early 1990s it became clear that students could easily prepare their own ETDs, and that the inherent complexities of SGML could be avoided where ETDs are concerned. Thus in 1994, Virginia Tech, as part of a pilot project, began to convert some of the printed theses and dissertations received from its graduate schools to PDF.
In 1996 the pace of ETD activities gained momentum when the U.S. Department of Education funded a three-year nationwide effort to extend the concept of ETDs across the country. The aforementioned pilot project at Virginia Tech led to a mandatory requirement that all theses and dissertations submitted after 1996 be only in electronic form. Thereafter the concept of ETDs spread to Canada, the United Kingdom, Germany, and other countries. And, last but not least, to coordinate all these efforts, in 1996 the free, voluntary federation called the Networked Digital Library of Theses and Dissertations (NDLTD) was established and quickly began to expand.
Today NDLTD has more than 196 member institutions, 172 of which are academic institutions in the U.S. and abroad. (4) To date, these institutions have put a combined total of more than forty thousand ETDs online. Metadata records (similar to the cataloging information one might find in a library catalog, including title, year, author, abstract, and descriptors) for these documents can be found in the OCLC-based NDLTD Union Catalog. (5) These records were harvested from the servers of thirty-eight NDLTD member institutions--those institutions that support the Open Archives Initiative (OAI) protocol for metadata harvesting. Access to these documents can be categorized as open, restricted, fee-based, and mixed. Whereas open access permits the entire ETD to be freely viewed by anyone with access to the Web, restricted access contains the flow of information to the campus network of the home institution. Mixed access prohibits readers from viewing a portion of the ETD. And fee-based access is a proprietary arrangement whereby individuals must purchase the thesis or dissertation, which they order from an institution in book form or printable PDF. Decisions regarding the accessibility of ETDs are made by the student-authors--who may or may not want their work published on the Web--or the institutions that host the servers where the documents reside. Unfortunately, less than half of the music-related ETDs in the NDLTD Union Catalog are freely accessible to the world at large. (6)
In addition to NDLTD, there are several commercial enterprises that provide access to ETDs. The most notable of these is ProQuest Digital Dissertations, known informally as UMI (the brand name for the archival and retrospective products of the ProQuest Information and Learning Company). For a price, UMI provides full-text access to more than 100,000 electronic theses and dissertations. (7) The German firm Diplom.de, founded in 1997, has nearly 11,000 ETDs for sale (8); and Dissertation.com, a Florida-based company also founded in 1997, has approximately 500 dissertations in its catalog. (9)
For the sake of clarity and perspective it should be noted that in 2000-2001 more than 44,800 doctoral degrees and ten times that many master's degrees were awarded by colleges and universities in the United States, (10) and that the majority of these degrees required a master's thesis or doctoral dissertation. Of these doctoral degrees, 791 were in music. (11) It should also be noted that the number of graduate degrees awarded by American colleges and universities over the past ten to fifteen years has been rather consistent. Hence, since 1987 when the idea of ETDs was first conceived, only a small percentage of theses and dissertations have been published electronically. If one compares this number to the proliferation and exponential growth of the World Wide Web since its inception just twelve years ago (1992), and the wealth of high-quality, academically sound Web sites that one can access today, it becomes abundantly clear that our ability to create and disseminate digital information is grossly underutilized where ETDs are concerned.
THE NATURE OF ETDS
At the present time there are two standard types of ETDs: author-created documents consisting of a text file converted to PDF, HTML, or XML and submitted (typically) over a network connection with related metadata; and electronic files created by scanning the pages of a paper thesis or dissertation. The latter are usually created by university or service company staff, and are commonplace in retrospective projects where a university wishes to share or sell its research. (12)
As for the basic technology that supports most ETDs, more than a decade after they were first introduced, PDF files remain the predominant file format. (13) They capture formatting information from a variety of desktop publishing applications, making it possible to send formatted documents and have them appear on the recipient's monitor or printer as they were intended. But PDF files require much more storage space than other formats (such as HTML and XML); they do not easily support full text searching; and, most importantly, they are a proprietary format that requires specialized software to be read. Colleges and universities routinely use PDF files for e-reserves, to create digitized copies of printed articles and chapters of books to be distributed over the Web. PDFs are also used for e-books. (14)
The advantages of author-created ETDs over printed theses and dissertations are numberous, but can be summarized as follows:
* ETDs support highly expressive multimedia supplements;
* ETDs are cost effective in comparison with paper theses and dissertations; (15)
* ETD activities teach students about electronic publishing and information technology, thus helping them to function more effectively in the Information Age;
* ETDs promote easy access to current research, and improved visibility for authors, which, in turn, facilitates a quantum leap forward in scholarly communication.
Today, surprisingly few music-related ETDs include multimedia supplements. In terms of content and form, most ETDs are virtually identical to printed theses and dissertations; and other than the possibility of electronic access, they offer very few advantages over their paper counterparts. While academe's conservatism and the relative newness of ETDs are largely to blame for this, the modus operandi of North American universities and their relationship to UMI also contribute to the under-utilization of multimedia supplements. Of the eighty-seven music-related ETDs with unrestricted access in the aforementioned NDLTD Union Catalog, only six include multimedia supplements. (16) As meager as this number may be, it compares favorably to the quantity of ETDs with multimedia supplements on UMI's server, where none of the more than 3,000 music-related ETDs include multimedia supplements. (17)
This paucity of media files can be explained two ways: In the case of academic institutions that mount ETDs on their own servers, the use of multimedia supplements is generally permitted (18) but not encouraged because faculty advisors and dissertation committee members are largely operating within the old paradigm, that being the print world. Their concept of ETDs is generally limited to providing better access to information, not additional or better information. UMI, on the other hand, discourages the use of multimedia supplements in ETDs because of licensing and copyright restrictions. Whereas many academic institutions provide free access to ETDs, UMI sells them. And whereas free downloads from a university Web site are considered "fair use," downloads from UMI are business transactions (see "Copyright and Fair Use" below). Presently there are more than a thousand academic institutions represented in UMI's ProQuest Digital Dissertations database. (19) Institutions that encourage the use of multimedia supplements in ETDs do so knowing that UMI will not distribute the supplements via the Web. The importance of this can hardly be exaggerated because in each and every instance the multimedia supplements are an integral part of the dissertation. (20) Moreover, UMI is selling downloads of these theses and dissertations as though they were complete. (21)
So what are multimedia supplements and what do they do? MP3 is the file format of choice for audio files embedded in ETDs. It is a network-friendly, nonproprietary file format requiring relatively little storage space. MP3 players can be downloaded for free on the Web, while high quality MP3 rippers (22) can be downloaded for as little as twenty dollars. Quick Time and MPEG Movie Player have been used by a number of authors who incorporated video clips into their theses and dissertations, and Sibelius Scorch (23) files were used by at least one dissertator to illuminate orchestration discrepancies in various nineteenth-century editions of Robert Schumann's Symphony in D Minor, op. 120. (24)
Apple Quick Time, MPEG Movie Player, and the two leading music notation program players (Sibelius Scorch and Finale Viewer), like MP3 players, can be acquired for free. The widespread availability of these applications can ensure future access to the contents of ETDs that rely on them if the authors and academic institutions that archive these documents take steps to make the software available in the future. This can be accomplished, in part, by downloading and maintaining archival copies of freely distributed applications, and then bundling them with corresponding media files on the host server. Thus, when a user clicks on a link to a multimedia supplement, the application will automatically run. The archiving institution will be stuck with the unenviable task of maintaining computer platforms and operating systems capable of running the bundled applications, but doing so would not be antithetical to the mission of many university archives and libraries where superseded media are routinely supported. (25)
COPYRIGHT AND FAIR USE
ETDs and paper theses and dissertations are afforded equal U.S. copyright protection under the law, regardless of the author's nationality or domicile and whether or not the copyright is registered. While university policies vary, it is the custom that the author of a thesis or dissertation is the owner of the copyright to that work, be it a bound volume or an ETD.
When multimedia supplements are utilized in an ETD a number of copyright issues arise having to do with the content of the supplement and the software required to read the file. There are two ways to look at these issues: (26) One can think in terms of commercial publishing and the laws, restrictions, and protections associated with that paradigm, or one can think in terms of not-for-profit scholarly publishing and fair use. Today, the vast majority of North American academic institutions adhere to the commercial paradigm--that is, a broad interpretation of the law--that UMI follows. (27) This requires that authors obtain written permission from the copyright holders of all materials they utilize, including images, audio excerpts, video clips, and computer applications. ("Ubiquitous free-ware" such as Netscape and Adobe Acrobat are exempted from this regulation.) In the context of our present economic and legal climate, this can be especially problematic. One need only look at the music industry's reluctance to allow anything they produce to be used without "just" (monetary) compensation to realize the magnitude of the problem.
The difficulty in obtaining permission to use materials deemed essential to scholarly enterprise contributed to the creation of our present laws governing copyright and fair use. Section 107 of the U.S. Copyright Law defines fair use in terms of the scope and purpose of that use. It states that copyrighted material can be used if
* use of the material has a legitimate, not-for-profit educational purpose, including commentary, criticism, news reporting, teaching, scholarship, or research;
* one considers the nature of the copyrighted work used (the scope is broader for factual works like textbooks, journal articles, bibliographies, treatises; and narrower for creative ones like novels, plays, and music);
* one considers the amount of the copyrighted work used (small portions of the total work generally weigh in favor of fair use, but use of larger amounts may be fair under certain circumstances); and
* one considers and moderates adverse effects on the market for, or value of, the work; difficulty getting permission is taken into account under this factor. (28)
If these criteria are met, use of copyrighted material in an ETD may be considered a fair use. In the final analysis it should be clear that the copyright holders' interests will not have been harmed by the reproduction and dissemination of their works, and that the author of the ETD has taken adequate steps to protect the copyright holders' interests. (For example, utilizing small parts of a work converted to a highly compressed electronic format with appreciably less fidelity ensures that the copied material will not be considered a substitute for the original product.) The University of Texas at Austin recently adopted this interpretation of fair use, and it is hoped other academic institutions will follow suite. This interpretation will provide students and their faculty advisers with additional impetus to include multimedia supplements in ETDs.
METHODS OF DEVELOPING DIGITAL ARCHIVES
There are basically three ways to create digital archives of the kind and scale we are discussing. (29) The primary differences among these approaches are the issue of centralization versus decentralization, the level of cooperation between organizations supplying the information, and the nature and level of human involvement in the developmental process. We have what can be described as the UMI approach, the NDLTD/OAI approach, and the commercial search-engine approach to developing large-scale digital enterprises, each having distinct attributes as well as inherent advantages and disadvantages.
The UMI approach is centralized and proprietary. UMI acquires the ETDs through a process whereby graduate students and institutions submit documents they wish to have included in the UMI archive and database of dissertations, called ProQuest Digital Dissertations. This approach affords product uniformity, tight control over issues having to do with access and content, and--financial considerations aside--access to a rather large quantity of information. But as UMI is a business, its approach to ETDs is profit driven; which is to say, you get what you pay for and only what you pay for (after the first twenty-four pages of a dissertation, which can be accessed free of charge). (30) Providing access to information in this manner facilitates commerce, first and foremost, not scholarly communication and learning.
The NDLTD/OAI approach is based on the harvesting of data from Web sites of a loosely federated group of institutions. To build an ETD archive using this approach, metadata must be collected from the sites of member institutions and integrated into a central system. This is, by and large, a three-step process. First you "crawl" the selected ETD sites to harvest (collect) pages containing ETD metadata. Then you parse the extracted pages to cull the relevant data. And after that is done, you make the data available through a standard protocol from which you build services, including a search interface.
The previously described OCLC-based NDLTD Union Catalog was developed in this manner, primarily by harvesting data from sites that utilize a standard metadata format called Dublin Core, (31) and support a simple metadata harvesting protocol developed for the Open Archives Initiative (OAI). (32) But this approach to building a digital archive is fraught with problems stemming from the large measure of autonomy that NDLTD member institutions retain, the decentralization of their collections, minimal interoperability among different digital library systems/protocols, and a lack of homogeneity with regard to the application of metadata, etc. (33) Because autonomy, decentralization, minimal interoperability, and heterogeneity are hallmarks of the NDLTD/OAI approach, projects of this type tend to be problematic and slow to develop, and access to information is uneven and often impeded.
NDLTD is working hard to overcome these limitations. They are developing an experimental system called MARIAN that allows for harvesting data from NDLTD member sites using a variety of standard protocols, not just the OAI protocol. But this system has yet to be perfected. The primary advantage to building collections the NDLTD/OAI way is that the operating costs are distributed amongst a large group of institutions.
A third approach to developing large-scale digital enterprises is the commercial search engine approach, which is all about gathering information that is openly accessible on the Web. This process is completely automated and therefore yields a poorer quality of service than can be achieved with the other approaches. Data collected in this manner lacks structure, consistency, and provenance; and because commercial search engines make no assumptions about the data they collect, the information they deliver is unfiltered, uneven, and poorly organized. Commercial search engines also lack authority control, thesauri, well-developed taxonomies, and the type of indexing that facilitates accurate, in-depth searching. As information retrieval systems go, they are cumbersome and unsophisticated. That said, there is very little cost to gathering information this way, so the quantity of data can be very great. When there is little or no human involvement in the developmental process, this is the best that can be done.
While each of these approaches to developing digital archives has distinct advantages and disadvantages, they all share one common shortcoming: none of them attempt to aggregate the information they collect with similar information. Why must ETDs be archived separately and not with other documents containing scholarly information when it is now widely accepted in library circles that organizing materials by medium in an online environment impedes access to information? In the twenty-first century we seek to aggregate information, not scatter it about an information universe that already is far too fragmented and chaotic.
There are a number of ways of accomplishing this objective, the most unambiguous of which is to have our libraries send up-to-date bibliographic information to OCLC for every item and every type of resource they collect, including ETDs. But OCLC is what it is: a bibliographic utility riddled with duplicate records, holes, and inconsistencies; and its database is not developed or managed in a way that facilitates the collection and maintenance of links to online resources. Moreover, its public catalog, WorldCat, is not freely accessible; similar to UMI's dissertation database, WorldCat is available only to users affiliated with a member library that subscribes to WorldCat.
In a more perfect world, the process of making ETDs accessible via OCLC would work like this:
* A graduate student creates an ETD and accompanying metadata, including descriptors and an abstract, and uploads the document to a server at the student's home institution via a network connection.
* The home institution (i.e., the library's technical services unit) creates a bibliographic record for the ETD, assigns a URL that is entered in the 856 field of the MARC record, uploads the bibliographic record to OCLC, and maintains the server on which the ETD resides.
* The researcher, for whom all this is being done, retrieves the ETD via a hypertext link in WorldCat, which has been accessed through an institutional affiliation, public library, or state-wide library system.
This is how the process should work. But if you have ever searched for ETDs in WorldCat you know that is not how it is. What you find instead is a hodgepodge of misinformation. There are bibliographic records for printed dissertations with ETD equivalents, but no mention of, or link to, the ETDs. You find notes stating that access to a particular ETD is restricted to the campus network of the home institution when, in fact, access to the document is unrestricted. You find a lot of broken links. And, worst of all, you do not find bibliographic records for a multitude of ETDs that will never be entered in the OCLC database. This includes ETDs created at non-OCLC member institutions in Europe and elsewhere, and ETDs created (by the author or home institution) after their print equivalents were cataloged and submitted to OCLC.
So, how best to create a reliable, readily accessible, aggregated resource for scholarly information on the Web? One approach to developing digital collections is described in my recent article in this journal entitled "The Economics of Information: DW3 and the Case for Creating a Music Megasite." (34) There it is argued that the best way to develop a large-scale, scholarly enterprise for Web-based resources in music is to form a consortium of large academic libraries with a mandate to create and maintain a comprehensive database of music-related links--a well-organized, well-funded project where tight control can be exercised over issues having to do with access, content, and presentation. (35) This enterprise would be developed by individuals as opposed to automated processes, and employ a high level of technology, a content management system or relational database that allows the developers to take advantage of enabling technologies such as XML (eXtensible Markup Language), (36) and robust metadata schemes that make possible the sophisticated search capabilities scholars require. This method of developing a large-scale enterprise is labor-intensive and expensive, but it would yield the highest quality service that one could hope to obtain, and the benefits of free, universal access.
It seems that most ETDs in the OCLC-based NDLTD Union Catalog are not indexed by Google and other popular Internet search engines, and those that are indexed are difficult to find if the search does not benefit from the incorporation of specific information, such as the author's name or the exact title of the document. Whether by design--because the robots (spiders) of commercial search engines are not being given access to the servers of NDLTD member institutions--or omission, this is a mistake that must be addressed. Inhibiting the flow of information in this manner thwarts scholarly communication because it impedes access to scholarship.
Conversely, a vertical index (or "vortal") like that which is described in "The Economics of Information" would foster scholarly communication by providing an unprecedented level of access to resources that otherwise might not see the light of day, resources such as the theses and dissertations in the NDLTD Union Catalog. Unlike information providers such as OCLC and Google, vertical indexes are subject-specific. Hence, they can achieve a level of depth and comprehensiveness with regard to Web content that cannot be achieved by larger, horizontal enterprises. Moreover, the contents of a vertical index can be indexed by commercial search engines if, indeed, the vertical index is constructed in a way that facilitates that process. And interoperability between vertical indexes can also be achieved.
Lest the importance of commercial search engines be overlooked, consider this: In June of 2000 I put up a Web site dedicated to the Spanish pianist and composer Isaac Albeniz. (37) The site is comprised of three essays, each being a revised chapter of my 1994 master's thesis. (38) Within six months of its inception the site was receiving one thousand hits a month; by spring 2001 the number increased to more than two thousand hits a month; and a year later the site was consistently receiving more than three thousand hits a month. (39) Is this an unusually high level of exposure for what is essentially a site devoted to scholarly communication? Probably. Are the ETDs on UMI's server and those of more than 196 NDLTD member institutions around the world receiving this kind of attention? Probably not. And it is likely that the majority of these theses and dissertations are never accessed for the simple reason that nobody can find them. My Albeniz site, on the other hand, has been indexed by numerous commercial search engines and is readily accessible to anyone with access to the Web; hence, the extremely high number of hits. The value of such exposure is immeasurable. (40)
THE FUTURE OF ETDS
Questions have been raised as to whether or not editors and publishers of commercial and not-for-profit academic presses will regard widely disseminated ETDs as prior publications and, as such, declare them ineligible for consideration as print publications. Two surveys of science journal editors and publishers--one conducted in 1999 by Joan Dalton of the University of Windsor, (41) the other in 2001 by Nancy Seamans of Virginia Tech (42)--indicate that this is by and large not the case. The majority of those surveyed by Dalton and Seamans (82 percent and 86 percent, respectively) indicated that a work derived from an ETD would be considered for publication; and editors from Elsevier and Academic Press (two industry giants that recently merged) are on record as saying that their companies encourage authors to link their journal articles to personal Web pages (i.e., their ETDs). (43) A survey in 2002 of humanities editors and publishers conducted by Bobby Holt of Virginia Tech yielded similar results. (44)
Considering this--and everything else that has been said in this essay--one might think that the future of ETDs is somewhat assured. But telltale signs of conservatism and intransigence abound. The preponderance and continued use of media-deficient PDF files and the adherence to an old interpretation and application of the copyright laws are demonstrative of the rather odd conservatism of existing ETD initiatives and the failure of academe to keep pace with changes in the environment at large. Why should an ETD be little more than an e-book equivalent of a printed thesis or dissertation--a digital representation of a linear document whose form has changed little in more than a hundred years? Why should it not look and function like the vast majority of Web pages and sites we access every day?
What metadata is to information retrieval, hypertext and multimedia supplements should be to the documents that comprise the next generation of ETDs. Hypertext supports nonlinear thinking and multidimensional representation of complex ideas, including ideas that can best, or only, be illustrated in nontextual mediums. A dissertator writing on music now has the option of incorporating audio files and video clips into a thesis, if doing so will illuminate the discourse. Sophisticated music notation programs such as Finale and Sibelius can be utilized in a similar way. These technologies have transformed our world and the information universe that virtually all students inhabit. So why not apply them to theses and dissertations, and apply them liberally? If we want to prepare the next generation of musicians, musicologists, and composers to function effectively in the Information Age we must move ahead.
1. Yale Fineman, "DW3 Classical Music Resources: Managing Mozart on the Web," portal: Libraries and the Academy 1, no. 4 (October 2001): 383-89; Fineman, "The Economics of Information: DW3 and the Case for Creating a Music Megasite," Notes 58, no. 3 (March 2002): 504-10; Fineman, "Electronic Theses and Dissertations," portal: Libraries and the Academy 3, no. 2 (April 2003): 219-27.
2. Based in part on Edward Fox, "A Brief History of ETD Activities: 1987-2000," chapter 1.4 in The ETD Guide, http://etdguide.org/ (accessed 24 February 2004).
3. A DTD states what "tags" and "attributes" are used to describe content in an SGML document, where each tag is allowed to be used, and which tags can appear within other tags. Tags are commands inserted in a document that specify how the document, or portion of the document, should be formatted. Attributes are characteristics such as underlining. (An underlined word would be said to have an "underline attribute.")
4. NDLTD members are listed at http://tennessee.cc.vt.edu/~lming/cgi-bin/ODL/nm-ui/members/index.htm (accessed 24 February 2004).
5. http://alcme.oclc.org/ndltd/servlet/OAIHandler?verb=ListSets (accessed 24 February 2004). Most of the 8,264 OCLC set records in the NDLTD Union Catalog are duplicate records of those listed by other NDLTD member institutions. Hence, at the present time there are approximately forty thousand set records in the NDLTD Union Catalog.
6. At the time this article was written in November 2003, there were 183 music-related ETDs in the OCLC-based NDLTD Union Catalog. Of these, 91 had restricted access, 87 unrestricted access, and 5 mixed access.
7. "About ProQuest Digital Dissertations," http://wwwlib.umi.com/dissertations/about_pqdd (accessed 24 February 2004).
8. Diplom.de, http://www.diplomica.com (accessed 24 February 2004).
9. http://www.dissertation.com (accessed 24 February 2004), not to be confused with the dissertation consulting service at http://www.dissertations.com. Data supplied by Jeff Young, CEO of Dissertation.com (August 2, 2003).
10. U.S. Department of Education, National Center for Education Statistics, Digest of Education Statistics, http://nces.ed.gov/programs/digest/d02/tables/dt257.asp (accessed 24 February 2004).
11. Ibid., table 255, "Bachelor's, Master's, and Doctor's Degrees Conferred by Degree-Granting Institutions, by Sex of Students and Field of Study: 2000-2001," http://nces.ed.gov/programs/digest/d02/tables/dt255.asp (accessed 24 February 2004).
12. Edward Fox, "What are ETDs?" chapter 1.1 in The ETD Guide.
13. PDF is a proprietary file format of Adobe Systems, Inc. Whereas HTML files and graphic file formats such as JPEG, GIF, and TIF are supported by numerous applications, PDFs require a specific application (Adobe Acrobat Reader) to be viewed. U.S. and Canadian colleges and universities rely heavily on PDF for storing and disseminating different types of information, and it would be disastrous if Adobe Systems were to go out of business or stop supporting this file format (i.e., product).
14. Early English Books Online (http://eebo.chadwyck.com/home) and Engineering Handbooks Online (http://www.engnetbase.com/default.asp) utilize PDFs, whereas netLibrary's (http://www.netlibrary.com) holdings are rendered in HTML (all accessed 24 February 2004).
15. Submission over networks costs nothing in comparison to the expense of printing, copying, binding, processing, and warehousing paper theses and dissertations because most institutions already have adequate network and computing capabilities to accomplish this. Likewise, most academic institutions also have sufficient resources for students to create ETDs.
16. The following ETDs contain audio files (WAV, MIDI, MP3) or video clips (Apple QuickTime): Boris Becker, "Wahrnehmung und Wirkung der Trommel: Perception and Effect of the Drum" (Doctoral diss., Universitat Duisburg, 2002); Ivica Bukvic, "Meditations 3: The Sea" (M.M. thesis, University of Cincinnati, 2000); Federico Garcia, "An Anatomy of the World for Voice and Six Instrumentalists, on Texts by John Donne" (M.A. thesis, University of Pittsburgh, 2003); Rebecca Gearhart, "Ngoma Memories: A History of Competitive Music and Dance Performance on the Kenya Coast" (Ph.D. diss., University of Florida, 1998); Sally Stephenson, "Portraits of the Songwriting Process in Elementary Classrooms" (Ed.D. diss., West Virginia University, 2001); Michael S. Taylor, "There Arises Light: A Work for Orchestra in One Movement" (M.M. thesis, University of Tennessee, Knoxville, 2002).
17. ETD downloads from ProQuest Digital Dissertations consist of nothing more than page images (what UMI calls "image-only PDF" files) that purchasers may print or read on a monitor.
18. Over 80 percent of NDLTD member institutions permit students to utilize multimedia supplements, but 95 percent of ETDS submitted are comprised of "plain vanilla" PDFs. For additional information on multimedia and ETDs, see Judith R. Edminster, "The Diffusion of New Media Scholarship: Power, Innovation, and Resistance in Academe" (Ph.D. diss., University of South Florida, 2002); electronic version at http://dmi.usf.edu/edminster/Dissertation/ (accessed 24 February 2004).
19. See "About ProQuest Digital Dissertations," n. 7.
20. An example of this inequity is the 2003 graduate thesis of Roger Luke DuBois, "Applications of Generative String Substitution Systems in Computer Music" (Ph.D. diss., Columbia University, 2003). This thesis includes both audio and video files that cannot be downloaded from the UMI Web site along with the text. The entire dissertation, including the multimedia supplements, can be downloaded free of charge from the author's Web site at Columbia University, http://www.music.columbia.edu/~luke/dissertation/ (accessed 24 February 2004).
21. UMI treats all theses and dissertations with multimedia supplements as "compound documents." This class of document is only distributed by UMI in hardcopy, as a printed volume with an accompanying CD-ROM. It would be most helpful if UMI would advertise this limitation within the search results, citations, and subsequent screens where business is transacted. Doing so would alert those who may want to download a thesis or dissertation that media supplements cannot be obtained via the Web. Presently, there is only one mention of this limitation on the ProQuest Digital Dissertations Web site; it appears on an introductory screen entitled "Guidelines for Submission of Dissertations and Masters' Theses in Electronic Format," http://wwwlib.umi.com/dissertations/about_etds (accessed 24 February 2004).
22. MP3 rippers convert regular audio CD tracks into computer-friendly WAV and MP3 file formats.
23. Scorch is a free Web browser plug-in that facilitates viewing, listening to, customizing, and printing of scores created with Sibelius music notation software on the Internet. See http://www.sibelius.com (accessed 1 March 2004).
24. Jean Marie Hellner, "Robert Schumann's Symphony in D Minor, op. 120: A Critical Study of Interpretation in the Nineteenth-Century German Symphony" (Ph.D. diss., University of North Texas, 2003); electronic version at http://www.library.unt.edu/theses/open/20031/hellner%5Fjean%5Fmarie/index.htm (accessed 24 February 2004). Most regrettably, these multimedia supplements were not included in the final version of the dissertation.
25. The International Piano Archives at Maryland, for example, supports every form of superseded media on which significant piano music has been recorded, including piano rolls. If these recordings can be reformatted (migrated to a modern medium), they are; and when they cannot be reformatted, the recordings are utilized in their original formats.
26. For an excellent discussion and legal opinion on fair use and ETDs see Tim Brace and Georgia Harper. "Lessons of a Thousand: Two Years of Required Electronic Dissertation Submission at the University of Texas at Austin." http://edoc.hu-berlin.de/etd2003/brace-tim/HTML/index.html (accessed 24 February 2004).
27. UMI's interpretation of copyright and fair use are articulated in a document entitled "Copyright Law and Graduate Research: New Media, New Rights, and Your New Dissertation" by Kenneth D. Crews, associate dean of the faculties for copyright management at Indiana University-Purdue University in Indianapolis, and professor in the School of Law, and the School of Library and Information Science, http://www.umi.com/hp/Support/DServices/copyrght/ (accessed 24 February 2004). This document discusses copyright law from the perspective of a commercial publisher, not an academic institution or an individual whose work may be published in accordance with fair use guidelines, which is something UMI cannot do.
28. Brace and Harper, part 2: "There are times when a requirement to get permission would make it impossible for us to use certain materials for educational purposes, either because permission is too difficult to obtain, too expensive, or even impossible. But that is when our reliance on fair use is on its firmest footing. In fact, that is the essence of fair use: to make it possible for educators to use others' works without their permission when our use furthers the goals of copyright without undermining the incentive the law provides to authors and their assignees."
29. Marcos Andre Goncalves, et al., "Web-DL: An Experience on Making Digital Libraries from the Web," unpublished paper, Virginia Tech (2002), 3. In this article the authors also state that there are three basic approaches to developing digital archives. But their categorization fails to consider other (equally, if not more viable) approaches to developing digital collections.
30. UMI charges academic users $25.50 to download a complete dissertation in PDF.
31. Dublin Core Metadata Initiative, http://dublincore.org (accessed 24 February 2004).
32. Open Archives Initiative, http://www.openarchives.org/ (accessed 24 February 2004).
33. Goncalves, 2.
34. Cited in n. 1.
35. We might look to the three most important bibliographic documentation projects in the field of music--RILM, RISM, and RIPM--for insight on how to form, fund, and manage a successful consortium or international collective.
36. Unlike PDF and HTML, which mix content and formatting together into static pages that are unwieldy and difficult to maintain, XML enables the creation, importation, indexing, maintenance, and delivery of content as discrete objects, separate from any particular format. This allows the use and reuse of content easily in many applications and formats. Separating content from presentation in this manner facilitates flexibility and interoperability that assures the long-term success of the enterprise.
37. The Life and Music of Isaac Albeniz, http://www.lib.umd.edu/PAL/YALE/albenizl.html (accessed 24 February 2004).
38. Yale Fineman, "Isaac Albeniz and the Andalusian Musical Idiom" (M.A. thesis, Tufts University, 1994).
39. On the Duke University server where this site resided before being moved to the University of Maryland, a successful request for a page with a distinct URL constitutes one hit. My Albeniz site is comprised of six pages: an index page; three separate but related essays, each with its own URL; a page for footnotes and one for links. Hence, when I say that the site receives more than 3,000 hits a month, that number is the total number of pages requested in a particular month.
40. This site has brought me in touch with a wide array of scholars and performers, including the pianist Francisco Aybar, a renowned interpreter of Spanish music for whom I have written program notes (Carnegie Hall recital, 19 January 2003).
41. Joan Dalton, "Electronic Theses and Dissertations: A Survey of Editors and Publishers," unpublished paper presented at the "Third International Symposium on Electronic Theses and Dissertations: Applying Media to Scholarship" (ETD 2000), 16-18 March 2000, http://dmi.usf.edu/Conference/ (accessed 24 February 2004).
42. Nancy H. Seamans, "Electronic Theses and Dissertations as Prior Publications: What the Editors Say," Library Hi Tech 21, no. 1 (March 2003): 56-61.
43. Gail McMillan, "Do ETDs Deter Publishers? Coverage from the 4th International Symposium on ETDs," College & Research Libraries News 62 (2001): 620-21.
44. The results for all three surveys are available at the Digital Library and Archives database of the University Libraries of Virginia Tech, http://lumiere.lib.vt.edu/surveys/results/ (accessed 24 February 2004).
Yale Fineman is music librarian and head of reference and circulating collections at the Michelle Smith Performing Arts Library, University of Maryland. Portions of this article were presented as part of a panel discussion on ETDs at the 2002 annual meeting of the American Musicological Society. A revised version of that presentation was later published as "Electronic Theses and Dissertations," portal: Libraries and the Academy 3, no. 2 (April 2003): 219-27; electronic version at http://muse.jhu.edu/journals/portal_libraries_and_the_academy/toc/pla3.2.html (accessed 24 February 2004). I would like to thank all those who generously contributed their time and expertise to the writing of the present article, especially Georgia Harper, attorney, Office of General Counsel, University of Texas: Anne Bowden, university counsel, University of Maryland; and H. Robert Cohen, founder and general editor, Retrospective Index to Music Periodicals (RIPM).
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 1, 2004|
|Previous Article:||Forgery in the Music Library: a cautionary tale.|
|Next Article:||Necrology index.|