Theoretical Writings

Librarians and library users alike have long recognized the access problems associated with both printed and recorded music. Donald R. Wakeling's 1945 article humorously described several problems from the music librarian's point of view: misattribution of works, frequent borrowing or reworking of material from already-existing works, and the frequent need for the librarian to recognize requested materials by sight or from the (usually atrocious) singing of the patron. [1]

It is particularly significant that when Wakeling's article was reprinted in a reader on current topics in music librarianship, the situation had still remained unchanged, even after 32 years.[2] And, despite rapid growth in the application of technology in libraries since 1973, Wakeling's article continued to accurately describe the situation. Writing in 1989 on authority control of music under AACR2, Kären Nagy described the same problems -- 46 years after Wakeling![3]

A solution to the problems Wakeling and Nagy described has long been available, however. According to Brook,[4] in his well-known book on thematic catalogs, the earliest known example of a thematic library catalog was begun in 1720 and the earliest published library catalogs containing thematic indexes were published in 1843.

Interest in computer thematic indexing among musicologists began after the Second World War when there was a growing recognition of the benefits of technology in many scientific and academic endeavors. A symposium on musical representation,[5] held in 1965, addressed the benefits and problems of translating conventional musical notation into a machine-readable format. Several possible systems were discussed, including DARMS, Plaine and Easie Code, ALMA, and others. Articles by LaRue and Logemann[6] and by Zimmermann[7] each advocated cooperative effort to achieve widespread availability of thematic indexing and listed the benefits, ranging from identification of unfamiliar melodies to musicological research in such a database.

It was not until 1982, however, that Miller and Miller proposed online thematic indexing, calling for the addition of 2 fields to the MARC record.[8] In their scheme, one field would contain an encoded, alphanumeric incipit, perhaps in Brook's Plaine and Easie Code.[9] This field would only be descriptive, however, because it could not be used to retrieve the same melody in differing keys. Therefore, a second field would be used for either a transposed or intervallic representation of the incipit.

Anticipating objections due to additional costs, Miller and Miller stated that since music cataloging requires on average twice the time necessary to catalog a book, two extra fields would not add significant time. Likewise, Roughton and Tyckoson[10] argued that the addition of another coded field has a negligible effect on storage requirements in either bibliographic records or online catalog indexes.

Unfortunately, the music library community has not adopted thematic indexing for its catalogs on a large scale. The Music Library Association appears never to have officially addressed the possibility of thematic indexing. It was conspicuously absent from a report on music and automation.[11] In addition, a report on improved access for music described the inadequacy of the Library of Congress Subject Headings for treating multi-faceted work such as music, and the inherent problems of the MARC format for music that hamper Boolean searching,[12] but failed to mention any possibility of content-based indexing. The report compares the merits of natural language versus controlled vocabulary but does not include any discussion of the questions 'What is the natural language of music?' or 'What is music about?'

So in 1993, due in part to the library profession's reluctance or inability to address them, the problems described by Wakeling still exist. Jan LaRue, one of the earliest proponents of computerized thematic indexing, a participant in the 1965 symposium on music representation, and a researcher in the computerized analysis of musical style, describes the current situation as an "ancient crisis." [13]

Attention has returned in recent years to the issue of musical representation. As evidence of this renewed interest, two issues of an important music journal were dedicated to representation and scoring. An overview of the topics of hierarchy and structure, extensibility, protocols and encoding mechanisms, and others were offered by Dannenberg[14] in the first issue while the second contained an in-depth description of the HyTime/SMDL representation.[15] Standard Music Description Language (SMDL) seemd particularly attractive due to its relationship to Standard Generalized Markup Language (SGML) and the tagged-text method which has become popular in recent years. Another indicator of renewed interest was the recent release of the Humdrum Toolkit. Not only was it freely available, text-based, and highly extensible, but it also provided tools for searching and manipulating the encoded data.[16]

Perhaps, then, there is still reason to hope that the "ancient crisis" can be resolved. Carol Mandel, writing on possible enhancements to the online catalog, stated that

      [o]ur users are, in fact, asking for indexing
     information in the library catalog.  Respondents to
     CLR's Online Catalog User Study ranked access to book
     indexes and tables of contents as the second most desired
     enhancement to the online catalog.[17]

It appears that the indexing information required for music is the musical text itself. Several methods are available for providing thematic indexing. It can be done, but should it be done? Mandel wrote

     Eventually the decision [whether to provide these
     enhancements] will be made by a jury of library users.  But  
     the librarian's role is not passive. ... Should we enhance     
     the MARC record to improve subject access?  We won't know     
     until we try.[18]

Systematic Research

A review of the literature revealed several reports of research directly and indirectly related to the provision of online thematic indexing. Basart reported the results of an informal survey of librarians and publishers taken in the 1980s[19] in which she found that music librarians, when asked what they needed in the way of reference tools, gave high priority to thematic indexes covering major 19th-century composers. In addition, they identified electronic databases as highly desirable. The most urgent need, however, appeared to be a post-1800 international bibliography of printed music indicating which pieces were arrangements. While online thematic indexing was not mentioned specifically, it could be an appropriate remedy for each of these needs.

A survey of faculty members conducted at the graduate degree-granting centers of the State University of New York in 1992 found that

     High levels of use of library online catalogs and
     online index/abstract databases loaded on the catalog
     indicate that these resources are good targets for
     continued expansion and expenditure of increasingly
     scarce funds.[20]

Only 11 percent of the faculty reported a lack of interest or need as an obstacle to their uses of electronic resources. Furthermore, the authors unequivocally stated that

     Faculty readiness, in terms of necessary equipment
     and interest, to access to electronic information
     resources, is almost universal.[21]

Presumably, a similar level of interest exists at other locations and among users of music libraries. Yet, virtually no research has been done to discover who the users of music libraries are and what services they desire. Only a single example of a user survey of music libraries was found.[22] However, there was no indication the survey was ever actually conducted.

On the other hand, Troutman's research indicated almost universal dissatisfaction among the members of the MLA electronic discussion list with the access provided to music by current OPACs.[23] In proposing a solution, however, she rejected the notion of the special status of music materials and advocated a continued reliance on uniform titles.

Burbank and Henigman's research indicated a problem with this approach, however. Their survey of OPAC vendors found that the basic symbols of music, i.e. '#' and 'b', so often necessary for uniform title searches, received exasperatingly poor treatment. The majority of catalogs (54%) eliminated the symbols from the search string. Of these, 8 percent used '#' as the truncation symbol, further complicating the problem.[24]

Temperley discussed the problems encountered in preparing a thematic index to a limited repertoire -- printed sources of hymn-tunes with English language texts from the Reformation through 1820.[25] The problems included incipits that were too short for adequate identification, melodies with the same notes but different meters, and similar tunes with different incipits due to pickup notes, altered notes, or melismatic passages, especially where the first note of the melisma was not a main note of the melody.

An early implementation, pre-dating the use of the MARC format for music, of thematic indexing in a computerized catalog was reported by Schiødt.[26] An incipit was included for all works, using the "Plaine and Easie Code." For those items with a text, the incipit was coded beginning at the start of the text and a text incipit, in another field, was aligned with the music. For instrumental works, the main theme was encoded. If some doubt existed regarding where the main theme commenced, the most characteristic figuration was encoded starting from the beginning of the piece.

Finally, Paisley's research on identifying "unknown" creators[27] was found by the researcher to be exemplary, the kind of research that might be conducted on a much larger scale if online thematic indexing were more readily available. Paisley analyzed the use of two-note melodic intervals in works by Bach, Haydn, Mozart, Beethoven, and Brahms, successfully demonstrating that statistical methods could be used to sharply differentiate each composer from the others. In addition, Paisley showed that stylistic differences existed which could be used to determine authorship of a set of "unknown" works.

The widespread dissatisfaction with current access methods for music, the high level of interest in computerized tools among many users, and the renewal of interest in the digital representation of music indicate that now could be the most opportune time for libraries to implement thematic indexing. The first step, however, must be an exploration of the need for thematic indexing among the users of music research libraries.

Chapter Two Notes

1. Donald R. Wakeling, "Some Trials of a Music Librarian," The Music Review 6 (1945): 13-16.

2. Donald R. Wakeling, "Some Trials of a Music Librarian," in Reader on Music Librarianship, ed. Carol June Bradley. (Washington: Indian Head Editions, 1973): 136-138.

3. Kären Nagy, "Music Authority Control: A Public Service Perspective," in Authority Control in Music Libraries, ed. Ruth Tucker, MLA Technical Report No. 16 (Canton, MA: Music Library Association), 15.

4. Barry S. Brook, Thematic Catalogues in Music: An Annotated Bibliography. (Hillsdale, NY: Pendragon Press, 1972), xii.

5. Barry S. Brook, ed., Musicology and the Computer: Three Symposia (New York: City University of New York Press, 1970), passim.

6. Jan LaRue and George Logemann, "EDP for Thematic Catalogues," Notes 22, no. 3 (1965/66): 1179-1186.

7. Franklin B. Zimmermann, "Melodic Indexing for General and Specialized Use," Notes 22, no. 3 (1965/66): 1187-1192.

8. Karen Miller and A. Patricia Miller, "Syncopation Automation: An Online Thematic Index," Information Technology & Libraries 1, no. 2 (Sept. 1982): 270-274.

9. Brook, Barry S., "The Simplified Plaine and Easie Code System for Notating Music, a Proposal for International Adoption," Fontes Artis Musicae 12 (May-Sept. 1965): 156-160.

10. Karen G. Roughton and David A. Tyckoson, "Browsing with Sound: Sound-Based Codes and Automated Authority Control," Information Technology & Libraries 4, no. 2 (June 1985): 130-132.

11. Lenore Coral, et al. "Automation Requirements for Music Information," Notes 43 (Sept. 1986): 14-18.

12. "Improving Access to Music: A Report of the MLA Music Thesaurus Project Working Group," Notes 45 (April 1989): 714-721. 13. Jan LaRue and David Cannata, "An Ancient Crisis in Music Bibliography: The Need for Incipits," Notes 50 (Dec. 1993): 502-518.

14. Roger B. Dannenberg, "Music Representation Issues, Techniques, and Systems," Computer Music Journal 17 (Fall 1993): 20-30.

15. Donald Sloan, "Aspects of Music Representation in HyTime/SMDL," Computer Music Journal 17 (Winter 1993): 51-59.

16. David Huron, The Humdrum Toolkit. (Menlo Park, CA: Center for Computer Assisted Research in the Humanities, 1994), 7-37.

17. Carol A. Mandel, "Enriching the Library Catalog Record for Subject Access," Library Resources and Technical Services 29 (Jan./March 1985): 10.

18. Mandel, 15.

19. Ann Basart, "Reference Lacunae: Results of an Informal Survey of What Librarians Want" in Foundations in Music Bibliography, ed. Richard Green (New York: Haworth Press, 1993): 365-384.

20. Judith A. Adams and Sharon C. Bonk, "Electronic Information Technologies and Resources: Use by University Faculty and Faculty Preferences for Related Library Services," College and Research Libraries 56 (March 1995): 129.

21. ibid.

22. Susan M. Clegg, "User Surveys and Statistics -- The Opportunities for Music Libraries," Fontes Artis Musicae 32 (Jan. 1985): 69-75.

23. Leslie Troutman, "The Online Public Access Catalog and Music Materials: Issues for System and Interface Design," in Advances in Online Public Access Catalogs. (Westport, CT: Meckler, 1992), 9-37.

24. Richard Burbank and Barbara Henigman, "Music Symbols and Online Catalogs: A Survey of Vendors and an Assessment of Retrieval Capabilities," Information Technology and Libraries 11 (Sept. 1992): 203-209.

25. Nicholas Temperley, "The Problem of Definitive Identification in the Indexing of Hymn Tunes," in Foundations in Music Bibliography. ed. Richard Green. (New York: Haworth Press, 1993), 227-239.

26. Nanna Schiødt, "MUSICAT: A Method of Cataloguing Music Manuscripts by Computer, as Applied in the Danish RISM Manuscript Project," Fontes Artis Musicae 23 (July-Sept. 1976): 158-167. 27. William J. Paisley, "Identifying the Unknown Communicator in Painting, Literature and Music: The Significance of Minor Encoding Habits," Journal of Communication 14 (1964): 219-237.

Next section
Previous section
Home page