BCC94/WGFAM/1 DISCUSSION PAPER FACETED ACCESS TO MUSIC: POSSIBILITIES AND RAMIFICATIONS

BCC94/WGFAM/1 DISCUSSION PAPER FACETED ACCESS TO MUSIC: POSSIBILITIES AND RAMIFICATIONS presented by the Working Group on Faceted Access to Music February 1994 This paper addresses the following questions: *Why does the Working Group on Faceted Access to Music exist? *What is faceted access? *How can faceted access improve subject retrieval? *How does a "thesaurus" relate to faceted access and why is one needed? *What work has been done so far on the Music Thesaurus? *How will we use the thesaurus and how will it be implemented? *What questions still need to be answered about the possibilities and ramifications of implementing the Music Thesaurus? 1. Why does this working group exist? In the 1989 article "Improving Access to Music," the Music Library Association's Music Thesaurus Project Working Group discussed problems with subject access to music, some resulting from inadequacies in the Library of Congress Subject Headings (LCSH), and others caused by flaws in the design of the USMARC format relating to the 04x fields. The recommendation of the Music Thesaurus Project Working Group was that a thesaurus for the discipline of music be created as a solution to these problems. In 1991, with funds from the Council on Library Resources, Harriette Hemmasi and her colleagues James Anderson and Fred Rowley at Rutgers University began preliminary work on the creation of a music thesaurus, and Hemmasi has since then continued work on building the thesaurus. At the 1992 Music Library Association Annual Meeting, Jerry McBride distributed a discussion paper, "Changes to 04x Fields in MARC Records for Bibliographic Description." In this paper, he outlined problems in subject access to music resulting from the design of the 04x fields and LCSH, and presented several possible solutions. The late Kevin Freeman, at the 1993 Annual Meeting, gave a paper entitled "Faceted Access to Music: Proposals on 04x and 658 MARC Fields," in which he suggested that the "adoption of a system of faceted indexing for musical materials" would solve the problems outlined by McBride. Freeman stated that "there must be a consensus in the music library community" before such a system can be adopted, and proposed the creation of a working group on faceted access. His proposal was taken by the Subject Access Subcommittee to the Bibliographic Control Committee, which agreed to create this working group. The Working Group on Faceted Access to Music is charged with facilitating the adoption of faceted indexing for musical materials. The Working Group recognizes that a system of faceted indexing for music cannot be adopted until an appropriate tool, such as a music thesaurus, is available. One of our specific charges is to explore ways to promote the understanding of faceted access among music librarians. We hope that this discussion paper will increase understanding of the issue and will promote widespread discussion among the MLA membership. There will be a forum for discussion at the open meeting of the working group at the 1994 Annual Meeting in Kansas City on Friday, March 4, from 2:30-3:30. The members of the working group are listed at the end of this paper, and we welcome your comments, concerns, and questions. 2. What is faceted access? Library of Congress subject headings frequently combine together into one heading terms representing different concepts, such as "Music and anthropology", "Opera-- Dramaturgy", and "Chiming clocks--germany." Generally, these

concepts are not differentiated from each other within the heading. In a faceted-access system, on the other hand, the concepts associated with a particular discipline are identified and organized into unique, discrete categories called facets. (Barnett and Petersen p 15) Each concept is denoted by a term, a "word or expression that has a precise meaning in some uses, or is peculiar to a science, art, profession or subject." (Foskett pp 68-69) These terms are listed in a thesaurus, which organizes them hierarchically according to facet, showing each term's relationship to other terms within the discipline. The conceptual context that is thus provided for each term is one of the great advantages of a faceted-access system: it enables both searchers and indexers to easily understand in what context any term is used in a particular discipline. 3. How can faceted access improve subject retrieval? Brad Young has noted that subject access to music requires "two things now generally lacking in standard subject cataloging:" 1) non-topical access points, and 2) multi-element access points. (Young p 184) Music librarians are particularly aware of the need for nontopical subject access. The need for multi-element access points is met by those online library catalogs that allow keyword searching of subject headings; if the desired facet of information is brought out anywhere in the subject heading string, it is now available to users through keyword searching. However, facets of information required by music users often do not appear in subject headings. An example of this problem is given in the 1989 report of the Music Thesaurus Project Working Group: "A user should be able to locate Schubert's "Der Hirt auf dem Felsen" by searching for any combination of the following: high voice, piano, and clarinet." However, the Library of Congress subject heading for this work is "Songs (High voice) with instrumental ensemble." The phrase instrumental ensemble is not specific enough to meet the needs of the user searching by exact instrumentation. The 048 field of the USMARC format was created specifically to provide access via instrumentation. However, this field is not used by many libraries because 1) it requires time-consuming coding instead of natural language, 2) the codes are not comprehensive (for example, no code exists for contrabassoon), and 3) many library catalogs do not index this field. McBride's discussion paper thoroughly examines problems with the 04x fields. For those systems that do index this field, an additional problem remains: the inability to link the 048 field to a particular work in an analyzed collection, resulting in "false drops" (unwanted retrievals of bibliographic records containing the desired search terms, but not in the desired context). The adoption of a faceted-access system, such as the Music Thesaurus, will not solve the problem of false drops in catalog records for analyzed collections. Implementation of a linking device in the USMARC format is necessary to eliminate this problem. Related to the need for nontopical access is the problem of access to the object as subject. Our current system is inconsistent, at best, in allowing users to distinguish between physical description and topical content. For example, LCSH creates a subtle distinction between opera as a musical form and works about opera. The plural heading "Operas" is assigned to works that are manifestations of operas, either printed or recorded. The singular heading "Opera" is assigned to works about opera. However, this logic is not followed throughout LCSH. Though musical settings of Robert Browning's poetry can be found under "Browning, Robert, 1812-1889-- Musical settings," a search for information about musical settings of Robert Browning's poetry will not be successful if sought under "Browning, Robert, 1812-1889-- Musical setting." Another approach to this problem is evident in a recent subject heading change: "Madrigal" has been discontinued in favor of "Madrigals--History and criticism." The growing choice of USMARC fields for coding object characteristics is evidence of the need for indexing this information, but is also creating confusion among catalogers, system designers, and users about how best to distinguish between object and subject. (Dooley and Zinkham pp 44-45) The adoption of a faceted indexing system for music would necessarily involve grappling with this problem and implementing a consistent solution to it. 4. How does a "thesaurus" relate to faceted access and why is one needed? The report of the Music Thesaurus Project Working Group quotes the American National Standards Institute's definition of a thesaurus as "a compilation of words and phrases showing synonymous, hierarchical, and other relationships and dependencies, the function of which is to provide a standardized vocabulary for information and retrieval systems." (MLA p 720) Foskett defines a thesaurus as "a list of terms showing their classification according to the ideas they represent... Such a list can only be compiled by concept indexing." (Foskett p 115) A thesaurus is constructed using facet analysis to identify concepts and denote them by terms, and then to determine the relationships among those

terms. The thesaurus makes these relationships among terms evident to the users of the thesaurus. Why are these relationships so important? Because the success of an information retrieval system is dependent upon its syndetic structure, the organizational framework that allows links to be established among related entities. Smiraglia writes that it is the connections among headings and the entities they represent that "makes collocating devices out of what would otherwise be mere finding lists." (Smiraglia p 164) Collocation enables the user to see a given work as a member of a class of related works. The need for adequate syndetic structure was brought to light when the creators of the Art and Architecture Thesaurus attempted to organize the terms related to art and architecture in the Library of Congress Subject Headings and other subject lists into all the possible categories and hierarchies necessary to cover the field. They were struck by the presence of large gaps throughout the system. As Toni Petersen explained, "When the subject terms of a field of knowledge are arranged conceptually rather than alphabetically, the limitations of the alphabetically based indexing systems clearly emerge. It is no wonder that art librarians have been complaining so strongly about LCSH. The words they need to describe their field are not all present." (Petersen 1983 p 209; Petersen 1990 p 650) The reasons for the gaps in LCSH coverage include: 1) LCSH is an alphabetical list of subject headings; 2) its purpose is not to represent the knowledge domain of a particular subject or discipline; 3) LC subject headings are created only as needed (as items are added to the Library of Congress collection); 4) there is no way to find conceptual gaps in a list of headings organized alphabetically rather than conceptually. Besides showing us the gaps in music terminology that exist in LCSH and enabling us to fill those gaps, a thesaurus that conceptually arranges the music terms in LCSH will enable us to see the entirety of our discipline--to find the terms that are already there, in their context--which will be a great benefit to our users. Since 1974, international standard thesaurus guidelines have provided rules for constructing terms, establishing hierarchical relationships, and assigning related terms. (Taylor pp 454-455) Mary Dykstra emphasizes that LCSH consists of "conglomerations of words that describe subjects," rather than thesaurus "terms." She writes, "It is extremely important to grasp the fundamental distinction between subject heading and term because this distinction forms the basis of where LCSH and thesauri part company with each other." The international thesaurus standards also clearly specify "exactly how terms may relate to one another. In other words, once the basic building blocks are established, there are rules for how these building blocks may be used in the construction of larger relational structures." (Dykstra p 44) Dykstra provides a summary of the two hierarchical relationships allowed by the thesaurus standards; these hierarchical relationships are denoted by the codes "Broader Term" and "Narrower Term." Besides the relationships designated by "Broader Term" and "Narrower Term," thesaurus standards also include instructions for establishing the relationships between terms that are coded "Use," "Use For," and "Related Term." Although LCSH borrows thesaurus terminology such as "Broader Term" and "Narrower Term," its headings are not terms, and their relationships are not constructed according to international guidelines for thesaurus construction. The resulting inadequate syndetic structure of LCSH is evident in the results of a study by Karen Markey and Diane Vizine-Goetz indicating that "nearly 40 percent of LC topical subject headings have no Broader Terms and that a substantial number of them do not have Narrower Term references (75.5 percent of all headings)." (Whitehead 1990 pp 81-82) For example, "Poetry and music" is a nonpreferred heading; we are told instead to use "Music and literature," which means that to find information about poetry and music we must wade through all the citations to works about music and all kinds of literature. 5. What work has been done so far on the thesaurus? Because the use of the Library of Congress Subject Headings is so widespread, and so many bibliographic records contain those headings, the objective of the Music Thesaurus Project is to base the terminology of the Music Thesaurus on the vocabulary in LCSH. The creation of a music thesaurus from LCSH vocabulary will assist in identifying and filling gaps in the existing syndetic structure for music, and will allow missing terms to be supplied. The resulting thesaurus will be "tied both conceptually and through a shared vocabulary directly to LCSH." (Hemmasi) Using thesaurus construction software (ARIS, or Anderson Rowley Information Systems), the Music Thesaurus Project created a database of LC music subject headings through September 1992. Hemmasi then used the ARIS software to "deconstruct" the subject headings, factoring out the separate concepts represented in each heading and denoting them

by terms taken from LCSH vocabulary whenever possible. "While preserving the original structure of the music headings, the database provided new ways of searching and viewing the pre- coordinated LC subject headings, individual terms within those headings, and heading/term relationships." (Hemmasi) Hemmasi identified and ranked the concepts represented in the database, and used this conceptualization to design a hierarchical structure of five facets: Agents Forms/Genre Geo-cultural attributes Sound devices Other topics Hemmasi then sorted the 12,000 subject headings into categories that correspond to the five facets. In the process of developing the forms/genre facet for vocal music, Hemmasi has expanded the tentative outline for the thesaurus as follows: Code A Agents musicians E Events ceremonies* fasts/feasts/festivals* holidays/seasons* services* F Forms/genre instrumental vocal* sacred* secular* G Geo-cultural attributes languages* locations religions* S Sound devices instrumental vocal* T Texts sacred* secular* O categorized Other topics terms/facets not yet *in progress

6. How will we use the thesaurus and how will it be implemented? Though it is too early in the development of the thesaurus to know the answer to this question, Harriette Hemmasi has provided the following examples of how the Music Thesaurus might be implemented, using the codes for facets from the above outline, and assuming use of subfield delimiters $c=facet, $a=primary term, $b=secondary term, and $n=number. Ex. 1. Work: Psalm 23 for voice, clarinet and percussion Sacred songs with instrumental ensemble Psalms (Music)--23rd Psalm Faceted Text: Psalm 23 Approach: solo voice clarinet percussion $c t $a Psalm 23 $c s $b voice $c s $b clarinet $c s $b percussion Ex. 2. Work: Sacred cantatas for solo soprano voice Solo cantatas, Sacred (High voice) Faceted: Forms/genre: cantatas (sacred) Approach: soprano $c f $a cantatas (sacred) $c s $b soprano Ex. 3. Work: Christian songs for high voice with chamber orchestra orchestra Sacred songs (High voice) with chamber Faceted Forms/genre: songs (sacred) Approach: Geo-cultural: Christian high voice chamber orchestra $c f $a songs (sacred) $c g $b Christian $c s $b high voice $c s $b chamber orchestra Ex. 4. Work: Portuguese love songs for SSAA with piano Choruses, Secular (Women's voices, 4 parts)

Love Songs with piano Faceted Forms/genre: love songs Approach: Geo-cultural: Portuguese soprano Number: 2 alto Number: 2 piano $c f $a love songs $c g $b Portuguese $c s $b soprano $n 2 $c s $b alto $n 2 $c s $b piano Ex. 5. Work: Psalm 100 in German for chorus (SATB) with horn and organ Choruses, Sacred (Mixed voices) with instrumental ensemble Psalms (Music)--100th Psalm Faceted Text: Psalm 100 Approach: Geo-cultural: German chorus (SATB) horn organ $c t $a Psalm 100 $c g $b German $c s $b chorus (SATB) $c s $b horn $c s $b organ Ex. 6. Work: Ascension Day songs for tenor and bass with 2 flutes and continuo Ascension Day music Sacred duets with instrumental ensemble Faceted Forms/genre: songs (sacred) Approach: Events: Ascension Day tenor flute Number: 2 harpsichord violoncello $c f $a songs (sacred) $c e $b Ascension Day $c s $b tenor $c s $b flute $n 2 $c s $b harpsichord $c s $b violoncello

Ex. 7. Work: Islamic chants in Kurdish for Ramadan Season for 3 tenors Chants (Islamic) Ramadan Choruses, Sacred (Men's voices, 3 parts), Unaccompanied Faceted Forms/genre: chants Approach: Geo-cultural: Islamic Kurdish Events: Ramadan tenor Number: 3 $c f $a chants $c g $b Islam $c g $b Kurdish $c e $b Ramadan $c s $b tenor $n 3 7. What questions still need to be answered about the possibilities and ramifications of implementing the Music Thesaurus? There are many unresolved issues at this early stage in the development of the Music Thesaurus. All music librarians are encouraged to bring their questions, general and specific, to the open meeting of our working group on Friday, March 4, at 2:30. At this meeting, we will discuss those questions, and other issues raised in this paper. Harriette Hemmasi will be joining us for this discussion. It is critical at this step that the music library community participates in a discussion about faceted access and the thesaurus, if we want to achieve significant improvement in subject access to music in the near future. Please contact any member of the working group with your comments and questions: Amanda Maple, Chair St. Olaf College internet: maple@stolaf.edu office: 507-646-3362 Kathy Glennan University of Southern California internet: glennan@calvin.usc.edu office: 213-740-4021 David Thomas University of Pittsburgh internet: dhthomas@icarus.lis.pitt.edu office: 412-624-9475 Stephen Yusko Library of Congress internet: yusko@mail.loc.gov office: 202-707-5182 WORKS CONSULTED

American National Standards Institute. Guidelines for Thesaurus Structure, Construction and Use. ANSI-Z39.19-1980. Quoted by MLA Music Thesaurus Project Working Group,. p. 720. Barnett, Patricia J. and Toni Petersen. "Extending MARC to Accommodate Faceted Thesauri: The AAT Model." In Beyond the Book: Extending MARC for Subject Access. Ed. by Toni Petersen & Pat Molholt. Boston, Massachusetts: G.K. Hall, 1990, pp. 7-23. British Standards Institution. Guidelines for the Establishment and Development of Monolingual Thesauri. BS 5723: 1979. (now superseded by BS 5723: 1987) Quoted by Dykstra, pp. 44-45. Dooley, Jackie M. and Helena Zinkham. "The Object as "Subject": Providing Access to Genres, Forms of Material, and Physical Characteristics." In Beyond the Book: Extending MARC for Subject Access. Ed. by Toni Petersen & Pat Molholt. Boston, Massachusetts: G.K. Hall, 1990, pp. 43-80. Dykstra, Mary. "LC Subject Headings Disguised as a Thesaurus." Library Journal 113/4 (1 March 1988): 42-46. Foskett, A.C. The Subject Approach to Information. 4th ed. Hamden, Connecticut: Linnet Books, 1982. Freeman, Kevin A. "Faceted Access to Music: Proposals on 04X and 658 MARC Fields." Paper presented at the 1993 Annual Meeting of the Music Library Association. Hemmasi, Harriette. "From Deconstruction to Reconstruction: The Case of LC Music Headings in the Music Thesaurus." Summary of her presentation to the Library of Congress Cataloging Forum June 23, 1993, distributed over MLA-L July 7, 1993, by Kenneth Valdes of the Library of Congress Music Subject Group. International Organization for Standardization. Documentation: Guidelines for the Establishment and Development of Monolingual Thesauri. ISO 2788, 1974. (now superseded by the 1986 edition) Quoted by Dykstra, pp. 44-45. Larson, Ray R. "Between Scylla and Charybdis: Subject Searching in the Online Catalog." Advances in Librarianship 15 (1991): 175-236. Library of Congress Music Subject Group. "Improving Subject Access for Music Materials: A Proposal." April 28, 1993. Markey, Karen and Diane Vizine-Goetz, "Characteristics of Subject Authority Records in the Machine-Readable Library of Congress Subject Headings." Dublin, Ohio: OCLC Online Computer Library Center, Office of Research, 1 August 1988. Quoted by Whitehead (1990), pp. 81-82. McBride, Jerry. "Discussion Paper: Changes to 04X Fields in MARC Records for Bibliographic Description." Distributed at the 1992 Annual Meeting of the Music Library Association. MLA Music Thesaurus Project Working Group. "Improving Access to Music: A Report of the MLA Music Thesaurus Project Working Group." Notes 45 (June 1989): 714-721. Petersen, Toni. "The AAT: A Model for the Restructuring of LCSH." The Journal of Academic Librarianship 9 (Sep. 1983): 207-210. Petersen, Toni. "Developing a New Thesaurus for Art and Architecture." Library Trends 38 (Spring 1990): 644-658. Smiraglia, Richard P. Music Cataloging: The Bibliographic Control of Printed and Recorded Music in Libraries. Englewood, Colorado: Libraries Unlimited, 1989.

Taylor, Arlene G. Introduction to Cataloging and Classification. 8th ed. Englewood, Colorado: Libraries Unlimited, 1992. Whitehead, Cathleen K. "The Art & Architecture Thesaurus as an Alternative to LCSH." In Cataloging Heresy: Challenging the Standard Bibliographic Product. Ed. by Bella Hass Weinberg. Medford, NJ: Learned Information, Inc., 1992, pp. 59-74. Whitehead, Cathleen. "Mapping LCSH into Thesauri: The AAT Model." In Beyond the Book: Extending MARC for Subject Access. Ed. by Toni Petersen & Pat Molholt. Boston, Massachusetts: G.K. Hall, 1990, pp. 81-96. Young, Brad. "Recent Trends in Access to Music Materials." In Beyond the Book: Extending MARC for Subject Access. Ed. by Toni Petersen & Pat Molholt. Boston, Massachusetts: G.K. Hall, 1990, pp. 171-186. Last updated February 4, 2000