A Revival of the Music Conspectus A Multi-Dimensional Assessment for the Score Collection

Abstract

With an innovative use of the music conspectus, the Hong Kong Baptist University library conducted a score collection assessment to identify not only the strengths and weaknesses of the collection, but also problems with the choice of score publishers and formats in the acquisitions process. Because of its flexible application, this modified music conspectus can be easily adopted by libraries of all sizes and libraries that use any classification system. This article provides a detailed description of the techniques used, and highlights the findings and benefits received, as well as actions done following the project.

The Hong Kong Baptist University (HKBU) is a medium-size public-funded tertiary institution with a full-time student enrolment of around 8,000 and is one of the three universities in Hong Kong that offers music programs. Serving a small student body of around 230 music students of undergraduate and graduate levels, the HKBU library's music collection contained over 15,000 volumes of scores as of June 2007 in addition to books and audiovisual materials. In order to understand the current situation of the score collection in Western art music published in Western languages, an assessment was conducted between summer 2007 and spring 2008. With an innovative and modified use of the music conspectus initially developed by the Research Libraries Group (RLG), the library was able to identify not only the strengths and weaknesses of the collection, but more importantly, problems with the choice of score publishers and score formats in the selection and acquisitions process. Because of its flexible application, this modified music conspectus can be easily adopted by libraries of all sizes and libraries that use any classification system. This article provides a detailed description of the preparation, techniques used, and findings of the assessment, and highlights the benefits received and actions done following the project.

Collection Background

The score collection of Western art music in the HKBU library comprises scores in all formats such as full scores, miniature scores, piano reduction scores, solo instrumental parts, etc. With a short collection history of about fifty years, selecting scores is primarily the responsibility of the music faculty who make decisions on what the library should acquire based on the faculty's and students' teaching, research, study, and performance needs. Faculty members are regularly sent "yellow slips" or approval plan notification slips, publishers' catalogs, and new title announcements, and then forward their requests to the library for orders to be placed. Consequently, the content of the score collection reflects, to a large extent, the faculty's interests or what was presented to them in publishers' catalogs. As Elizabeth Henry, Rachel Longstaff, and Doris Van Kampen observed, the music areas in which faculty members are more vocal tend to be better represented in the collection.¹ Also, there has been little input from the library, and there is no effective approval plan to complement faculty's selection. Hence, the selection process lacks a systematic approach to developing the score collection as a whole, and is therefore susceptible to holes and gaps in many areas.

Literature Review

Originally developed by the Research Libraries Group (RLG) in the late 1970s, the conspectus was a tool that gave an overview and a comparison of the existing collections showing where the strengths were and recording future collecting intensity among the RLG conspectus participating member institutions.² As Ferguson, Grant and Rutstein have explained, the goal was to "improve the stewardship of funds through better communication among those building collections to acquire, make accessible, and preserve the world's scholarly production for the national community."³ By making the collecting activities a coordinated plan, unnecessary duplication of research materials could be avoided such that a larger scope of library materials could be made available to users through the interlibrary loan system.⁴ The use of the conspectus was soon adopted by the Association of Research Libraries (ARL) in 1983 for its North America Collections Inventory Project (NCIP), and later by other regional consortia such as the Library and Information Resources for the Northwest (LIRN), and the New York Metropolitan Reference and Research Library Agency (METRO).⁵ The Music Program Committee of the RLG also began to create a music conspectus in the early 1980s, and in 1986, the Music Library Association (MLA) proposed to use the music conspectus to gather information from libraries of all sizes and types to form the National Music Collection database.⁶

Much literature has been published on the topic of the conspectus methodology. With regard to the music conspectus in particular, Jane Gottlieb compiled a book titled Collection Assessment in Music Libraries which included papers originally presented at the 1991 MLA annual meeting.⁷ In the book, Elizabeth Davis provided guidelines on evaluating the collection using the music conspectus in the METRO project,⁸ and Peggy Daub supplied a very detailed paper about its application, brief results of and benefits received by various institutions.⁹ In Daub's survey, most of the music librarians who had used the music conspectus agreed that its values had accurately represented their collections. Some indicated that through its use they were able to identify weak areas in their collections, and that it helped them write stronger collection development policies. Others opined that knowing the conspectus values of other peer institutions aided them in making justifications for increased funding.¹⁰

Nonetheless, music librarians also criticized the challenges in using this assessment tool. In the same survey, Daub revealed that music librarians found the Library of Congress (LC)-based subject lines did "not represent useful categories that would be used in collection evaluation and development," but were only quantitative shelflist measurements that were to give a quick overview of the music collection.¹¹ This argument was also echoed by librarians using the conspectus in non-music fields. For example, Richard J. Wood stated that the LC classification numbers on the conspectus worksheets failed to embody the total collection,¹² while Larry R. Oberg pointed out that the gaps between the LC-based conspectus lines was one of its problems.¹³ In the survey conducted by Mary H. Munroe and Jennie E. Ver Steeg, respondents complained about how imprecise any classification scheme was in their conspectus studies.¹⁴

This deficiency of the conspectus was in fact even more prominent in the field of music where publications are quite distinctive compared to materials in other disciplines. As Kent Underwood stated, "real differences in content do tend to accompany differences in format … [and] the different formats are created and collected for different purposes."¹⁵ Yet, little attempt has been made to examine this aspect of the music conspectus. Similarly, Lenore Coral argued that the LC classification scheme in the music conspectus did not provide the kind of detail that would describe actual music collecting activities, nor clarify which composers' works, which editions, which genres, or which periods or geographical areas are collected.¹⁶ As such, the music conspectus has indeed left many important areas untouched.

Though the use of the conspectus was quite popular in the 1980s and '90s, many librarians found using it laborious and time-consuming. Thus, variations in conspectus methods were used. In 1995, Howard D. White created the "Brief Test" which was based on the idea of the conspectus, but with a goal to simplify the entire process by assessing only as few as forty titles selected by subject experts. These forty titles were grouped into four different conspectus levels (with ten titles for each level) from Level 1 to Level 4 (with Level 0 "Out of Scope" and Level 5 "Comprehensive," excluded), based on the ranking of the holdings counts retrieved from the Online Computer Library Center (OCLC). The library collection was then checked against this final conspectus value-ranked title list, and the library could claim the highest conspectus level in which at least fifty percent of the titles were owned.¹⁷

Later, this "Brief Test" evolved into the "Coverage Power Test" which was designed "to test the entire collection of each library against the entire literature," and was aimed to rectify some issues, for example, the possible inconsistency problem¹⁸ and the sensitivity of results¹⁹ due to the small number of sample titles chosen for each conspectus level. Instead of having a subject expert prepare the forty-item list, a list of titles in the "entire literature" of a specific subject was retrieved from OCLC based on a certain call number range. This list would be ranked from high to low according to the holdings counts. Similarly, the same process would be done for a list of titles in the same call number range for the library collection being assessed. Comparisons would then be made between the holdings counts of the "entire literature" and those of the library collection assessed, and the conspectus values would be assigned to the library collection based on the percentage coverage of the entire literature in the subject.²⁰

In recent years, OCLC has also offered a service called WorldCat Automated Analysis (WCA), which allows libraries to analyze their collections according to size, coverage, publication date, language, format, and audience based on the data found in WorldCat. It also facilitates peer comparison with two to five libraries and checks for collection overlap and uniqueness. Because the whole WCA process is automated, librarians with little knowledge in the subject area studied can still easily carry out the assessment, and worries about biases in compiling the core list such as what happened with the conspectus or Brief Test method were now eliminated.²¹

However, when looking back at the music field, not much development or application of new assessment tools has been found in recently-published music literature, with the exception of projects using circulation statistics, interlibrary loan statistics, or preservation condition of scores. Although there were heated discussions on the Brief Tests and the WCA, Jennifer Benedetto Beals commented that these approaches might be more suitable to monographs than serial or multimedia materials,²² and that they are rather impossible to use in a music score collection. Consequently, since both the Brief Tests and the WCA rely on the accurate reporting of data and holdings in OCLC, libraries who do not have a consistent practice in doing so will find the two methods not feasible. In addition, especially for music scores, separate cataloging records have been created for bibliographically-similar editions and each has its own OCLC accession number. Since the WCA performs its analysis by matching accession numbers only, it is prone to produce doubtful results when reporting collection uniqueness and overlap.²³ Further, according to Darby Orcutt and Tracy Powell, a lower institutional reporting rate to OCLC was found for videos and other non-book format, thus making the results of the Brief Tests and the WCA unreliable.²⁴ Although the WCA does provide such details as the age and language of the collection, this information does not seem to be of much use as the publication year of scores is often ignored by music users, and the fact that scores are usually cataloged as items with "no linguistic content" has also made such language examination meaningless. Music users, on the other hand, are generally more concerned about the edition or the publisher of the score.

Therefore, to overcome the shortcomings of the music conspectus and other assessment methods, the HKBU library created a tool that would allow these methods to be used in a more comprehensive way when assessing its score collection. By modifying the RLG Music Conspectus and dividing the collection hierarchically, it has facilitated a less complicated application of the music conspectus for internal assessment purposes. Furthermore, a positive result of this new multidimensional approach to music assessment will be to make librarians rethink the benefits of the conspectus, and what it can do that other methods cannot.

Assessment Preparation

Before the project began, many decisions were made with respect to the scope and the methodology.

Defining the scope

Because of the small number of students studying Chinese music in HKBU, the scope of this assessment covered music scores published in Western languages only. Thus, scores published in Chinese or other Asian languages were excluded. In addition, only those music categories that the HKBU Music Department needed for its curriculum and research were considered. For example, wind band music for which the department does not have an individual course was left out of the project. Complete editions, sets containing a comprehensive collection of works by a specific composer, were also treated separately using a simple benchmarking exercise to compare holdings against other local academic music libraries. As such, they have been excluded from this study.

Defining the purpose and choosing the appropriate assessment method

Due to the small size and numerous gaps in the collection, it was deemed not worthwhile and too costly to use the automated evaluation analysis services (WCA). As a result, a conspectus project was considered. However, unlike earlier projects that aimed to obtain an overview of a national collection or to compare holdings among a group of libraries, the current assessment study focused on comparing the library's music score holdings against a core list so that the results could serve as internal guidelines for future collection development.

Adopting and Modifying the Music Conspectus

Although many libraries have used the music conspectus successfully, it was quite difficult for the HKBU library to carry out such a task. In the original music conspectus, the M schedule of the LC classification for scores was divided into over fifty conspectus lines according to subject for the purpose of comparison and analysis between each music group (see table 1).²⁵ Though this LC-based conspectus was theoretically usable in libraries that used other classification systems, the employment of the music conspectus by the HKBU library, which uses the Dewey Decimal Classification (DDC), was not easy. Because of the major revamp of the music section of the DDC in past years, most of the older scores were not retrospectively reclassified to mirror these changes. So, scores of the same music genre might be classed in different places, making it rather impossible to do the assessment by strictly following the classification numbers in the conspectus lines.

Also, with the small size of the collection, the meticulous division of the classification schedule in the music conspectus was considered to be too complex. Hence, all these lowered its usability.

Nonetheless, the concept of the music conspectus was adopted for the project. Rather than splitting up the classification schedule into numerous segments as the original conspectus did, a few broad music categories were identified based on music types, namely "Orchestral," "Concerto," "Chamber Music," "Instrumental," "Voice/Choral," "Opera/Musical," and "Anthology." Each music category was then subdivided by the music genre (see table 2). For instance, the "Orchestral" category was further broken down into "Symphonies," "Overtures, Suites, Tone Poems, etc.," "String & Chamber Orchestra," and "Ballet." Then, for some genres that were especially important to the HKBU music users, in order to allow for a more refined analysis, these were further split into smaller subjects according to their instrumentation or ensemble type (see table 3). By using this strategy, the application of the conspectus was not bound by the classification system or the call number attached to the score, but was based on the genre of the music itself. Therefore, not only could this modified music conspectus be used in non-LC libraries, it could also solve the problems caused by the inconsistent use of classification numbers as a result of the redesign of a classification schedule.

Once the framework of the modified music conspectus was completed, the Western score collection was checked against a core list. Two numerical values on a scale of 0 to 5, with 0 being "Out of Scope" and 5 being "Comprehensive Level," were then assigned to each music genre assessed. The first value was the Existing Collection Strength (ECS) which described the collection level of a particular portion of the collection at the time of assessment, and the second value was the Desired Collecting Intensity (DCI) which indicated the desirable level which the collection should ultimately achieve to adequately support users' needs. While the scores of the ECS were assigned by the music liaison librarian, the DCI scores were provided by the conductor of the university orchestra, who oversaw the performance activities in the music department. By involving a faculty member, this enabled the library to gather a more objective opinion about how the collection should develop using an expert who works with music students and professors on a daily basis, and best knew their musical needs.

Table 1.

Excerpt of the Conspectus lines of the original Music Conspectus (Source: Jane Gottlieb, ed., Collection Assessment In Music Libraries [Canton, MA: Music Library Association, 1994], 82)

Compiling the Core Title Lists and Checking Holdings

Similar to other assessment projects, a core title list was compiled based on standard bibliographies such as A Basic Music Library: Essential Scores and Sound Recordings (BML) published by the American Library Association in 1997,²⁶ and other sources, including audition lists of major music schools or professional orchestras, repertoire requirements of important international music competitions, and curriculum and course syllabi. The music faculty was also consulted and a list of the major works of thirty-eight contemporary composers was added to complement the core list in order to ensure an adequate coverage of contemporary and twentieth-century music in the assessment. The Western score holdings were then checked against this core list by the music liaison librarian or a part-time student worker studying in the music department.

Table 2.

Worksheet for HKBU's modified Music Conspectus

Table 3.

Further subdivisions of the genre "Chamber Music"

More than a Conspectus Exercise: Publisher and Format Evaluation

While many conspectus studies primarily or solely involved a yes-or-no title check against a core list or the holdings of other institutions, the HKBU library further employed a multidimensional technique to identify not only what the library owned (or the number of titles), but also to see if score publishers and formats (whether full scores, miniature scores, piano reduction scores, etc.), available for use were sufficiently fulfilling users' needs.

In the music industry, a work in the public domain, such as a Mozart piano sonata, can be published by many companies. While some offer "urtext" editions or include critical commentary in performance scores, some provide reprints of other editions, or add numerous editorial notes or interpretation marks in the music. Though there is no hard-drawn line of good and bad, musicians generally have a preference for different editions or publishers of certain composers/types of works. Thus, having the right edition by more highly regarded publishers for users is an important matter in good music-collection management.

Apart from the quality of publishers, it was also of interest to look at the availability of score formats in the library. Music publications are different from other library materials in many ways, and music scores may come in many versions, with each serving a different purpose.

Therefore, putting into consideration all of the above, an additional step was taken to record the name of the publisher for each score assessed, and the score format found for each title. Such careful scrutiny allowed the library to know whether scores produced by the "preferred publishers" were purchased. Through this extra effort, the library was able to obtain a distribution of all score formats acquired for each type of music.

Analysis Results

With only one music liaison librarian working on this project while engaging in other duties, such as cataloging and library instruction, and with one part-time student helper working in the summer, this project took about nine months to complete. After checking holdings against the core list and examining the publishers and formats of each score in the library, many valuable findings resulted.

Strengths and weaknesses

Like other conspectus studies, the strengths and weaknesses of the collection were identified. It was evident that the strongest parts of the HKBU library's Western score collection were in the orchestral and opera/musical areas, and the weakest part was in the chamber music section. The breadth and depth of other parts of the collection differed widely. There was broad coverage of orchestral works, but in contrast, there appeared to be an imbalance in the collection of solo works for different instruments (e.g., more core titles available for piano, and fewer for percussion or brass).

Variety of score publishers

With a multi-angle approach, the analysis results uncovered issues relating to the choice of music score publishers. Recording the name of the publisher of each title assessed resulted in knowing that a significant portion of the scores held were published by "less preferred" publishers, even when better alternatives were available. For example, the library owned two sets of scores and parts to Franz Schubert's Piano Trio No. 1. While the highly preferred editions for music users would be the urtext published by Bärenreiter or Henle, neither of these were acquired. On the other hand, reprint editions with substantial amounts of interpretative markings, were purchased instead. Though it would not be possible to know the history or cause to such acquisition decisions, this demonstrated a need for better quality control and clearer guidelines in the selection process.

Suitability of score formats acquired

When studying score formats, hidden phenomena which were unknown in the past were revealed. For chamber music works, it was found that oftentimes only scores were available without their corresponding performance parts. Over seventy percent of the chamber music items were full, study, or miniature scores, and merely thirty percent were performance parts. With this knowledge, it was a good indicator that the library should begin buying the missing instrumental parts which are crucial to chamber music study. Moreover, there seemed to be a pattern for buying miniature scores rather than full scores, as demonstrated by the fact that sixty-two percent of the orchestral works were in miniature score format, and only thirty-one percent were in full score format. Again, the reasons behind these acquisition decisions remain a mystery. Nonetheless, this has raised the question of whether this was a result of the faculty's selection bias or real users' needs. Other important findings in the score assessment included the obscure presence of a few number of score and parts for large orchestral works, and the absence of the corresponding full or study scores to concertos, for which piano reduction scores and solo parts sets were bought. By looking at these, it became apparent that revised collection development guidelines are needed so that the appropriate or preferred formats of scores will be acquired for certain types of music. The acquisitions and collection scope may also need to be redefined. For example, some formats such as the score and parts sets for large orchestral works which often contain over sixty instrumental parts should perhaps be housed in a separate performance library where direct supervision and proper management of the parts could be done by orchestra staff.

Discussion and Recommendations

After the assessment, the library began to acquire additional knowledge on what the Western score collection contained from different perspectives. By knowing the strengths and weaknesses of the collection, the library is now able to accommodate changing needs more quickly. The effort in evaluating the score publishers and formats also proved to be worthwhile without requiring much extra time, for the final all-around picture produced was instrumental in detecting flaws during the selection and acquisitions process. As Marcia Pankake, Karin Wittenborg, and Eric Carpenter noted, librarians needed to know the cause of weak selection practices and to act upon them.²⁷ Therefore, drawing on these findings, areas for improvements were identified and two sets of follow-up actions were performed as a result.

Externally, a score enhancement project was initiated immediately after the assessment, and ten weak areas were selected for prioritized collection development with approval and financial contribution from the library and the music department. Informal discussions with music faculty and students were conducted to ascertain if music users have a preference for full, study, or miniature scores. This created a more casual channel for users to freely express their opinion and reasons for certain preferences. A formal music user survey was also conducted to gather statistical information about music users' library use behavior, their perceived importance of music materials, and their collection development preferences. The survey results were invaluable in helping to understand library use patterns of each music user group and their real music needs.²⁸ These additional activities allowed for the direct involvement of users in collection building and also facilitated the creation of a truly user-centered collection.

Follow-up actions were also done internally. A list of "preferred" music publishers for different types of music or composers was compiled by the music liaison librarian for the technical services staff to follow in case such order details were not provided by the faculty requester. A training workshop was also given to the staff to introduce them to the differences between and purposes of various score formats in order to understand the logic of why certain materials should be chosen over others. This way, staff would not blindly follow the guidelines provided, but would be able to make sensible judgments based on music users' needs. Furthermore, music orders submitted to the acquisitions section could now be looked over quickly by the music liaison librarian before sending out to vendors to ensure that the best possible or necessary score formats and music publishers had been chosen. A plan to fully update the collection development policy is also underway, aiming to provide clearer guidance on the consistent selection of appropriate materials that support the research, teaching, study, and performance needs of music users.

Conclusion

Music score publications are complex, and the existence of a diverse range of scores for the same musical work goes beyond mere reproduction. The variations in formats, publishers, or editions are of great concern to music users. Hence, the assessment of a score collection should not be just a title-checking procedure, but should employ a more qualitative approach that can actually guide collection development activities.

Tailor-made for music scores, this new modified music conspectus turned the collection inside-out, and revealed many selection and acquisitions loopholes that one would easily miss in daily work. Its separation from the classification schedule also enhanced its usability in non-LC settings, and its application can be straightforwardly extended to libraries that have not been able to keep up with the changes in classification. Since it is genre-based, libraries will have the flexibility of doing a simple broad assessment based on a few large music categories and genres that are particularly needed by users, or a comprehensive in-depth analysis by adding more refined music categories to the conspectus list or further subdividing each music genre into smaller subsets according to instrumentation or ensemble type. Consequently, conducting a conspectus project is no longer only for large universities or consortia, but can also be carried out by smaller libraries, where money and staff may be limited.

There are many ways to evaluate score collections, but this is the first attempt to incorporate a multidimensional concept in music collection assessment, and as such, there is more to be explored. Music users are very specific about what they need in regard to formats, editions, and quality, and the assessment tool for music scores should reflect this need.