Addressing Ethnic/World Music in the Variations2 System




Variations2 was based upon concepts and search requirements associated with Western Art music. This emphasis was selected because it reflected the majority of holdings in the Indiana University William and Gayle Cook Music Library.


Presently, Variations2 catalogers enter information for four types of records: Contributor, Work, Container, and Instantiation. Due to search needs associated with Western Art music, Variations2 tends to center around the concept of the work: the abstract idea of a composition as first realized by its creator(s). For additional information on Variations2 record types and relationships, please see the Data Model, Data Dictionary, Jenn Riley’s Variations2 presentation slides, and other documents accessible through the Variations2 Metadata research page. Screenshots of the cataloging interface are also available online.


Ethnic music requires information and search options somewhat different from that for Western Art music and the present version of Variations2 does not adequately allow for the cataloging and searching of such music. To identify issues that should be addressed to make Variations2 more accommodating to ethnic music, an attempt was made to catalog several different ethnic recordings within the current system. The remainder of this document identifies some of the problems and issues that came up during this work.


Selected Records


Eight commercial recordings and one field recording were cataloged. The commercial recordings ranged from releases highlighting specific performers to compilations of various field recording excerpts. Both the bibliographic record and physical items were consulted when cataloging the commercial recordings. The one field recording was a relatively small collection housed in the Archive of Traditional music and cataloging was therefore based solely upon the bibliographic record. Selected recordings contained ethnic/world music from the following geographical regions: Afghanistan, Africa, Australia, Bulgaria, Cape Breton, East Asia, India, Indonesia, and Norway. Following is some basic bibliographical information on each record:


  1. [Afghanistan, Baluchis, 1976], 1 sound tape reel (field recording housed at the ATM – call # ATL 18343 & 77-107-F ATL 18343)
  2. Blueprint / Natalie Macmaster, 1 CD (Rounder:11661-7056-2) (music library – call # frontlog 5640230)
  3. De l'aube à la nuit, 1 CD (Auvidis Ethnic:B 6748) (music library – call # frontlog AJF1565)
  4. Gamelan of Surakarta, 1 CD (JVC:VICG-5263 & VICG-5263-2) (music library – call # CD ZL1825 .J4-4)
  5. Instruments de musique du monde. 1 CD (Chant du monde:LDX 274 675 & CM 251) (music library – call # CD ZL1627.11)
  6. Le mystère des voix bulgares. Vol. one, 1 CD (Elektra/Nonesuch:9 79165-2) (music library – call # frontlog 6109456)
  7. Sjugurd og trollbrura, 1 CD (Finnskogen Kulturverksted:FiKCD 1958) (music library – call # frontlog 5265995)
  8. The Garland encyclopedia of world music. volume 1, Africa, 1 CD (Garland) (music library – call # frontlog AJM3745)
  9. The rough guide to Australian Aboriginal music, 1 CD (World Music Network:RGNET 1026 CD) (music library – call # frontlog 4790091.1)





A number of issues and problems became apparent once an attempt was made to catalog ethnic music within Variations2. These issues fall under the following headings discussed below: “formats”, contributors, works, subject headings, instrumentation, form/genre/style, language, geographical location, function/context, themes/subjects/topics, and provenance. In order to better accommodate ethnic music, most suggested improvements fall into the following broad areas: the creation of new field types, the development of search mechanisms for existing and new fields, and the reevaluation/restructuring/development of controlled vocabulary lists and method of input.




Variations2 users are able to restrict searches by the format of the original object. However, ethnic recordings and related materials may be presented and organized in “formats” not associated with Western Art music, such as field recordings and field notes. New terms may need to be identified and added to the Format field controlled vocabulary list. In addition, an examination will have to be made of these additional “formats” to determine the best manner in which each may be presented within the system.




Contributor names are searchable in Variations2 according to assigned roles. In the search interface, these roles are grouped into the following three broad categories: Creator/Composer, Performer/Conductor, and Other Contributor.


New ethnic music-related roles will need to be identified and added to the existing controlled vocabulary list for contributor roles. Entries on the current list were derived from MARC relator terms, but do not include all terms from the MARC list. The small sample upon which this document is based demonstrated that the following roles were not represented: ethnographer/recorder, researcher, collector, and donor.


Unlike most Western Art recordings and scores, the amount of information provided for ethnic music contributors may vary widely. Whereas full names may be listed, often first names, descriptive identifications (i.e. school children, priests, village members), and ethnic groups/tribes/castes are presented instead. A decision must be made on whether contributor records should be made for all such entries, or if there is another manner in which this information may be handled for ethnic recordings. In addition, “traditional” is often provided in place of a creator name and a decision must be made as to whether this should be dealt with in some manner.


Finding authority records for ethnic music containers brought to light some patterns. Records existed for almost every identified ethnographer. Records also existed for a few performers who had released solo or small ensemble recordings. In general, however, records did not exist for most performers and identified composers (this was the case for a commercial Cape Breton fiddling recording, as well as an ethnographic compilation). Regarding Latin American and Hispanic names, it was nearly impossible to determine whether a name associated with the recording being cataloged was represented in the authority file, particularly when several individuals with similar roles appeared to have the same first and last names.




The concept of a “work” in ethnic music is problematic. Some recordings, such as a compilation of Cape Breton fiddle tunes, may provide clear titles for each tune. However, in this tradition, these tunes may be organized in sets which may also be assigned a title. The question becomes whether a work record should be created for the set, each individual tune, or for all entities and the relationships indicated in some manner. In order to best represent and search such music, the manner in which Variations2 presently deals with work relationships must be reevaluated and changed.


In some instances, no clear title is presented. Rather, the “title” might be consist of a combination of instrument, ethnic group, function, context, etc. terms (i.e. initiation song, excerpt from a specific event in a ceremony). Should work records be created for such entries or may this information be entered and searched through some other manner? For field collections, titles might be supplied by catalogers. If so, how should such entries be formatted?


There is also the question of how to deal with collection level versus single level works. One could decide that a field recording should be represented as a collection level work record and therefore each “track” would be represented by a lower node within the work structure. If this was done in the current system, it is impossible to directly relate subject headings, form/genre/style terms, contributors, date/place of recording, and other information to each track. This issue provides another reason for reassessing Variations2 work records and relationships.


Subject Headings


Assigning subject headings proved to be somewhat difficult at times. Bibliographic records for compilation recordings may include more general subject headings, but in some instances more specific headings might be more appropriate for individual tracks. Another problem dealt with the fact that while the cataloger was well-versed in subject headings for Western Art music, such was not the case for ethnic music. Initially the cataloger searched IU music library holdings for viable subject headings, then switched to searching holdings for the Archive of Traditional Music, which proved a more helpful source. However, the ATM assigns a number of “local subject headings,” the exact meanings of which were not always clear. To make the most of entries in this field, some discussion must be had regarding different practices and types of subject headings for ethnic music.


For the nine preliminary records, “world music” was entered as a subject heading in order to retrieve all ethnic recordings cataloged in the Variations2 database.




Currently, instrument terms are selected from a controlled vocabulary list derived from Library of Congress Subject Headings. The Variations2 system does not accept terms not included in this list. The list does not allow for direct additions by catalogers and presently, content in this field is not searchable by end-users.


Most of the instruments encountered while cataloging the nine ethnic recordings were not on the existing Variations2 instrumentation list. The closest match to a given term (i.e. “string—ethnic”) may not be adequate when searching for ethnic music. Either many terms must be identified and added to the existing list and hierarchy, or a means must be developed by which catalogers may add new terms to the list and hierarchy as needed. Another option is to create a separate list for ethnic instrument terms. However, present day Western Art music composers may incorporate such instruments in their compositions, so all terms must be accessible regardless of musical tradition.


In some recordings and tracks, instrumentation may not be indicated. For such cases, a decision must be made as to whether catalogers should leave the instrumentation field empty, guess, or include a broad instrumentation term.




Form, genre, and style are indicated in Variations2 through a controlled vocabulary list derived from Library of Congress Subject Headings and style terms found on Catalogers may not make direct additions to the list, but any term added to the field will be accepted, whether it is on the list or not. Contents in this field are presently not searchable and are considered optional.


Again, most of the terms encountered while cataloging the nine recordings were not on the Variations2 list. It became obvious that terms from certain regions, such as Afghanistan and Australia, are not well represented in LCSH. Whereas catalogers may enter new terms in the form/genre/style field, a mechanism should be developed through which terms may be added to the existing list and hierarchy, in part to deal with alternate spellings.


Form, genre, and style terms ranged from precise terms to descriptive statements incorporating ethnic group, region, and techniques/styles of playing. In one instance, members of one ethnic group were identified as performing in the style of another ethnic group. Some discussion must go into deciding exactly what type of information is acceptable here, particularly if certain elements (ethnic group, region) are also indicated elsewhere in the system.


Liner notes often define unfamiliar form, genre, and style terms. It would be helpful if such information could be included either within the system or in accompanying cataloging documentation.




Language issues relate to two types of information: language used in a performance and language in accompanying materials (program, liner, or field notes).


In regard to languages associated with a performance, catalogers again select languages from a controlled vocabulary list. Only terms found on the list may be selected and catalogers may not directly add terms to the list. Presently, language terms are not searchable.


In regard to the nine ethnic music recordings, most of the languages were not on the existing Variations2 language list. Therefore, either more work must be done to identify a broader spectrum of languages or a means must be developed by which catalogers may directly add new terms to the list.


In some instances, language of performance was not indicated. A determination must be made as to whether it is best to leave the field blank in such situations or whether a general region-based term should be indicated (e.g. “African”).


Accompanying materials may be in a language completely foreign to the cataloger and anyone in his or her vicinity. In such cases, it is extremely difficult to determine performer names, form/genre/style terms, what various dates represent, etc. Whereas not much can be done about this if no one knowledgeable about a particular language is accessible, it should be noted that this is more of a problem for ethnic music than it is for Western Art music recordings.


Geographical Location


Currently, four fields exist in which geographic location-related information may be entered: Place of Origin in contributor records, Place of Composition in work records, Place of Publication in container records, and Place of Performance in instantiations. Whereas Place of Publication entries are imported from MARC bibliographic records, all other location-related entries are directly entered by catalogers and are considered optional.  No controlled vocabulary list exists and catalogers transcribe information as found on the item or in the consulted resource. Information entered in these fields will display in record information screens but is not searchable.


On ethnic recordings, geographical location may be expressed in several ways. Sometimes specific countries, cities, or towns are identified. In other instances, a more descriptive identification of settlement, region, or something along the lines of “within the border lands [of several countries]” is provided. Geographical location may also be associated with ethnic or cultural groups in statements such as “lands of the Shona people.” These varied manners of identifying location, and perhaps others not encountered in this small sample, should be considered when determining the best way to deal with geographical locations associated with ethnic recordings.


Additional questions on this topic include where this information should be recorded (instantiation, work, or some yet to be determined record type/field), and whether a controlled vocabulary or some other specified source should serve as the basis for entries.


Function or Context


Music presented on ethnic recordings may have been performed for a specific function or within a particular context. Presently, the only fields in which such information might be indicated are in the Work Title, Form/Genre/Style, and Notes fields.




Tracks or groups of tracks may be associated with certain topics, subjects or themes. For instance, several recordings grouped tracks according to associations with deities, holidays, and instrument classification. This information is rarely built into the work title or found in the notes field. No uniform manner currently exists in which such information may be traced or searched.


Provenance Statements


The nine recordings cataloged identified three different types of provenance statements. The first type of statement identifies the library collection with which a given recording is associated. In the IU music library, such information is traced in the container record Notes field. The second type of statement identifies that the recording or track is “from the collection of” or was “collected by” a specific individual. For example, compilation recordings including selections from ethnographic collections at various institutions might identify the name of each collection. Presently, such information could be added to the instantiation record Notes field. The last type of provenance statement consists of a story detailing the manner in which a recording was obtained. In one instance, liner notes described how a track was from a recording that had circulated and been copied over and over again for many years. The particular version from which the track on the recording was derived was purchased from an identified individual in 1941. Again, such information would currently be entered in the instantiation Notes field.


Contents in the various Notes fields display in record information windows, alongside any other information entered as a note. Under current rules, the inclusion of such information is considered optional.



Time-Consuming Elements


The most time-intensive aspects of cataloging ethnic music in Variations2 were (in no particular order):


  1. Assigning subject headings
  2. Dealing with recordings presented in a completely unfamiliar language
  3. Dealing with contributors when (1) it was unclear whether an authority record matched and (2) the contributor happened to be a generic term such as “priest” or “people from the village”
  4. Assigning Form/Genre/Style terms and deciding whether to include ethnic/cultural group names in this field
  5. Instrumentation




Caitlin Hunter