College & Research Libraries News

An automated finding aid for special collections

By Lucy Shelton Caswell Curator Library for Communication and Graphic Arts Ohio State University

Karen A. Smith Special Collections Cataloger Library for Communication and Graphic Arts Ohio State University

and Hannah Thomas Manuscripts Cataloger Library for Communication and Graphic Arts Ohio State University

You may recognize this common reference desk scenario: The caller has an urgent request, “The song is from an old movie and I think the words are something like ‘You must remember this, a kiss is still a kiss…’ Do you have the sheet music?” An alternate scene: The researcher is preparing a book on the centennial of the Statue of Liberty. He asks to see all the editorial cartoons in the collection which depict this image. A third request: The professor stops by to say that her class needs to see several examples of historic photographs and she especially needs to have three or four cyanotypes on view for her students to see. Or: A scholar from England needs to know the names of Jerome Lawrence’s correspondents when the American Playwrights Theatre was established. How can these information needs be met?

Developing an automated finding aid to help libraries answer requests such as these is the challenge being undertaken at the Ohio State University Libraries (OSUL). While materials representative of other cultures and time periods are collected, four of the special collections and one of the departmental (branch) libraries at Ohio State have especially strong holdings of materials representative of several aspects of 19th and 20th- century American culture. They are located at different sites on the Columbus campus and each is charged with collecting in a separate area: American fiction, the theatre arts, music, and the graphic arts. All relate to the American experience, and the intellectual links between the collections are well- defined. In fact, one of the primary goals of the proposed database is to increase awareness of the inter-related nature of the special collections at the Ohio State University. The diverse physical locations of these special materials compound the traditional problems of organizing large collections of non-book materials to promote ready access to them for a broad range of scholarly purposes.

Work on creating an automated finding aid to solve these problems began in earnest early in 1987, and a structure for the proposed database has been developed. The automated finding aid project (known as SCDB—the Special Collections Database1) has the following goals:

1) Intellectual and physical control of the collections at the item level. What do we have and where is it?

2) Easy access to information about Ohio State’s special collections by students and scholars, both locally and, potentially, nationally and internationally. How can we better inform current and potential researchers about the resources available at Ohio State?

3) Information for future collection development and management. How can the special collections at Ohio State grow to complement and enhance currently held materials and to fill gaps in present holdings?

Several assumptions have been made about SCDB as it relates to OSUL’s online public access catalog and circulation system, the Library Control System (LCS), and OSUL contributions to the OCLC online union catalog. All books and serials in these collections are to be fully cataloged and, therefore, represented on LCS and OCLC. Collection level bibliographic records for discrete collections held by these special collections libraries (e.g., for the papers of Walt Kelly) will be created in the suitable MARC format and represented on LCS and OCLC. Where appropriate to meet local and/or national scholarly purposes, subcollection or item-level MARC records will also be created (e.g. ‚ for a daguerreotype self-portrait of Matthew Brady). All names used in SCDB will conform to conventions of the Anglo-American Cataloguing Rules, second edition, and will be verified against existing headings on LCS.

SCDB is intended to provide both item-level description of and varied means of access to materials through index fields in the database. For the purpose of SCDB, an item is one unit as defined by the person processing the collection and may vary from one motion picture lobby card to a folder of correspondence or a collection of drawings. SCDB is planned to improve upon the large number of typewritten finding aids (cards, notebooks, etc.) now available in these libraries and to provide a single information source for the content of these collections.

When discussing the design objectives of SCDB, ease of input and ease of retrieval of information were most important. Under the supervision of collection curators, advanced students or graduate associates will determine how to represent and input the data, so the number of fields had to be limited and format-specific data entry work forms developed. The current version of a scheme for general data description and analysis is shown in Figure 1 and an explanation of the field abbreviations shown in Figure 2.

A key feature of SCDB is the availability of non- traditional library access points, but it will not be necessary to describe each item using every field available on the data entry form. Many fields are to be repeatable, for example, the personal name/cre- ator field, PERN, may contain several names and lifedates, qualified by relator codes to indicate persons associated with the item (such as the actors, producer, director and lighting designer for a given play). The relator codes to be linked with PERN have been adapted from the RBMS list of relator terms (C&RL News, October 1987, pp. 553-57). Similarly, TTLE functions to allow the title of the parent work or the first line of the verse of a song to be related to the item.

Three analytical fields have been created. TOPC may be used to indicate what the item is about: the cartoon is about the National Recovery Act, inflation, or Roosevelt’s election to a fourth term. The LCSH field is provided for those times when one or more Library of Congress subject headings might be used. SHWN is to list what is depicted in graphic materials: Rudolph Valentino is on the poster; a snake is in the cartoon.

The designers of SCDB believe that access by type of material is also important, so the “7” fields were created. A second use for these fields is to describe components of an item (such as the case for a tintype). MATT will describe the general class or genre of an item (dance score, editorial cartoon, poster, etc.). MATD specifically describes the object at hand. (A comic strip might be in the form of an original drawing, proof, color guide proof, engraver’s proof, proof on newsprint or tearsheet. A poster might be a one-sheet, three-sheet, lobby card or window card.) MEDM is to describe the format or technical process used for the object’s creation. (The original cartoon mentioned above might be ink and crayon on coquille board.) VIEW will describe commonly used image categories (half-view portrait, landscape, etc.).

The library record section of the data entry form (fields 16-27) will be visible by password to staff only. The acquisition and disposition records are intended to give better custodial information about an item and to provide centralized online acquisition information (donor records).

There has been much discussion about the MARC mapability of SCDB and possible future methods of linking it to LCS. The data fields of SCDB are intended to follow consistent definitions and usage which would make mapping to MARC for descriptive purposes a viable option. Another possibility would be to use LCS as a gateway to provide more general access to the SCDB database.

Descriptive fields

RECN 1.00 System-supplied record number (unique ID)
PERN 2.00 Personal name/creator, date and relator code
TTLE 3.00 Title and relator code
PUBL 4.00 Publisher
PRIN 4.10 Printer
PLAC 5.00 City/state/country of publication/printing
PLPF 5.10 City/state/country of performance
SITE 5.20 Site of performance
DAT1 6.00 Date (year/month/day) of publication, creation, copyright, performance, etc., and relator code
LOCO 7.00 Location code
MATT 7.10 Type of material (general object class or genre)
MATD 7.20 Nature of materials (specific object descriptor)
MEDM 7.30 Medium (technical process and/or physical material)
VIEW 7.40 View category of visual images
MATN 7.50 Notes relating to physical features of material
EXTN 7.60 Extent (number of pages, pieces, etc.)
SIZE 7.70 Dimensions in centimeters (length x width x depth)
PHYC 7.80 Physical condition
LANG 8.00 Language code
LLCS 9.00 LCS code indicating which library has item
LCSH 10.00 Library of Congress Subject Headings
TOPC 11.00 General topic(s) or subject(s)
SHWN 12.00 Person(s), place(s), thing(s) illustrated
NOTE 13.00 Notes
RSTR 14.00 Restrictions
LINK 15.00 Linking record to larger collection to which item belongs
Library records (not available to public):
SACQ 16.00 Source of acquisition (name(s), address, phone)
ACDA 17.00 Accession date
MACQ 18.00 Method of acquisition: purchase, donation, gift, transfer, loan
PRPR 19.00 Purchase price
APVL 20.00 Appraisal value
ASVL 21.00 Assigned value
PRES 22.00 Preservation status indicator
ACTN 23.00 Action log (actions taken, dates, etc.)
DAFA 24.00 Date of future action(s)
RESR 25.00 Researcher(s) who have used material
RSCD 26.00 Record source code (person who created record) plus year, month, day

FIGURE 2. Field abbreviations.

It must be emphasized that all collections and significant subcollections will be represented on LCS and OCLC in cataloging records in the appropriate MARC format. The special collections automated finding aid is not intended to function as a library catalog, but more as an index. For example, the traditional American library cataloging concept of “main entry” will not be used. “Main entry” and “added entry” names will all be treated identically in the PERN field. (Names used as subjects will be put into separate fields.) The physical description fields will allow for expanded description of and access to both graphic items (such as an am- brotype) and support formats (such as a leather case). In this instance, both case and ambrotype would be described fully. Also, multiple sizes of posters for a film could be included in the same record, but posters of each size would be individually retrievable. Vocabulary control will be maintained in many of the physical description fields.

As part of the planning process, the committee attempted to learn of current projects which would be instructive as models. A literature search was done. Database documentation proved informative from two photography collections: the Center for Creative Photography, University of Arizona, and the Photography Collection, The Humanities Research Center, University of Texas at Austin. A review of museum-based data systems was also informative, and the Smithsonian Institution was contacted for information regarding its in-house system. Various visual resources librarians and other special collection librarians were consulted.

Discussion of the STAR system was initiated with the Getty Center for the History of Art and the Humanities which has used STAR for several years. In addition, an early prototype of the database was set up on Ingres at Ohio State.

Now that the plan for SCDB exists, the next step is to talk to vendors to determine whether or not any currently available software can implement it, before requests for funding can go forward. SCDB is still in the project definition stage and its final format is heavily dependent upon the software system chosen. Topping the list of products under consideration are BASIS, STAR, ARTIS and Ingres. We seek the reaction of our colleagues to the brief description of SCDB in this article and would be most grateful to receive any comments or suggestions for its improvement. Please send responses to Lucy Caswell, The Ohio State University, 242 West 18th Ave., Columbus OH 43210-1107 (614) 292-0538. We would be especially interested to know about other libraries which have created automated finding aids.


  1. Nena L. Couch, Thomas F. Heck, Susan J. Logan, Geoffrey D. Smith, and Robert A. Tibbetts are also in the working group which conceived and designed SCDB (Special Collections Database).
Copyright © American Library Association

Article Views (Last 12 Months)

No data available

Contact ACRL for article usage statistics from 2010-April 2017.

Article Views (By Year/Month)

January: 2
February: 2
March: 1
April: 11
May: 3
June: 2
July: 2
August: 2
September: 1
October: 7
November: 1
December: 0
January: 1
February: 3
March: 3
April: 8
May: 4
June: 5
July: 9
August: 2
September: 4
October: 3
November: 7
December: 4
January: 0
February: 0
March: 0
April: 0
May: 0
June: 0
July: 0
August: 7
September: 4
October: 5
November: 1
December: 5