The Virtual Library of Virginia
Search the VIVA website:

Analytics for Chadwyck-Healey poetry databases


All VIVA public institutions and two independent colleges have access to the Chadwyck-Healey poetry databases: English Poetry, English Verse Drama, American Poetry, and African-American Poetry. Analytics in 8,145 MARC bibliographic and 2,072 MARC name authority records are available to VIVA libraries, along with a tool kit for customizing the bibliographic records for local loading protocols and cataloging policies.


1996: VIVA licensed the Chadwyck-Healey poetry databases. The University of Virginia's Electronic Text Center loaded and processed the full text databases and made them available to faculty, students, and staff of authorized VIVA institutions. Vendor supplied MARC records, generated from TEI headers, were delivered on 8mm tapes as part of the license agreement.

1997: UVa loaded a set of test records into VIRGO, then loaded all 8,167 analytic records in December 1997. After loading the entire MARC tape, UVa exported four sets of records for the four poetry databases in the SIRSI flat ASCII format. Only the College of William and Mary, another SIRSI library, was able to make use of this format.

1998: The VIVA User Services Committee Task Force on Cataloging and Intellectual Access (VIVACat) investigated the possibility of providing author and title access by loading the vendor supplied MARC records in a variety of online library systems at VIVA institutions.

Jackie Shieh (UVa) analyzed MARC record structure and quality. She and Ed Summers (ODU) wrote scripts to bring the master records up to minimal MARC/AACR2 standards. Summers then customized them for record loading protocols, holdings record formats, and local cataloging policies at VCU, ODU, and JMU.

During JMU's bibliographic record load, original cataloger Judy Anderson found discrepancies with local headings. VIVACat recognized the problem and agreed to distribute authority control work among five institutions. UVa, the only NACO library represented, researched names not found in the initial pass through. Summers extracted data from the resulting spreadsheets to reprocess the master bibliographic records and flip 100/700 headings to the established form of name.

1999: Summers and colleagues released an open source Perl module for manipulating MARC records with simple scripts (

Shieh left for the University of Michigan and began testing C-H records there, while completing some of the original authority work. Summers left ODU for private sector employment in NYC.

2000: Now at Averett College, and without programming support or access to the poetry databases, VIVACat chair Elaine Day processed the C-H master records with LC's maker/breaker script and MS Word global edit to make final changes for original authorities, then stored the records on a VIVA server.

2001: Day processed SOLINET/netLibrary records for Averett with a combination of the MARC Perl Module, maker/breaker script, MS Word global editing and MarcEdit, which at that time had limited capabilities.

2002: Day customized a second SOLINET/netLibrary collection using only MarcEdit and no original scripting.

Day announced availability of the records and editing tools to VIVA libraries. Six VIVA libraries expressed interest in loading the C-H bibliographic and/or authority records.

Three former Task Force members (Shieh, Summers, Day) drafted an article on the project for a Haworth Press special edition.


Files Available from VIVA Office

  • Bibliographic Files (MARC, ASCII)
    • African American Poetry (98 records)
    • American Poetry (1,288 records)
    • English Poetry (4,474 records)
    • English Verse Drama (2,285 records)
  • Authority Files (MARC, Excel spreadsheets)




Why would we want to add these records to our local OPAC? Aren't they too specific? ("granularity" problem)

We recommend that reference and catalog librarians discuss the wisdom of loading the poetry database analytics in local catalogs. This collection differs from many other digital resources in that:

  • The Chadwyck-Healey poetry databases are digitalized monographs. The analytic bibliographic records provide access analogous to monographic tables of contents. When feasible, library catalogs have provided author and title access to contents of anthologies.
  • The databases are static and access is permanent, in contrast to dynamic web or digital library content requiring ongoing catalog and URL maintenance.
  • Users are accustomed to searching for names of poets and literary criticism of poems in the library catalog. They may be pleasantly surprised to find the full text of a poem rather than an anthology with unspecified contents.



Three of the record sets are large. We recommend beginning with African American Poetry (98 records), and breaking the other sets into sections for PC editing with limited memory.

The bibliographic records meet minimal standards. We cannot guarantee that they are free of errors, such as incorrect filing indicators in MARC 245. Use your library system or other MARC viewing/editing software to assess the records before loading in your catalog.

The authority records were downloaded from OCLC in 1999. They have not been synchronized with the authority file since that time. The bundled authority records will probably duplicate some names already in your library system. Be aware of how your system handles duplicate authority records before loading them.

We recommend entering a "hook" for retrieval of records associate with each of the four databases (keyword phrase or other common element). In most systems, you can also create holdings records by inserting the appropriate MARC tag and subfield contents in the bibliographic records prior to loading.

Elaine Day (AU, will answer general questions about the project or download files. Please contact Barbara Anderson (VCU, with questions on specific records.


VIVA is funded by the Virginia General Assembly and the VIVA member institutions, and is sponsored by the State Council of Higher Education (SCHEV).