Society of American Archivists
Technical Subcommittee on Descriptive Standards

Comments on the General Revision of ISAD(G)
12 October 1998

General Comments:

  1. We would like ISAD(G) to contain clearer information and guidance on the application of the standard to various formats of multi-level description. ISAD(G) as currently structured seems to us to provide guidance for formulating archival descriptions using a hierarchical chain of MARC-based or MARC-like catalog records. We feel that it is critical that ISAD(G) be applicable to a variety of forms in which archival descriptions may appear.
    1. Regarding the application of ISAD(G) to catalog records, we feel that ISAD(G) does not currently detail a mechanism for establishing links between the various hierarchical descriptions that together constitute a description, at various levels, of a single unit of archival materials. Such a mechanism would need to carry the linkage information across a variety of information-processing systems.
    2. Regarding the application of ISAD(G) to other archival description formats, we find ISAD(G) less clear on the application of the standard for structuring finding aids of the type commonly known as registers, inventories, or guides. Specifically, we would like ISAD(G) to contain some guidance on and examples of its application to the creation of the container lists common in many archival finding aids.
  2. We would like to see the introductory information for ISAD(G) more clearly address a question likely to be raised as archival descriptions are increasingly machine processed and output in various formats for end-user consumption. For whom or what is the information in an ISAD(G)-compliant archival description intended? Specifically, as the automated processing of archival descriptions becomes increasingly diversified beyond the MARC world, would a description really be ISAD(G) compliant if it contained some of the ISAD(G)-delineated elements in human-readable form and others as machine-processable attributes that would not necessarily be displayed to the human end user of the descriptive product? One currently available ISAD(G) to EAD crosswalk maps some ISAD(G) elements to the textual content of EAD tags and others to attribute values on EAD tags, the latter not necessarily information that will ultimately be displayed to end users by a particular delivery system. Would a description using such a crosswalk be considered ISAD(G) compliant?
  3. We feel that the ISAD(G) Introduction currently seems to suggest a preferred order for ISAD(G) elements, but is unclear regarding the basis upon which that order has been established (why that particular order and not others that seem viable?) If ISAD(G) is primarily about the standardized capture of descriptive information so that it can be presented clearly and unambiguously to end users of descriptive products, we wonder about the efficacy of the "areas of description" concept--which seems to have its origins in ISBD--for organizing this information as opposed other models such as the one promulgated in EAD? We feel that it would be helpful for ISAD(G) to contain more variety in examples of the application of the standard, including ones that address the finding aid model as opposed to the catalog record model. Since ISAD(G) promulgates multi-level description, we feel that any recommendations regarding minimal sets of elements and preferred order need to be made for, at the very least, the standard variety of levels contained in the current 3.1.4 example: fonds, series, sub-series, file, and item.
  4. Beginning in the glossary and continuing throughout ISAD(G), we find the term "unit of description" to be unclear. Any particular element within ISAD(G) or within a finding aid might be considered a "unit of description." We would suggest something like "unit being described" as a clearer term choice for the intended meaning of "unit of description" as conveyed in the ISAD(G) glossary.
  5. We feel that it is important that the ISAAR(CPF) and the ISAD(G) data structures be more explicitly integrated. One way of accomplishing this might be a complete integration of ISAAR(CPF) into ISAD(G), so that authority data could be carried directly in the descriptive record as well as in a separate authority record. Another might be a linking mechanism that would make the relationship between an ISAAR(CPF)-compliant authority record and the appropriate element in an ISAD(G)-compliant descriptive record more explicit regardless of the information system processing the records.

Comments Regarding Specific ISAD(G) Elements:

  1. We are troubled by the muddying in ISAD(G) of the distinction between information contained in the elements "title" (3.1.2) and "name of creator" (3.2.1). We feel that the name of the creator of any unit being described is too significant to simply be buried within the title element and that the ISAD(G) rule for 3.2.1 should not relegate this element to secondary status vis--vis the "title". We also feel that the "name of creator" field, regardless of whether or not that creator's name appears in the title, is important as a place to incorporate the authority controlled version of the creator's name. Some examples of the usage of this element should also be provided in the ISAD(G) documentation.

Support for the Canadian Council of Archives Comments:

The following references the item numbers contained in the "General Comments" section of the document Comments on the General Revision of ISAD(G), Canadian Committee on Archival Description, Canadian Council of Archives, September 1998.