Advocates of a "metadata systems approach" to the description of electronic records argue that metadata's capacity to provide descriptive information about the context of electronic records creation will obviate, or reduce significantly, the need for traditional archival description. This article examines the assumptions about the nature of archival description and of metadata on which metadata strategies are grounded, for the purposes of ascertaining the following: whether the skepticism concerning the capacity of traditional description to meet the challenges posed by the so-called "second generation" of electronic records is justified; whether the use of metadata as archival description is consistent with their nature and purpose; and whether metadata are capable of servinng archival descriptive purposes.
Critical Arguements
CA "Before the archival profession assigns to traditional archival description the diminished role of "added value" (i.e. accessory) or abandons it altogether, the assumptions about the nature of archival description and of metadata on which metadata strategies are grounded ought to be carefully examined. Such an examination is necessary to ascertain the following: whether the skepticism concerning the capacity of traditional description to meet the challenges posed by the so-called "second generation" of electronic records is justified, whether the use of metadata as archival description is consistent with their nature and purpose, and whether metadata are acapable of serving archival purposes."
Phrases
<P1> In an article published in Archivaria, David Wallace summarized recent writing on the subject of metadata and concluded that "[d]ata dictionaries and the types of metadata that they house and can be built to house should be seriously evaluated by archivists" because of their potential to signficantly improve and ultimately transform traditional archival practice in the areas of appraisal, arrangement, description, reference, and access. <warrant> <P2> In the area of description, specifically, advocates of "metadata management" or a "metadata systems approach" believe that metadata's capacity to provide descriptive information about the context of electronic records creation will obviate, or reduce significantly, the need for traditional description. <P3> Charles Dollar maintains that archival participation in the IRDS standard is essential to ensure that archival requirements, including descriptive requirements, are understood and adopted within it. <warrant> <P4> According to David Wallace, "archivists will need to concentrate their efforts on metadata systems creation rather than informational content descriptions, since in the electronic realm, archivists' concern for informational value will be eclipsed by concern for the evidential value of the system." <warrant> <P5> Charles Dollar, for his part, predicts that, rather than emphasize "the products of an information system," a metadata systems approach to description will focus on "an understanding of the information system context that supports organization-wide information sharing." <P6> Because their scope and context are comparitively narrow, metadata circumscribe and atomize these various contexts of records creation. Archival description, on the other hand, enlarges and integrates them. In so doing it reveals continuities and discontinuities in the matrix of function, structure, and record-keeping over time. <P7> Metadata are part of this broader context, since they constitute a series within the creator's fonds. The partial context provided by metadata should not, however, be mistaken for the whole context. <P8> Metadata, for example, may be capable of explaining contextual attributes of the data within an electronic records system, but they are incapable of describing themselves -- i.e., their own context of creation and use -- because they cannot be detached from themselves. For this reason, it is necessary to describe the context in which the metadata are created so that their meaning also will be preserved over time. <P9> A metadata system is like a diary that, in telegraphic style, records the daily events that take place in the life of an individual as they occur and from the individual's perspective. <P10> Archival description, it could be said, is the view from the plane; metadata, the view from the field as it is plowed. <P11> While a close-up shot-- such as the capture of a database view -- may be necessary for the purposes of preserving record context and system functionality, it does not follow that such a snapshot is necessary or even desirable for the purposes of description. <P12> Because the context revealed by metadata systems is so detailed, and the volume of transactions they capture is so enormous, metadata may in fact obscure, rather than illuminate, the broader administrative context and thereby bias the users' understanding of the records' meaning. In fact, parts of actions and transactions may develop entirely outside of the electronic system and never be included in the metadata. <P13> If the metadata are kept in their entirety, users searching for documents will have to wade through a great deal of irrelevant data to find what they need. If the metadata are chopped up into bits corresponding to what has been kept, how comprehensible will they be to the uesr? <P14> The tendency to describe metadata in metaphorical terms, e.g., in relation to archival inventories, has distracted attention from consideration of what metadata are in substantial, concrete terms. They are, in fact, records created and used in the conduct of affairs of which they form a part. <P15> The transactions captured by metadata systems may be at a more microscopic level than those captured in registers and the context may be more detailed, given the technological complexity of electronic record-keeping environments. Nevertheless, their function remains the same. <P16> And, like protocol registers, whose permanent retention is legislated, metadata need to be preserved in perpetuity because they are concrete evidence of what documents were made and received, who handled them, with what results, and the transactions to which they relate. <warrant> <P17> While it is true that metadata systems show or reveal the context in which transactions occur in an electronic system and therefore constitute a kind of description of it -- Jenkinson made the same observation about registers -- their real object is to record the fact of these transactions; they should be, like registers, "preserved as a [record] of the proceedings in that connection." <P18> Viewing metadata systems as tools for achieving archival purposes, rather than as tools for achieving the creators' purposes is dangerous because it encourages us to, in effect, privilege potential secondary uses of metadata over their actual primary use; in so doing, we could reshape such use for purposes other than the conduct of affairs of which they are a part. <P19> Metadata strategies risk compromising, specifically, the impartiality of the records' creation. <P20> For archivists to introduce in the formation of metadata records requirements directed toward the future needs of archivists and researchers rather than toward the current needs of the creator would contribute an element of self-consciousness into the records creation process that is inconsistent with the preservation of the records' impartiality. <P21> If the impartiality of the metadata is compromised, their value as evidence will be compromised, which means, ultimately, that the underlying objective of metadata strategies -- the preservation of evidence -- will be defeated. <P22> None of these objections should be taken to suggest that archivists do not have a role to play in the design and maintenance of metadata systems. It is, rather, to suggest that that role must be driven by our primary obligation to protect and preserve, to the extent possible, the essential characterisitcis of the archives. <P23> The proper role of an archivist in the design of a metadata system, then, is to assist the organization in identifying its own descriptive needs as well as to ensure that the identification process is driven, not by narrowly defined system requirements, but by the organization's overarching need and obligation to create and maintain complete, reliable, and authentic records. <P24> That is why it is essential that information holdings are identified and described in a meaningful way, organized in a logical manner that fascilitates their access, and preserved in a manner that permits their continuing use. <P25> Record-keeing requirements for electronic records must address the need to render documentary relationships wisible and to build in procedures for authentication and preservation; such measures will ensure that record-keeping systems meet the criteris of "intergrity, currency an relevancy" necessary to the records creator. <P26> In other words, effective description is a consequence of effective records management and intelligent appraisal, not their purpose. If the primary objectives of metadata are met, description will be fascilitated and the need for description at lower levels (e.g., below the series level) may even be obviated. <P27> Metadata systems cannot and should not replace archival description. To meet the challenges posed by electronic records, it is more important than ever that we follow the dictates of archival science, which begin from a consideration of the nature of archives. <P28> Archival participation in the design and maintenance of metadata systems must be driven by the need to preserve them as archival documents, that is, as evidence of actions and transactions, not as descriptive tools. Our role is not to promote our own intersts, but to deepen the creator's understanding of its interests in preserving the evidence of its own actions and transactions. We can contribute to that understanding because we have a broader view of the creator's needs over time. In supporting these interests, we indirectly promote our own. <P29> To ensure that our descriptive infrastructure is sound -- that is to say, comprehensible, flexible, efficient, and effective -- we need equally to analyze our own information management methods and, out of that analysis, to develop complementary systems of administrative and intellectual control that will build upon each other. By these means we will be able to accomodate the diversity and complexity of the record-keeping environments with which we must deal.
Conclusions
RQ "Since 'current metadata systems do not account for the provenancial and contextual information needed to manage archival records,' archivists are exhorted [by Margaret Hedstrom] to direct their research efforts (and research dollars) toward the identification of the types of metadata that ought to be captured and created to meet archival descriptive requirements. "
SOW
DC Dr. Heather MacNeil is an Assistant Professor at the School of Library, Archival, and Information Studies at the University of British Columbia. Dr. MacNeilÔÇÖs major areas of interests include: trends and themes in archival research & scholarship; arrangement and description of archival documents; management of current records; trustworthiness of records as evidence; protection of personal privacy; interdisciplinary perspectives on record trustworthiness; and archival preservation of authentic electronic records
