The Evolution of Archival Cataloging Standards and Metadata Practices

The history of archival cataloging standards and metadata practices is a story of continuous adaptation. From handwritten card catalogs to linked data ecosystems, the methods archivists use to describe, organize, and share historical records have transformed dramatically. This evolution reflects not only technological progress but also a deepening understanding of how context, provenance, and user needs shape the preservation of collective memory. As archival collections grow in size and complexity, the standards that govern their description must balance precision with accessibility, ensuring that records remain discoverable and meaningful across generations. The shift from siloed, institution-specific approaches to globally interoperable frameworks has been driven by collaboration among professional organizations, technological innovators, and user communities, each contributing to a richer, more connected archival landscape.

Early Archival Practices: The Foundation of Description

The Era of Manual Systems

In the early 20th century, archivists operated in a largely analog world. The Principles of Archival Description—provenance, original order, and context—were already well established, thanks to pioneers like the Dutch manual of 1898 (Muller, Feith, and Fruin) and Sir Hilary Jenkinson’s A Manual of Archive Administration (1922). Cataloging was a labor-intensive process executed by hand. Finding aids were often typed or handwritten lists stored in binders or card cabinets. Each record or series received a minimal set of descriptive elements: creator, dates, physical extent, and a brief scope note. This approach worked well for small, homogeneous collections but could not scale as archives grew and diversified. For example, the U.S. National Archives, established in 1934, faced an immediate backlog of billions of pages of federal records, highlighting the inadequacy of manual methods for large-scale government archives.

The Role of Provenance and Original Order

Provenance—the principle that records should be maintained according to their creator’s organizational structure—became the bedrock of archival practice. Original order required that the sequence of records as created be preserved. These concepts, though timeless, were applied inconsistently across institutions. Without a shared standard, a researcher moving from one repository to another might encounter wildly different descriptive practices. The need for harmonization became increasingly apparent, especially as governments and universities began to invest in large-scale archival programs after World War II. The first formal efforts to codify these principles came from the Dutch manual and later from the British and American archival traditions, but it took decades for these to converge into a single international framework.

Development of Standardized Frameworks

The Push for International Consensus

By the mid-20th century, the archival profession recognized that to ensure consistent access and exchange of information, agreed-upon rules were essential. The International Council on Archives (ICA), founded in 1948, spearheaded efforts to create a common language for archival description. In Canada, the Rules for Archival Description (RAD) emerged in the 1980s, providing a detailed code for describing records at multiple levels (fonds, series, file, item). RAD was influential but primarily focused on Canadian contexts. At the same time, the General International Standard Archival Description (ISAD(G)), first published in 1994, offered a more globally applicable framework. ISAD(G) defined a set of 26 descriptive elements organized into seven areas (such as identity, context, content, and conditions of access) and introduced the multilevel descriptive model that remains central today. The standard was later complemented by the International Standard Archival Authority Record for Corporate Bodies, Persons, and Families (ISAAR(CPF)), which provided a structure for authority records that linked creators to their records.

National Standards and Their Impact

Other countries developed their own standards, often aligned with ISAD(G). The United States used Anglo-American Cataloguing Rules (AACR2) for bibliographic materials but created separate rules for archives: the Archives, Personal Papers, and Manuscripts (APPM) manual (1983) and later Describing Archives: A Content Standard (DACS) (2004). The UK adopted the Manual of Archival Description (MAD) and then ISAD(G) via the National Council on Archives (NCA) Rules. These efforts established a vocabulary and structure that could be shared, even if implementations varied. The emphasis on multilevel description ensured that users could drill down from a broad collection overview to the most granular item, all while maintaining contextual relationships. For instance, the fonds-level entry might describe the complete body of records from a university department, while series-level entries break it into administrative files, student records, or research reports, each with its own scope and date range.

The Rise of Digital Metadata Standards

Encoded Archival Description (EAD)

The transition to digital finding aids began in earnest in the 1990s. The Encoded Archival Description (EAD) standard, developed by the Library of Congress and the Society of American Archivists, revolutionized archival access by encoding finding aids in XML. EAD allowed archivists to mark up hierarchical descriptions (using tags for elements like <c01>, <unittitle>, <unitdate>) in a machine-readable format. This made it possible to search, display, and exchange finding aids across the internet. The standard continues to be maintained and widely used, though its complexity has led to simplified derivatives like EAC-CPF (Encoded Archival Context–Corporate bodies, Persons, Families) for authority records. EAD quickly became the de facto standard for sharing archival descriptions online, with implementations at major institutions like the Library of Congress, the British Library, and university archives worldwide. However, its steep learning curve and verbose XML syntax prompted calls for more lightweight alternatives.

Dublin Core and Interoperability

Simultaneously, the Dublin Core Metadata Initiative (DCMI) provided a minimalist metadata schema of just 15 elements (e.g., title, creator, date, subject, description). Originally developed for web resources, Dublin Core was quickly adopted by archives and museums for digital object descriptions because of its simplicity and cross-domain flexibility. While not as rich as EAD or ISAD(G), Dublin Core became the backbone of many digital library projects, such as the Open Archives Initiative Protocol for Metadata Harvesting (OAI-PMH). This allowed repositories to expose metadata to aggregators like the Digital Public Library of America (DPLA) and Europeana. Its lightweight nature made it ideal for rapid deployment but also required careful mapping from more detailed archival standards. In practice, many archives maintain a hybrid approach: a full EAD finding aid for internal use, with a Dublin Core record exported for external aggregation.

Specialized Standards: VRA Core, MODS, and METS

As digital collections expanded, dedicated standards emerged for specific media types. VRA Core (Visual Resources Association) offered a metadata schema for visual cultural works, with elements to describe the work itself, its images, and provenancial history. MODS (Metadata Object Description Schema) provided a richer alternative to Dublin Core for bibliographic objects, often used by libraries integrating archival materials. METS (Metadata Encoding and Transmission Standard) enabled the packaging of descriptive, administrative, and structural metadata for complex digital objects (e.g., digitized books with page-turners). These standards coexisted with, and sometimes supplemented, archival standards like EAD, creating a layered metadata environment. For example, a digitized photograph collection might use VRA Core for the image description, METS for technical and structural metadata, and EAD for the overall collection arrangement. This layered approach, while powerful, introduced challenges in maintaining consistency across multiple schemas.

The Impact of Digital Repositories and Aggregators

The proliferation of digital repositories in the 2000s further shaped metadata practices. Platforms like ContentDM, DSpace, and Fedora provided out-of-the-box support for Dublin Core and METS, enabling smaller institutions to publish digital collections without deep metadata expertise. Aggregators like the Digital Public Library of America (DPLA), launched in 2013, harvested metadata from hundreds of partners and normalized it into a shared profile based on Dublin Core and the Europeana Data Model (EDM). This forced archives to improve the quality and consistency of their metadata to ensure their collections were discoverable in a national or global context. The DPLA’s hub model encouraged the development of state or regional service hubs that provided training and tools for metadata creation, further spreading best practices.

Modern Practices and Interoperability

Linked Data and the Semantic Web

The early 21st century brought a paradigm shift toward linked data and semantic web principles. Instead of siloed XML files, archivists began to express metadata using the Resource Description Framework (RDF), treating each entity (people, places, organizations, records) as a URI (uniform resource identifier) connected by relationships. This approach, championed by initiatives like the Linked Data for Archives (LD4A) and the International Council on Archives’ Records in Contexts (RiC) project, allows archives to interlink their descriptions with other data sources—such as Wikidata, VIAF (Virtual International Authority File), or GeoNames—creating a web of contextual knowledge. For example, a fonds creator can be linked to his authority record, which in turn links to biographical resources, museum collections, and even secondary literature. The Social Networks and Archival Context (SNAC) project provides a concrete demonstration, building a graph of over 6 million biographical entities linked to archival materials across hundreds of repositories.

Records in Contexts (RiC)

Perhaps the most ambitious modern standard is Records in Contexts (RiC), a conceptual model and ontology developed by the ICA’s Expert Group on Archival Description (EGAD). RiC replaces the multilevel fonds-series-item hierarchy with a graph-based model where any entity (Record, Agent, Function, etc.) can be related to any other. It integrates descriptive, contextual, and relational information into a single framework. The standard is published as both a conceptual model (RiC-CM) and an OWL ontology (RiC-O). While adoption is still growing, RiC represents the cutting edge of archival description, promising truly interoperable, open, and extensible metadata. For instance, under RiC, a record can be directly linked to the function that created it, the agent who used it, and the event that led to its creation, all without requiring a rigid hierarchical container.

Tools and Platforms for Modern Metadata

Contemporary archival systems—such as ArchivesSpace, AtoM (Access to Memory), and CollectiveAccess—support these standards natively. ArchivesSpace, for example, uses EAD, DACS, and MODS as its core descriptive schemas, while AtoM maps to ISAD(G), RAD, and Dublin Core. Many of these systems also allow export of RDF/XML and link to external authorities. Additionally, the International Image Interoperability Framework (IIIF) has transformed access to digital images, providing an API that allows archival materials to be displayed, annotated, and compared across repositories without requiring metadata duplication. IIIF works in concert with METS and MODS to deliver rich, interoperable viewing experiences. The IIIF community continues to grow, with major archives like the National Archives UK and the Bibliothèque nationale de France adopting the framework for their digital surrogates.

Future Directions

Artificial Intelligence and Automated Metadata Generation

The explosion of born-digital records has made manual description impractical at scale. Artificial intelligence (AI) and machine learning (ML) are beginning to address this gap. Natural language processing (NLP) can automatically extract dates, names, and subjects from text documents, generating preliminary metadata. Image recognition can identify visual content, while classification algorithms can assign archival series codes. These tools are not yet fully reliable—especially for complex, handwritten, or highly contextual records—but they are rapidly improving. Initiatives like the Archives’ Machine Learning Pipeline at the U.S. National Archives and other experimental projects show promise for reducing backlogs and enhancing discoverability. In the future, AI might not only generate metadata but also detect relationships between records across collections, surfacing connections that human catalogers might miss.

User-Centered Design and Discoverability

Future standards will likely prioritize the user experience more explicitly. Current finding aids, even when encoded in EAD or published via linked data, can still be opaque to non-specialist researchers. Efforts are under way to create simplified, faceted browsing interfaces that surface archival descriptions in ways similar to library catalogs or search engines. The Social Networks and Archival Context (SNAC) project, for instance, builds a user-friendly graph of historical figures, connecting them to records scattered across hundreds of repositories. This moves archival description from a passive storage tool to an active research pathway. Additionally, the rise of wikidata as a collaborative authority file offers a low-barrier way for archives to enrich their metadata by linking to a globally maintained knowledge base.

Balancing Detail with Accessibility

A persistent challenge is the tension between exhaustive description (which enables precise research but requires significant labor) and minimal description (which is cheaper but may miss vital context). Future standards will need to support lightweight profiles that can be enriched over time, perhaps through crowdsourcing or automated enrichment. The DACS Single Level Description option already allows archives to describe only the fonds level, leaving lower levels for later. Similarly, RiC’s graph model allows for incremental addition of relationships. As archival professionals debate these trade-offs, the goal remains constant: to ensure that historical records are not just stored but are actively understood and used. The use of minimal and extensible metadata profiles, as advocated by the W3C’s Data on the Web Best Practices, can help archives allocate resources efficiently while still enabling rich discovery.

Ethical Considerations in Metadata

Modern archival standards are also grappling with ethical dimensions. Indigenous communities, for example, have demanded culturally sensitive metadata that respects traditional knowledge and control. The Traditional Knowledge (TK) Labels and Local Contexts initiative provide a framework for adding provenance and restrictions that go beyond standard access conditions. Linked data and open metadata can also inadvertently expose sensitive information; future standards will need to incorporate automated redaction, access tiers, and granular permission controls. The CARE Principles for Indigenous Data Governance (Collective Benefit, Authority to Control, Responsibility, Ethics) are increasingly influencing archival metadata design, pushing standards beyond technical interoperability toward social accountability. These principles require that metadata about Indigenous collections be co-created with community input and that access be managed according to cultural protocols.

Conclusion

The journey from handwritten catalogs to linked-data graphs reflects the archival profession’s enduring commitment to making records accessible across time and technology. Each new standard—whether ISAD(G), EAD, Dublin Core, or RiC—has addressed its era’s challenges: consistency, machine-readability, interoperability, and context. Looking ahead, the fusion of AI, semantic-web principles, and user-centered design promises to further transform how we describe and discover historical materials. The core archival values of provenance, original order, and context remain constant, even as the tools for implementing them evolve. By embracing both innovation and ethical rigor, archivists will continue to connect people with the evidence of the past, ensuring that memory is preserved not in isolation, but as part of a living, interconnected web of knowledge. The collaboration between the archival community and allied fields—libraries, museums, computer science, and indigenous studies—will shape the next generation of standards, making archival description more equitable, efficient, and expansive.