Aldus@SFU

The Wosk–McDonald Aldine Collection

Wrestling with Metadata

Minghui Zhou • September 30, 2025 • data, project

My work as a research assistant on the Aldus@SFU Project involves working to create more comprehensive metadata records for more than 120 volumes published by the Italian Renaissance printer-publisher Aldus Manutius (c. 1451–1515).

For those unfamiliar with the concept of metadata, a metadata record is a combined set of metadata elements and their values, used to describe and manage an individual resource. A metadata element is a property or piece of information that captures a specific characteristic of the resource; it is also commonly called a “metadata field.”

My works sounds straightforward – fill in those values for a particular set of metadata elements. In practice, though, I quickly learned that metadata is rarely simple, especially when working with centuries-old books full of metadata inconsistencies, ambiguities, and discrepancies.

What metadata are we collecting?

At its core, all metadata serves a purpose. As Miller (2011, 1) puts it, “Metadata is data or information that enables people to perform certain functions in relation to the information resources that the metadata is about.” That means the real question is not just what data we record, but why and for whom. Who are the audiences for these records? What will they need to engage with the resources in a meaningful way?

In our case, with the book metadata, we wanted to capture information that matters to both researchers and general users of digital collections. Some of the main elements I’ve been collecting include:

Each of these metadata elements supports different functions. For example, links to parallel digitized collections allow users to compare versions of the same Aldine title across institutions.

The IIIF Manifests were surprisingly easy to collect thanks to the DetektIIIF tool. This convenient browser extension (highly recommended) allows one to instantly identify the Manifest URL of a digital resource. Name authority records were another important piece of information to capture, as they ensure consistency and accuracy in identifying the people associated with the resource. VIAF, in particular, offers access to name authority files from major libraries around the world, helping us standardize references across borders and reduce ambiguities.

The challenges of working with metadata

Of course, much of the work is more complex than it seems at first. Metadata is supposed to bring clarity, but often it creates new puzzles. Authority files, for example, are designed to establish a single, consistent heading for a person, organization, or subject. In reality, they can be inconsistent, especially in VIAF, where I often found multiple records for the same author, which may have different dates, name variants, or other details. In other words, VIAF aggregates records from multiple national libraries, which enriches the data but also introduces redundancies. Therefore, not only may you find multi-national VIAF records for the same author that provide a similar amount of detail, a country-specific record may contain details absent from other, more general multi-national records. So part of the job is balancing completeness with consistency, and making judgment calls about which record to highlight as the most representative.

In general, names with dates after the tenth century are usually the easiest to verify and often do not require much verification. Names with dates before that, or without clear dates usually need more thorough verification in VIAF, as these entries often contain a lot of discrepancies in name forms and dates of birth/death. In those cases, I have had to dig deeper, comparing co-author fields, checking the title sections, and looking for any contextual clues hidden in the records. Often times I needed to research the author independently to piece together the most reliable match.

Work Cited:

Miller, Steven J. Metadata for Digital Collections. Neal-Schulman Publishers, Inc. 2011.