More Information
Increased Findability and Expanded Contextualization
Most MEAP teams have a hard time finding the right scale for metadata description at the beginning of the project. While they work to create meaningful metadata for each object, it is hard to define best practices that work across a collection. This is especially true for teams who may be digitizing for the first time and have never created metadata at the object level before.
In this new monthly column, "Focus on Metadata," we share recommendations based on our work with project teams around the world. Today, we focus on Subject Terms.
Applying the right Subject Terms can increase findability and invites a richer, contextual understanding of how one individual object might fit into a broader historical event or regional concern. Identifying subject terms is an important part of the metadata creation process. It is also one that can be difficult to navigate—especially for teams working with rare and unique collections. Go into too much detail, and you run the risk of creating subject terms that only apply to one specific object in your collection. Remain too abstract, and you may not provide enough detail to separate one object from another if the same three terms apply to every object in the collection.
The points below are meant to help MEAP project teams identify the right Subject Terms for their project and to apply them with consistency in mind.
For more details about MEAP metadata requirements and recommendations:
The MEAP Metadata Handbook introduces the four types of subjects—Subject.topic, Subject.name, Subject.geographic, and Subject.temporal. We encourage project teams to explore existing controlled vocabularies for each subject field. However, most MEAP grantees opt to develop local vocabularies that specifically describe their collection and reflect their communities' perspective.
As you define your own local practices, consider these general guidelines for thinking about and creating subject terms.
Subject terms are (generally) plural:
The question “What is the object about?” is a helpful guide for thinking about subject terms because answers to the question tend to follow the desired format for subject terms: plural.
A wedding photograph might show a bride and groom in formal dress, but this description is not how we would answer the question “What is the photograph about?” In this case, the photograph is about weddings; it is about brides; and it is about grooms. All three of those terms would work as subject terms for this object. (And all three are present in the UCLA Digital Library: Brides, Grooms, Weddings.) And even if this particular photograph only depicts one bride, one groom, and one wedding, the plural terms brides, grooms, and weddings are more appropriate subject terms because they speak to the categories of brides, grooms, and weddings, rather than individual examples from those categories.
Subject terms should be simple but specific:
Guiding questions like what, when, where, or who an object is about can also help teams figure out how simple or complex, detailed or specific their subject terms should be. For example, just because a magazine is printed in May 2010 does not mean it is about May 2010. Even though it’s simple, May 2010 would not be a very useful term because it’s not specific enough to help users understand why the date is significant.
Instead, for this hypothetical example, we see that the magazine’s May 2010 issue includes a series of articles marking the 200th anniversary of the May Revolution (Revolución de Mayo) in Buenos Aires, which makes for a much better subject.temporal term. Revolución de Mayo (1810) is a perfect answer to the question “When is the object about?” It is simpler than saying the magazine issue is about the 2010 commemoration of the 1810 Revolución de Mayo and it is more specific than saying “May Revolution.” The language (Spanish) and the date (1810) provide users with specific clues to help them understand that this Revolución de Mayo is not the South Korean May Revolution of 1961.
Subject terms should be even simpler for digital collections:
While the previous example leads us to a specific historical event that can become a subject-term for the object (in this case for the subject.temporal field), not all objects are as easy to decipher as others. When working with objects from a specific collection, it can be especially difficult to know whether the object is about the motion picture industry in Los Angeles (Motion picture industry–California–Los Angeles) or whether the object is about the Motion picture industry and Los Angeles, Calif. When creating subject terms for digital collections, the latter approach (two simpler terms) is often more useful than the former (one complex term).
For analog collections in which objects relating to the Motion picture industry–California–Los Angeles and Motion picture industry–India–Mumbai would be shelved together, complex terms that unite topic and location can help users find comparable and related objects because those objects are stored together. Browsing the shelf would allow a user to find one set of books shelved next to another. However, for digital collections in which objects are not stored physically, complex terms can prevent users from finding relevant and related objects.
In most digital collections platforms, users do not browse subject terms alphabetically or systematically. Instead, they search for keywords sporadically and click on the ones that are relevant to their interests. As a result, if users search for “film industry hollywood,” they might stumble across the specific subject heading Motion picture industry–California–Los Angeles and then click the link to find related objects with the same subject heading. However, assuming that this subject term applies to all relevant objects in the collection could cause them to miss objects that have been grouped under more specific terms, like those related to African American motion picture actors and actresses--California--Los Angeles or African American motion picture actors and actresses--California--Los Angeles.
When creating metadata for digital collections, Including each element as a separate term (“Motion pictures,” “Actors and actresses,” “African Americans,” “Los Angeles”) makes it easier for users to navigate. And because complex terms can cause users to miss related objects, the Digital Public Library of America recommends separating, rather than connecting, subject terms for digital collections (unless following a specific controlled vocabulary as the terms above do).
Subject terms should be adjective free:
Recognizing that attaching detailed descriptors to subject terms can hinder users’ ability to access objects in the collection fully, it’s also a good idea to omit unnecessary adjectives when creating subject terms. Even if the couple in the first example’s wedding portrait is a very happy couple, including the adjective “happy” to the subject term “couples” can make it harder for users to access all images of couples. Happy couples, sad couples, married couples, divorced couples are all couples but the extra adjectives make it harder for users to find a variety of couples when using the subject term facets.
Of course, this general recommendation to avoid adjectives does not preclude the use of culturally relevant terms like mithuna (auspicious couple) and dampati (loving couple) to describe cultural heritage objects. When talking about artistic forms and representative genres, these adjectives can be useful and relevant to include because they serve as analytical categories (rather than simple cataloger observations).