Final Project

Final Project

Catalog Records and Legacy Data

As you can see from the Cataloger game, creating catalog records can be tricky. Looking specifically at the Americas Collection of the UPenn Museum digital collection, we can see two issues; legacy data taking control over the catalog records and the complete authority of catalog records are held by UPenn staff.

Legacy data is information stored in an obsolete or old format or computer system. When museums, libraries, and archives use legacy data to fill or create catalog records, we see outdated terms and bias descriptions being used.

Museum curators, librarians, and archivists are trained to be the authority of catalog records, displays, and information. That is, the people who create catalog records, displays, and information control what is available, how it is interpreted, and how is it accessed.

Museum scholars like Hannah Turner, Benjamin Filene, and James B. Gardner note how past museum theories and practices do not fit in our digital and modern lives. James B. Gardner argues that museums should advocate for both history and visitors by negotiating the gap between historians and the public’s understanding of the past. Benjamin Filene compares strategies of public historians and non-public historians in creating museum displays and other public history projects. Both Gardner and Filene argue for a shared authority.

Hannah Turner looks specifically at the National Museum of Natural History’s anthropology catalog to understand the history of museum management. Her findings suggest the museum catalog is a socio-technical information infrastructure, that is catalog records are generated by the relationship of individuals (catalogers) and technologies (card catalogs, for example).

The Literature

Catalogers are painfully aware of “legacy data”, a term used by Turner to describe metadata that was directly migrated to computerized systems from typed and handwritten records with no attention to accuracy. Hannah Turner, “The Computerization of Material Culture Catalogues: Object and Infrastructure in the Smithsonian Institute’s Department of Anthropology” Museum Anthropology 39 no. 2: 169. During her research of the National Museum of Natural History’s anthropology catalog, Turner noted how object naming and classification did not change “despite decades of post-colonial research and revision”.Turner, “The Computerization of Material Culture” Museum Anthropology 39, no. 2: 170. Looking at UPenn’s digital collection, the terms used as object names tend to be Anglo terms.

This data visualization shows us the relationship of Object Names to a state in the U.S. in which the object originates from. How many Object Names are unfamiliar to you? While familiarity is important in catalog records, omitting indigenous names shows a bias towards English speakers.

This data visualization shows the relationship between Object Name and the country each object originates from. How do these Object Names compare to those from the United States? There is a larger variety of object from South America and Central America. However, English names are still heavily used.

This data visualization shows all Object Names (some where grouped together based on likeness, i.e. vase and vase fragment). Almost all of these Object Names do not include indigenous names.

Another part of The Cataloger was matching descriptions with objects. While that is not typically the case for museum catalogers, in most cases they either create their own description, use the information available upon object’s arrival to the museum, or they will provide the most basic descriptions as possible. As you noticed in the game, these descriptions vary from extremely detailed, like the Huipil blouse to basic details like the human figure. Take a look at this graph of Descriptions used.

This graph shows us what terms are used together with most often. If we scroll through the years, we see the descriptive terms most frequently used together change over time. These bag of terms are mostly used together to describe objects in the collection. Looking closely at these terms, we see more descriptions of shapes, size, materials, and texture are added to objects in later years. Overall, there is a little description of how an object was used.

In Gardner’s article, he encourages museum professions to share authority with the visitor in museum exhibits by being transparent about museum practice and theory, “The public needs to understand how museums have shifted. . .”James B. Gardner, “Contested Terrain: History, Museums, and the Public” The Public Historian 26 no. 4: 15. The same should be applied to digital collections. Take this hat for example. When I look up the word hat in the catalog, this item shows up immediately. However, when I search basket, this hat is still on the list of potential items. Looking at the hat it does resemble a basket.

This may be an issue with the digital collection platform. Yet, it still begs the question, was this object once considered a basket then changed to a hat? If UPenn was transparent about any changes or even how objects are categorized (even an explanation on how objects are named or more province of an object) may better explain why this object is a hat.


Many museums are looking to expand descriptions by providing crowdsourcing opportunities to researchers and museum visitors. Flickr is a popular platform for museums to allow social tagging and provide more contextual information. Let’s compare this Karuk Basket Flickr page to the catalog record in the digital collection. The Flickr page includes one unique piece of information that is not in the catalog record, a concise description.

If this information was included in the catalog record, do you think visitors and researchers would be more likely to use it? Unlike other descriptions used in the catalog records that we saw in the Descriptions Graph, this description used adjectives like “elegant”. Also note, the creator’s full name is used in the Flickr description along with her role as a “weaver”. In the catalog record, we only see her abbreviated name, “E. Hickox” as the “Maker”.

Filene’s article looks at public history projects like StoryCorps created by a non-public historian in comparison to traditional museum exhibits created by public historians. Similar to Gardner, Filene argues that traditional public history projects (i.e. museum displays) limit curators in creating innovative projects like StoryCorps. He encourages public historians to go beyond the boundaries of the physical museum and create projects like non-public historians, “what really inspires people to engage with the past and care about history- is emerging from outside public history’s professional realm. Benjamin Filene, “Passionate Histories: “Outsider” History-Makers and What They Teach Us” The Public Historian 34, no. 1: 30.   

In a perfect world, museum catalogers could add this descriptive metadata to reflect the Flickr page. In the meantime, social media like Flickr allows users to add Tags and comment on images. By giving museum visitors opportunities online through social media to add descriptions, museums can go beyond the boundaries described by Turner, Gardner, and Filene. By sharing the authority between public historians and the public, legacy data could slowly become less of an issue.


The data used for this project comes directly from the UPenn Museum digital collections data set. For the purpose of this project, objects from the Americas collection was used. The UPenn Museum encourages researchers to use their data sets for projects as well as sharing those projects with the museum.

Turner, Hannah, “The Computerization of Material Culture Catalogues: Objects and Infrastructure in the Smithsonian Institution’s Department of Anthropology” Museum Anthropology 39 no. 2 (2016): 163-177.

Filene, Benjamin, “Passionate Histories: “Outsider” History-Makers and What They Teach Us” The Public Historian 34, no. 1 (2013) 11-33.

Gardner, James B., “Contested Terrain: History, Museums, and the Public” The Public Historian 26, no. 4 (2004) 11-21.