Reframing 'Voodoo' Across Systems

Rethinking Archival Description at the Getty Research Institute


Words by Emily Benoff & Kelli Yakabu

INTRODUCTION:
ORGANIZING FUGITIVE KNOWLEDGE

Rooted in embodied forms of refusal and survival shaped by the experiences of enslaved Africans, the concept of fugitivity frames movement itself as a political act.

By challenging the presumed fixity of dominant historical narratives and the power structures that sustain them, fugitivity understands knowledge as a continually evolving communal resource that affirms multiple ways of knowing and being. In our work processing collections and developing controlled vocabularies at the Getty Research Institute (GRI), we often confront a structural tension between the fluid, shifting nature of lived history and the institutional need to stabilize meaning through standardized description. This tension becomes especially clear when we develop strategies for search and discovery that amplify marginalized perspectives and remediate harmful narratives.

EMILY BENOFF:
REPARATIVE METADATA: REMEDIATING “VOODOO" IN THE AAT

Since 2018, the GRI’s multi-project African American Art History Initiative has worked to create a center for the study of African American art through tailored acquisitions in library and special collections. In parallel, the Getty Vocabulary Program–the team I work on, which develops the multilingual and polyhierarchical Getty Vocabularies–began adding controlled terminology to meet the descriptive needs arising from this work. The Getty Vocabularies–primarily ULAN, TGN, and AAT–offers shared terminology for describing people, places, and concepts across art and cultural heritage. Their compliance with international metadata standards has made them widely used across GLAM institutions as tools for interoperability. Yet, knowledge organization, which tends to crystallize language around dominant attitudes, reinforces deep-seated histories of race, class, and gender-based violence.

In our descriptive work, we must aim to represent complex histories and shifting language without erasing their difficult pasts. To begin approaching this challenge, I took up the project of remediating terminology for one African diasporic tradition shaped by fugitive forms of knowledge and long constrained by colonial misunderstanding: “voodoo.” In the United States, “voodoo” is often used as a pejorative umbrella term for Black religious and spiritual practices and is frequently misassociated with witchcraft and satanism. In reality, “Voodoo” refers to a group of distinct religions–most notably Vodún, Haitian Vodou, and Louisiana Voodoo–each with its own history and practices. 

Vodún, a diverse set of spirit and ancestor practices, is the traditional religion of the Fon and Ewe peoples of Dahomey (modern-day Benin). It is often cited as a key ancestor of Haitian Vodou and Louisiana Voodoo, which developed syncretically as enslaved practitioners adapted their beliefs in the Americas during the Atlantic slave trade. When I began this project, however, the AAT only included the term “voodoo,” with terms like Vodou and Vodún treated as non-preferred equivalents. This continued a long history, dating to the 18th century, of popular culture demonizing African and African diasporic traditions in service of Western supremacy. In response, my remediation efforts centered on using the hierarchical structure of the thesauri to more accurately disambiguate the various religious traditions wrongfully grouped under the “voodoo” record. 

My first step was to look closely at how African syncretic religions came to be misrepresented under the term “voodoo.” The Getty Vocabularies rely on “authoritative” sources to establish credibility for their terms, so I challenged myself to rethink what should be considered authoritative. Instead of relying only on scholarly publications, I looked to community-based resources like Digital Benin, which includes Indigenous classifications, oral histories, and local knowledge. From there, I untangled the primary religious traditions often flattened under “voodoo”: Vodún, Vodou, and Voodoo. Although their names sound similar, they refer to distinct practices with their own histories and therefore deserve their own AAT records. 

Practiced for millennia and still widely observed in Benin, Vodún is a decentralized tradition with diverse practices, including ancestor veneration, deity ceremonies, and animal sacrifice. Developed later during the Atlantic slave trade, Vodou is the Afro-Creole religion of Haiti, created as enslaved West Africans blended Vodún practices with elements of Catholic ritual. It became a source of community, identity, and resistance for enslaved people in Saint-Domingue. At the same time, it was met with violence and repression from colonial institutions, and was often practiced privately, taking different forms across communities, families, and regions. In the nineteenth century, Haitian migration to southern Louisiana carried Vodou into the region, where it blended with Native American traditions, Afro-Creole practices, and folk healing to form Louisiana Voodoo. 

Because Vodún is the historical root of both Vodou and Voodoo, I structured the AAT entries to reflect that relationship. Vodún became the parent term, with Vodou and Voodoo added as its child terms, making the connections between the three visible within the hierarchy. Representing these traditions, however, presented an immediate challenge: their names are often used interchangeably in literature, and shaped by long-standing misconceptions. Some sources still refer to Vodún as “Voodoo,” while others use “Vodun” to describe Haitian Vodou. This inconsistent usage makes it difficult to determine the full range of non-preferred terms needed to guide users toward the correct record. 

Part of my goal was to help users arrive at the right term–even if they started with the wrong one–and to help them notice distinctions they may not have been aware of. I first captured as many non-preferred terms as possible for each religion, knowing that people often search with inconsistent spellings or inherited misunderstandings. I then used qualifiers (the parenthetical phrases that follow an AAT term) to link those variants to the correct preferred record. For example, I added Voodoo (Haiti) as a non-preferred form of Vodou, so a search for “voodoo” also returns Vodou and prompts users to choose the Haitian tradition if that’s what they meant. Likewise, Voodoo (Mississippi River Valley) directs users to Louisiana Voodoo. In practice, these variants catch fuzzy searches, while the qualifiers make the differences clear. Each term’s free-text scope note guides readers through common misidentifications and serves as a pedagogical tool. Additionally, I used the AAT’s related-term relationships to add “distinguished from” links between the records. 

AAT record for Vodún, highlighting the Related concepts section with relational links used to distinguish similar terms. Screenshot by the author, published with permission of the Getty Vocabulary Program. © Getty Research Institute. Data licensed under ODC‑BY.

So as not to erase the term’s history within the AAT, I kept the legacy “voodoo” term under the parent Religions, with the qualifier “(generalized religious reference).” I’ve marked the term as pejorative and used its scope note to explain how the word has been used to misrepresent African and African diasporic religions. 

The Getty Vocabularies organizes knowledge through shifting relationships between recommended language, legacy terms, and related concepts. Taking advantage of this structure, I began to see how political acts of fugitivity could emerge through the user experience of navigating the AAT online.

While many users initially approach the thesaurus through inherited or familiar classifications, the vocabulary’s design encourages movement across terms and hierarchies toward more precise descriptors, unsettling dominant narratives that attempt to fix and ahistoricize marginalized histories.

The polyhierarchical structure of the AAT also meant that revising these “voodoo” records set off a chain reaction, requiring updates to other entries connected to the term. One of the clearest examples was “voodoo dolls.” This concept has been tied to the generic “voodoo” record, even though my research showed these dolls are not features of Vodún, Vodou, or Louisiana Voodoo. Instead, the association is rooted in racist popular culture from the nineteenth to twenty-first centuries. In reality, the objects people call “voodoo dolls” align more closely with binding-figure traditions found in ancient Greco-Roman and Egyptian cultures. To resolve this confusion, I added the more accurate term binding figures as the parent for “voodoo dolls” and flagged “voodoo dolls” with an avoid-use qualifier. I also added culturally specific terms–such as minkondi (African binding figures) and kolossoi (Greco-Roman binding figures)–which are often wrongly grouped under the “voodoo doll” label.

AAT hierarchy for Dolls (figurines), with updated placement of binding figurines, kolossoi, minkondi, and voodoo dolls in the Getty Vocabularies’ taxonomy management system. Screenshot by the author, published with permission of the Getty Vocabulary Program. © Getty Research Institute. Data licensed under ODC‑BY. 

The immediate next step is to reconcile inconsistencies between the AAT “voodoo” terms and related entries in other Getty Vocabularies, especially ULAN—for example, clarifying how the ULAN role “voodooists” maps to the specific diasporic traditions defined in AAT (Vodún, Vodou, or Louisiana Voodoo). This example shows how descriptive remediation work is fundamentally generative, producing new knowledge that ripples beyond any individual term, eventually breaking beyond the Getty Vocabulary ecosystem itself.

KELLI YAKABU:
TAXONOMY IN ACTION: PROCESSING THE ROBERT FARRIS THOMPSON PAPERS

The GRI acquired the Robert Farris Thompson papers in 2019. Thompson, who passed away in 2021, was Professor Emeritus in the History of Art and African American Studies at Yale University from 1965 to 2010, where his scholarship focused on the art history and cultural complexity of the Afro-Atlantic world. His multidisciplinary approach combined his interests of art history, musicology, religion, and anthropology. The collection contains hundreds of research notebooks, manuscript drafts, photographs, teaching materials, audiovisual materials, and thousands of books from his personal library with extensive annotations and inserted ephemera. The collection and its finding aid will be available to researchers in 2027.

One focus of Thompson’s research and teaching was on Haitian Vodou, which he discussed in his seminal work Flash of the Spirit as well as Face of the Gods: Art and Altars of Africa and the African Americas. Many of his research notebooks, manuscript drafts, and notes also contain writings and drawings related to Vodou. In addition, he studied West African Vodún and its influence on and links to Vodou and Voodoo in the American South. 

When I first began processing the collection and learning about Thompson’s scholarship, the new terms had just been released in the AAT. The thorough and nuanced definitions between the religions meant that in item- and folder-level descriptions, I could call out which specific religion Thompson was writing about. Being able to differentiate between the different religions will greatly assist researchers when it comes to discovery and access. While I will use all three subject terms in the collection’s finding aid, future collections can choose which of these subject terms are the most accurate. This is especially true at the GRI with its African American Art History Initiative, where Vodún, Vodou, and Voodoo-related materials are more likely to be found in incoming collections.

Two drawings by Thompson of vèvè Simbi Makaya. Simbi Makaya is a powerful and respected lwa (spirit) in Haitian Vodou who is often associated with serpents, fresh water, and sorcery. © Robert Farris Thompson. Getty Research Institute, Los Angeles (2019.M.21).

These subject terms can also affect researchers by shifting their own understanding of Vodún, Vodou, and Voodoo. A finding aid, like the Getty Vocabularies, may be a researcher's first encounter with these terms, and can be a learning experience for them; I even used these terms in the AAT as a starting point when I was first learning about the collection. Also like the Getty Vocabularies, finding aids can sometimes go through reparative work, a fugitive practice itself, to better describe materials through more accurate and caring language. As more collections receive reparative description work, stronger and more nuanced connections can be made between different archival collections through these linked subject headings. In the case of the Thompson papers, which are being processed for the first time, I already feel empowered to better describe this collection where all three religions are widely documented.

CONCLUSION

Taken together, our work remediating the Getty Vocabularies and processing the Thompson papers shows the theoretical and practical challenges of describing fugitive knowledge–knowledge that resists classification–while still upholding our responsibility to improve its discoverability and access. Though completed at different times and places, our efforts show how controlled vocabularies function as both tools for standardization and frameworks that support archival processing. Yet, they also reveal reparative work itself as a fugitive practice, one that is, like fugitivity, a continuous endeavour rather than a static end goal. As the GRI continues its targeted acquisitions in African American and African diasporic heritage, we are applying lessons from this case study and beginning a cross‑departmental effort for Special Collections archivists to add terminology to the Getty Vocabularies as they process materials, improving descriptive work for future collections.

◡◠◡◠EVOLUTION◡◠◡◠
↯ More Stories ↯

Acid Free

Editors 
Akosa · Stephanie Becker · Elena Diebel · Catherine Falls · Jenny Galipo
Sarah Jardini · Lisa Kahn · SL Leimbach · Vern Molidor · Kelli Yakabu

Editor-in-Chief
Jen Neville

Design
Zaira Torres

acidfree.la@gmail.com   ·   @laacollective