It was once accepted practice to call married women by their husbands’ names, with the honorific “Mrs.” attached—for example, “Mrs. Franklin Delano Roosevelt.” For decades, this practice extended to archival finding aids and catalog records created at the Columbia University Rare Book and Manuscript Library (RBML). However, this is problematic from both a feminist and an information retrieval perspective. Calling a married woman by her husband’s name in a finding aid erases her identity and substitutes her appearance in the archival record with her husband’s.

At best, this additional layer of ambiguity hides potentially relevant materials from researchers. Twenty-six files in 10 different RBML archival collections containing material related to Eleanor Roosevelt called her “Mrs. Franklin Delano Roosevelt.” How many 21st-century researchers studying Eleanor Roosevelt search for her using those terms?

At worst, calling a married woman by her husband’s name in a finding aid is flat-out wrong. Naming conventions are culturally specific. Despite the western tendency to call her “Madame Chiang Kai-Shek,” the First Lady of the Republic of China from 1928 to 1975 was legally named Soong May-ling. Naming conventions also change over time. The photographer Jeanne Moutoussamy-Ashe chose to hyphenate her last name when she married tennis star Arthur Ashe in 1977. Calling her “Mrs. Arthur Ashe” is incorrect.

Like many of our colleagues across the profession, RBML archivists adapted to working remotely this spring and summer by auditing and remediating issues with existing archival metadata. In May, former Herbert H. Lehman Curator Tamar Evangelestia-Dougherty, now Associate University Librarian at Cornell, noted our “Mrs. Franklin Delano Roosevelt” problem on Twitter. This project took shape at her urging.

Columbia University Libraries uses an externally hosted instance of ArchivesSpace to store and manage archival metadata. A nightly automated script runs across our published finding aids, compares them to an exported Encoded Archival Description (EAD) dataset of finding aid components marked “publish” in ArchivesSpace, and updates the published finding aids with any changes made to the EAD since the script last ran. An additional benefit of this process, developed by Special Collections Analyst David W. Hodges and RBML Head of Archives Processing Kevin Schlottmann, is that we always have a full and up-to-date EAD dataset that we can search as desired, for example using targeted xQueries against specific types and levels of EAD elements.

To locate finding aid components in which married women were called by their husbands’ names, Schlottmann queried collection-level scope and content notes and file-level unit titles for the text string “Mrs.” Metadata associated with Agents (name authority and relationship records in ArchivesSpace) is also exported weekly via the ArchivesSpace API and written into a Google Sheet. Schlottmann searched this sheet for the “Mrs” text string as well. He pasted the results of all three of these queries, as well as the collection ID number, ArchivesSpace unique identifier, title, and Columbia University Libraries repository code, into three tabs of a shared Google Sheet.

Identifying components with the text string “Mrs” only targets potential instances of married women called by their husbands’ names; manual review is still necessary to confirm their presence. Public Services Assistant Vianca Victor and I chipped away at this work over the summer. Titles of published works (Mrs. Dalloway) and names of fictional characters (Mrs. Doubtfire) both appeared in our results. After beginning the project, we learned that Columbia University Libraries staff had already completed some of this work in 2008. These remediated elements also appeared, though they were easy to distinguish: they took the form “Eleanor Roosevelt (Mrs. Franklin D.)” or “Mrs. Franklin D. Roosevelt (Eleanor).” Perhaps it was a sign of the times that we called these results “false positives.”

The Rare Book and Manuscript Library was closed from March 16 until September 8. We did not have access to the collections—though even if we had, we would likely not have found the women’s own names in many of the original documents. Instead, we turned to digitized archival materials and online reference sources. The following were especially useful:

• The New York Times: Obituaries and marriage announcements were key. We found that these were often the only places in the Times where married women were referred to by their own full names. Otherwise, the paper referred to them by their husbands’ names until approximately the mid-1970s. (Amelia Earhart famously objected to this practice.) Oddly—and sadly—we found a handful of obituaries where the deceased was only identified by her husband’s name, with no mention of her family of origin or her life before marriage. One example is the Vassar College professor of economics Mrs. Arthur Hutchinson. We hope that the Vassar College Archives holds a record of Professor Hutchinson’s full name.
• Wikipedia: Biographical articles frequently include the name(s) of the subject’s spouse(s). While Wikipedia also has a gender bias problem, we found that many of the married women we identified using Wikipedia actually had biographical articles of their own. Thanks to the WikiProject Women for this pleasant surprise!
• Find a Grave and Ancestry.com: Many of the women we identified were socially prominent, but some were not. These resources were especially helpful for identifying women whose lives were not extensively covered in the New York Times or other major newspapers.
• The Columbia Spectator and Barnard Magazine: Excellent for identifying women in Columbia- and Barnard-related collections. College alumni magazines in general were useful resources; they tend to include married women’s maiden names because those were the names the women used as students.
• SNAC (Social Networks and Archival Context): SNAC aggregates biographical information from archives and other freely available resources such as Wikipedia. The content is reliable and the emphasis on context is especially useful for making sure that the Mrs. John Smith in question is the correct Mrs. John Smith. (We hope to contribute some of our findings from this project to SNAC as well.)

We tracked our work by adding columns to the Google Sheet. Schlottmann added a column for our suggested new scope and content note, Agent record, or unit title. When we finished the project, Kevin posted the data in this column to the corresponding field of the associated JSON object in ArchivesSpace via the API. Victor and I also found that adding columns for the woman’s full name, the source we used, and any additional notes were helpful for our own reference.

In the end, we were able to identify the full names of 34 out of 37 women called by their husbands’ names in collection-level scope and content notes, and 26 of 28 in Agent records. In file-level unit titles, we identified 1,257 individual women who were previously called by their husbands’ names. These names appeared in 1,692 file titles, meaning that many of the women we identified appeared in our finding aids at least twice.

I generally followed the parenthetical structure used in the 2008 remediation project when revising finding aid components to include women’s full names. While not strictly egalitarian—to my knowledge, none of our finding aids have ever referred to “Franklin Delano Roosevelt (Mr. Eleanor)”—this structure was a quick way to preserve information about the relationships between subjects of our collections. I made exceptions for a few women who were sufficiently well-known that their spouses’ names are also general knowledge. Eleanor Roosevelt, of course, was one. Yet I still feel uneasy about the implication that only very famous married women deserve to be described independently from their husbands.

Finding aid components where both spouses were addressed by the husband’s name relieved this particular stressor, though they could be tricky to revise in a way that made the marital relationship clear without sounding stilted or breaking up text strings that could function as search terms. The most striking example of this was a file titled “Dr. and Mrs. McIntosh.” I changed it to “Drs. Rustin and Millicent Carey McIntosh,” as both spouses held terminal degrees, and Millicent Carey McIntosh was dean and the first president of Barnard College. Few researchers are likely to think of her as “Mrs. Rustin McIntosh.” Overall, I hope we can implement a better way to convey information about marital relationships soon.

Victor and I made a dent in our finding aids, but much work remains to be done. There are still 1,236 of our file titles that still call women by their husbands’ names. Without dates—which are often not supplied in our older finding aids—calling women by their husbands’ names makes it difficult or impossible to distinguish between the wives of men who married more than once. Does Mrs. Eugene O’Neill refer to Kathleen Jenkins, the magazine writer Agnes Boulton, or the actor Carlotta Monterey? We will need to review the original materials to find out. Other barriers included husbands’ names that were very common or were misspelled. It was nearly impossible to identify married women when we were unable to identify their husbands.

It was also difficult to identify women who were not wealthy or socially prominent. We were not very successful at identifying the married women who were paid pensions by Andrew Carnegie. As such, we recognize that while this project increased the visibility of archival records of some women, it reinforced the marginalization of working class and poor women, women of color, and people in non-heteronormative relationships in the archival record. These erasures deserve to be addressed. We plan to apply some of the techniques this project used in future reparative reprocessing work, though we also recognize that work has a much broader scope.

The opportunity to practice using archival and archives-adjacent resources outside the RBML was an additional benefit of this project. Without physical access to our collections, we had to get creative about locating information in other ways. While our reading room has now reopened to current Columbia University affiliates who comply with all university COVID-19 safety requirements, unaffiliated users are not currently able to visit us in person. The research skills we refreshed and strengthened while identifying married women in our finding aids will also help us provide efficient and useful alternatives to in-person visits for our users.

Celeste Brewer is a processing archivist at the Rare Book and Manuscript Library, Columbia University. Her professional interests include reparative reprocessing, transparency in archival processing, accessibility, and archival ethics and accountability. She encourages everyone to review the Archives for Black Lives in Philadelphia Anti-Racist Description Working Group’s Anti-Racist Description Resources (October 2019) and incorporate its recommendations into their work.