Activists have worked to recover women’s history since the late 1800s, their efforts boosted by each wave of the feminist movement of the last century. Repositories, such as the Harvard Radcliffe Institute’s Schlesinger Library, Duke University’s Sallie Bingham Center for Women’s History and Culture and Smith College’s Sophia Smith Collection of Women’s History, specifically collect papers by and about women. But many nationally prominent centers, including the Smithsonian, are faced with a challenge of colossal proportions, even if they have significant funding to work with.
To make women discoverable (in other words, to add gender to the descriptive text of their existing entries), staff at the Smithsonian are poring over millions of pages of archival records and searching for women who were either purposefully dismissed or overlooked. In addition, data scientists are building new tools for finding these hidden histories. The Smithsonian told me that custom-made machine learning models will allow metadata fields to be searched by pronouns, speeding the slow and painstaking (human) task of record-locating and updating that otherwise might take decades.
Their work is already paying off. One discovery involves Margaret Moodey, an accomplished scientific researcher whose identification of gems and fossils is foundational to the National Museum of Natural History. Museum records from early in the 20th century, written and maintained by mostly male museum staff, documented her job title as “secretary” and “clerk,” not “scientist,” making her impossible to find in archival searches for that term.
Elizabeth Harmon, digital strategy specialist at the Smithsonian American Women’s History Museum, found a paper trail of Moodey’s contributions only when she examined the museum’s original geology department records as part of the metadata overhaul effort.
Yale University Library and others have taken on this big data challenge, too, by repairing records to include full name information for women. Yale’s metadata coordinator ran a search to locate every record in which a woman was identified only by her husband’s name, such as “Mrs. Charles Healy.” Researchers next examined biographical documentation, such as marriage announcements and obituaries, to provide names and identifying details. Yale’s project found more than women’s names in the process; it determined women’s nationalities and occupations, among other insights.