Friday, April 1st, 2022
The Virtual International Authority File: Where Women (May) Disappear
Although it is the prototype for a data discovery tool, LWP relies in no uncertain terms on other proprietary data tools like the Virtual International Authority File (VIAF). What we’ve learned, as a team, is that the VIAF is no less fraught in terms of its inclusions, its omissions–what Graban and Sullivan might call its “exclusionary logics” (“New Rhetorics of Scholarship,” 2018), and what LWP understand as “escapes.” In fact, the VIAF is a clear example of LWP’s examination into the ways in which “metadata movements reveal or conceal topical shifts over time” (LWP, “Critical Locations”). VIAF is an international publication metadatabase that aggregates records from databases across the world. It hopes “to lower the cost and increase the utility of library authority files by matching and linking widely used authority files and making that information available online,” and as such, it draws data from over 40 organizations and 30 countries, including contributions from the United States Library of Congress. Like other collaborative data repositories that rely on the shared holdings of other institutions (e.g., the HathiTrust, Google Books, etc.), the VIAF’s output is susceptible to errors in communication that occur across these systems, in spite of its goal to act as “a trusted global authority file.” In fact, since joining the project in August 2021, Ashley has received helpful results from VIAF for little more than 57% of the time; 43% of the time is marked by surprising inconsistencies, if not factual errors, in metadata.
Two cases from the LWP project’s data sets make this point succinctly: Mary Rieser Davidson [1926–2014] and Rose M. Kavana [1867–?]. Both were teachers and both were published authors, yet Rieser’s and Davidson’s publication histories were either non-existent or hidden within the VIAF, rendering them potentially less visible in casual Web-based searches. Situated half a century apart, Davidson and Kavana each represent one of VIAF’s central dilemmas: the dilemma of mismatched nomenclature, owing to name changes, marital status, and engendered data practices that privilege surnames and initials; and co-authorial obscurity, owing to some of the same factors as above, their roles as second or third author, or because they published under more than one alias. Whether their names and publications were intentionally or unintentionally unrecorded, the result is the same: we form our data reliances based on assumptions about what data identifications are available. As the world’s most comprehensive authority file, the VIAF is, at present, the most reliable tool we have for tracing publication histories of most known authors.
More than a simple omission, their under-representation in a global named authority file may well have epistemically harmful implications, reflecting what Catherine D’Ignazio and Lauren Klein might call “unexamined configuration[s] of structural privilege and structural oppression” (Data Feminism). Data results like ours reported here, as D’Ignazio and Klein and others have claimed, argue the need to critically investigate those structures and systems on which we have built our sense of historical possibility. Historically, Klein and D’Ignazio argue, “data have been used as a weapon by those in power to consolidate their control—over places and things, as well as people” (Data Feminism, 14–17). Investigating the inconsistencies that result from one form of consolidation may help us deflect or avoid another kind of consolidation that has more harmful exclusionary effects, and that often occurs when we try to trace our subjects primarily as text objects.
Recently, Ashley embarked on a series of test cases to achieve this kind of tracing, beginning with a subject she discovered at the University of Kansas.
Mary Rieser Davidson taught within the English and Women’s Studies Departments at the University of Kansas from about 1979–1996, where a stable archival collection holding her personal papers, syllabi, and professional work is housed in the Kansas Kenneth Spencer Research Libraries. A 1983 version of her curriculum vitae reports that she had published 4 articles, 2 book chapters, and at least 10 reviews of plays, yet Davidson has no record in the VIAF. A recent VIAF search yielded 43 results for a “Mary Davidson” and 2 results for a “Mary R. Davidson” (see Fig. 1). Knowing that her middle initial “R.” stands for “Rieser” and not “Richmond,” Ashley was able to eliminate one of the two Mary R. Davidsons. She eliminated the other one after noticing their birth and death dates did not match with Rieser Davidson’s (1885–1973, vs. 1926–2014). None of the other 43 options matched because their publications didn’t align with the topics that Rieser Davidson researched and taught throughout her career.
Despite her stable archive and recorded publications, Davidson’s omission from the VIAF could be caused by any number of nuanced factors over which we have no control, and into which the VIAF cannot offer clues. If she is represented somewhere in the record, the VIAF’s naming network doesn’t allow for her ready identification.
Ashley performed a second trace on another author, Rose M. Kavana, who was a teacher at Medill High School in the early 20th century and co-author with Arthur Beatty of Composition and Rhetoric Based on Literary Models. Ashley was unable to pull up VIAF records for “Rose M Kavana” when conducting a meta-search, but she could locate trace evidence of Kavana in other records, namely those tied to “Arthur Beatty.” “Beatty, Arthur, 1869-1943” yielded 14 entries documenting his work, one of which reported Composition and Rhetoric based on literary models, where—when she selected “co-authors” for Beatty’s publications—Ashley discovered an obscure reference to “Kavana, Rose M. (Rose Mary,), 1867- (1)” (see Fig. 2).
Of course, the “co-author” record tied to Beatty’s VIAF data did not link to a discrete record for Kavana. After additional searching for “Kavana, Rose M. (Rose Mary,), 1867- (1)”, “Rose M Kavana”, and “Rose Mary Kavana,” Ashley found no other records that referenced Kavana’s a publication with Beatty.
It is only in nested data where we finally find an explicit and discrete reference to Rose Mary Kavana’s co-authorship status. From Kavana’s record, Ashley navigated to the “About” section and was redirected to Kavana’s WorldCat entry, which identified her as the co-author of Composition and Rhetoric based on literary models. Once in WorldCat, Ashley discovered at least one single-authored publication, The Elements of English Composition, and four other publications that showed Kavana’s involvement.
In this case, Kavana’s authorial and co-authorial presence can ultimately be traced, but not discretely—certainly not independently of her involvement with a male counterpart—and not in such as a way as to demonstrate the full scope of her publication record, which includes data for at least 6 publications not displayed in her co-author’s record on the VIAF. These inconsistencies in the way that Kavana’s records are labeled, compared to the records of others on VIAF, are troubling for how they obscure her disciplinary legacy and subsume her intellectual influence to a categorization and labeling of data that is tied to invisible (or unarticulated) conditions of status.
To be clear, errors and inconsistencies in the way that the women’s data are stored, cataloged, aggregated by, and represented on tools like VIAF are not solely the “fault” of VIAF or its reach. But they are problematic in that they reflect fairly normative and standardized ways of searching. As a result, Davidson and Kavana appear to leave weak intellectual legacies given how their contributions become un-authorized within the very same open and democratizing system that should hold them aloft.We might ask ourselves, what occurs before, during, and after the labeling of these records and how might even user-generated metadata still be determined by these kinds of representational omissions of female authors? How might the valuable data points located under “additional information” or treated as auxiliary become more visible and earn a more discrete place in the VIAF universe (and that of the powerful international federations that help to constitute it) ? Can such a broadly open and collaborative tool operate in such a way as to eventually overcome those dark corners?
— T. Graban and A. Pendleton
