1 of 15

Reparative Processing: A Case Study in Auditing Legacy Archival Description for Racism

slides @ goo.gl/uwjQpg

Kelly Bolding

Project Archivist, Americana Manuscript Collections

Princeton University Library

@okbolding

Good morning, and thanks for being here. My name is Kelly Bolding and I’m a processing archivist at Princeton University. I’ll be addressing the theme of this panel by talking about a project we recently initiated to audit and remediate racist description in our legacy finding aids. I will be talking about how I arrived at this work, my own conceptual approach to it, as well as some specific strategies that have emerged from early case studies.

At Princeton, I’m responsible for the arrangement and description of manuscript collections that deal with early American history during the 18th and 19th centuries. This role involves redescription of older collections, as well as processing new acquisitions, including many related to the history of slavery in the Americas.

In my opinion, the truly radical work related to ethical archival practice is happening in the realm of community archives rather than in predominantly white institutions like my own. Learning from that work, I believe that all of us still have a responsibility to leverage the resources of our institutions to address what Bergis Jules describes as “failures of care around the legacies of marginalized people in the archives.”

2 of 15

Michelle Caswell

“Archives, Records, and Memory”

class @ UCLA

(Fall 2016)

graphic design by

Gracen Brilmyer

3 of 15

“the way in which black women of the Americas have been historically misrepresented or not represented at all can be connected to recruitment and retention problems in the archival profession.”

- Kellee E. Warren, “We Need these Bodies, but Not their Knowledge: Black Women in the Archival Science Professions and their Connection to the Archives of Enslaved Black Women in the French Antilles”

4 of 15

“Is the descriptive language I am using respectful to the larger communities of people invested in this record?”

- Michelle Caswell & Marika Cifor, “From Human Rights to Feminist Ethics: Radical Empathy in the Archives.”

“caring for the people who interact with your words.”

- Alex Kapitan, Radical Copyeditor

5 of 15

iterative processing

reparative processing

Updating racist description in finding aids is based in the principle that archival description is iterative, that is, its process is phased, and its product is always undergoing revision as new knowledge and time become available. Archivists are already engaged in the clean-up of messy legacy metadata. I’m proposing that remediation of racist description should become part of routine data management workflows. The process of replacing racist description itself will also be iterative since respectful language changes over time and mistakes will be made.

Iterative processing is also the term I use to refer to the strategies for prioritizing labor resources that are what I think of as defining an MPLP approach to archival processing. Reparative processing is the concept and practice I offer in response to MPLP’s call for prioritization. Reparative processing means centering past failures of care and the possibility of uncovering hidden voices as a factor in prioritizing which collections -- or portions of collections -- should receive work above and beyond baseline processing tiers.

I should also clarify that reparative processing is insufficient to count as actual reparations, which would involve direct payment to those harmed by the legacies of slavery, white supremacy, and mass incarceration. It is, rather, inspired by the concept of reparations and its strategic redistribution of resources as a means of redressing past injustices.

6 of 15

XQuery survey

As a first step towards strategically auditing our legacy archival description, we’re repurposing a method we’ve used in the past to search our EAD data when surveying for born-digital media. This method involves using XQuery, a scripting language for querying large sets of XML data, along with a regular expression that matches a specific pattern of words we want to look for. The output of the XQuery survey is a list of descriptive components that should be reviewed as a part of the description audit. While we started out looking for problematic words used to refer to race and ethnicity, we quickly realized we would need to cast a much wider net since certain non-derogatory descriptors were often used incorrectly, and many words needed to be evaluated in context. Our current approach is to identify components where people from marginalized communities are described and then assess whether or not our description is doing them justice or harm. While this approach generates some false positives, it is largely manageable in our institutional context because so many of are collections are so white.

7 of 15

describing

whiteness

examples from early American history collections

Describe white ethnicities (ex. Scottish American)

Call it like it is (settlers, colonists)

“The first white” ≠ “The first”

Use active voice to describe oppressive relationships

Denormalize colonial place names

8 of 15

“archivists often write massive memorials and monuments to wealthy, white, cisgendered and heterosexual men, including selective details about the creator that have minimal bearing on the records, and instead serve to valorize and venerate white western masculinity.”

- Jarrett M. Drake, “RadTech Meets RadArch: Towards A New Principle for Archives and Archival Description”

9 of 15

what is your <bioghist> doing?

https://github.com/kellybolding/scripts/blob/master/terms_of_aggrandizement.xquery

10 of 15

Name the subjects of records to the extent you name creators

11 of 15

12 of 15

describing

(not erasing)

racism

(is context)

Revise problematic language in archivist-supplied titles and notes

Retain* (but contextualize) creator-supplied racist folder titles

quotation marks
<scopecontent> or <processinfo> note
@type = “original” or “supplied”

Describe racist collecting policies & practices that shaped collections

esp. “assembled” topical collections
<scopecontent> note

*This is one approach. Others may make more sense in other contexts.

Changing the language we use to describe archival collections does not mean we’re making archival collections not racist. Instead, it means considering racism as important context for interpreting archival collections and describing that context as we would other information that affects users’ understanding of a body of records.

With this in mind, we plan to retain creator-supplied racist folder titles but clearly indicate when language came from the creator, either through the use of quotation marks or in a processing or scopecontent note. Another aspirational goal is to transparently describe racist collection development policies and practices that led to a certain group of materials being aggregated by a repository. While potentially risky, describing these practices also holds our institutions accountable for changing them.

13 of 15

be transparent / be accountable

preserve evidence of racism in archival description for future study

via version control or maintaining copies of legacy finding aids
record that revisions were made in <processinfo> note

provide mechanisms for users to report problematic description

Suggest a Correction button

Returning to Caswell’s action item, before we begin strategically updating racist language in finding aids, we need to ensure we have a method in place to “keep a note of [our] changes so that users can examine the history of the finding aid as an artifact.” At Princeton, we are indicating when description has been revised in a processing note, but we’re also lucky to have a version control system already in place that has kept track of every change to our finding aids since 2012. For finding aids that precede EAD conversion, we’re considering establishing a collection of legacy finding aids that is open to researchers.

Another aspect of accountability is making sure our description is assessed by outside communities without putting an uncompensated burden on those communities to do our work for us. Right now, we’re experimenting with using the “Suggest a Correction” feature of our finding aids site to encourage users to alert us to problematic description, but community input is an area where we still have much work to do.

14 of 15

inspirations*

Other folks out there leading ethical description work:

Annie Tang Dominique Luster

Dorothy Berry

Gloria Gonzalez Jasmine Jones

Rachel E. Winston

Sam Winn

Conversations with these folks:

Princeton colleagues: Elvia Arroyo-Ramírez, Faith Charlton, Valencia Johnson, & Alexis Antracoli

Archives for Black Lives in Philly, including Katy Rawdon

Alison Clemens

*All errors are my own.

15 of 15

Works Cited

Caswell, Michelle. "Teaching to Dismantle White Supremacy in Archives," The Library Quarterly 87, no. 3 (July 2017): 222-235. https://doi.org/10.1086/692299

Caswell, Michelle, and Marika Cifor. “From Human Rights to Feminist Ethics: Radical Empathy in the Archives.” Archivaria. 81 (Spring 2016), pp. 23-43.

Drake, Jarrett M. “RadTech Meets RadArch: Towards A New Principle for Archives and Archival Description.” Delivered at the Radcliffe Workshop on Technology & Archival Processing, April 4-5, 2016, in Cambridge, MA. https://medium.com/on-archivy/radtech-meets-radarch-towards-a-new-principle-for-archives-and-archival-description-568f133e4325

Jules, Bergis. “Confronting Our Failure of Care Around the Legacies of Marginalized People in the Archives.” Delivered at NDSA annual meeting, November 9, 2016, in Milwaukee, WI. https://medium.com/on-archivy/confronting-our-failure-of-care-around-the-legacies-of-marginalized-people-in-the-archives-dc4180397280

Kapitan, Alex. Radical Copyeditor (blog). https://radicalcopyeditor.com/

Warren, Kellee E. 'We Need these Bodies, but Not their Knowledge: Black Women in the Archival Science Professions and their Connection to the Archives of Enslaved Black Women in the French Antilles', Library Trends, vol. 64/no. 4, (2016), pp. 776-794.