About This Case Study

Every item in an archive comes wrapped in words somebody wrote: a title, an abstract, a scope note, subject headings — the description a researcher reads before, and often instead of, the thing itself. In 2019, Archives for Black Lives in Philadelphia (A4BLiP), a collective of Philadelphia-area archivists, published "Anti-Racist Description Resources," a working document that turned a growing professional argument into concrete recommendations: describe enslaved people by name rather than as property; name enslavers as enslavers; flag racist language from a record's creators rather than reproducing it as neutral fact; audit the legacy finding aids already on the shelf. The premise is the puzzle. Description looks like the most neutral genre in the archive — mere labeling, the phone book of history. The guidelines' starting point is that it never was: every description takes a perspective, and for records of slavery, segregation, and dispossession, the perspective on file has usually been the enslaver's, the official's, the institution's.

The philosopher Miranda Fricker gives the injustice here a name — two, in fact. Testimonial injustice is credibility deflated by prejudice: a speaker's word discounted because of who they are. Hermeneutical injustice is structural: an experience left unintelligible because the shared vocabulary for it was never built. A catalog description is where an archive can commit both — it decides whose perspective a record is described from, and what its vocabulary can say — and reparative description is the practice of rewriting it. The analytical move is to read a description as an instrument: to ask what work its words do, whose personhood they grant or withhold, whose perspective they encode as neutral — and then, because description is authored, to re-author it. This case study is a practicum: rather than analyzing a completed case, you will do the practice. You will learn the standard, select one real description from the National Archives' holdings, audit it line by line, and draft a reparative revision with a memo defending your choices.

Before You Begin

Have ready:

• Archives for Black Lives in Philadelphia (A4BLiP), "Anti-Racist Description Resources," 2019. Visit the project site and locate the Anti-Racist Description Resources document — the recommendations are your audit standard.
• "Archival Research," from the Transformative Research Toolkit, Othering & Belonging Institute, UC Berkeley — the working method for reading finding aids, absence, and bias in the record. Explore the toolkit module.
• DocsTeach, the National Archives' education platform, where you will select the document whose description you audit. Visit DocsTeach.
• A note-taking surface set up for two columns — the original description on one side, your revision on the other — plus space for the audit itself.

A note on language: This exercise requires reading racist language — in the words of the people who created the records, and sometimes in descriptions that quote or preserve them. The guidelines you will read address exactly this: when such language is evidence to be flagged, and when it is the describer's own voice to be replaced. Read it as period evidence of who described whom, and take breaks as you need them.

Before you begin, write down, in a sentence or two each:

• Recall a time an official description of something you know well — your neighborhood in a news story, your family in a form, your community in a database — got it wrong while being technically accurate. What did the description choose to say, and who was it written for?
• Can a description be accurate and unjust at the same time? What would it take to fix one — without erasing the evidence of how the institution used to describe?

The Exercise

Phase 1: The Standard (12–15 minutes)

Read the A4BLiP recommendations, and skim the Othering & Belonging Institute module for its guidance on reading finding aids and archival absence. Then build your instrument: extract five or six of the recommendations into a checklist of testable questions — each one something you can ask of any description and answer with pass, fail, or unclear. ("Does the description name the enslaved, or count them as property?" "Is the creator's racist language quoted and flagged, or reproduced as the archive's own voice?") Note, too, the distinction the guidelines draw between a creator's language kept as flagged evidence and the describer's own voice — repair is not the same as deletion.

For groups: Divide the guidelines among members; each returns with two checklist questions; merge them into one shared rubric of no more than eight items.

Phase 2: The Record (10–12 minutes)

Go to DocsTeach and select one primary-source document connected to the histories this unit examines — slavery, segregation, immigration exclusion, Indigenous dispossession, any era the questions reach. Browse or search by era and topic; each document is presented with a title and descriptive text, and the documents come from the National Archives' holdings, so where the platform points onward to a fuller catalog record, follow it — the fuller the description, the better the audit. Then: copy the description you will audit into your notes verbatim — title, descriptive text, any subject terms — with its URL and the date you accessed it; and read or view the document itself closely enough to know what the description is describing. Resist, for now, the urge to fix anything; the audit comes next.

For groups: Pairs scout candidate documents; the group settles on one primary record and one backup, chosen for descriptions rich enough to sustain a full audit.

Phase 3: The Audit (12–15 minutes)

Work through the description you copied, in four passes:

1. Naming. Who in the document is named in the description, and who appears as a role, a number, or a possession? Whose full name is present — and where a name is absent, what is the absence doing?
2. Perspective. Whose point of view does the description take — the record's creator (the official, the agency, the enslaver) or the people the record is about? Test it: rewrite the description's first sentence from the other party's perspective, and note everything that had to change.
3. Vocabulary. Mark every term that carries period judgment or euphemism. For each: is it the creator's language, quoted and flagged as evidence — or the describer's own voice, presented as neutral?
4. Absence. What does the document show that the description does not say — coercion, violence, resistance, names? Then run your Phase 1 checklist item by item: pass, fail, unclear.

Note where the description's own authorship is invisible — when it was written, by whom, under what rules; the gaps are part of what you are reading.

For groups: Audit the shared record individually against the rubric first, then reconcile line by line. Where two auditors disagree about a term, the disagreement is data — record it.

Phase 4: The Revision and the Memo (15–18 minutes)

Now do the work. Draft the reparative revision: a new title and descriptive paragraph for your document, applying your checklist — naming where the original counted, re-centering perspective, flagging creator language as evidence rather than speaking it. Then write the justification memo, a paragraph to half a page, defending the revision change by change: what you altered, which guideline warrants it, and what you deliberately kept — including original language retained as flagged evidence — and why. Then test where the practice strains:

• What could your revision not repair? Separate what description can fix — perspective, naming, euphemism — from what it cannot: what was never collected, and what the document itself did to the people in it.
• Who has the authority to make your revision real? You are auditing from outside the institution and, quite possibly, outside the communities documented. What would responsible re-description require beyond a better paragraph — consultation, shared authority, even a community's right to refuse description altogether?
• The strain test. Where does the reparative frame fit well, and where does it strain? Consider a description so thoroughly cleaned that it hides how the institution used to speak; or a record whose subjects might not want it surfaced at all. Can repair itself do harm — and how would your memo answer that charge?

For groups: Split the group — half drafts the revision from the reconciled audit, half drafts the memo. Swap, challenge each other's choices, and finalize both together.

Closing Reflection

In two or three sentences, complete this thought:

The most consequential choice in the description I audited was __________. My revision changed __________. The hardest thing to hold steady, after this exercise, is __________.

Write the most specific version you can; the value is in naming exactly which words did the work, exactly what you re-authored, and exactly where your own authority ran out.

A Note on Modes

Solo mode. Work the phases in order. The two artifacts to keep are your completed audit — the checklist with its pass, fail, and unclear marks — and your revision with its memo: together, a before-and-after of one description, with the reasoning on record.

Group mode (3–6 people). Designate a timekeeper. Phase 1 is divided and merged; Phase 2 is a collaborative selection; Phase 3 runs as individual audits reconciled line by line; Phase 4 is split, swapped, and finalized together. The closing reflection works well as individual writing after the session ends.

After the Case Study

• The program behind the practice: Hughes-Watkins, Lae'l. "Moving Toward a Reparative Archive: A Roadmap for a Holistic Approach to Disrupting Homogenous Histories." Journal of Contemporary Archival Studies 5 (2018). Read the article. The audit scaled up to a whole institution — acquisition, description, and outreach as one reparative program; it appears in the unit's Further Readings.
• The concept's source: Fricker, Miranda. Epistemic Injustice: Power and the Ethics of Knowing. Oxford: Oxford University Press, 2007. Publisher page. Testimonial and hermeneutical injustice in full, with a standard for what epistemic justice would require; it appears in the unit's Key Scholarship.
• The right to refuse: Carter, Rodney G. S. "Of Things Said and Unsaid: Power, Archival Silences, and Power in Silence." Archivaria 61 (2006): 215–233. Read the article. The argument that some silences are chosen, and that the power to withhold belongs to the documented — the question your memo's authority section opens; it appears in the unit's Key Scholarship.