Equity Language Commons

A cross-referenced omnibus of progressive equity-language guidance: every source organization's rule side-by-side, per term, with attribution.

🌐 Live site: equitylanguagecommons.org

📖 Browse the commons: Chapters · Glossary · Sources

🤝 Want to contribute? Start at the Contribute page: there are four paths in (suggest, discuss, PR, email) and a step-by-step onboarding section for people new to GitHub.

---

What this is

Every progressive equity-language style guide says something a little different. Sierra Club has a position. NABJ has one. Native Governance Center has one. The Trans Journalists Association has one. Diversity Style Guide aggregates them. Each one is one author's view at one moment in time.

The commons puts them next to each other.

For every term: Latinx, tribe, transgender, unhoused, disability, on and on: every source organization's ruling appears together: their direct quote, their year, their page reference, and a position badge that summarizes where they landed (use, use-with-care, avoid, evolving, etc.). A short synthesis paragraph identifies the cross-source pattern: where guides agree, where they diverge, what's shifted over time.

The commons does not issue its own rulings. It surfaces what the source guides say so readers can see the multi-source picture and decide what fits their audience.

What this is not

• Not a single-author style guide. That's a separate genre. SumOfUs's A Progressive's Style Guide (Hanna Thomas + Anna Hirsch, 2016) is the spiritual predecessor: a single-author synthesis. The commons preserves SumOfUs as one source among many.
• Not exhaustive. The commons covers progressive equity-language guides specifically. Brand identity guides, journalism style references for grammar/AP usage, and clinical/academic terminology guides are out of scope unless they intersect with equity language.
• Not a clearinghouse. Each source guide remains on its publisher's site (or, where the publisher has gone dark, in our private mirror). The commons cites; it does not republish.

How it works

You read a source guide       The commons reads 30+ source guides
     ↓                                  ↓
You see one rule                You see every rule side-by-side
     ↓                                  ↓
You apply it                    You make a more informed choice

The data model centers on terms, sources, and chapters:

• A term page (e.g. /terms/transgender/) lists every source organization's treatment of the term, with attribution, quote, and position badge. A synthesis paragraph identifies cross-source patterns. Audience notes give per-audience operational guidance.
• A source page (e.g. /sources/nlgja/) describes one organization, lists every term in the commons that cites it, and provides access-posture information (hosted publicly, private mirror with link-out, or link-out only).
• A chapter page (e.g. /chapters/sexuality-and-gender-identity/) bundles related terms with a lede, cross-cutting principles drawn from the actual term patterns, and a chronology.

The site also ships with:

• A Glossary Index covering 1,154 primary entries from the source corpus, with aliases grouped under their canonical entries.
• A SQLite build-time index at /data/elc-index.sqlite for anyone who wants to run faceted queries against the commons without scraping.

Current coverage

Item	Status
Indexed commons terms	139
Published chapters	9
Source organizations	37
Glossary Index	1,154 primary entries
SQLite index	Live

Stack

• Astro 7 + TypeScript strict mode: content collections with Zod schemas, static-site generator
• Pagefind: client-side search at build time
• Cloudflare Pages: hosting, free tier
• Cloudflare Email Routing: hello@equitylanguagecommons.org forwarded to maintainer inbox
• Python 3.11+ stdlib only: the build pipeline (PDF extraction, coverage matrix, scaffolders, enrichers, glossary index, SQLite index)
• GitHub: version control, Discussions, Issues, PRs

No databases, no runtime dependencies, no JavaScript framework. The whole site is static HTML + CSS with a single client-side script for search.

Running locally

You'll need Node 20+, Python 3.11+, and poppler (for PDF extraction).

git clone https://github.com/jordankrueger/equity-language-commons.git
cd equity-language-commons/site
npm install
npm run dev   # http://localhost:4321/

For production preview:

npm run build
npm run preview

For the public build:

# Run from the repo root
./scripts/build-glossary-index.py    # Glossary data
./scripts/check-glossary-index.py    # Glossary integrity
./scripts/build-sqlite-index.py      # SQLite index
python3 scripts/lint-content.py      # Content checks
cd site && npm run build             # Production site

Copyrighted source PDFs and full-text extractions are kept in a private maintainer archive and are not part of the public repository. Propose new sources through an issue so a maintainer can archive and verify them.

Detailed contribution workflow in CONTRIBUTING.md.

Editorial principles

Three rules govern every piece of content in the commons:

1. Self-identification is primary. When a guide's rule or the commons's synthesis touches on identification, defer to how the subject identifies. Never override stated preference with a default.
2. No blame-leaning language about source-guide authors. Every style-guide author was doing the best they could with what was available at the time. Use neutral chronological framing ("pre-dates X," "earlier than"): never "outdated," "behind," "hasn't aged well."
3. Editorial synthesis is welcome: editorial judgment of individual authors is not. Identify patterns and trends across the corpus. Don't characterize earlier guides as failures.

Quote attribution rules:

• Every direct quote ≤50 words (fair-use safety margin)
• Every quote cites org, year, source URL, page/section, and confidence level
• Every claim in synthesis must be traceable to a source in the corpus

Full editorial voice rules in CONTRIBUTING.md.

License

The commons consists of three layers, each with its own license:

• Content (synthesis, chapter intros, schema, per-term composition): CC BY 4.0. Share, adapt, and reuse freely with attribution.
• Code (Astro components, build pipeline, deploy tooling): MIT. Standard open-source software license.
• Source-guide quotations: © respective publishers. Used under fair use (each quote <50 words with full attribution). Reuse beyond fair-use citation requires permission from the original publisher.

Full text in LICENSE.

Credits and acknowledgments

The commons cites work from dozens of organizations doing the on-the-ground work of equity language. A full list, with current access status, lives at /sources/. The corpus would not exist without:

• Hanna Thomas + Anna Hirsch: A Progressive's Style Guide (SumOfUs, 2016): the spiritual predecessor.
• Native Governance Center, NAJA / Indigenous Journalists Association, NABJ, NLGJA, Trans Journalists Association, GLAAD, HRC, interACT: primary identity-led style references.
• Sierra Club, NCDJ, APA, Race Forward, Diversity Style Guide, Racial Equity Tools, Color of Change, Define American, Immigrant Defense Project, Dart Center / GCJT, World Food Program USA, Words About War Matter: primary movement and journalism style references.
• The maintainers of the Conscious Style Guide and Diversity Style Guide: peer aggregators whose work is the prior art for this kind of cross-reference project.

Maintained by Jordan Krueger.

Code of Conduct

This project follows the Contributor Covenant 2.1 with a project-specific note about terminology debates. Reports to hello@equitylanguagecommons.org.

---

A side project, not a CampaignHelp deliverable. Built as a gift to the community of progressive communicators who do this work for a living.